Reconstruction of a Ross lost Cambrian Series 2 mixed siliciclastic–carbonate platform from carbonate clasts of the Shackleton Range, Antarctica

ABSTRACT The presence of archaeocyath-bearing clasts from Cenozoic tills and Cambrian Mount Wegener Formation reveal erosion of a hidden Cambrian carbonate platform in Shackleton Range, Antarctica. We provide microfacies, paleontological, diagenetic and tectonically induced fabric data from carbonate clasts which, in addition to available geochemical and geochronological data from Shackleton Range, allow the paleoenvironmental reconstruction of a lost Cambrian Series 2 mixed siliciclastic–carbonate platform that was developed and eroded during the Ross orogeny. Carbonate production was dominated by non-skeletal grains in possibly restricted platform-interior and oolitic shoal complex settings, while open subtidal sub-environments (calcimicrobe carpets, calcimicrobe–archaeocyath patch reefs, muddy bottoms) were dominated by a diverse calcimicrobe assemblage and/or by secondary to accessory heterozoan assemblage (archaeocyaths and other sponges, chancelloriids, hyoliths, coralomorphs, trilobites, echinoderms). We describe a Botoman assemblage with 34 archaeocyathan species among 12 existing archaeocyathan genera. A new archaeocyath family Shackletoncyathidae is proposed. New species (Rotundocyathus glacius sp. nov., Buggischicyathus microporus gen. et sp. nov., Paragnaltacyathus hoeflei, Shackletoncyathus buggischi gen. et. sp. nov., Santelmocyathus santelmoi gen. et sp. nov., Wegenercyathus sexangulae gen. et sp. nov.) and Tabulaconus kordae coralomorph are reported from Antarctica for the first time. Archaeocyathan fauna share few species with contemporary fauna of South Australia (9) and even fewer with the Antarctic platforms of the Shackleton Limestone (2) or the Schneider Hills limestone (1). Similarity is greater with Antarctic allochthonous assemblages of Permo-Carboniferous tillites from Ellsworth Mountains (2), Cenozoic deposits from King George Island (4) or Weddell Sea (1). The Shackleton Range lost/hidden platform shows a distinct entity related with its tectonosedimentary evolution, in a possible back-arc basin on the Mozambique seaway during the E and W Gondwana amalgamation, which distinguishes it from those developed on the palaeo-Pacific margin of the E Antarctic craton.

the TAM and in the Ellsworth Mountains (EW) at the WAntarctica region.
The intraplate TAM belt separates the E from W Antarctica regions (Fig. 2a) that present clear differences in their geological evolution. The E Antarctic Craton (EAC) is formed by the amalgamation of Precambrian terranes during different Precambrian-Cambrian orogenies (e.g., Boger 2011). The geologic record of the TAM belt (see Goodge 2020)

includes a Mesoarchaean and
Paleoproterozoic basement (part of the EAC). This basement was rifted and developed an Andean-style Gondwana convergent margin with carbonate deposition (PM and cTAM in Fig. 1), basin inversion and clastic molasse-type sedimentation during the late Neoproterozoic to early Palaeozoic Ross orogenic cycle. Intra-cratonic and foreland basin sedimentation and mantle upwelling (Gondwana breakup) continued during Late Palaeozoic and Early Mesozoic times, and experimented rift-shoulder uplift   Krohne et al. (2016) after Tessensohn et al. (1999a) and Kleinschmidt et al. (2002). Rock outcrop localities with archaeocyaths are indicated with numbers in bold: 1 = Otter Highlands; 2 = Haskard Highlands, 3 = La Grange Nunataks; 4 = Herbert Mountains; 5 = Pioneers Escarpment; 6 = Read Mountains; 7 = Du Toit Nunataks; 8 = Stephenson Bastion. Abbreviations: OHT = Otter Highlands Thrust; MWT = Mount Wegener Thrust. (C) Schematic n-S cross-section through the centre of the Shackleton Range. Tectonic units I-IV according to Tessensohn et al. (1999a), who interpreted the Shackleton collisional orogen as a result of the final amalgamation between E and W Gondwana during Late Precambrian-Cambrian time. I = Proterozoic basement; II = Ophiolitic complex; III = low-grade metasedimentary units with Cambrian fossils (including archaeocyaths) and Ross deformation ages; IV = Proterozoic basement (East Antarctic Craton) and autochthonous sedimentary cover. related to Cenozoic extension (Goodge 2020). On the other hand, the W Antarctica extensional province is formed by Palaeozoic-Mesozoic microcontinental blocks (outboard terranes) as the Ellsworth-Whitmore Block (EWB), Thurston Island, Marie Byrd Land and the Antarctic Peninsula (Dalziel & Elliot 1982;EWB, TI, MBL, AP in Fig. 1) and both the Weddell Sea and WAntarctic rift systems (Fig. 2a). From ∼500 Ma, the WAntarctica region is a tectonically active margin between the subducting paleo-Pacific oceanic plate and the EAC (Jordan et al. 2020 and references therein) (Fig. 1). Thus, the best known Cambrian carbonate successions in Antarctica (TAM and EW) were not part of the same carbonate platform, nor they did share a common tectonosedimentary evolution during the Cambriansomething that is key when establishing the origin of allochthonous archaeocyathbearing clasts in the neighbouring regions of southern Gondwana (Fig. 1) On the other hand, the record of Antarctic allochthonous archaeocyathan assemblages is abundant, scattered and, at times, very far from the hypothetical source areas, and some of them have not yet been studied. However, because less than 2 % of the bedrock is exposed in Antarctica, archaeocyath-bearing clasts are key records that give us a chance to solve many geological problems. For instance, in the TAM, the presence of archaeocyath-bearing clasts in Cambrian conglomerates (e.g., Starshot Formation, Douglas Conglomerate) was useful to establish maximum depositional ages, lithostratigraphic relationships and relative timing of deformation, uplift and erosion of the Cambrian Series 2 carbonate platform (Rowell et al. 1986(Rowell et al. , 1988Myrow et al. 2002a). In recent tills, the presence of archaeocyathbearing clasts derived from Cambrian polymictic conglomerates would indicate that the Douglas Conglomerate could be hidden under the ice at least as far S as the Beardmore Glacier ).
In the EW (EWB in Fig. 1), the Heritage Group contains thick carbonate successions such as the middle Cambrian Drake Icefall Formation and the middle-upper Cambrian Minaret Formation . However, the only record of lower Cambrian carbonate sedimentation occurs as reworked carbonate clasts in the lower Cambrian conglomerates of the Kosco Peak Member (Castillo et al. 2017) from the Heritage Group and in Permo-Carboniferous tillites of the Whiteout Conglomerate (Buggisch & Webers 1992). In fact, the Whiteout Conglomerate represents the Late Palaeozoic Gondwana glaciation sedimentation in W Antarctica (Matsch & Ojakangas 1992). But the source area of the carbonate clasts is unknown in the EW; specifically, the archaeocyath-bearing clasts give a lower Cambrian age (Debrenne 1992) and no clasts believed to be from the Minaret Formation are found (Buggisch & Webers 1992).
In Late Palaeozoic Gondwana glaciation deposits outside Antarctica, the presence of archaeocyath-bearing clasts ( Fig. 1) allows paleobiogeographic correlations between Antarctica and other erratic assemblages in S Gondwana where the record of Cambrian carbonate successions is unknown, as in the Falkland Islands (Stone et al. 2012), South Africa (Debrenne 1975), Namibia (Perejón et al. 2019) and Argentina (González et al. 2013). In this regard, the autochthonous archaeocyathan assemblages from the lower Cambrian Shackleton Limestone (TAM ;Hill 1964b;Debrenne & Kruse 1986, the Schneider Hills limestone (Pensacola Mountains (PM), TAM; Konyushkov & Shulyatin 1980;Debrenne & Kruse 1989), the middle Cambrian Nelson Limestone (PM; Wood et al. 1992) or the upper Cambrian Minaret Formation (EWB; Debrenne et al. 1984;Henderson et al. 1992) are, so far, essential to establish the likely source areas of allochthonous records. But, are these the only Cambrian carbonate successions, or are there other records hidden under the ice? And, if so, how can we infer and differentiate them? The Shackleton Range (SR in Fig. 1) is interpreted as a collisional orogen due to the final amalgamation between E and W Gondwana during Late Precambrian-Cambrian times (Tessensohn et al. 1999a), but the presence of shallow-water marine successions from the Cambrian is unknown in this sector. However, glacial erratics with Cambrian brachiopods from a locally sourced area were the first fossils collected by the Shackleton Range geological expedition in recent moraine deposits (1970Clarkson 1971;Thomson 1972). And later, during the German geological expedition GEISHA (1987)(1988), Cenozoic glacial erratic archaeocyath-bearing clasts were discovered. Höfle & Buggisch (1995) suggested that a major expansion of the Antarctic Ice Sheet carried these erratic carbonate clasts from their hypothetical source area: the Nelson Limestone in the PM, the southernmost part of the TAM (PM in Fig. 1). Some years later, during the EUROSHACK expedition (1993)(1994), a sampling of new archaeocyath-bearing clasts from the Mount Wegener Formation was carried out.  suggested an early Cambrian Atdabanian age for the synorogenic upper slope to basinal deposits of the Mount Wegener Formation on the basis of the potassium-argon (K-Ar) ages of detrital muscovites and the presence of the trace fossil Oldhamia, calcimicrobes and archaeocyaths. Therefore, the carbonate clasts from the Mount Wegener Formation come from the erosion of shallow marine deposits; although geochemical and paleocurrent data suggest a northern source area, its final provenance is not yet determined .
The aims of this paper are: (a) to analyse the microfacies of the Cambrian carbonate clasts from the Shackleton Range (Cenozoic tills and Cambrian Mount Wegener Formation); (b) to identify the diagenetic processes that carbonate clasts and host rock conglomerates have undergone, from the shallow marine platform stage up to the very low-grade metamorphic overprint and nappe-tectonic stage; (c) to analyse the taxonomic relationships of the archaeocyath specimens; (d) to determine the age of archaeocyathan assemblage to set the minimum age of the carbonate platform they derived from; (e) to establish possible palaeobiogeographic correlations between the archaeocyathan assemblage from the Shackleton Range sector and other coeval assemblages; (f) to give new age constraints of the Mount Wegener Formation; (g) to figure out the different depositional subenvironments of the lost carbonate platform in the Shackleton Range sector; and (f) to compare this lost carbonate platform and its tectonosedimentary evolution with other contemporary Antarctic platforms.

Geological setting
The Shackleton Range (lat. 80°-81°S, long. 19°-31°W) is the most prominent massif at the Coats Land region of Antarctica, extending for about 200 km in an E-W direction. The range is bounded by the Slessor and Recovery glaciers that drain the E Antarctic Ice Sheet into the Filchner Ice Shelf (Weddell Sea) (Fig. 2a, b). Above permanent snow and ice, the closest outcrops/blocks are the Whichaway Nunataks to the S (130 km), the Argentina Range at the PM to the SW (240 km) and the Theron Mountains to the N (150 km). The first geological survey of the western part of the Shackleton Range was carried out by Stephenson (1966) during the Trans-Antarctic Expedition (1955)(1956)(1957)(1958). Later, the geology of the range was successively studied by numerous field parties from the UK, USSR, Germany, Italy and the USA (see Clarkson 1995;Kleinschmidt 2007;Fig. 3 and references therein).
The geology of the range consists mainly of medium-highgrade amphibolite facies of the Proterozoic Shackleton Range Metamorphic Complex (SRMC) (Clarkson 1972), which is made of infracrustral rocks from the Neoarchaean-Paleoproterozoic Stratton Group (Tessensohn & Thomson 1990) and the Paleoproterozoic-Mesoproterozoic Read Group (Pankhurst et al. 1983(Pankhurst et al. , 1985Fig. 3). The snow and ice cover made it difficult to establish lithostratigraphic relationships between the scattered rockoutcrops; in fact, only two sedimentary contacts have been observed between the Proterozoic SRMC and the younger sedimentary successions, disregarding the Cenozoic deposits (Fig. 3). However, the nappeand thrust-tectonic soon became apparent and finally well established (Marsh 1983;Buggisch et al. 1990;Brommer 1998;Kleinschmidt et al. 2001). Therefore, the Shackleton Range is subdivided into different tectonostratigraphic units developed during the Ross orogeny (Buggisch et al. 1994b;Buggisch & Kleinschmidt 1999;Tessensohn et al. 1999a;Fig. 2c, A-C in Fig. 3).
The metasedimentary to metavolcanic supracrustal rocks of the Pioneers Group (Tessensohn & Thomson 1990;Roland et al. 1995) and the rocks of the Ophiolitic Complex (Talarico et al. 1999) with eclogite facies metamorphism (Schmädicke & Will 2006) occur in the northern part of the range. Both appear tectonically interleaved with the Stratton Group ( Fig. 3; Schubert et al. 1995), forming the Northern Belt (B in Fig. 3). In the southern part of the range, the Read Group (Tessensohn & Thomson 1990;Olesch et al. 1995), which is part of the EAC, is unconformably overlaid by the Watts Needle Formation (Marsh 1983), interpreted as Criogenian-Ediacaran marine mixed platform sedimentation (A in Fig. 3). The boundaries between the northern and southern belts are thrusts (OHT and MWT in Figs 2b,3), which bound the Mount Wegener Nappe (Kleinschmidt & Buggisch 1994;Buggisch & Kleinschmidt 2007 and references therein). This allochthonous tectonic unit comprises very-low-to low-grade metasedimentary rocks (Clarkson 1972;C in Fig. 3) as the ?Mesoproterozoic to Neoproterozoic Wyeth Heights and Stephenson Bastion Formations, and the Cambrian Mount Wegener Formation, which hosts allochthonous archaeocyath-bearing clasts in conglomerates and olistoliths . The Shackleton Range is a collisional orogen related to the sinistral collision between the E Antarctic and Kalahari Cratons with the closure of the Mozambique Ocean during the final amalgamation of the E and W Gondwana (Talarico et al. 1999;Tessensohn et al. 1999a;Kleinschmidt et al. 2001). This late Pan-African event produced the westward overthrusting of the Northern Belt and the southward overthrusting of the Mount Wegener Nappe over their foreland, the EAC (Read Group) and its sedimentary cover (Watts Needle Formation) according to Buggisch & Kleinschmidt (2007).

Stratigraphic record of the allochthonous archaeocyaths in the Shackleton Range
In the Shackleton Range, the ice divide separates tributary glaciers that flow northwards to the Slessor Glacier from those that flow southwards to the Recovery Glacier. The archaeocyathbearing clasts are found at the S of the Fuchs Dome (central ice cap) and the Shotton Snowfield as glacial erratics in Cenozoic tills and conglomerates hosted by the Cambrian Mount Wegener Formation (Fig. 4a, b).
The glacial erratic archaeocyath-bearing clasts ( Fig. 4a-c) were discovered in the Shackleton Range during the German geological expedition GEISHA (1987)(1988). At that time, the presence of archaeocyath-bearing clasts in the conglomerates of the Cambrian Mount Wegener Formation was unknown (Fig. 4b,d) and knowledge about the uplift of the Shackleton Range and the glaciology of the region was limited. Höfle & Buggisch (1995) analysed the glacial morphology and till deposits from the Shackleton Range. Some of the data and conclusions are summarised below. The glacial erratic archaeocyath-bearing clasts occur in two different situations (Fig. 3): as Cenozoic tills on top of the Stephenson Bastion table mountain (Fig. 4c) and as part of a subrecent moraine near Du Toit Nunataks (Fig. 4b).
The archaeocyath-bearing clasts consist of wackestone, floatstone and boundstone with Epiphyton, Renalcis, Batenevia ramosa Korde and trilobites. Many of the outcrops at the Shackleton Range are table mountains, relics of an exhumed peneplain (Stephenson 1966;Skidmore & Clarkson 1972;Marsh 1985). The scarce till deposits on the table mountains occur in places protected from erosion. In fact, easily weathered fossiliferous limestone clasts appear in hollows and depressions. On the relict till deposits on top of the Stephenson Bastion table mountain (Fig. 4c), the archaeocyath-bearing clasts are subsidiary clasts while the reddish-brown ?Beacon Supergroupsandstone derived clasts and local bed rock fragments (greenishgrey quartzite derived from the Stephenson Bastion Formation) are dominant. Subglacial erosional forms on the Stephenson Bastion table mountain (Fig. 4c) suggest that overriding ice covered the western part of Shackleton Range and flowed NW (Höfle & Buggisch 1995;Kerr & Hermichen 1999;Sugden et al. 2014). Höfle & Buggisch (1995) hypothesised that the Cambrian exotic erratics on the Stephenson Bastion were generated during a major expansion of the Antarctic Ice Sheet at the end of Miocene. They suggested the Whichaway Nunataks as source area for the exotic ?Beacon Supergroup erratics. For Cambrian archaeocyath-bearing clasts, they suggested the Whichaway Nunataks and the PM, which have early to middle Cambrian limestones. Kerr & Hermichen (1999) interpreted the southeastern flows as an expansion of the Recovery Glacier and those from the SW as an expansion of the Filchner-Ronne Ice Shelf. They concluded that there is no evidence for extensive glacial modification of the Shackleton Range plateau by the E Antarctic Ice Sheet or by Quaternary ice. Cosmogenic nuclide data indicate that the glacial overriding of the higher summits of the Shackleton Range and deepening of the Slessor and Recovery glaciers troughs was earlier than 2.5 Ma, and significant erosion occurred in the mid-Miocene maximum (Sugden et al. 2014).
The archaeocyath-bearing clasts occur in conglomerates and boulders of the Mount Wegener Formation, which is part of the Mount Wegener Nappe.  reported archaeocyaths from E of the Read Mountain at the Oldhamia, Trueman Terraces and the Swinnerton Ledge ( Fig. 4d). They described bioclastic rudstone and grainstone, archaeocyath/calcimicrobe boundstone microfacies with diverse biota such as archaeocyaths, trilobites, mollusc, echinoderms and calcimicrobes (Epyphyton, Korilophyton, Renalcis granosus Vologdin, Renalcis seriata Korde, Girvanella, Batenevia and Subtifloria) and concluded that the carbonate clasts represent the breakup and destruction of shallow-water carbonate deposits.

Mount Wegener Formation
Broadly, the Mount Wegener Formation comprises shales/slates, greywackes and conglomerates, and shows folding and metamorphic overprint that increases from SE to NW (Buggisch et al. 1994b). The total thickness of the Mount Wegener Formation probably exceeds 1000 m, and stratigraphy (see Fig. 5), facies, geochemistry and provenance analysis was carried out by  with data obtained during the EUROSHACK expedition (1993)(1994). The main results and conclusions are summarised below. The facies of the Mount Wegener Formation are interpreted as marine clastic deposits from upper slope to deep basin set up in synorogenic conditions, the presence of plagioclase and volcaniclastics support an intracontinental back-arc environment (Buggisch et al. 1994a). The clast composition of the conglomerates shows evidence of repeated erosion and redeposition in shallow-water conditions before their final sedimentation as marine slope deposits. Reworked sedimentary rocks are dominant, while basementderived rocks are rare. The most abundant clasts are sandstone, carbonate and conglomerate. The sandstones (greywackes) from the Mount Wegener Formation present intermediate proportions of silicon dioxide/aluminium oxide and low potassium oxide/sodium oxide ratios like those observed in the passive continental margins. Indeed, isotope geochemistry (ε Nd,530Ma values) suggests that the supracrustal rocks of the Pioneers Group (the Northern Belt, a Ross/Pan-African overprinted basement; Fig. 3) could be a potential source of sediments from the Mount Wegener Formation. The K-Ar dating ages of detrital muscovites between 572 and 534 Ma also suggest basement rocks with Pan-African cooling histories as the main source.
The age of the Mount Wegener Formation was interpreted by Buggisch et al. (1994b) as early Cambrian based on the presence of fossils (Oldhamia cf. antiqua and Oldhamia cf. radiata, Epiphyton sp. and ?Botomaella sp.) and on the rubidium-strontium (Rb-Sr) analysis (561 ± 18 Ma and 539 ± 9 Ma Figure 4 (A) Location of glacial erratics and in situ archaeocyath-bearing clasts in the Shackleton Range. The main rock outcrops (1-8), some tributary glaciers (S1-3, R1-2) and the current direction of ice flows are shown (modified from Höfle & Buggisch 1995). Box outlines area of (B). Abbreviations: 1 = Otter Highlands; 2 = Haskard Highlands; 3 = La Grange Nunataks; 4 = Herbert Mountains; 5 = Pioneers Escarpment; 6 = Read Mountains; 7 = Du Toit Nunataks; 8 = Stephenson Bastion; S1 = Blaiklock Glacier; S2 = Straton Glacier; S3 = Gordon Glacier; R1 = Cornwall Glacier; R2 = Glen Glacier. (B) Partial geological map of the southern Shackleton Range (modified from Clarkson et al. 1995). Black stars: in situ archaeocyath-bearing clasts are found in Cambrian marine slope deposits of the Mount Wegener Formation (Buggisch et al. 1994a;. Green stars: glacial erratic archaeocyath-bearing clasts occur as part of pre-Quaternary till deposits (Stephenson Bastion) and recent frontal moraines (Du Toit Nunataks). Boxes outline areas of (C) and (D). Abbreviations: OHT = Otter Highlands Thrust; MWT = Mount Wegener Thrust. (C) Detail from (B) showing the location of glacial erratic archaeocyath-bearing clasts according to Höfle & Buggisch (1995). The late Precambrian Stephenson Bastion Formation emerges in a plateau landform. Archaeocyaths occur as part of pre-Quaternary till deposits on the plateau. The pre-Quaternary till deposits on the Stephenson Bastion plateau, erosive structures and forms (striated rock surfaces, roches moutonnées, etc.) suggest that the overriding ice flowed northwestward (see the text). The overriding ice directions are shown according to Kerr & Hermichen (1999) and Sugden et al. (2014). (D) Detail from (B) showing the location of the sampling sites with archaeocyaths within the Mount Wegener Formation. The exact location of the samples with archaeocyaths from Oldhamia Terraces is unknown. Key: 1 = conglomerates; 2 = greywackes; 3 = shales; 4 = olistoliths; 5 = Oldhamia ichnotaxon. Geological map and sedimentary palaeocurrents of the Mount Wegener Formation according to . The black bar represents the stratigraphic section (see isochrones; Fig. 3 and references therein). The structural data and K-Ar analysis on phyllosilicates (2-6 μm whole rock fraction) indicate that the low-grade metamorphic overprint and the southwards transport of the Mount Wegener Nappe occurred around 490 Ma as a result of the Ross (Pan-African) orogeny (Buggisch et al. 1994b).

Materials and methods
The carbonate clasts and archaeocyath-bearing limestone clasts samples described here were collected during the GEISHA (First German geological expedition to the Shackleton Range) and EUROSHACK (European Antarctic Geological Expedition) expeditions. These geoscientific land expeditions took place in the Shackleton Range during 1987Range during /1988Range during and 1994Range during /1995 GEISHA was jointly supported by the Alfred Wegener Institute for Polar and Marine Research (AWI) and Federal Institute for Geosciences and Natural Resources (BGR), whereas EUROSHACK was under the leadership of the BGR and British Antarctic Survey (BAS).
In the GEISHA expedition, about 300 samples were collected to assess the degree of metamorphism, and approximately 200 thin sections were studied with this purpose (Buggisch et al. 1994b). In the EUROSHACK expedition, about 150 samples  . Facies codes according to Pickering et al. (1986): A1.1 = stratified gravels; B2.1 = parallel-stratified sands; B2.2 = cross-stratified sands; C2 = organised sand-mud couplets; F1.1 = rubble; F2 = contorted/disturbed strata. Only the stratigraphic position of the samples with archaeocyaths (ESH Trueman Terraces) and those with probable lateral equivalence (ESH Swinnerton Ledge) is shown. (B-E) Thin-section photomicrographs (plane-polarised light) of polymictic conglomerates of the Cambrian Mount Wegener Formation in Oldhamia Terraces (B) and Trueman Terraces (D) and conglomerate clast of the Cenozoic tills on the Stephenson Bastion (E) (see Fig. 4). (B) Calcimicrobe-rich limestone pebble floating in a sandy matrix. Clasts and sandy matrix are cut by a late vein fracture system (LV) that is cut by an incipient rough cleavage fabric (white arrows). (C) Detail Epiphyton and cements crossed by a late cleavage fabric (white arrow). (D-E) Note the similarities in clast composition and tectonic deformation (white arrows: cleavage, irregular sutured grain boundaries) in both conglomerates. In these examples, conglomerate clast from the Cenozoic tills (E) show more continuous cleavage traces around strongly tectonically oriented grains.
were taken from 43 different stratigraphic levels and/or localities from the Mount Wegener Formation. They were used to analyse facies, provenance, geochemistry and K-Ar dating. Most of these samples come from conglomerates or limestone clasts within the conglomerates .
A total of 237 petrographic thin sections have been reviewed in this study (GEISHA and EUROSHACK samples): 45 from the Stephenson Bastion and Du Toit Nunataks and 188 from the Read Mountains (Fig. 4b). GEISHA thin sections are from the Stephenson Bastion (9) and Du Toit Nunataks (2). EURO-SHACK thin sections are from the Stephenson Bastion (34) and the Mount Wegener Formation (189) in the eastern part of the Read Mountains: Oldhamia (91) and Trueman (50) terraces, Swinnerton Ledge (47). In addition, four thin sections of indeterminate origin from EUROSHACK have been reviewed.
Out of 237 petrographic thin sections, 223 were medium in size (5 × 5 cm 2 ) and 14 were small (2.7 × 4.8 cm 2 ). The microfacies and palaeontological analyses were performed with petrographic and binocular microscopes. The samples were classified according to the schemes of Dunham (1962), Embry & Klovan (1971) and Wright (1992). The peloid subcategories are according Flügel (2004) and the sediment grain sizes according Wentworth (1922) size classes. The percentage of allochems was semiquantitatively analysed by visual estimation charts (several authors in Flügel 2004). The basic types of porosity and poresize classes were described following the classification of Choquette & Pray (1970). Selected thin sections were stained with alizarin red sulphur and potassium ferricyanide to analyse their staining response (calcite/dolomite and iron content, respectively) according to Dickson (1965Dickson ( , 1966. The classification of dolomite crystal fabrics follows the Sibley & Gregg (1987) scheme and the crystal-size scale for carbonate rocks is according to Folk (1962). A total of 367 archaeocyath specimens and one coralomorph have been recognised from 46 thin sections. Among archaeocyathan specimens, 189 were classified following the systematics of the Treatise on invertebrate paleontology (Debrenne et al. 2015).

Carbonate clasts record from the Cambrian Mount Wegener Formation and the Cenozoic tills
The analysed carbonate clasts from the Cambrian Mount Wegener Formation and the Cenozoic tills (Stephenson Bastion and Du Toit Nunataks) show equivalent components and microfacies. The analysed carbonate clasts are accessory to primary components in the sandy polymictic conglomerates and breccias from the lower Cambrian Mount Wegener Formation and from the Cenozoic tills. The carbonate clasts correspond to limestones and dolostones (see sections 5.1 and 5.2) where the main allochems are non-skeletal grains, skeletal grains and calcimicrobes. The non-skeletal grains include aggregate grains, mud peloids (micritic intraclasts), bahamite peloids (round micritic grains), algal peloids (calcimicrobe-derived micritic grains) and different types of ooids (types 1-3). Very fine to finegrained size superficial ooids with quartz nuclei are type 1 ooids. Medium-sized and very coarse to granule-sized ooids are type 2 and type 3, respectively. The carbonate clasts from the Mount Wegener Formation and Cenozoic tills also show equivalent early to late diagenetic phases and tectonic-induced fabrics (Fig. 5). The main diagenetic processes recorded in carbonate clasts and their host rocks (conglomerates) are described in section 6.
In addition to the similarities observed in the microfacies, diagenetic and tectonic processes, the size distribution of the archaeocyathan cups has been analysed to assess whether or not the allochthonous archaeocyaths from the Cenozoic tills (Stephenson Bastion and Du Toit Nunataks) are derived or not from the Mount Wegener Formation (Fig. 6). The diameter of the archaeocyathan cups is measured in transverse section from the thin sections. A total number of 189 specimens have been analysed (see section 7) and 181 specimens have been measured from the Stephenson Bastion/Du Toit Nunataks (70) and the Mount Wegener Formation (111). Some archaeocyaths cups are distorted by tectonic stresses (see section 6). In these cases, the minimum diameter has been selected to reduce bias due to tectonically strained archaeocyath cups and characterise the size distribution of archaeocyaths. The archaeocyathan fauna from the Cenozoic tills shows the same average diameter as those observed in the Mount Wegener Formation (3.25 mm), as well as equivalent median values (2.65 and 2.85 mm, respectively; see Fig. 6a, b). Therefore, the analysis of cup sizes supports a single source for the archaeocyaths from the Cenozoic tills and the Mount Wegener Formation. In addition, we have grouped all the measured archaeocyaths into a single set to analyse how the cup sizes are distributed. We have selected three diameter size ranges: small (0.2-2.9 mm), medium (3.0-5.9 mm) and large (6.0-14 mm). In the Shackleton Range, more than half of the specimens are small in size (53 %), and the rest are mainly medium in size (34 %), while the largest sizes are minority (13 %) (Fig. 6c). Most archaeocyaths have a size diameter ranging from 10 to 50 mm (Rowland 2001). Furthermore, typical conical cups of Ajacicyathidae are around 5-15 mm (Debrenne et al. 2012). Cordie & Dornbos (2019) have measured the traditional morphometric characters of more than 1000 archaeocyaths, mainly from the orders Ajacicyathida and Archaeocyathida. They have reported an average diameter of 10.6 mm and a median value of 8.67 mm (0.78 mm minimum to 74 mm maximum). Cordie & Dornbos (2019) reported an average diameter of 10.6 mm from more than 1000 measured archaeocyaths. Therefore, the archaeocyaths from the Shackleton Range record are small compared to published data. This trend in a smaller sized fauna for Antarctic record has also been observed in Cambrian helcionelloids from the Shackleton Limestone (central TAM) that were compared with helcionelloid mollusc assemblages from the Ajax Limestone (Australia) and the Bastion Formation (Greenland) (Jackson & Claybourn 2018).
All these data from the analysis of microfacies, diagenetic processes and archaeocyathan diameter sizes confirm that Cenozoic carbonate erratics from the Stephenson Bastion and Du Toit Nunataks are derived from the Cambrian Mount Wegener Formation. Thus, the analysis of carbonate clasts from the Shackleton Range provides crucial information about the history of a hidden Cambrian carbonate platform from which they were derived.

Dolostone clasts microfacies (DM)
The analysed dolostone clasts (Figs 7, 8a) have been arranged in three main categories according to the percentage of quartz grains, the proportion of non-skeletal grains or the predominance of a dolomite crystal fabric as: (DMa) dolomitic sandstones to sandy dolostones (two and four subcategories, respectively); (DMb) aggregate-grain-to ooid-rich dolostones (three subcategories); and (DMc) dolostones.
The presence of very well-rounded quartz sand grains is characteristic in the dolomitic sandstones to sandy dolostones. Dolomitic sandstones range from fine to medium very well-sorted dolomitic sandstone (DMa1) to coarse, poorly sorted intraclastic dolomitic sandstone (DMa2) (Fig. 7a). In the first one, quartz grains reach up to 65 % of the rock volume. In the second one, they represent 40-50 %, while subrounded dolomicrite and dolospar intraclasts reach 30 % of the rock volume.
The sandy dolostones are coarse intraclastic dolorudites (DMa3), medium oolitic-intraclastic dolorudites (DMa4) and dolomicrites (DMa5-6), and their quartz grain content is in the range 1-30 % of the rock volume. In the sandy, coarse, very poorly sorted massive clast-supported intraclastic dolorudites (DMa3) (Fig. 7b), subangular to well-rounded intraclasts (up to 1.7 cm) are polygenic. Dolomicrites and silty dolomicrites are dominant intraclasts (up to 60 % of the rock volume), while dolograinstones rich in ooids, aggregate-grains or algal peloids are minor intraclasts. Oolitic dolograinstone intraclasts are made up of different types of ooids (types 1-3). In addition, fragments of type 3 oolitic cortices, abraded micritic-coated clasts and orange fibrous cement crusts also occur. Dominant coarse, very well-rounded quartz (up to 1 mm) and minor fine to medium subangular quartz sand grains are up to 25 % of the rock volume. Calcimicrobe intraclasts with Proaulopora and ?Renalcis are accessories.
Aggregate grain-to ooid-rich dolostones (DMb) are dominated by aggregate grains or ooids or a mixture of non-skeletal grains. Different cements and replacements of dolomite and silica occur in primary interparticle porosity and secondary selective leaching of the cortex and nuclei of non-skeletal grains, resulting in varying degrees of preservation of the original fabrics (see section 6). Compound grains, 200 μm to 1 mm in size, with grape-like to lobate shapes and irregular dark micritic envelopes, are forming aggregate grain dolowackestones to dolopackstones (DMb1) (Fig. 7e) with very accessory small superficial ooids (type 1). Laminar micritic concentric type 3 ooids (up to 5 mm, 'giant ooids' sensu Sumner & Grotzinger 1993 rather than pisoids) can form loosely packed oolitic dolowackestones to packed dolograinstones (DMb2) (Fig. 7f). These microfacies show moderately sorted to well-sorted simple ooids, with both spherical and ellipsoidal shapes. Symmetrical cortices predominate and broken cortices that act as nuclei are very common. Very well-rounded quartz grains (650 μm) are accessory in matrix and within compound ooids. Oolitic dolograinstones with very wellsorted silicified type 2 ooids (medium-grained sand size) also occur. The very poorly sorted aggregate grain-oolitic dolograinstone (DMb3) (Fig. 7g) shows a distinctive regular to irregular fenestral fabric (up to 1.5 cm wide) that is filled with geopetal infills and cements. In this microfacies, ooids are mixture of 1-2-3 types with different preservations. In the fenestral aggregate grain-oolitic dolograinstone are found large compound intraclasts, up to 5 mm wide, with aggregate grains and peloids as intracomponents and multiple micritic envelopes.
Dolostones with cemented non-fabric-selective porosity also occur (DMc). They are very fine to finely crystalline dolomites with intercrystalline mesopores and very large megapores (16 mm × 41 mm, vug to channel types). Large vuggy to channelised cavities can reach up 25-30 % of the rock volume and display multi-episodic fillings ( Fig. 8a; see section 6).
The archaeocyath-bearing microfacies are clearly dominated by regular archaeocyaths (149 specimens versus 40 irregular specimens). In addition, in the archaeocyath cementstone the presence of irregular archaeocyaths is totally accessory. Regarding archaeocyaths distribution, a total number of 16 families of archaeocyaths occur in the boundstones (see Appendix 1). Rotundocyathus glacius, Cadniacyathus sp. and Archaeopharetra irregularis are the dominant species. Fifteen families are represented in the calcimicrobial boundstone with archaeocyaths, six in the calcimicrobe-archaeocyath boundstones and seven in the archaeocyath cementstones. Ajacicyathidae is the dominant family in the archaeocyath cementstones and calcimicrobial boundstones with archaeocyaths. The calcimicrobe-archaeocyath boundstones are dominated by Densocyathidae and Ajacicyathidae. Common families in all boundstones are Ajacicyathidae, Densocyathidae, Shackletoncyathidae, Loculicyathidae and Archaeopharetridae, while Kaltatocyathidae are only found in the calcimicrobe-archaeocyath boundstones.
The presence of skeletal components, other than archaeocyaths, in the rest of the analysed microfacies (limestone-or dolostone-derived clasts) is very rare in terms of volume and/or frequency. The archaeocyath floatstone consists of poorly sorted reworked archaeocyaths (25-35 % of the rock volume), hyoliths (1-5 % rock volume), allochthonous calcimicrobe remains (Proaulopora, Subtifloria) and calcimicrobial boundstone intraclasts (Renalcis-or Epiphyton-dominated microframes). The archaeocyath debris varies from septum to large cup remains. The archaeocyathan walls are mostly free of calcimicrobe encrustations but are unusually encrusted by Girvanella. Eleven families of archaeocyaths are represented in the archaeocyath floatstones -Bronchocyathidae and Ajacicyathidae are the dominant ones. Bronchocyathidae, Kymbecyathidae, Dictyocyathidae and Copleicyathidae are only found in the floatstones (see Appendix 1). Thalamocyathus trachealis is dominant in the floatstones.
6. Principal diagenetic processes and tectonic fabrics recorded in the carbonate clasts from the Cambrian Mount Wegener Formation and the Cenozoic tills 6.1. Early diagenetic phases The dolomitic sandstones, sandy dolostones and aggregate grain-to ooid-rich dolostones display evidence of early marine phreatic cementation, such as yellow fibrous to bladed isopachous rims (M1RD1) and pore-filling equant cement (M2RD1) (Fig. 9a, f). Meteoric vadose diagenesis processes, such as partial dissolution producing secondary moldic porosity, are common in aggregate grain-to ooid-rich dolostones. Some samples show partial to complete intraparticle dissolution and internal collapse that produced geopetal infills in aggregate grains and ooids ('half-moon' ooids sensu Wherry 1916). The evidence of selective leaching (oomoldic porosity and selective destructive replacive fabrics of some cortical laminae; Figs 7f, 9c) suggests a differentiation of the primary mineralogical composition of the ooids (high magnesium calcite, aragonite). Likewise, fenestral aggregate grain-oolitic dolograinstones display dark microcrystalline crusts and meniscus fabrics with distinctive rounded pores (Fig. 7g) produced during meteoric vadose diagenesis (Dunham 1971;Longman 1980). Many dolostone microfacies exhibit early pervasive mimetic dolomitisation of allochems (e.g., concentric ooids, aggregate grains), micrite and early marine phreatic and vadose cements (Figs 7, 9a-f). Early mimetic dolomitisation (RD1) predates compaction and is followed by different stages of dolomitisation and silicification linked with secondary porosity and fractures (Figs 9a-f; see section 3). In the aggregate grain-to ooid-rich dolostones an early authigenic and diagenetic phase of silica occurs as chalcedony void-filling cement (Schc) within the remaining interparticle and/or within the secondary moldic intraparticle porosities (e.g., oomoldic porosity, Fig. 9b). Chalcedony cement post-dates early mimetic dolomitisation (RD1) and predates all other fracture-related cements. Some cortices and nuclei from ooids and aggregate-grains exhibit replacive cryptocrystalline and microcrystalline silica fabrics (RS1c-m in Fig. 9c, d, f).
The calcimicrobial boundstone with archaeocyaths, calcimicrobe-archaeocyath boundstone and archaeocyath cementstone exhibit the largest fabric-selective porosity that is filled with early marine phreatic cementation. The inter-, intra-skeletal and growth framework meso-and megapores are filled with yellow, inclusion-rich, fibrous calcite cements (M1Cc) and bladed to equant non-ferroan calcite mosaics (M2Cc) that precede the formation of stylolites. In the archaeocyath cementstones, the isopachous marine fibrous calcite crusts (up to 2.5 mm thick) constitute around 30-40 % of the rock volume (Fig. 9g). Early marine phreatic cementation occluded a significant part of the primary mesopores in the calcimicrobe-and archaeocyathbearing clasts. Nevertheless, the larger fabric-selective porosity registered successive diagenetic phases (see below).

Breakup and brecciation of platform and downslope transport of carbonate clasts
Irregular vugs with very angular silt to sand-sized crystal and quartz grain sedimentary fillings post-date early marine phreatic cements (M1Cc and M2Cc) in boundstone microfacies (Fig. 9h). Furthermore, the orientations of the geopetals within irregular vugs and the calcimicrobe microframework show inconsistent polarity relationships between them (Fig. 9i). Therefore, these irregular vugs cut and eroded the early pre-existing marine cements and the calcimicrobe microframework, suggesting meteoric diagenesis related to the breakup and sedimentary brecciation of the carbonate platform. These sedimentary fillings are equivalent to those observed in enlarged fenestral megapores developed in the peloid-intraclastic-bioclastic grainstone clasts (Fig. 9j).
The enlarged fenestral megapores are filled with two sedimentary fillings (SI1, SI2 in Fig. 9j) and drusy mosaic cement with crystals showing compositional zoning. Drusy zoned mosaic cement begins with non-ferroan calcite (pink-stained) and ends with ferroan calcite (mauve-stained) (Fig. 9k). However, not all megapores are occluded by the complete succession of drusy zoned mosaic cement. The sedimentary fills correspond to an early very fine-grained geopetal crystal silt (SI1) that is followed by a second fill (SI2). The latter fill has angular silt to sand-sized quartz grains and eroded crystals derived from the non-ferroan calcite drusy mosaic (Fig. 9l). The SI2 is observed in those fenestral pores close to the margins of the limestone clast. Furthermore, SI2 is equivalent to the surrounding conglomerate Megapores are filled with sedimentary fillings (SI1 and SI2) and (K) drusy mosaic with compositional zoning starting with non-ferroan calcite and ending with fracture-related ferroan calcite cement. White box outlines area of (L). (L) Sedimentary filling (SI2) is composed of sandy-silty matrix (identical to the host rock conglomerate matrix) and eroded crystals derived from the non-ferroan calcite cement. matrix (Fig. 9j). Thus, SI2 was infiltrated into the still-open fenestral pores during sedimentation of the carbonate clasts as part of the matrix-supported polymictic gravels on the marine slope, after the early non-ferroan calcite drusy mosaic. The succeeding ferroan calcite cement is fracture-related cement and occludes the remaining fenestral porosity (Fig. 9k).

Late diagenetic phases and mechanical overthrust-emplacement processes
In general, clasts in the sandy polymictic conglomerates from the Mount Wegener Formation and Cenozoic tills show dominant tangential contacts due to their deposition as matrix-supported gravels. However, the contacts between sandy matrix and clasts are sutured due to tectonic deformation (Fig. 5d, e). In carbonate clasts, mechanical compaction features were inhibited mostly by early cementation (M1, M2), dolomitisation (RD1) and silicification (Schc and RS1c-m) (Fig. 9). Thus, loosely packed fabrics are the most common, except for sandy intraclastic dolorudites (Fig. 7c), where tangential and long contacts dominate and there are also concavo-convex contacts between quartz and dolomite grains. Evidence of chemical compaction occurs as simple to sutured stylolites in some limestone clasts. These stylolites post-date both shallow marine cements (M1Cc, M2Cc) as well as geopetal sedimentary infillings produced during the downslope transport of the clasts.
Burial dolomites correspond to three different stages of dolomitisation (D2, D3 and D4). Stage D2 consists of coarsely crystalline (0.25-0.5 mm up to 1 mm), planar-s to planar-e type dolomite (types according to Sibley & Gregg, 1987). D2 occurs mainly as replacive/cement dolomite in the interparticle, secondary intraparticle, intercrystal and fracture-associated porosities of the aggregate grain-to ooid-rich dolostones. D2 post-dates silica Schc cement and RS1c-m replacement (Fig. 9c, f). Stage D2 is also the first dolomite cement in the vug to channel cemented dolostones where the large megapores are lined by multiple isopachous layers of cement crusts (1.5-2.5 mm thick) (Figs 8a, 10a, b). In some cases, cavity-cement crusts may start with early yellow inclusion-rich fibrous cement (like those previously described as M1Cc) but now being a dolomite, so it could represent early mimetic dolomitisation of a fibrous marine aragonite/calcite cement precursor (M1RD1). However, most mesopores, megapores and fractures are rimmed by a first generation of cloudy to clear, coarsely crystalline (>0.25 mm) rhombic D2 dolomite (Fig. 10a). The boundary between the rhombic crystals D2 and the next stage D3 can be locally irregular and rich in black hydrocarbon residues. Stage D3 corresponds to a black inclusion-rich zoned, fibrous to elongated-bladed dolomite crust (up to 2 mm thick) (Fig. 10a). In some cavities, the D3 dolomite crusts show a clearer, less inclusion-rich epitaxial late D4 stage (Fig. 10b). There are traces of bitumen after the growth of D4 dolomite. Therefore, the hydrocarbon migration postdates the D2 and D4 dolomite stages and was coeval with the D3 dolomite stage. In addition, the partial migration of the trapped hydrocarbon in the D3 cement left an open microporosity that was filled with the fracture-related cement. Stage D3 is not observed in the aggregate grain-to ooid-rich dolostones ( Fig. 7e-g). However, a possible equivalent to stage D4 is recorded as a fracture-related phase of brown, thick twinning, very coarsely crystalline (1-4 mm), planar-s dolomite in the aggregate-grain to ooid-rich dolostones (Fig. 9e, f) and in the archaeocyath cementstones and floatstones (Fig. 9g). The abundance of black bitumen inclusions in D3-D4 suggests that they are burial diagenetic dolomites. The occurrence of black bitumen in mesopores and megapores from dolostones and the observed cross-cutting relationships suggest that hydrocarbon migration predates the void-late-fracture-related remnant cements (LVA, late cement vein system A) (Figs 8a, 10c). The appearance of black bitumen residues is recorded in thin hair-like fractures, intercrystalline mesopores and megapores in dolostones, calcimicrobial boundstones and polymictic conglomerates.
The late cement veins correspond to different mineral-filled fracture systems (LVA and LVB, respectively). The LVA system consists of quartz-calcite veins and the LVB system consist of (ferroan saddle dolomite)-quartz-ferroan calcite-quartz veins. LVA and LVB fluids precipitated in remnant open porosity in dolostone and limestone clasts and host rock conglomerates. Cross-cutting relationships, mineralogies and fabrics support LVB postdating LVA. In the LVA system (Fig. 9d, f), dissolved silica fluids first produced chert and microquartz replacement from the host rock and later precipitated as megaquartz mosaics within the remaining porosity and fractures. Second, dissolved non-ferroan to very slightly ferroan carbonate fluids precipitated as very to extremely coarse poikilotopic calcite cement with thin twin planes (and minor thick twin planes) in pores and fractures. In the LVB system, the fracture walls are sometimes lined with ferroan saddle dolomite relics (turquoise/greenish stain and curved faces; FeD in Fig. 10d) and, generally, by megaquartz crystals with euhedral terminations, which are embedded by coarsely to very coarsely crystalline zoned ferroan calcite with tabular thick twin planes (FeCc in Fig. 10d-g). Dissolved silica fluids from the LVB system also produced partial and selective host rock cryptocristalline and microcrystalline silica replacement (RS2c-m), syntaxial overgrowths of detrital quartz grains and megaquartz mosaics in remnant porosity (Sqtz).
Some limestone clasts show obliterative diagenetic fabrics and they have been classified here as sparstone clasts. However, the distribution pattern of clots, the size of cavities and the appearance of archaeocyath ghosts suggest that they were calcimicrobe-archaeocyath-bearing microfacies that underwent neomorphism and partial replacement processes that were associated with fracture-related fluids during burial (Fig. 10h, i). In fact, replacive microdolomite crystals occur in the limestone clasts associated with hair-like fractures with ferroan carbonate fluids from the LVB system. This replacive microdolomite is observed in host rock, calcimicrobes, archaeocyaths and shallow marine cements (Fig. 10j, k).
Some samples from the Cambrian Mount Wegener Formation and the Cenozoic erratics show evidence of low-grade tectonically induced fabrics. Furthermore, in the Mount Wegener Formation, the tectonically induced fabrics occur close to the overthrust. Buggisch et al. (1994b) described how the sandstones and polymictic conglomerates from the Mount Wegener Formation are tightly folded and deformed above the decollement zone. They observed how the limestone clasts were extremely stretched while feldspars were brittly deformed and described the growth of illite/sericite, chlorite and minor biotite. They recognised two stages of deformation: a first isoclinal folding with penetrative schistosity and distinct crenulation cleavage, and a later recumbent open folding development during the final emplacement of the Mount Wegener Nappe.
We have recognised in the carbonate clasts several features associated with the low-grade tectonically induced fabrics produced during the deformation and emplacement of the Mount Wegener Nappe. In some samples, archaeocyaths and cements can be distorted by the effects of plastic deformation (Fig. 10l), as well as flattened ooids produced by pervasive tectonic shear. Calcite cements associated with late cement veins (LVA and LVB) present different degrees of twin lamellae development. The non-ferroan to very slightly ferroan coarsely poikilotopic calcite cement that precipitated from the LVA system (PkCc in Figs 9d, 10c, h) exhibits mostly type I and rare type II calcite twins. However, the coarse to very coarse crystalline zoned ferroan calcite cement that precipitated from the LVB system exhibits types II (tabular thick) and III (tabular thick curved) calcite twins ( Fig. 10d-g). Calcite twin morphologies can be used as a low-temperature geothermometer based on data collection and analysis from Ferrill et al. (2004). These authors correlate the average calcite twin with the temperature of deformation, so that the thin twins dominate below 170°C, and the thick twins dominate above 200°C. On the other hand, late cement vein systems (LVA and LVB) show the coexistence of calcite and quartz in the veins. Experimental work on calcite and quartz solubility shows that the quartz solubility increases with rising temperatures while the solubility of calcite decreases; between 150°C and 300°C, all proportions are possible (Sharp 1965). These temperature values are in agreement with the very low-grade metamorphism conditions reached by the Mount Wegener Formation at the southern Read Mountains according with the illite crystallinities of°Δ2θ > 0.25 (Buggisch et al. 1994b).
The analysed conglomerates and carbonate clasts may show different degrees of cataclastic deformation ( Fig. 5b-e), from minor to dense anastomosing cleavage and incipient brecciation to strongly sheared fabrics ( Fig. 10m) with finer seams of solution residues. This deformation post-dates all observed cementation and replacement phases, including late cement veins (LVA and LVB systems; Fig. 10e-g). K-Ar analysis in phyllosilicates (2-6 μm fraction) indicates that low-grade metamorphic overprint and southwards transport of the Mount Wegener Nappe were around 490 Ma (i.e., Furongian) as a result of the Ross (Pan-African) orogeny (Buggisch et al. 1994b).
Description. One-walled cup 1.2 to 1.7 mm diameter. The wall is 0.08 to 0.12 mm thick and bears scarce single-pore tumuli with 0.22 to 0.28 mm diameter.
Remarks. This species is characterised by the irregularity of the size and spacing of its tumuli pores.
Remarks. The narrower spacing of intervallar bars and the values of IK distinguish from Dokidocyathus simplicissimus present in Nimrod Glacier (Debrenne & Kruse 1986) and other Australian described species. Although for similar diameters the intervallar bar spacing is in the range of the Australian species zero, the small size of the cups does not allow us to assign them to a specific species.
Remarks. The porosity of the walls is characteristic of this genus and the branching cups are similar to the transverse section assigned to K. gregarius (Gravestock 1984, fig. 31g). This is the first recorded occurrence of the Kaltatocyathus from Antarctica.
Description. Cup 6.3-8 mm in diameter with intervallum 1.49 mm in width. Outer wall with pore canals (diameter 0.20 mm, lintels 0.12-0.32 mm, wall thickness 0.36 mm). Inner wall with simple, rounded to elliptical pores (diameter 0.20 mm, lintels 0.12 mm, wall thickness 0.12 mm). The presence of possible scattered bars in the intervallum is not confirmed.
Remarks. The diameter of the Byrd Glacier specimens is bigger than specimens described here, but the wall porosity is very similar between all specimens. The bars are rare, scattered and even not visible in some sections (Debrenne & Kruse 1986, fig. 6).
Remarks. The presence of pectinate tabulae and the inner wall with short spines allows us to assign our specimen to the species from Gnalta Shelf.
Remarks. The Shackleton Range specimens are smaller than those described from the Gnalta Shelf and N. cf. lawrencei from Ajax Mine, but the porosity of the walls and septa are similar. The radial and intervallar coefficients vary proportionally during the growth of the cups. The porosity coefficient in both walls is greater in the smaller cups and decreases when the diameter of the cup increases, as observed in Shackleton Range specimens, until it stabilises since it presents little variability from diameters of 4 mm with coefficient values from 1 to 1.5, as observed in N. lawrencei and N. andersoni from Australia. This pattern does not hold for N. cf. lawrencei. The septal porosity coefficient tends to decrease with increasing cup diameter, with values >1, except for some N. lawrencei specimens (∼0.6 outer wall Ø/l); this trend is not followed by N. cf. lawrencei.
Remarks. The poor state of preservation of the cups, affected by recrystallisation and deformation processes, does not allow us to assign them to a described species.
Occurrence.  Diagnosis. Outer wall with two to six rows of diaphragm pores. Inner wall with one row of simple pores per intersept, without small spines. Septa completely porous, with two to five rows of pores.
Occurrence Remarks. The poor preservation and the fragmentary state of these specimens do not allow us to assign them to a specific genus and species.
Remarks. The poor state of preservation of many of these specimens and high degree of fragmentation, making many of them incomplete, do not allow us to assign them to some Australian or Antarctic species of Cadniacyathus. There are some differences with the Australian species Cadniacyathus asperatus such as the non-bulging outer wall and its smaller number of pores and the septal porosity.
Occurrence. Diagnosis. Outer wall with several simple pores. Inner wall with one row of pores per intersept, bearing upwardly projecting cupped bracts. Septa with two to six rows of pores, stirrup pores are absent. Remarks.

Buggischicyathus
Perejón, Menéndez & Moreno-Eiris differs from Dailycyathus Debrenne (1970) by the absence of stirrup pores and the porosity of the septa. From Deceptioncyathus Gravestock (1984) for the absence of synapticulae. From Leptosocyathus Vologdin (1937) for not having upwardly S scales in the inner wall and the different septal porosity. Diagnosis. Cup 2.8-10 mm in diameter. Outer wall with two to four pore rows, inner wall with one pore row per intersept, bearing upwardly projecting cupped bracts. Septa completely porous with rounded to elliptical pores and two to six pore rows.
Remarks. The septa of our specimens have numerous pores of small size, unlike the other genera of the Densocyathidae that have only one pore in the inner wall, besides other notable differences, such as Dailycyathus in the septal porosity and absence of stirrup pores; it differs from Deceptioncyathus by not presenting synapticulae and from Leptosocyathus by not having scales on the inner wall.
Remarks. The poor state of preservation does not allow us to observe clearly the number of pore rows of the internal wall or the type of bracts.
Remarks. The wall and septal porosity are characteristic of T. trachealis, a common species in the Australia-Antarctica province. It differs from Gordonicyathus because the latter presents a greater number of pores in the septa, in cups of similar diameter.
Remarks. The fragmentary nature of the specimens does not allow us to assign a specific species. Several species of Baikalocyathus have been described in Australia. It is the first find of Baikalocyathus in Antarctica.
Occurrence. Allochthonous clasts: Antarctica, Shackleton Range, Oldhamia Terraces, carbonate clasts from the Mount Wegener Formation. Cambrian Series 2, Botoman. Since in Australia it has been recorded during the Atdabanian age, all the species described from Mount Scott Range appear in the Lower Faunal Assemblage II (Gravestock 1984) or the equivalent Spirillicyathus tenuis Zone (Zhuravlev & Gravestock 1994 (Fig. 14e) Etymology. In memory of Hans-Christian Höfle, German glaciologist who sampled and studied the Cenozoic erratics of all this Antarctic material.
Diagnosis. Outer wall with two rows of simple pores. Inner wall with straight canals, each canal span several intersepts. Septa completely porous with five or more rows of pores. Synapticulae.
Remarks. Although the cup is fragmented, it is possible to recognise the internal wall so characteristic of Gnaltacyathus, but the presence of synapticulae suggests that it is a different genus.
Remarks. Ladaecyathus jagoi Debrenne & Kruse (1986) has been described in Nimrod Glacier, but the bad preservation of our material does not allow us to assign it to a specific species.
Occurrence. Etymology. In memory of Werner Buggisch, German geologist who sampled and studied this Antarctic material.
Diagnosis. Cup 0.8-3.2 mm in diameter. Outer wall with two to three rows of multiperforated tumuli per intersept. Inner wall with two to three rows of pores per intersept, bearing supplementary tubular bracts on central cavity side. Septa completely porous.
Remarks. The porosity of the inner wall and the septa are the distinctive character to distinguish our material from the other genera belonging to the other families of the Lenocyathoidea.
Occurrence. Diagnosis. Outer wall with multiperforated tumuli. Inner wall with several rows of pores per intersept, bearing possibly upwardly projecting, S-shaped scales. Septa aporose to sparsely porous.
Remarks. This genus is characterised by inner wall with one to two rows of pores per intersept, bearing S-shaped scales. Septa aporose to sparsely porous.

Santelmocyathus santelmoi Perejón, Menéndez & Moreno-Eiris
gen. et sp. nov. (Fig. 15a, b) Etymology. From the San Telmo Spanish ship, possibly the first to reach the coast of the Antarctic continent in 1819.
Diagnosis. Cup 2.8-8.3 mm in diameter. Outer wall with one to three rows of multiperforate tumuli. Inner wall with one to two rows of pores per intersept, bearing possibly upwardly projecting, S-shaped scales. Septa aporose to sparsely porous with one to three rows of pores. Radial coefficient RK 2.9-3.8.
Remarks. Santelmocyathus santelmoi differs from Shackletoncyathus buggischi in the inner wall, by having scales and not bracts, and the septal porosity, as it occurs in another new genus of family Shackletoncyathidae.
Remarks. The canals of the outer wall and the simple pores of the inner wall are characteristic of Fallocyathus. The type of preservation of cup sections does not allow us to assign them to a specific species.
Remarks. The specimens are intensely affected by recrystallisation and tectonic deformation processes, which does not allow assigning them to a specific species. Antoniocoscinus retifer is present in Ajax Mine.
Remarks. Measurements and coefficients of Shackleton specimens fall within the range of the species Erismacoscinus bilateralis described by Kruse & Debrenne (2020), which includes some specimens of Gordon's 1920 E. endutus and E. fultus from Weddell Sea.
Remarks. The porosity of both walls is similar to Retecoscinus apart from the presence of tabulae with slit-like pores, but the designation is doubtful due to the poor preservation of this specimen.
Occurrence. Diagnosis. Outer wall with normal pores. Inner wall with several rows of pores per intersept, bearing S-shaped bracts; septa completely porous; retiform tabulae with subpolygonal pores.
Remarks. Wegenercyathus shares the same septal porosity as Rudanulus, the difference is that Rudanulus has a longitudinally plicate outer wall and inner wall with scales. The new genus differs from Pilodicoscinus in that the latter has plicate outer wall, cupped inner wall bracts and aporose to sparsely porous septa. Wegenercyathus shares the same septal porosity as Yhecyathus, the difference is that the latter has plicate outer wall, and cupped inner wall bracts.
Diagnosis. Outer wall with two or three rows of simple pores per intersept. Inner wall with several rows of pores per intersept, bearing S-shaped bracts; septa completely porous; retiform tabulae with hexagonal pores.
Remarks. In addition to the differences already indicated with the other genera of Rudanulidae, the presence of retiform tabulae with large subpolygonal pores distinguishes it, since these hexagonal pores are unusual.
Remarks. The porosity of both walls is characteristic of C. convoluta, including the other species described for the genus that are currently synonymous, which allows us to assign our specimens.
Remarks. The poor state of preservation of the specimen does not allow the observation of all the diagnostic characters to be able to assign it to a described species, although our material may have slightly wider intervallum than Putapacyathus regularis does, but it is a small difference, and it is found on trend lines of the intervallar coefficient.
Remarks. The poor preservation of our material does not allow us to assign it to a specific species. It is the first occurrence of this genus in Antarctica.
Remarks. The difference between P. parvus and P. sarmaticus is the absence or presence, respectively, of diaphragms over the outer wall pores. The Shackleton Range specimens present exostructures especially developed in the apical part. Vesicular tissue occupies the intervallum and central cavity in some cups, as well as the specimens described by Wrona & Zhuravlev (1996) from King George Island.
Remarks. The poor preservation and small size of our material do not allow us to assign it to a specific species.
Remarks. Our material presents the porosity of inner wall, septal porosity, synapticulae and septal arrangement similar to the specimen described by Debrenne 1992 from EW, which in Kruse & Debrenne (2020) is included in the synonymy as ?G. graphica.
Remarks. Different species of Archaeocyathus genus have been identified in a wide geographic range: Europe, Africa, Asia, America, Australia and Antarctica. Note among these the assignment to Archaeocyathus sp. has been made in material from South Africa, main Karoo Basin, Zwartskraal, Dwyka tillites (Debrenne 1975;Debrenne & Kruse 1989) and in Antarctica, Ellsworth Mts, Heritage and Sentinel Ranges, erratic clasts in the Permo-Carboniferous Whiteout Conglomerate (Debrenne 1992).
Remarks. Our specimen presents a skeletal structure very similar to Metacyathellus caribouensis (Handfield) in Handfield (1971, p. 64, pl. 11, fig. 2a) from Canada. Metacyathellus lairdi (Hill 1964b) has been described in Antarctica, Nimrod Glacier (Hill 1964b). The poor preservation of our material does not allow us to assign it with certainty to a specific species.
Occurrence. Allochthonous clasts: Antarctica, Shackleton Range, Stephenson Bastion, Cenozoic glacial erratic tills. Cambrian Series 2, Botoman. Remarks. The small size of the cups and their state of preservation do not allow us to assign these samples to a specific genus and species.
Description. Solitary cup, 4 mm in height and 2.4 mm in diameter. Wall 0.01-0.04 mm thick with irregular undulations in the apical part. Inner cavity is crossed by thin tabulae (0.02 mm), most of which are complete, flat or arched upwards when they are recurved. Tabulae spacing ranges from 0.20 to 0.40 mm. Septa are not visible. The upper edge of cup is preserved.
Remarks. Our specimen shows the arrangement, distance and thickness of tabulae similar to the others described previously. Mansy et al. (1993) figured and do not describe a small specimen like our cup. The preservation of the skeletal parts of our specimen is clearly different from that observed in the neighbouring archaeocyath cups (in the same thin section), being a differentiating feature of its primary skeletal composition. It is the first occurrence of this genus in Antarctica.

Biostratigraphical and palaeobiogeographical correlations
The stratigraphic distribution of the archaeocyathan genera from the Shackleton Range ranges from Tommotian 1 to Toyonian 3 (Siberian stages, Cambrian Stage 2 to Stage 4) according to Debrenne et al. (2015) (see Appendix 3). There are no biozones based on archaeocyaths in Antarctica. The Antarctic archaeocyathan fauna has usually been compared with Australian species (see Debrenne & Kruse 1989;Wrona & Zhuravlev 1996) since many species are common and support the concept of a unified Australia-Antarctica province (Australian archaeocyath fauna from Arrowie and Stansbury basins, Gnalta Shelf and fauna from Antarctica, South Africa and Falkland Islands, according to Kruse & Shi in Brock et al. 2000). Thus, we compare the Shackleton Range species with the Australian biostratigraphic schemes. In Arrowie and Stansbury Basins, Australia, Zhuravlev & Gravestock (1994) designated three biozones (Warriootacyathus wilkawillinensis, Spirillicyathus tenuis and Jugalicyathus tardus) and informally named two younger biozones ('Syringocnema favus beds' and ' Archaeocyathus abacus beds').
Kaltatocyathus gregarius and ?Baikalocyathus sp. are reported in the Atdabanian Spirillicyathus tenuis Zone in Australia (Gravestock 1984), but they coexisted with Botoman taxa in the Shackleton Range (see Appendices 2 and 3). Thus, they expand their stratigraphic ranges in the Australia-Antarctica province. Debrenne & Kruse (1986) also reported the presence of Kymbecyathus avius from samples of central TAM and it was assigned questionably to the Atdabanian Jugalicyathus tardus Zone by Zhuravlev & Gravestock (1994), and may possibly be of Atdabanian age, rather than stated Botoman age, due to structural complexity (Kruse & Debrenne 2020, p. 53). In the Shackleton Range, K. avius occurs in a Cenozoic glacial erratic sample from Stephenson Bastion with other undetermined taxa. Therefore, it would have an uncertain ?Atdabanian-?Botoman stratigraphic range in the Shackleton Range (see Appendix 3).
In the Shackleton Range, a total of 189 specimens have been identified in the Mount Wegener Formation from Trueman Terraces (41), Swinnerton Ledge (8), Oldhamia Terraces (66) and in the Cenozoic erratics from the Stephenson Bastion (70) and Du Toit Nunataks (4) (see Appendix 2). The Shackleton Range archaeocyathan assemblage comprises 34 different taxa corresponding to five new genera, six new species, ten specific species, seven doubtful genera, 14 sp. and four gen. and sp. indeterminate. A new archaeocyath family has been proposed. Therefore, the Shackleton Range archaeocyathan fauna is one of the most diverse records of allochthonous Antarctic assemblages described so far.
The Mount Wegener Formation and the Cenozoic erratics share the following 9 species out of the total taxa described:  Hill (1965) from the Whichaway Nunataks, and has subsequently been reassigned to ?Cadniacyathus curvatus by Debrenne & Kruse (1989).
The remaining newly reported genera from the Shackleton Range (?Baikalocyathus, ?Antoniocoscinus, ?Retecoscinus and Neoloculicyathus) have a wide geographical distribution. However, the Shackleton Range fauna have three species that have only been described in Australia -Kaltatocyathus gregarius (Arrowie Basin; Gravestock 1984), Nochoroicyathus hystrix and Nochoroicyathus lawrencei (Gnalta Shelf; Kruse 1982) (see Appendix 4). The Shackleton Range fauna show an extremely limited specific affinity with the autochthonous fauna from Antarctica. A total of 45 species have been previously described in the early Cambrian record of Antarctica, including 37 from the Shackleton Limestone, central TAM (Hill 1964b;Debrenne & Kruse 19861 and 2 in Fig. 20) and 17 from the Schneider Hills limestone, Argentina Range (Konyushkov & Shulyatin 1980;Debrenne & Kruse 1989;4 in Fig. 20). However, only Kymbecyathus avius (Shackleton Limestone) and Thalamocyathus trachealis (Shackleton Limestone and Schneider Hills limestone) are in common with the Shackleton Range assemblage (see Appendix 4).

Comparison with allochthonous archaeocyathan assemblages
The Shackleton Range fauna only share ?Ladaecyathus and ? Erismacoscinus with the allochthonous archaeocyathan record from the marine Cambrian El Jagüelito Formation in South America (González et al. 2011 and references therein).
The Mount Wegener Formation consists of marine clastic deposits sedimented from the upper slope to deep basinal settings in synorogenic conditions . The high proportion of plagioclase, volcanoclastic and unstable heavy minerals suggest that the lower Cambrian Mount Wegener Formation was deposited in a back-arc basin (Buggisch et al. 1990;Kleinschmidt & Buggisch 1994;Buggisch et al. 1994a). However, the carbonate platform that generated the clasts that eventually formed the conglomerates and breccias of this formation does not outcrop in the Shackleton Range. Furthermore, this concealed platform developed on the Northern Belt, northwards of the Read Group (part of the EAC) and its sedimentary cover (Watts Needle Formation) ( Fig. 3a; b in Fig. 21).
In the Northern Belt of the Shackleton Range, the Precambrian basement rocks with Pan-African overprinting and the northern neighbouring Coats Land Block (Kleinschmidt & Boger 2009;Loewy et al. 2011) are separated from the EAC (northernmost part of the Mawson Continent, Will et al. 2009;Boger 2011) by an E-W suture. Aeromagnetic data suggest that the E-W Shackleton Range suture could extend at least 500 km into E Antarctica and shift to an N-S orientation in the Recovery Lakes area (Golynsky et al. 2018). It should be noted that the Coats Land Block is separated from the Kalahari and Grunehogne cratons (K and G in Fig. 21) by the Grenvillian-age Maud Belt, which is interpreted as the continuation of the Namaqua-Natal Belt of southernmost Africa Wang et al. 2020 , fig. 11).
The sedimentation age of the Mount Wegener Formation has been estimated in different ways (Fig. 3). Shales give an Rb-Sr isochron age of ∼526 Ma that was interpreted as a pre-cleavage event, such as diagenesis of sediments (Pankhurst et al. 1983). K-Ar dating ages around 547-506 Ma (2-6 μm whole rock fraction) were interpreted as mixtures of inherent sedimentation/ diagenesis ages and around 490 Ma as the upper limit of deformation and metamorphism (Buggisch et al. 1994b). However, the K-Ar dating of detrital muscovites from greywacke turbidites of the Mount Wegener Formation ranges between 572 and 534 Ma, reflecting source rocks with different Pan-African primary cooling and exhumation histories; thus, the minimum cooling age limited a maximum age of sedimentation of the Mount Wegener Formation up to ca. 535 Ma . In fact, this minimum cooling age of detrital muscovites matches with the age of magmatism in the northern terrane, which is associated with subduction of oceanic crust by ∼530 Ma Will et al. 2009). Therefore, the Rb-Sr and K-Ar dating ages support an unknown Ediacaran-Terreneuvian rock source that does not crop out in the Shackleton Range (Figs 3, 21). The only known Ediacaran rocks are the EAC's autochthonous sedimentary cover, Watts Needle Formation. However, the paleocurrents of the Mount Wegener Formation point to a source area located to the N , whereas the paleocurrents from the Watts Needle Formation point to the S (Buggisch et al. 1990). Furthermore, the neodymium (Nd) isotope values of the Mount Wegener Formation are more like Pan-African Grenville-age basement rocks such as Pioneers Group, than those from the Read Group .
The fossil content of the Mount Wegener Formation was analysed in an incipient way (Buggisch et al. 1990(Buggisch et al. , 1994a, so the previous given age is a wide early Cambrian Atdabanian age . It should also be considered that there are two groups of fossils, one within the allochthonous carbonate clasts in the conglomerates, and another in the autochthonous presence of the Oldhamia ichnotaxon on the turbidite levels (Oldhamia cf. antiqua and Oldhamia cf. radiata according to Buggisch et al. 1990). The First Appearence Datum (FAD) of Oldhamia is placed in the Fortunian (Mángano & Buatois 2016). However, Herbosch & Verniers (2011) reviewed the biostratigraphic value of the cosmopolitan Cambrian Oldhamia ichnospecies and suggested that of 19 occurrences observed worldwide, 14 (Mount Wegener Formation included) occurred in a well-constrained time interval, ranging from the base of Stage 3 to the lower three quarters of Wuliuan. Furthermore, these authors conclude that Oldhamia taxa from the Mount Wegener Formation just above the archaeocyaths, have an age that could extend from the base of Stage 3 (appearance of trilobites) to the lower half of Stage 4 (high diversity of archaeocyaths). The stratigraphic section of the Mount Wegener Formation records around 770 m at the Trueman Terraces but probably exceeds 1000 m (see Figs 4d, 5), although the total thickness of the unit is unknown. Oldhamia is recorded in the intermediate levels of the Oldhamia Terraces. However, we cannot rule out the presence of archaeocyath-bearing clasts or ichnotaxa in the rest of the mapped thickness of the Mount Wegener Formation (see This LAD for O. radiata would be below the basal conglomerate of the Gull Lake Formation, in the transition between the Nevadella and Bonnia-Olenellus Zones. The basal limestone conglomerate contains archaeocyath-and Tabulaconus-bearing clasts that may represent debris flows from coeval platform deposits of the Sekwi Formation (Gordey & Anderson 1993). In fact, this basal limestone conglomerate is correlated by MacNaughton et al. (2016) with the regressive sandstone found within the Sekwi Formation in the Mackenzie Mountains, Northwest Territories. The Sekwi Formation contains trilobites from the Fallotaspis, Nevadella and Bonnia-Olenellus Zones (Fritz 1972) and three Botoman archaeocyath assemblages (archaeocyathan zonation for Laurentia according to Mansy et al. 1993, modified by McMenamin et al. 2000. Lower Botoman Ethmophyllum withneyi-Sekwicyathus nahanniensis belonging to the middle Nevadella strata and two middle Botoman assemblages Claruscoscinus fritzi-M. caribouensis and ?Pycnoidocoscinus serratus-Tabulaconus kordeae occurring within the middle Bonnia-Olenellus strata. Specifically, the carbon isotopic excursion cycle C recorded in the uppermost Nevadella Zone in the Sekwi Formation has been correlated with the VII positive excursion on the Siberian carbon isotope curve by Dilliard et al. (2007). Harvey et al. (2011) provided a maximum age of 514.45 ± 0.36 Ma for the Cambrian Stage 3-Stage 4 boundary with the zircon 206 Pb/ 238 U dating of an ash from the upper part of the trilobite Callavia biozone (England). On the Siberian Platform, He et al. (2019) correlated the boundary of the uppermost Atdabanian archaeocyath Fansycyathus lermontovae Zone and the lowermost Botoman trilobite Bergeroniellus micmaciformis-Erbiella Zone with this radiometric age. Thus, the age of the Atdabanian/Botoman boundary remains uncertain on the Siberian Platform. The VII positive C-isotope excursion is recorded in the basal Botoman trilobile B. micmaciformis-Erbiella Zone, around 514 Ma (according to He et al. 2019). Thus, in the Selwyn Basin, O. radiata was coeval with lower Botoman archaeocyaths but did not coexist with middle Botoman archaeocyath assemblages or with the coralomorph Tabulaconus kordeae.
It is noteworthy to mention that the Cambrian fauna of the Mount Wegener Formation shares three species with the fauna of the Selwyn Basin: O. radiata, O. antiqua and Tabulaconus kordeae. The archaeocyathan fauna from the limestone clasts of the Mount Wegener Formation provide a correlation with Botoman fauna (see section 8) for the carbonate platform from which they were derived. Specifically, the archaeocyath assemblage of the Shackleton Range shares taxa with the Syringocnema favus beds fauna (see section 8). This archaeocyathan fauna is timeequivalent to the trilobite Pararaia bunyerooensis Zone (see Jago et al. 2020 and references therein), and the radiometric ("chemical abrasion" or CA-TIMS method ) ages of three volcanic horizons within the P. bunyerooensis Zone from the Mernmerna Formation support ages of 514.46 ± 0.13 Ma, 514.56 ± 0.13 Ma and 515.38 ± 0.13 Ma (Betts et al. 2018). The archaeocyath-bearing clasts from the Mount Wegener Formation were deposited on upper slope to basinal settings after the breaking up and sedimentary brecciation of the carbonate platform, just after early marine phreatic to vadose diagenesis (see section 6). Therefore, at least the first 650-700 m of thickness (Fig. 5a) can be assigned to a maximum depositional age of terminal Stage 3 (∼515.5-514.3 Ma according to Australian data from Betts et al. 2018). The Shackleton Range record shows that O. radiata was partially contemporaneous with Tabulaconus kordeae and suggests certain diachronism with the O. radiata LAD from the Selwyn Basin. There are insufficient data to establish the final sedimentation age of the Mount Wegener Formation beyond the Wuliuan (according Herbosch & Verniers 2011) or beyond the early Guzhangian considering the LAD of Cambrian Oldhamia ichnospecies proposed by  Buggisch et al. 1990;Kleinschmidt & Buggisch 1994) and in the central Transantarctic Mountains sector (c-d, modified from Goodge 2020). In the Shackleton Range, the clasts from the Mount Wegener Formation and those from this unit present in the Cenozoic glacial tills suggest the existence of a volcanic arc and mixed sediment inputs of Ediacaran and Cambrian age (Terreneuvian Series 2). However, Terreneuvian Series 2 shallow-water deposits on the Proterozoic Watts Needle Formation or their deep-water equivalents are unknown. Low sedimentation rates or even erosion at the EAC margin during that time cannot be ruled out. In this study we propose the reconstruction of a lost Cambrian carbonate platform (see Fig. 22 MacNaughton et al. (2016). Therefore, the minimum age of the unit is given by the deformation and metamorphism ages around 490 Ma (Buggisch et al. 1994b) during Furongian.
In the Shackleton Range, the middle Cambrian erratics are fossiliferous shales and calcareous siltstones containing trilobites, obolid brachiopods, hyoliths and other molluscs, and are informally known as the 'trilobite shales' (Fig. 3; see Thomson et al. 1995 andreferences therein). Solov'ev &Grikurov (1979) described nine trilobite assemblages from these erratics. The presence of the brachiopod Notiobolus tenuis (Popov & Solov'ev 1981) was assigned to a pre-Drumian age based on its co-occurrence with Ptychagnostus gibbus and Ptychagnostus praecurrens (Popov et al. 2015). Solov'ev and Grikurov's trilobite associations have been grouped as Fauna 2 and correlated with gibbus to atavus Zones, from late Templetonian to early Floran Australian stages by Cooper & Shergold (1991). However, Lieberman (2004) suggested that some of the Solov'ev & Grikurov (1979) figured material might be referred only questionably to Ammagnostus laiwuensis, showing a broad range in South China from the upper Ptychagnostus atavus Zone to the Proagnostus bulbus Zone (Peng & Robison 2000), late Floran to early Mindyallan (Drumian-Guzhangian). Therefore, the trilobite shales were deposited during the Wuliuan-Drumian or even Wuliuan-?Guzhangian, depending on whether Lieberman's subsequent reassignment is correct. Trilobite distortion and K-Ar ages (2-6 μm) of shales around 463-455 Ma are interpreted as weak deformation and very low-grade metamorphism (Buggisch et al. 1994a). Thus, the stratigraphic position of this ex situ record has been placed under the sedimentary molasse deposits of the Ordovician Blaiklock Glacier Group ( Fig. 3; Thomson et al. 1995). The Blaiklock Glacier Group has yielded an Rb-Sr date of 482 ± 11 Ma isochron (Pankhurst et al. 1983), while the K-Ar date of micas from underlying leucogneiss, undeformed granitic clast and detrital micas from sandstones are around 516-498 Ma with palaeomagnetic declination data according to known Ordovician pole positions . Thus, the detrital micas from the Blaiklock Glacier Group support a Guzhangian-Furongian uplift and a history of exhumation for the northern basement of the Shackleton Range.
In summary, part of the upper slope to deep basinal deposits from the Mount Wegener Formation contains evidence of rock sources from late Ediacaran up to Cambrian Series 2, and locally sourced middle Cambrian erratics (trilobite shales) support that Cambrian shallow marine sedimentation continued during Wuliuan-?Guzhangian ages in the Shackleton Range sector (Fig. 21). However, the tectonic emplacement of the Mount Wegener nappe and its very low-grade metamorphic overprint is older (around 490 Ma) than the weak tectonic deformation and low-grade metamorphic overprint observed in the trilobite shales . This suggests that middle Cambrian shallow marine sedimentation did not develop in the same tectonosedimentary environment (Fig. 21).

Palaeoenvironmental reconstruction of the lost Cambrian platform
The analysis of the microfacies of carbonate clasts allows us to reconstruct different sub-environments of the lost carbonate platform from which they were derived. The analysed carbonate clasts belong to different groups of lithofacies (dolomitic sandstones to sandy dolostones, aggregate grain-to ooid-rich dolostones, dolostones, calcimicrobe-and archaeocyath-rich limestones)an outcome of the carbonate production in shallow waters from platform-interior ?restricted, oolitic shoal complex and open subtidal platform settings (Fig. 22). The high proportion of terrigenous sands in some dolostone lithofacies also suggests a mixed siliciclastic-carbonate platform attached to land.
The pervasive fabric-retentive dolomitisation observed in dolostone clasts (RD1 in Fig. 9a, c, f) could indicate conditions like those observed in present-day dolomites in restricted evaporative shallow-marine to supratidal environments (Tucker & Wright 1990 and references therein). Recent penecontemporaneous dolomites were found within microbial mats (Vasconcelos et al. 1995;Mazzullo 2000) and the recognised importance of low-temperature microbially mediated dolomite formation has strengthened in the last decades (Petrash et al. 2017 and references therein). It should be noted that microbially mediated dolomitisation produces small amounts of dolomite in modern shallow-marine sediments compared to other fossil examples (e.g., reflux dolomitisation by mesohaline brines is capable of pervasive dolomitisation of large areas of carbonate platform; see Machel 2004). However, DiLoreto et al. (2019) have found rhombohedral ordered dolomite in microbial mats in Qatari sabkhas dominated by filamentous anoxygenic photosynthetic bacteria. These authors propose that, in parallel to secular changes in ocean geochemistry, the evolution and structure of microbial mat communities may have favoured the predominant type of carbonate precipitation within the mat. Precambrian dolomites are primarily associated with shallow subtidal to intertidal facies related to microbial deposits, while limestones correspond to deeper waters facies (see discussion in Tucker 1992 and references therein). In Neoproterozoic oceans, the co-occurrence of marine aragonite, high magnesium calcite and dolomite (mimetic dolomitisation and primary cement) suggest extremely high magnesium/calcium ratios and marine anoxia (Hood & Wallace 2018). Furthermore, microbial sulphate reduction probably triggered precipitation of fibrous dolomites from euxinic porewaters (Hu et al. 2020). Evaporative conditions in combination with high rates of microbial activity (e.g., anoxygenic photosynthetic bacteria in biofilms) could have promoted the formation of extensive penecontemporaneous dolomites on those ancient microbe-dominated carbonate platforms, as suggested by Daye et al. (2019). Currently, we do not have any direct evidence of evaporite formation, since trace elements, fluid inclusions, oxygen and carbon isotopic data are not available; therefore, any dolomitisation model (microbial, evaporative, seepage-reflux, meteoric-marine mixing-zone) should be viewed with caution. Thus, the observed mimetic penecontemporaneous dolomitisation (RD1) seems to be a facies-selective process associated exclusively with platform-interior ?restricted and oolitic shoal complex settings, developed in a near-surface and/ or shallow burial diagenetic setting.

Platform-interior ?restricted setting: mixed sandy carbonate peritidal and storm-related deposits
The dolomitic sandstones (Fig. 7a) with beach rock cements are interpreted as mixed sandy carbonate shore/flat deposits. Their very well rounded and sorted fine to medium detrital quartz grains suggest long-term abrasion due to the transport of the wind by saltation and surface creep. Therefore, mixed sandy carbonate shore/flat deposits (7 in Fig. 22) could have been fed by coastal eolian deposits. Silty/sandy dolomicrites with irregular vuggy to channelised cavities filled with clastic fillings (Fig. 7d) could reflect tidal flat sedimentation of mixed sandy carbonates with alternation of low energy (fallout of the suspended load during slack-water periods), exposure and/or burrowing (development of vuggy porosity), and sedimentation of the bed load as clastic fillings (pelletoids, superficial type 1 ooids, sand grains) by the action of current or waves under moderate energy conditions. Sandy very poorly to moderately sorted dolorudites with a mixture of eroded and redeposited intraclasts (silty/sandy dolomicrites, sandy oolitic-type 2-compound intraclasts, oolitic-, aggregate-grain-and algal peloid-rich dolograinstones), fragments of type 3 oolitic cortices, abraded micritic coated clasts and orange fibrous cement crusts were produced during highenergy events such as strong storms (Fig. 7b, c; 6 in Fig. 22). The mixture of intraclasts can be correlated with alternating processes of sedimentation, erosion, reworking and deposition by waves, tides and/or storms in peritidal areas, where intraformational conglomerates are common (Flügel 2004). Indeed, the silty/sandy dolomicrite intraclasts could be rip-up clasts eroded from muddy and mixed tidal flat areas. Storms could produce the observed mix of ooids (types 1-3) that are derived from different sub-environments. Superficial ooids (type 1), mostly with quartz nuclei, point to sandy pelletal tidal flat environments, while larger micritic concentric ooids (type 2 and type 3) indicate high-energy environments such as oolitic shoals (see below), as suggested by the study of carbonate grain distribution developed by Steinhoff & Strohmenger (1996) in the Upper Permian Zechstein 2 of Germany. Therefore, the largest ooids were derived from active shoal settings, while the smallest were produced in moderate-to low-energy, platform-interior settings (8 in Fig. 22). In addition, large sandy oolitic (type 2) compound intraclasts (Fig. 7b), derived from oolitic shoals, also involve repeated reworking and sedimentation in storm-influenced platform-interior settings. The scarcity and low diversity of calcimicrobe remains (Proaulopora, Renalcis; Fig. 8d) in the storm-related deposits indicate that they occasionally colonised the platform-interior, while meanwhile a high siliciclastic input prevented the expansion of other calcimicrobes. Proaulopora remains have been described in fenestral, peloidal and microbial grainstones developed in high-energy peritidal settings of the lower Cambrian Láncara Formation, Spain (Álvaro et al. 2000).

Oolitic shoal complex
Another important group of carbonate clasts is formed by the aggregate grain-to ooid-rich dolostones. These are dominated by aggregate grains or ooids or a mixture of both, reflecting different but closely related sedimentary sub-environments in a shallow subtidal to intertidal shoal complex (9 in Fig. 22). The low detrital sand content suggests that sedimentation took place far from the influence of coastal siliciclastic input.
Ooid-rich dolostones have noticeable diagenetic and microfacies characteristics linked with their paleoenvironmental settings (e.g., mimetic dolomitisation, concentric micritic laminae, ooid sizes). The type 2-3 ooids of the studied carbonate clasts are laminar concentric dolomicritic ooids (Figs 7f, 9a, c, f). The micritic laminae are considered secondary microfabrics produced by physical-chemical and/or microbial processes. Laboratory experiments suggest that mimetic concentric dolomite ooids may be useful indicators of precursor aragonite ooids and early dolomitisation (Zempolich & Baker 1993). The role of agitation and abrasion can be correlated with the size of the ooid, the density of the bands and the cortical fabric (Medwedeff & Wilkinson 1983), so equivalent rates of carbonate precipitation and cortex abrasion could produce micritic ooids (Wilkinson et al. 1984). However, micritic laminae have also been explained as micritisation by endolithic cyanobacteria and final filling of microborings with random microcrystalline cement (Margolix & Rex 1971;Reid & MacIntyre 2000). In addition, research in Bahamian ooids shows that microbes do not play an important role in early ooid genesis, but modify the chemistry and microfabrics of the cortices through extensive microboring activity and biotic aragonite cementation associated with cyanobacteria Solentia sp. and Hyella sp. (Duguid et al. 2010). Some Cambrian ooids show different types of microbial activity as well, such as encrusting filamentous cyanobacteria (e.g., Girvanella-cortex ooids; Liu & Zhang 2012) or the presence of microbial microborings, filaments and extracellular polymeric substances or EPS (Tan et al. 2018).
The selection, composition (type 2-3 ooids) and presence of broken and overgrown ooids indicate that oolitic dolograinstones correspond to highly turbulent environments caused by the action of waves, tides and/or currents that form oolitic shoal deposits. The development of giant ooids (type 3) is favoured by high saturation of seawater carbonate, low supply of nuclei, high accretion rates, high current velocities and ramp-style architecture according to Sumner & Grotzinger (1993). Other experiments carried out by Trower et al. (2017) indicate that both precipitation and abrasion play a significant role in the final size of ooids. They correlated the mode of transport and the size of ooids, so that the transport as suspended load produces larger ooids than those dominated by bed load transport. In the studied carbonate clasts, the different ooids (types 1-3) occur in characteristic microfacies, the small superficial ooids with quartz nuclei (type 1) suggest platform-interior settings with high siliciclastic contribution, while medium and giant ooids correspond to high-energy oolitic shoal settings. The most common examples of giant ooids come from Precambrian, Cambrian, Lower Triassic and Jurassic platforms (Lehrmann et al. 2012 and references therein). The lower Cambrian Qingxudong Formation, Sichuan Basin, could be a counterexample for giant ooids developed in deepening and moderate-energy subtidal environments affected by episodic hydrodynamic events (Tan et al. 2018).
Those depositional textures with abundant dolomicrite and varying proportions of ooids and/or aggregate grains indicate low-to moderate-energy conditions in the vicinity of oolitic shoal settings. The loosely packed oolitic dolowackestones represent a mixture of reworked ooids (type 3-2) that were redeposited by waves/storms in low-energy, depressed, muddy areas that were adjacent to an oolitic shoal setting, similar to washover deposits (6 in Fig. 22). The aggregate grain dolowackestones and dolopackstones (Fig. 7e) with the finest size sediment represent low-energy, depressed, muddy areas in a protected subtidal backshoal setting. Nowadays, aggregate-grains are characteristic allochems in shallow marine environments on tropical and subtropical carbonate platforms. For instance, the Bahama Banks grapestones are indicative of uneven water turbulence, low sedimentation rates and low-nutrient environments (Illing 1954;Purdy 1963a, b;Winland & Matthews 1974) developed in the vicinity of the shelf-margin reefs, and grading into oolitic shoals and algal-foraminiferal sands (Flügel 2004 and references therein).
The aggregate grain-oolitic dolograinstone (Fig. 7g) shows a characteristic laminoid fenestral fabric that can result from multiple processes such as degassing of decaying organic matter and desiccation of microbial mats and lime mud, among others, which are commonly associated with modern peritidal environments (Scholle & Ulmer-Scholle 2003;Flügel 2004). The lack of quartz sand grains suggests that sedimentation took place far from the coastal siliciclastic inputs. The mixture of components (e.g., aggregate grains, medium and giant ooids, large compound intraclasts) reflects that the fenestral aggregate grain-oolitic dolograinstones developed in transition zones between aggregate-rich areas and active oolitic shoals. Finally, the observed diagenetic vadose fabrics (e.g., intraparticle secondary porosity as oomolds, internal geopetal infills, microcrystalline crust, meniscus cement; Figs 7f, g, 9b) indicate that they developed in intertidal conditions, likely on a partially/episodically subaerially exposed oolitic shoal setting.

Open subtidal platform setting: calcimicrobe carpets, calcimicrobe-archaeocyath patch reefs and storm-related deposits
The calcimicrobe-rich and/or archaeocyath-rich limestone clasts from the Mount Wegener Formation and Cenozoic erratics evidence different environmental conditions in an open subtidal platform (1-5 in Fig. 22). The co-occurrence of calcimicrobes and archaeocyaths points out photic, normal conditions. As above, the low content of detrital quartz sand grains suggests shallow subtidal conditions far away from the coarse detrital siliciclastic coastal inputs. Calcimicrobe-rich microfacies are derived from calcimicrobial boundstones (calcimicrobe carpets; 1 in Fig. 22) and/or calcimicrobial boundstones with varying proportions of archaeocyaths (calcimicrobe-archaeocyath patch reefs; 2-4 in Fig. 22). The diversity of archaeocyath-rich microfacies shows that the archaeocyaths colonised diverse subenvironments, forming calcimicrobe-archaeocyath patch reefs under varying conditions of water turbulence (calcimicrobearchaeocyath boundstones and archaeocyath cementstones), and living in open spaces (calcimicrobe-free spaces; 5 in Fig. 22), on muddy substrates that could be disturbed during storms (archaeocyath floatstones; 6 in Fig. 22).
The occurrence of calcimicrobes in different microfacies of the carbonate clasts allows us to reconstruct their distribution on the platform. Calcimicrobe carpets are dominated by low-diverse calcimicrobial microframeworks. The frequency and prevalence of Epiphyton and/or Angusticellularia suggest that they could easily colonise substrates forming carpets or meadows on the Cambrian seabed, where grazing pressure was low (Debrenne & Zhuravlev 1997). In contrast to Epiphyton and Angusticellularia, Proaulopora and Renalcis appear to have no problem forming carpets/meadows near to siliciclastic inputs. Extensive calcimicrobial colonisation of the substrate may inhibit larval settlement of sessile benthic organisms such as filter-feeding archaeocyaths, where competition for available 'open space' is important when grazing pressure is low. The growth rate of archaeocyaths is considered relatively slow so they were easily overgrown by calcimicrobes or buried by mud (Zhuravlev 2001 and references therein). In fact, in the Epiphyton-group-dominated bioherms, some archaeocyaths are smaller and have a thick exotheca (Zhuravlev 1996 and references therein).
In the Shackleton Range sector, the ajacicyathides proliferated in the open subtidal platform from muddy bottoms to calcimicrobe-archaeocyath patch reefs (see Appendix 1). They played different roles as minor framebuilders in the calcimicrobial boundstones with archaeocyaths or as main framebuilders in the archaeocyath cementstones. In Cambrian ecosystems, the mostly solitary ajacicyathides inhabited soft, muddy substrates with a high sedimentation rate (Debrenne 2007), as well as forming thickets and skeletal piles for the successive calcimicrobearchaeocyath colonisation (Zhuravlev 2001).
The maximum diversity in the number of archaeocyath families (75 %) and the presence of rare calcimicrobe intergrowths (highest calcimicrobe diversity) occurs in the calcimicrobial boundstone with archaeocyaths (2 in Fig. 22; see Appendix 1). It could represent a transition zone between low-diversity calcimicrobe carpets and other calcimicrobe-archaeocyath patch reefs, where the volume of sessile archaeocyaths increases (from 15 % to 50 %) but the diversity of families decreases (30 % and 35 %, respectively; see 3 and 4 in Fig. 22). Furthermore, the small size of the growth framework cavities in the calcimicrobial boundstone with archaeocyaths indicates a homogeneous and dense colonisation of the substrate by calcimicrobes. In the Shackleton Range, Epiphyton-Tarthinia-Girvanella or Epiphyton-Girvanella intergrowths are not common (Fig. 8f-i). Examples of Cambrian Series 2 intergrowths have been described in the Tarthinia-Epiphyton-Gordonophyton-Renalcis boundstone from Mongolia (Wood et al. 1993) and in dendritic thrombolites from North China (Lee et al. 2014). However, both examples lack Girvanella intergrowths and the latter lacks co-occurrence with archaeocyaths. Toyonian archaeocyathan-calcimicrobial reefs of South China display a diverse assemblage of calcimicrobes (Adachi et al. 2014), where Girvanella is found intergrowing with Epiphyton bushes and encrusting archaeocyaths. However, Girvanella crusts around archaeocyaths are very rare in the Shackleton Range sector.
The calcimicrobe-archaeocyath patch reefs with a lower diversity of archaeocyathan families (35 %) show larger growth cavities and an extensive development of early marine cement-supported fabrics, indicating that these patch reefs (calcimicrobe-archaeocyath boundstone and archaeocyath cementstone, respectively) developed wave-resistant frameworks under high-energy conditions. The peloid-intraclastic-bioclastic packstones to grainstones with remains of archaeocyaths could represent storm sheets derived from areas close to the calcimicrobe-archaeocyath patch reefs and surrounding muddy bottoms and deposited above the fair-weather wave base (FWWB), whereas the archaeocyath floatstones were deposited below the FWWB.
The muddy bottoms were colonised by a diverse heterozoan assemblage. Trilobites, brachiopods, hyoliths and echinoderms are common in wackestone to packstone pockets, thus likely living in the vicinity of the calcimicrobe-archaeocyath patch reefs on muddy subtidal substrates. There is a high diversity of archaeocyath families (55 %) in wackestone to floatstone, the presence of Bronchocyathidae, Kymbecyathidae, Dictyocyathidae and Copleicyathidae being exclusive, with Bronchocyathidae being the dominant family together with the conspicuous Ajacicyathidae.

Comparison with neighbouring Cambrian inboard successions of the EAC
The Cambrian carbonate platform inboard successions are restricted to the central and southern Transantarctic Mountains and the PM (TAM and PM in Fig. 21) and they record the transition from passive to active margin deposition (Goodge 2020 and references therein).
The best-known Cambrian Series 2 carbonate record corresponds to the lower Atdabanian to Botoman Shackleton Limestone (Byrd Group; Fig. 23 and references therein) in the central TAM (sector c-d in Fig. 21). The total thickness of the unit is unknown due to tectonic complication. The lower mixed siliciclastic-carbonate part is well recognised in the central Holyoake Range on the Errant Glacier side (∼133 m thick; Myrow et al. 2002a , fig. 4). The upper carbonate-dominated succession, around 225-320 m thick, has been analysed in the N and centre of the Holyoake Range Claybourn et al. 2019, respectively). The lower Shackleton Limestone consists of interbedded quartz sandstone with mud drapes and wave ripples and fine-grained dolomitic grainstone with hummocky crossstratification (Myrow et al. 2002a).
The upper Shackleton Limestone comprises supratidal (Burgess & Lammerink 1979) to deep subtidal deposits on a carbonate ramp (Myrow et al. 2002a) where isolated bioherms (<2 m) and biohermal complexes (20-50 m thick) proliferated . Rees et al. (1989) recognised four depositional subenvironments whose lithofacies associations share some characteristics with those recognised in the Shackleton Range. The intertidal association corresponds to dolomitic limestones (fenestral cryptomicrobial laminites, fenestral mudstone, parallel-and ripple cross-laminated peloidal grainstone and dolomitic intraclastic flat pebble rudstone), while in the Shackleton Range they correspond to mixed sandy penecontemporaneous dolostones with a diverse allochem association. The carbonate sand shoal association shows varied allochems (ooids, coated grains, oncoids, bioclasts, grapestones) in thick, parallel-, planar cross-and bimodal cross-stratified grainstone sequences, which are interbedded with bioclastic, intraclastic and peloidal grainstones as well as oncolite rudstones. However, oncolite rudstones and bioclastic grainstones are not found in the Shackleton Range sector. In the Shackleton Limestone, the microstructures of ooids are diverse (radial, concentric or composite versus concentric cortices) as well as their nuclei (peloids, trilobites, echinoderms versus quartz grains or broken cortices). However, the presence of giant ooids or sandy oolitic compound intraclasts is not reported in the Shackleton Limestone. And, although early diagenetic processes are recorded (isopachous cement, micritisation, oomoldic and shelter porosity, equant spar), neither mimetic penecontemporaneous dolomitisation nor early silicification processes are described in the Shackleton Limestone. In fact, dolomitisation of the carbonate sand shoal association is described as partial replacement of ooids by large dolomite crystals. The shallow-subtidal shelf association shows some common lithofacies such as peloid, intraclastic, bioclastic packstones to grainstones or oolitic grainstones derived from nearby sand shoals that are interpreted as storm deposits. However, bioturbated bioclastic wackestones and packstones or spongiostromata oncolite packstones are not recorded in the Shackleton Range sector. The archaeocyathan-microbial reef association is interbedded with burrow-mottled and oolitic limestones, and allochems from the carbonate sand shoals are also present in some of the reef complexes. This interdigitisation of facies and components is not observed in boundstone from the Shackleton Range sector. However, the diversity in calcimicrobes is much higher in the Shackleton Range, since only Renalcis, Epiphyton and Girvanella have been reported in the Shackleton Limestone. Rees et al. (1989) described three types of reefs, formed by Epiphyton-bearing boundstone, Renalcis boundstone and stromatactis-bearing boundstone. Only the Epiphyton-bearing boundstone, developed in moderately high-energy areas, share some characteristics with the analysed boundstone microfacies. In fact, Epiphyton and Angusticellularia are the main calcimicrobes in calcimicrobe carpets, as well as in calcimicrobe-archaeocyath patch reefs in the Shackleton Range sector. In the Shackleton Limestone, the dominant reef type is Epiphytonbearing boundstone with subordinate Girvanella and Renalcis. Archaeocyaths comprise one-third of its volume or are absent. This type of reef developed centimetre-to decimetre-scale, open growth-framework cavities colonised by calcimicrobes, lined by marine cement crusts with multi-episodic geopetal infills. However, boundstones of the Shackleton Range sector display millimetre-to centimetre-scale growth-framework cavities and lack multi-episodic geopetal infills (see mesostructure maps in Fig. 22).
In the central Transantarctic Mountains, Goodge (2020 and references therein) differentiated four main phases in the Cordilleran-type, active continental-margin Ross orogenic belt: • Early Neoproterozoic passive-margin phase represented by the Beardmore Group; The in situ Cambrian Series 2 carbonate record closest to the Shackleton Range crops out in the Argentina Range, PM (Fig. 23). Rowell et al. (1992a) referred to this 200-m-thick formation informally as 'Schneider Hills limestone' and suggested that it may have been part of the same broad carbonate shelf that the Shackleton Limestone deposited on the western passive margin of the E Antarctica basement (Nimrod Group) (c-d in Figs 21, 23). However, infracrustal rocks do not crop at PM and its Cambrian Series 2 deposits are not contiguous with those of the central TAM (Schneider Hills limestone is absent in the closest Neptune Range). In the PM the supracrustal rocks correspond to the Hannah Ridge Formation (Fig. 23). Detrital zircon data from this formation gave an age of latest Neoproterozoic or younger (Rowell et al. 2001;Goodge et al. 2004a). This formation underwent two phases of deformation prior to uplift, erosion and deposition of the Miaolingian Nelson Limestone, which contains supracrustal rock clasts with two cleavages in its basal conglomerate (Curtis & Storey 2003 and references therein). Evans et al. (2018) suggested that Cambrian Series 2 (and older) limestones may be covered by snow and ice at the Patuxent Range or were even removed by uplift and erosion in the Neptune Range during an early phase of Ross orogeny. Alternatively, the lack of detrital inputs of this age in the succeeding record could also reflect a local palaeotopographic high that avoid the sedimentation during Cambrian Series 2.
Data on the Schneider Hills limestone are scarce and there is no published stratigraphic section. Konyushkov & Shulyatin (1980) described the presence of calcimicrobes (Renalcis, Epiphyton), stromatolites and gave a list of genera and species of archaeocyaths, but they did not figure them. Later, Debrenne & Kruse (1989) revised the Antarctic Cambrian archaeocyaths and gave the Konyushkov and Shulyatin's assemblage a Botoman age. Rowell et al. (1992b) described a succession of several hundred meters, with a lower part containing bioclastic (archaeocyath-rich) and oolitic grainstones, bioturbated wackestones and packstones with scattered microbial reefs (0.5-10 m wide, up to 2 m thick). The top of the lower part is rich in camenellan tommotiids (Dailyatia Rowell et al. 1992b; Dailyatia icari sp. nov. Claybourn et al. 2021), and bradoriide Bicarinella evansi (Rode et al. 2003) and is overlaid by 200-m-thick Cambrian, massive boundstone with grainstone and burrowed flanking beds. Archaeocyaths are accessory components in the massive boundstone and flanking beds. The succession continues with 150-250-m-thick, thin-bedded bioturbated limestones with scattered small boundstone mounds and a second massive boundstone with small cavities (<1 cm) and flanking beds of alternating grainstone and burrowed packstone. Rowell et al. (1992b) compared the thick Schneider Hills limestone massive boundstones with those of similar size in Virginia, the shelfmargin skeletal algal reefs described by Barnaby & Read (1990).
Comparisons between the Shackleton Range and PM Cambrian Series 2 carbonate platforms are limited due to existing data from the PM. The carbonate clast microfacies from the Shackleton Range sector do not exhibit bioclastic (archaeocyath-rich) grainstone, stromatolite or bioturbated wackestones or packstones. Burrowed dolomicrite with sandy infillings occur in the Shackleton Range sector (Fig. 7d) but we have interpreted them as part of the sandy-mixed carbonate tidal flat (7 in Fig. 22). Carbonate microfacies in common between the two regions are boundstone and oolitic grainstone, but they are widespread in the Cambrian Series 2 carbonate platforms. The mesostructure of calcimicrobe-archaeocyath patch reefs from the Shackleton Range sector (Fig. 22) shows cavities larger than those of the Schneider Hills limestone. Therefore, only the comparisons between the archaeocyathan faunas in both sectors (see section 8) make paleobiogeographic sense.
In the PM, the Ross orogeny is characterised by three Miaolingian to ?Ordovician deformation events (Curtis et al. 2004 and references therein) constrained by radiometric and paleontological data (Fig. 23). The early contractional deformation event is around 505 Ma and produced folding, associated cleavages, emplacement of syntectonic granites (later Sherpan Peak phase) and exhumation of the Hanna Ridge Formation prior to deposition of the Miaolingian Nelson Limestone. It is followed by a middle Ross extensional event characterised by sedimentation of the Nelson Limestone that terminated by bimodal volcanism ca. 500 Ma (Gambacorta Formation). The Nelson Limestone is >400 m thick (Evans et al. 1995(Evans et al. , 2018 and is Drumian or Drumian to early Guzhangian in age according to paleontological data ( Fig. 23 and references therein), so the total thickness was deposited in a 4-5 Myr time interval.
The Weins Formation has an unconstrained post-Gambacorta Formation pre-Neptune Group age. The presence of Notiobolus sp. is recorded in the Weins Formation (Storey et al. 1996). This genus is of Wuliuan-Jiangshanian age, which is the combined range of its three extinct species (Popov & Solov'ev 1981;Holmer et al. 2001;Popov et al. 2015). Therefore, the Weins Formation has a ?Guzhangian-?Jiangshanian age. Finally, a late Ross contractional event, ?Furongian-?Ordovician (there are no fossils or radiometric dating), produced the moderate deformation and weak cleavage of the previous units and the syntectonic deformation of the basal part of the overlying Neptune Group (Storey et al. 1996).
In summary, the spatial and temporal tectonic variability along the palaeo-Pacific margin and the Mozambique seaway margin (sectors c-d and a-b, respectively, in Fig. 21) during the Ross orogeny controlled the development of carbonate platforms. Cambrian Series 2 carbonate platforms, several hundred meters thick with calcimicrobe-archaeocyath bioconstructions, were developed on both margins during tectonic quiescent intervals from Atdabanian to Botoman (TAM) or at least during Botoman (PM and Shackleton Range). In the central TAM, the occurrence of shallow-water mixed siliciclastic-carbonate deposits is reported in the lower Shackleton Limestone, below the positive carbon excursion at the base of the Atdabanian or Stage 3 (Myrow et al. 2002a); therefore, part of the mixed sedimentation could have started at the end of Terreneuvian in this sector. In the reconstructed lost Cambrian Series 2 mixed carbonate platform of the Shackleton Range sector, the terrigenous influence is interpreted as part of the sandy mixed carbonate tidal flat deposits. Botoman is the age obtained from the assemblage of archaeocyaths in the Shackleton Range (see section 8). However, an older age cannot be ruled out for the development of a mixed siliciclastic-carbonate platform though, due to the existence of Ediacaran-Terreneuvian detrital ages recorded in the slope to basinal deposits of the Mount Wegener Formation (Fig. 24).
Some important differences observed in the Shackleton Range sector are the presence of early diagenetic phases such as pervasive mimetic dolomitisation (RD1) and authigenic and diagenetic silicification (Sch, RS1c-m), which are not described in other Cambrian inboard successions. The observed extensive penecontemporaneous dolomitisation can be related with more restrictive conditions in the Antarctic Mozambique seaway margin than in the Antarctic paleo-Pacific margin. The latitudinal differences (see ∼520 Ma paleogeographic reconstruction by Merdith et al. 2017) and the character of the seaway margin itself could lead to more arid conditions that, along with microbial activity, favoured early dolomitisation. Likewise, in a back-arc setting, the weathering of the volcanic arc and/or its activity could provide large amounts of dissolved silica that would favour near-surface silicification processes in porous microfacies (e.g., diagenetic vadose fabrics of the oolitic shoal complex). In addition to silica-supersaturated pore fluids, near-surface silicification processes require unsaturated carbonate pore fluids and a pH below 9. There are different mechanisms that promote the replacement of carbonates by silica (see Hesse 1989), but it seems that the organic matter oxidation by sulphate-reducing bacteria would be a key reaction for the precipitation of silica in platform sediments (Noble & Van Stempvoort 1989).
In the Shackleton Range sector (Figs 21, 24), the synsedimentary brecciation of the mixed siliciclastic-carbonate platform took place during the terminal Stage 3 (∼515.5-514.3 Ma by correlation with the Australian data from Betts et al. 2018) or during the late Atdabanian to lowermost Botoman (∼514.45 Ma considering the correlation age suggested by He et al. 2019) as indicated by the fauna recorded in the platform-derived clasts in the slope and basinal deposits of the Mount Wegener Formation. The first 700 m of thickness of the Mount Wegener Formation were deposited contemporaneously or shortly after platform breakup, as indicated by the successive diagenetic phases observed and their cross-cutting relationships recorded in the carbonate clasts and polymictic conglomerates (see section 6). At the same time, in central TAM, there was the synchronous development of an erosional unconformity, the drowning of the Shackleton carbonate ramp and its burial by siliciclastic sedimentation (Holyoake and lower Starshot Formations), which led to the cessation of carbonate production in this sector (Figs 21,23).
It is unknown whether the synsedimentary platform brecciation led to the demise of the carbonate production in the Shackleton Range sector. Shallow marine siliciclastic sedimentation (trilobite shales) may have continued during the Wuliuan-? Guzhangian interval in a different tectonosedimentary setting (? African Mozambique seaway margin; a-b in Figs 21, 24). However, Cambrian carbonate production continued in the PM during Miaolingian (Nelson Limestone) and Furongian (subordinate oolitic limestones of the Weins Formation; see Storey et al. 1996;Fig. 23).
The lower Starshot Formation (central TAM, Fig. 23) has some similarities with the Mount Wegener Formation, regardless of their differing depositional environments (shoreline to deeper shelf versus slope to basinal deposits) and tectonosedimentary settings (forearc versus back-arc basin). The polymictic conglomerates from both formations reflect erosion of volcanic arcs and carbonate platforms, archaeocyath-bearing clasts included (Myrow et al. 2002a). The thickness of the Starshot Formation is unknownup to 3000 m (Laird 1963)and no depositional contacts have been described. It was deposited from shoreline to deep shelf (Laird 1963;Laird et al. 1971) and sandstone tempestite beds were deposited as wave-dominated turbidites (Myrow et al. 2002b). The basal part of the Starshot Formation is late Botoman in age (Myrow et al. 2002a). Therefore, the sedimentation of both units could be partially contemporaneous and under synorogenic conditions.
The total thickness of the Mount Wegener Formation is unknown but probably exceeds 1000 m , especially considering its diagenetic evolution. The Mount Wegener Formation underwent various burial diagenetic processes (see section 6) such as dolomitisation (D2-D4) and hydrocarbon migration before undergoing the final tectonic deformation. The latest burial processes recorded correspond to late-fracture-related cementation (LVA and LVB systems), low-grade tectonically-induced plastic and cataclastic fabrics and very low-grade metamorphic overprinting. The late-fracture-related cementation produced the precipitation of very to extremely coarse non-ferroan to slightly ferroan poikilotopic calcite, ferroan saddle dolomite, coarse to very coarse ferroan calcite and varied silica phases such as megaquartz mosaics (Sqtz), quartz overgrowths and chert and microquartz replacements (RS2c-m). The cataclastic fabrics related to the nappe tectonism post-date all the observed cement and replacement phases. The low-grade metamorphic overprint and the southwards transport of the Mount Wegener Nappe occurred around 490 Ma as a result of the Ross (Pan-African) orogeny (Buggisch et al. 1994b;Fig. 24).
Finally, the Blaiklock Glacier Group (Shackleton Range), the Neptune Group (PM), the Douglas Conglomerate and the upper Starshot Formation (central TAM) represent late to postorogenic siliciclastic sedimentation during the late Ross orogeny (Figs 23, 24).  and a secondary to accessory heterozoan assemblage. • In the platform-interior ?restricted setting, sandy mixed carbonate peritidal deposits were represented by non-skeletal grains such as mud peloids, bahamite peloids, superficial type 1 ooids and very well-rounded and sorted quartz grains. In the oolitic shoal complex, type 2 and type 3 ooids (giant ooids) were produced in high-turbulence conditions, while aggregate grains predominated in low-energy, depressed and protected backshoal settings.  (Shackleton Limestone, TAM) and Thalamocyathus trachealis (Schneider Hills limestone, Argentina Range). • The affinity between Shackleton Range fauna and assemblages from Permo-Carboniferous tillites is greater. Shackleton Range fauna shares two species and three genera with the EW (Antarctica); two species and one genus with the Falkland Islands (South America); two species and two genera with the Dwyka Group, main Karoo Basin (South Africa); one species and two genera with the Dwyka Group, Aranos Basin (Namibia) and one genus with the Sierras Australes (Argentina). • The affinity between the Shackleton Range fauna and assemblages from Antarctic Cenozoic deposits is greater, sharing four species and three genera with the King George Island; two species and five genera with the Wichaway Nunataks; and two species and two genera with the Weddel Sea faunas. • The reconstructed deposits of the hidden/lost Shackleton Range platform are similar to others produced in coeval Cambrian platforms along the paleo-Pacific margin (TAM), but show clear differences (component richness/diversity, diagenetic and tectonically-induced processes) due to its latitudinal situation and tectonosedimentary evolution. • The lost carbonate platform deposits underwent early marine phreatic and meteoric vadose diagenetic processes before the breakup and synsedimentary brecciation of the platform. Mimetic penecontemporaneous dolomitisation and early authigenic and diagenetic silicification are recorded only in platform-interior ?restricted and oolitic shoal complex deposits. The breakup and downslope transport of the carbonate clast are recorded as breccia fabrics, irregular vugs with crystal silt and angular quartz grain sedimentary infillings, and geopetal infills oriented inconsistently with the growth polarity of calcimicrobe microframeworks. • At least the first 700 m of synorogenic upper slope to basinal deposits of the Mount Wegener Formation were deposited at the same time or shortly after the platform breakup around ∼515.5-514.3 Ma as indicated by the presence of the terminal Stage 3 archaeocyath-and Tabulaconus-bearing platform-derived clasts. The presence of Cambrian Oldhamia ichnospecies could suggest that sedimentation continued up to Wuliuan and even earliest Guzhangian. • The Mount Wegener Formation underwent various burial processes such as dolomitisation (D2-D4) and hydrocarbon migration prior to late-fracturerelated cementation (non-ferroan to slightly ferroan poikilotopic calcite, ferroan saddle dolomite, ferroan calcite, megaquartz mosaics), low-grade tectonically induced plastic fabrics (distorted archaeocyaths, flattened ooids, different types of twin lamellae in late-fracture-related cements) and, finally, cataclastic fabrics that post-dated all observed cementation and replacement phases. • In Furongian times, very low-grade metamorphic conditions and tectonically induced features developed with the thrusting of the Mount Wegener Nappe over its foreland (EAC) around 490 Ma during the Ross (Pan-African) orogeny (Buggisch et al. 1994b).

Acknowledgements
We want to thank the work done by Carlos Alonso Recio with the photography of archaeocyaths and montage of the plates (Photography Lab, Paleontology Area, Universidad Complutense de Madrid, UCM). We thank Fernando Cebrián for the graphic design of Shackletoncyathus buggischi. Assistance with the references provided by Pedro Martín Duque was greatly appreciated (Library, Faculty of Geology, UCM). We would like to extend our thanks to Aitor Antón, Juan Carlos Salamanca and Beatriz Moral for their work with the samples (Thin Section Lab, Stratigraphy Area, UCM). This paper is a contribution to the Spanish project CGL2017-87631-P. We especially want to thank all the reviewers for their thorough and accurate comments and recommendations. We would also like to acknowledge the corrections and comments made by Alistair John MacGowan and Iván Moreno González. Thanks to your help this work has improved substantially.