Milani&de Wit Karoo-parana

doi:10.1144/SP294.17 2008; v. 294; p. 319-342 Geological Society, London, Special Publications

E. J. Milani and M. J. De Wit

infills flanking the Gondwanides: du Toit revisitedsequences of South America and southern Africa and their basin Correlations between the classic Paraná and Cape�Karoo

Geological Society, London, Special Publications

serviceEmail alerting

article to receive free email alerts when new articles cite thisclick here

requestPermission to seek permission to re-use all or part of this article click here

SubscribePublications or the Lyell Collection

to subscribe to Geological Society, London, Specialclick here

Notes

Downloaded by Petrobras on 1 February 2008

London © 2008 Geological Society of

http://sp.lyellcollection.org/cgi/alerts

http://sp.lyellcollection.org/cgi/reprintpermission?citation=Milani+and+De+Wit+294+%281%29%3A+319

http://sp.lyellcollection.org/subscriptions

Correlations between the classic Parana and Cape–Karoo

sequences of South America and southern Africa

and their basin infills flanking the Gondwanides:

du Toit revisited

E. J. MILANI1 & M. J. DE WIT2

1Petrobras Research Center, Exploration R&D, Tectonics Group,

950 Horacio Macedo Av., Cidade Universitaria, 21941.915, Rio de Janeiro, RJ, Brazil

(e-mail: [email protected])2AEON and Department of Geological Sciences, University of Cape Town,

Rondebosch 7701, South Africa.

Abstract: Early during the twentieth Century, pioneering correlations between the Palaeozoic–Mesozoic basins of South America and southern Africa were used by Alexander du Toit to supportthe initial concepts of continental drift and the proposal of a united Gondwana continent. New stra-tigraphic tools and data can now be used to further tease out similarities and differences to recon-struct the detailed histories of these, the Parana and Cape–Karoo basins. In turn this knowledgecan be used also to increase our understanding of the origin and evolution of Gondwana.Recent advances in tectonics and stratigraphy showed that both basins evolved together along acommon early Palaeozoic Gondwana margin facing the Panthalassa. Thereafter, this marginwas transformed into a series of linked foreland basins coupled to the evolution of the Gondwa-nides. In detail, the foreland successions differ considerably due to spatial and temporal differ-ences in tectonic histories along the Gondwanides. Only towards the end of the Palaeozoic didboth basins evolve and merge into a single continental-scale, and truly intracratonic, terrestrialGondwana basin that persisted until the early Cretaceous. This shared history was once again dis-rupted in the Early Cretaceous during the opening of the South Atlantic Ocean.

The two largest, long-lived and once contempora-neous Phanerozoic sedimentary basins of Gondwanaoccur in South America and southern Africa, and areknown as the Parana and Cape–Karoo basins,respectively. These basins now flank oppositemargins of the South Atlantic Ocean. As early as1916, Juan Keidel recognized their geological simi-larities and, in 1927, Alexander du Toit published thefirst stratigraphic account of these similarities fol-lowing his lengthy sojourn through South America,in a publication ‘A geological comparison of SouthAmerica with South Africa’, sponsored by the Car-negie Institution of Washington. Du Toit’smandate was to test Alfred Wegener’s then highlycontroversial concept of continental drift (Wegener1912) rooted in a low-resolution correlationbetween these two basins. Du Toit found the bio-and litho-stratigraphy of the South American rocksequences of the Parana Basin in Brazil and thedistant flanking mountain of the Sierra de laVentana in Argentina to be remarkably similar tothose that he had himself mapped out so carefullyfor many years in the Cape–Karoo Basin and its per-ipheral Cape Fold Belt mountains of southern Africa(Fig. 1). Many geologists have remarked on the

similarities ever since, often incrementally improv-ing on the correlations that du Toit synthesized inhis famous 1937 book ‘Our Wandering Continents’.By the 1990s modern stratigraphic tools further con-firmed many of these similarities, but also started topinpoint some substantial differences. Here weexplore some of these recent results in detail. Inorder to fully appreciate the correlations and theshared evolution of the two basins, we also brieflyhighlight stratigraphic sequences of basins adjacentto the Parana Basin that are traditionally treatedseparately, and yet clearly have a common historywith parts of the Parana Basin and the largerframework of basin evolution along Gondwana’sPanthalassan margin.

The present Parana Basin (Fig. 2), with a surfacearea of about 1.1 � 106 km2, is the remnant of a vastsedimentary basin of central-eastern South Americathat preserves a Phanerozoic stratigraphic record ofalmost 400 million years, ranging from Late Ordo-vician to Late Cretaceous times, with a maximumcumulative thickness, including Mesozoic igneousrocks, of about 7 km. Six supersequences (majorunconformity-bounded, second-order allostrati-graphic units, in the sense of the cratonic sequences

From: PANKHURST, R. J., TROUW, R. A. J., BRITO NEVES, B. B. & DE WIT, M. J. (eds) West Gondwana:Pre-Cenozoic Correlations Across the South Atlantic Region. Geological Society, London, Special Publications,294, 319–342. DOI: 10.1144/SP294.17 0305-8719/08/$15.00 # The Geological Society of London 2008.

of Sloss 1963) record successive phases of sedimentaccumulation that alternated with substantial timesof erosion in the Parana Basin during the Phanero-zoic (Milani 1997). The two lower supersequences,Ordovician–Silurian and Devonian in age, docu-ment two Early Palaeozoic transgressive–regressivemarine cycles. A substantial period of erosion (up to50 Ma long) preceded the deposition of the thirdsupersequence, which spans the range Carbonifer-ous to Lower Triassic. The three upper superse-quences are Mesozoic continental sedimentarypackages associated with abundant igneous rocks.

The lowermost sediments are represented by theOrdovician–Silurian Rio Ivaı Supersequence. Thiscomprises poorly preserved relics, up to 300 mthick, of a marine package of siliciclastic rocks(Alto Garcas Formation in Brazil, Caacupe Groupin Paraguay) deposited in a series of regionalSW–NE orientated troughs and overlain by diamic-tites (Iapo Formation) that record in South Americathe widespread Ashgill glaciation of Gondwana

(Milani et al. 1996). These basal units are overlainby early Silurian (Llandovery) shales of the VilaMaria Formation (Brazil) and the equivalentVargas Pena Formation (Paraguay), which rep-resent second-order maximum flooding conditionsfor the Ordovician–Silurian cycle.

The second supersequence in the Parana Basin isDevonian in age. It is represented by the continentalto shallow-marine Furnas Formation, an extensiveblanket of white, kaolinitic sandstones up to250 m thick, in turn overlain from Pragian timesonwards by neritic shales of the Ponta Grossa For-mation. In the central Parana Basin this Devonianpackage is up to 850 m thick, but this increaseswestward to reach a few kilometres in thickness inDevonian depocentres in Argentina and Bolivia(Gohrbandt 1993).

The third supersequence of the Parana Basinrepresents the classic Carboniferous–PermianGondwana section. The sequence is up to 2.5 kmin thickness, including a maximum of 1.5 km of

Fig. 1. Present-day knowledge of the West Gondwana pre-break up configuration, showing Pan-African–Brasilianostructures and provinces (modified after Powell 1993). The ‘Limit of Mesosaurus’ (found in the Permian Irati andWhitehill beds) and the ‘Cape foldings’ (Permian–Triassic Sierra de la Ventana and Cape fold belts) supported duToit’s regional correlations between the Cape–Karoo and Parana basins and were his main criteria for a ‘SuggestedContinental Restoration under the Displacement Hypothesis’, first published in the pioneering ‘A geologicalcomparison of South America with South Africa’ (1927).

E. J. MILANI & M. J. DE WIT320

glacial diamictites, sandstones and shales of theItarare Group. The latter define three major cyclesof de-glaciation and sedimentation patterns thatrecord a retreat of the ice cap towards the southwhile Gondwana was moving to the north (Franca &Potter 1988). By Artinskian–Kungurian times theclimate had changed to allow deposition of

extensive coal beds (Rio Bonito Formation),followed by a significant thickness of Kazaniansiltstones and shales: the Palermo Formation,which represents the maximum palaeobathymetricconditions of the entire Carboniferous–Permianpackage. Overlying this is a package of blackshales with very high organic carbon content,

Fig. 2. Palaeo-geological reconstruction of Carboniferous–Permian basins of SW Gondwana, drawn over thepresent-day geography of Africa and South America (modified after Veevers et al. 1994; Lopez-Gamundı et al. 1994;de Wit et al. 1988). Basins: 1, Parana; 2, Karoo; 3, Sauce Grande; 4, Precordillera–Paganzo; 5, Tarija. Bold dashedline is the northern limit of the Gondwanides, as represented by de Wit et al. (1988). A–B, C–D and E–F are thegeological cross-sections shown in Figure 3. Bold dots mark the section of Figure 6. Z–Z

0is the position of the

cross-section presented in Figure 9.

PARANA AND CAPE–KAROO BASIN CORRELATIONS 321

cprado

Highlight

which includes the Mesosaurus fauna-bearing IratiFormation, deposited during increasingly arid con-ditions (Araujo et al. 2001). This manifests thebeginning of the drying-up of the ‘Parana sea’, acycle that ended in the continental red beds of theRio do Rasto Formation, and was followed by ter-restrial sedimentation until the Cretaceous.

Towards the south, about 1000 km apart fromthe main present-day Parana Basin in Brazil, isthe Sauce Grande Basin in Argentina (Fig. 2),adjacent to the Sierra de la Ventana fold belt(Lopez-Gamundı & Rossello 1998). Representinga stratigraphical record ranging from Ordovicianto Permian (including pre- and syn-orogeniccycles) and exhibiting an outstanding expressionof the Late Permian southern Gondwana orogen-esis, both the basin and the fold belt are keyelements for inter-regional correlation and forthe understanding of this sector of SW Gondwanapalaeomargin, in spite of the uncertain geody-namic relationship between the Parana andthe Sauce Grande basins due to the lack ofpresent-day geological continuity between the twoareas (Fig. 3).

The Cape–Karoo Basin (Fig. 2) stretches acrossmuch of South Africa, and northwards intermit-tently across into Namibia and Zimbabwe, with anoriginal area in excess of 1 � 106 km2. The basinwas once almost certainly more extensive, withlocal remnants preserved as far as central Africa

and Madagascar. In this regional context the rocksspan a history of almost 400 Ma. The main depo-centres are, however, confined to South Africawhere the total thicknesses can reach over 10 km,and the preserved stratigraphic record spans justover 300 Ma. The stratigraphic record of the basinis classically divided into two supergroups, tra-ditionally treated as having formed in two separatesuccessor basins, in turn subdivided into a numberof groups. These supergroups are, in fact, first-orderallostratigraphic units that include a series of super-sequences, in a stratigraphical framework and hier-archy similar of that described for the Parana Basin.

The lowermost of these is the Cape Supergroup,which ranges in age from late Mid Cambrian(c. 500 Ma) to Late Devonian (c. 360 Ma) and com-prises a number of second-order units with welldefined marine transgression–regression sequences(Broquet 1992), devoid of any volcanic material.The overlying supergroup is known as the KarooSupergroup. This starts with an extensive section ofglacial sediments, with up to seven major iceadvance–retreat episodes that represent c. 50 Maof the predominantly Carboniferous–Early Permian‘Dwyka’ Gondwana glaciation (Opdyke et al.2001). In the southern part of the basin, thelowermost diamictite contact is gradational withthe uppermost units of the underlying Cape Super-group. The Dwyka is relatively abruptly terminatedby a thin and marine transgressive unit of black

Fig. 3. Schematic, geological cross-sections showing the tectonic and stratigraphic configuration of three sites alongSW Gondwana margin. Note distribution of major supersequences: in sections A–B and C–D Ordovician–Silurianand Devonian packages thicken towards the palaeo-margin of the continent, whereas the Carboniferous–LowerTriassic section is clearly controlled by the evolving orogen. Source of data: A–B modified after Duane & Brown(1992); C–D and E–F modified after Franca et al. (1995).


cprado

Highlight

cprado

Highlight

cprado

Highlight

shales and sporadic carbonates (Prince AlbertFormation), followed by organic carbon anduranium-rich pyritiferous lacustrine deposits (White-hill Formation), in turn transitionally overlain by athick package of upward-coarsening turbidite depos-its (Wickens 1992). The turbidites grade into acomplex diachronous delta–shoreline section thatrecords a change in depositional style between thelower and upper Karoo Basin (Rubidge et al.2000). The uppermost Karoo rocks represent amajor terrestrial cycle of sedimentation formedunder increasingly arid conditions (Decker & deWit 2005), culminating with extensive dune-sandsthat in turn are overlain abruptly by the Early JurassicDrakensberg lavas of the Karoo continental floodbasalts (c. 182 Ma, Duncan et al. 1997), best pre-served in Lesotho. Farther north, in Namibia andsouthern Angola, aeolian sedimentation continuedup until the early Cretaceous before being coveredby the c. 134 Ma Etendeka basalts and rhyolites, asmall African remnant outlier of the great Paranacontinental flood basalt province that virtually termi-nated sedimentation in the Parana Basin of SouthAmerica.

Thus in both Gondwana basins sedimentationwas terminated by the same large igneous eventthat heralded the initiation and progressiveopening of the South Atlantic Ocean, whose proto-continental margins south of the Walvis Ridgebecame the foci of extensive seaward-dippingbasalt sequences (Stern & de Wit 2004). Duringthis transition phase the former intracontinental vol-canic and sediment depocentres of the contempora-neous Parana and Karoo basins shifted to theirrespective continental margins flanking the SouthAtlantic (Turner et al. 1994), ending their long-lived shared Palaeozoic–early Mesozoic history.

Classically, the Parana and Karoo basins wereconsidered typical intracratonic basins in whichall stratigraphic features represent the interplaybetween tectonics (subsidence-uplift) and sedimentaccumulation rates (e.g., Rust 1975; Tankard et al.1982; Zalan et al. 1990; Cole 1992) to produce anapparent framework of sequences similar to thosedescribed by Sloss (1963) for the North Americanintracratonic basins. Alternatively, sequence strati-graphies of the Gondwana basins have been ana-lysed using the Vail et al. (1977) model ofeustatic sea-level changes (e.g., Broquet 1992;Pereira et al. 2005). None of these models canfully account for the observed stratigraphy in theParana and the Cape–Karoo basins (e.g., Cloetinghet al. 1992; Milani 1997; Catuneanu et al. 1998;Milani & Ramos 1998).

The evolution of the two Gondwana basinsunder scrutiny can be summarized in its simplestform as follows. First, throughout the Palaeozoic,outward growth of continental lithosphere within

the Gondwanides along the southern margin ofGondwana was an important process throughwhich the Parana and Cape–Karoo basins becameprogressively isolated from their once conterminous‘south’-facing continental-margin position, open tothe Panthalassa Ocean throughout much of pre-Carboniferous Palaeozoic times (Zalan et al.1990; Milani 1992; de Wit & Ransome 1992;Catuneanu et al. 1998). This evolving tectonicframework resulted in progressive hinterlandmigration of the depocentres, to finally becometrapped as an intracratonic basin within the heart-land of Gondwana by Late Permian times.

In this broad geodynamic framework, weemphasize correlations in order to highlight simi-larities and differences between subsidence andsedimentation histories of the Parana and Cape–Karoo basins on a Gondwana scale. Subsidenceanalysis of both basins reveals the existence ofnotably synchronous episodes of accelerated subsi-dence in both the foreland and intracratonicsettings, suggesting that these areas have had acommon evolutionary history sharing not onlyregional sedimentary environments but also mech-anisms of subsidence. Following the opening ofthe South Atlantic Ocean, terrestrial sedimentationcontinued to dominate the interior of both conti-nents. Whilst the most continuous subsequent strati-graphic record is preserved in the dominantlymarine sections around the margins of bothcontinents, here we restrict ourselves to theonshore stratigraphic record of both basins upuntil about the time of first rifting in the SouthAtlantic. The later is best signalled onshore by theSerra Geral–Etendeka Large Igneous Province.

Regional tectonic framework

The tectonic evolution of the Parana and Cape–Karoo basins, and their respective similarities anddifferences, are influenced on a first order basis bytwo major lithospheric domains (Fig. 4).

The Gondwana shield

This corresponds to the core of the palaeo-continent, being a complex collage of Proterozoiclithospheric blocks and Archaean cratons weldedtogether along Neoproterozoic–Lower Palaeozoicorogens that are commonly referred to asPan-African and Brasiliano in Africa and SouthAmerica, respectively (see Introduction to thisvolume).

The central portion of the Parana Basin containsthe greatest thicknesses of almost all of thePhanerozoic supersequences, with a maximumcumulative thickness of about 7 km. Both the


cprado

Highlight

cprado

Highlight

cprado

Highlight

Ordovician–Silurian and Devonian packagesthicken towards the west, but the Carboniferous–Permian sections attain maximum thickness inthe central-northern region of the basin and thinto the south. About 80 deep exploration boreholeshave delineated an underlying shield with cratonicnuclei, traversed by Brasiliano belts (Cordaniet al. 1984; Zalan et al. 1990; Basei et al. 2000).Milani (1997) and Milani & Ramos (1998) inter-preted new gravity and magnetic data togetherwith borehole information on basement rocks andtheir ages and emphasized the role of reactivationof the Brasiliano belts during the Palaeozoictectonic and sedimentation history of the ParanaBasin. Across the central Parana Basin, geo-physical data reveal a persistent SW–NE trendof basement anomalies (Marques et al. 1993 –unpublished Petrobras internal report cited by

Milani 1997; Milani & Ramos 1998). Seismic andwell data confirm this to be a central rift filledwith Ordovician–Silurian sedimentary rocks(Milani 2004). A deep borehole in this rift pene-trated a lower magmatic unit (the Tres Lagoasbasalt, Mizusaki 1989 – unpublished Petrobrasinternal report cited by Milani 1997) and associatedvolcaniclastic rocks intercalated with the sediments.The occurrence in this basin of Early Palaeozoicigneous material (443 + 10 Ma; Ar–Ar, Milani2004) suggests a phase of significant tectonic exten-sion during the initial subsidence history ofthis basin.

In southern Africa, the Cape Basin package isthickest (c. 6 km) just north of the tectonic frontof the Cape Fold Belt; and from there, the sequencesof the Cape Basin thicken southwards farther still toa maximum of about 8 km (Broquet 1992). Towards

Fig. 4. Regional tectonic setting of the southern margin of Gondwana during the Phanerozoic. A long-lived zone ofconvergence and recurrent collisional tectonics developed (the Gondwanides) due to the interaction between thepalaeocontinent and the oceanic floor of Panthalassa (arrows). Compiled from Powell (1993) and de Wit et al. (1988),after the concept of du Toit (1927). Inset shows a summary of the classical Palaeozoic orogenic periodsrecognized in the South American segment of SW Gondwana (after Ramos 1988). Numbers 1 to 3 correspond tosome areas inside the Gonwanides used as references in the regional subsidence and/or stratigraphic analysis,respectively: Bolivian Tarija basin, Precordillera–Paganzo and Sauce Grande–Sierras Australes of Argentina.


the north, the Cape sequences thin rapidly, and areoverlapped by the Karoo sequences (Fig. 3). TheDwyka glacial deposits also thin towards thenorth; this has been attributed to the originaltopography and basin shape at the time of theirdeposition (du Toit 1937). About half of the base-ment underlying the main Cape–Karoo Basin iscomprised by the Archaean Kaapvaal Craton inthe north, the rest of the basin to the southbeing underlain by the Mesoproterozoic Natal–Namaqua metamorphic complex (Pitts et al. 1992;Eglinton & Armstrong 2001). The crustal and litho-spheric thicknesses and the chemistry of these twobasement terranes are distinctly different (Fouchet al. 2004), but subsidence rates and sedimentthickness of the overlying basin do not vary signifi-cantly across this basement boundary (Cloetinghet al. 1992; Catuneanu et al. 1998). Recent magne-totelluric and seismic reflection sections in thesouthern half of the Karoo Basin confirm that theinternal stratigraphy is undisturbed across thisboundary (Branch et al. 2007; Weckmann et al.2007). The Mesoproterozoic basement below theregion overlapping the tectonic front of the CapeFold Belt contains the largest continental magneticanomaly in the world (the Beattie anomaly, e.g.,Pitts et al. 1992; Harvey et al. 2001; Lindequeet al. 2007; Stankiewicz et al. 2007; Weckmanet al. 2007), and a deeply buried Mesoproterozoicpalaeosuture has been invoked for this east–westtrending anomaly. A possible extension of theanomaly has recently been recognised just northof the Sierra de la Ventana in Argentina (Villaret al. 2005).

To the north and west, the main Karoo Basin isseparated from the areas in Namibia with compar-able stratigraphic units below the Etendeka floodbasalts by a major Neoproterozoic province, theDamara Belt. Along the west coast of southernAfrica, to the west of the north–south branch ofthe Cape Fold Belt, the Cape–Karoo Basin overliesparts of the southern arm of the Damara Belt in theform of the Neoproterozoic Gariep–MalmesburyBelt and its flanking remobilised Mesoproterozoicbasement (Frimmel & Frank 1998). To the southand east of the Cape syntaxis, Mesoproterozoicbasement probably extends under the entire east–west trending section of the Cape Fold Belt. Pre-viously it had been assumed that the basementhere was Neoproterozoic, but many of these rocksequences are now known to be isolated sedimentswith mafic igneous rocks that were deposited inlocal rifts below the lowermost thermal subsidencesequences (e.g., the Table Mountain Group) of theCape Supergroup (Barnett et al. 1997), indicatingthat, as in the early Parana Basin, here too signifi-cant tectonic extension occurred during the initialsubsidence of the basin.

The orogenic ‘Gondwanides’

This name is used to describe an extensive belt ofcontemporaneous Phanerozoic orogens and relatedbasins (Fig. 4) flanking the southern border ofGondwana (Keidel 1916; du Toit 1927, 1937) alsoreferred to as ‘Samfrau geosyncline’ by du Toit(1937) and clearly displayed on the GeologicalMap of Gondwana by de Wit et al. (1988).

It is now known that the basement to the Gond-wanides comprises a mosaic of smaller crustaldomains (or blocks) with distinct geologicalhistory and geophysical–geochemical character-istics (e.g., Vaughan et al. 2005). It is not the objec-tive of this paper to describe these variations indetail. Suffice it to say that the lithosphere of theGondwanides is chemically and physically distinctfrom that of the interior of the Gondwana shield.

A second order difference between the outboardedge of the Parana and the Cape–Karoo basins isrelated to regional, along-strike contrasts in Palaeo-zoic tectonothermal evolution of the Gondwanides.Some of this has been attributed to an inherited‘orocline’ along the continental margin of Gond-wana. This oroclinal bend may have ‘shielded’southern Africa from direct convergent tectonism,in contrast to the margin of the South Americansector of Gondwana (e.g., Johnston 2000). Thisdifference in tectonothermal history along theGondwanides reveals itself in the different earlyforeland subsidence histories of the respective fore-land basins, especially their Palaeozoic strati-graphic infills, as described below.

Throughout most of the Palaeozoic the southernmargin of Gondwana, particularly the sector thatnow corresponds to the Andean border of SouthAmerica was the focus of active convergencebetween the Gondwana shield and the oceanic litho-sphere of Panthalassa (Bahlburg & Breitkreuz1991; Gohrbandt 1993; Vaughan et al. 2005). Inmore detail, a succession of orogenic cyclesmarks the Palaeozoic history of the southwesternedge of the Gondwana shield in South America(Fig. 4). These include three major tectono-sedimentary/magmatic cycles: the Pampean (earlyCambrian), the Famatinian (Ordovician to Devo-nian) and the Gondwanic (or Gondwanian: Carbon-iferous to early Triassic) cycles (e.g., Ramos 1988,1990; Pankhurst & Rapela 1998; Rapela 2000). ThePampean cycle is related to the final assembly ofGondwana, with sediment provenance from theKalahari and Natal-Namaqua areas and with tec-tonics dominated by transpression (Rapela et al.2008). The Famatinian cycle encompasses twopulses of compressional deformation and associatedphenomena, referred to as the Ocloyic and thePrecordilleran orogenies. The Gondwanic cycleincludes the Chanic and the Sanrafaelic (Sierra de


la Ventana) orogenies. These important orogenicepisodes have been related to accretion of a seriesof exotic terranes to the margin of Gondwanaduring the Palaeozoic (Ramos 1988, 1990, 2005;Pankhurst et al. 2006); they are well documentedin South America but not all found in southernAfrica. The mid-Ordovician orogenic peak of theOcloyic orogeny has been ascribed to a more exten-sive Laurentia–Gondwana continent–continentcollision (Dalla Salda 2005), but again there is noevidence for this in southern Africa.

Farther outboard from the present South Africancoast, accretion tectonics along the Gondwanideswas probably not as important, although it is lesseasy to verify, because this area has been severelydisrupted during subsequent Gondwana break-upinto a number of microplates that are now dispersedbetween southern Africa and Antarctica (e.g.,Vaughan & Livermore 2005). The precise recon-struction of these relatively small microplates hasnot yet reached a consensus, and most of the perti-nent Palaeozoic tectonic history is preserved only inthe poorly exposed stratigraphy of the AntarcticPeninsula. Nevertheless, within the simple recon-struction framework that we use for our analyseshere, correlations can be established with reason-able confidence. In southern Africa only the equiv-alent of the Sanrafaelic orogeny is well developedas the Cape orogeny that was active during the time-span from Late Permian to Early Triassic (Halbich1992). This tectonism also affected the interior ofthe Gondwana shield by reactivating old structuresin the basement below the Parana Basin andsurrounding areas (Zalan et al. 1990; Daly et al.1991; Milani 1997; Trouw & de Wit 1999).

One first order similarity of the Palaeozoic sedi-mentary geology along southwestern Gondwana(Fig. 5) is the vast volume of mature ortho-quartzitic siliciclastic rocks (de Wit & Ransome1992; Franca et al. 1995). It is said that uponviewing these siliciclastic rocks in South Africa,the legendary F. J. Pettijohn commented: ‘this isthe biggest pile of sand I have ever seen’ (ArthurFuller, pers. comm. 1990). A significant exceptionto this is found in the Argentine Precordillera.There, a thick succession of carbonates, bearingtypical species of the Cambrian Olenellus fauna(Borrello 1965) is exposed in association withMesoproterozoic basement (c. 1 Ga, Grenvillianage), separated from adjacent regions by major tec-tonic sutures (Ramos et al. 1986; Astini et al. 1996).This has led to the interpretation that the Precordil-lera was an exotic terrane that originated in Lauren-tia, rifted and drifted away from it, and finallydocked against Gondwana (Ramos et al. 1986;Astini et al. 1995; Astini 1996; Astini et al. 2005;Dalla Salda 2005; Thomas & Astini 2005), andthat the collision during Middle to Late Ordovician

times created the structures of the Ocloyic orogenyin Argentina and Bolivia. Nothing similar to thiscarbonate sequence occurs inside cratonic Gond-wana; this is definitively a phenomenon related tothe history of the palaeo-border of the continent.

A second, major cycle of deformation is recog-nized in the Precordillera region as the Precordil-leran orogeny (Furque 1965; Astini 1996). Thisepisode induced important deepening in the fore-land basin and accumulation of up to 2200 m ofEarly to Middle Devonian turbidites known as thePunta Negra Formation. Astini (1996) ascribesthis Precordilleran orogeny to the collision of asialic block known as Chilenia (Ramos et al.1984). At this time a considerable amount ofthermal subsidence is recorded in the lower clasticsequence of the Cape–Karoo Basin, including thefirst marine sediments of the Cape Supergroup(the Bokkeveld Group, Broquet 1992). Thisheralds the first significant marine incursion alongthe South African sector of the Gondwana continen-tal margin and a possible response to far field plateboundary stresses in South Africa at that time (e.g.,Cloetingh et al. 1992). Throughout the Devonian,the source for these sediments was in the north.Thus, although these deposits of the Cape Super-group represent the lateral equivalent of thedeeper water Punta Negra Formation forelandbasin turbidites in South America, their sourceregions lay in opposite directions. In the ParanaBasin, the remarkably rapid deepening inpalaeo-environmental conditions experiencedby the Devonian sea during Pragian times canbe attributed to craton-ward propagation of aGondwanide-related flexural down-warp (Fig. 6;Milani & Ramos 1998).

The precise timing of collision of Chileniaagainst the complex and segmented margin ofGondwana is still debated. Previously it waslinked to the Late Devonian–Early CarboniferousChanic orogeny (Ramos et al. 1984; Ramos1988). Adjacent to the Chanic deformation zone,Early Carboniferous tectonism produced transten-sional subsidence along older sutures in the forelanddomain (Fernandez-Seveso & Tankard 1995) toaccommodate the lower section in the Paganzo,Rıo Blanco and Calingasta–Uspallata basins ofWestern Argentina. It is not known how far theChanic-induced subsidence reached inside SWGondwana. No corresponding Lower Carboniferoussedimentary section could have been recorded inthe Parana Basin due to the lack of depositionalspace; at that time important ice caps were locatedright over the basin (Franca & Potter 1988).

The Carboniferous–Permian subsidence cycleterminated during the Sanrafaelic orogeny, and thesubsequent unconformably overlying uppersequences represent the climax of are-volcanism


Fig. 5. Summary of sedimentological and stratigraphic aspects of basins discussed in this paper (for location see Fig. 2;see Legend for lithologies and facies). Main source of data for Bolivia, Sempere (1995); Precordillera–Paganzo,Ramos (1990), Fernandez-Seveso & Tankard (1995), and Kokogian et al. (1993); Sauce Grande, Lopez-Gamundı et al.(1994, 1995) and Iniguez et al. (1989); Cape–Karoo: Veevers et al. (1994) and Cole (1992), and Parana, Milani(1997). Time-scale after Gradstein et al. (2004).


Fig. 6. Regional correlation of gamma ray borehole data showing the distribution of Devonian strata in southern SouthAmerica (Milani 1997). Ages for the Argentinian units are from Barrett & Isaacson (1988), and for the ParanaBasin from Melo (1988). (I) During Lockovian to Pragian times, a stable substratum led to the development of a wideblanket-like, continental to shallow marine sandy platform (Santa Rosa–Furnas formations). Pragian timesmarked the beginning of an accelerated cycle of subsidence (II), leading to a major Devonian flooding. The subsidenceplots show an important break both in the foreland and in the cratonic domains during Pragian times, and thisevent of accelerated subsidence is likely to have caused Early Devonian simultaneous drowning of the entire area.(III) The pattern of higher rates of subsidence continued up to Frasnian times. For location see Figure 2.


and continental volcaniclastic sedimentation (theChoyoi) between 275–250 Ma (Lopez-Gamundıet al. 1994). Pankhurst et al. (2006) ascribe muchof the intense silicic magmatism, in the south atleast, to slab tear-off and lower crustal meltingsubsequent to Mid Carboniferous collision of asouthern Patagonian continental block. During thissame time period in South Africa and Namibia,many thin rhyolitic–andesitic air-fall tuffs are rep-resented as bentonite layers throughout the KarooBasin (e.g., Cole 1992). U–Pb dating on zirconsfrom tuffs in the Collingham Formation showsthem to be especially common at around 270–280 Ma, but their total age range is 250–300 Ma(Bangert et al. 1999). Several workers have linkedthese Karoo tuffs to volcanic activity along theactive convergent margin of the Gondwanides inArgentina (e.g., Cole 1992). In South Africa, theseash-fall deposits predate and/or overlap with thefirst episodes of deformation phases of the CapeFold Belt, and record the onset of the forelandbasin deposition in southern Africa, with sedimentsdominantly sourced for the first time from the southand east (Cole 1992; Cloetingh et al. 1992;Catuneaunu et al. 2002). Similar bentonitic layers(air-fall tuffs) in the Permian sequences of theParana Basin have also been dated to rangebetween 278 and 299 Ma (Guerra-Sommer et al.2005; Santos et al. 2006). The bentonites offer arobust basis for potential detailed chronostrati-graphic correlation between the Palaeozoicsequences of the two basins.

Stratigraphical record and subsidence

analysis

The early foreland domains: Palaeozoic

depocentres along the Gondwanides

Three selected areas (Fig. 4) highlight variousaspects of the early Palaeozoic foreland basin evol-ution marginal to the cratonic Gondwana: the TarijaBasin of Bolivia, the Precordillera–Paganzo basinand the Sauce Grande Basin flanking the SierrasAustrales (Sierra de la Ventana) Fold Belt of Argen-tina. We compare these to the Palaeozoic sectionsof the Cape–Karoo Basin and the Cape Fold Beltin southern Africa, where true foreland basin depo-sition did not occur until the late Palaeozoic(Fig. 5).

The Palaeozoic successions in Bolivia are thickand widespread. They comprise five major superse-quences known as the Tacsara, Chuquisaca, Villa-montes, Cuevo and Serere allostratigraphic unitsthat include dominantly latest Cambrian to theEarly Triassic terrigenous clastic sediments(Sempere 1995). Pennsylvanian to Lower Permian

Copacabana limestones (Dıaz-Martınez 1995) inthe Cuevo Supersequence, are among the scarceoccurrences of carbonate rocks within the Gondwa-nides of South America. Another is found in thePrecordillera region of Argentina, where a surpris-ingly thick preserved Cambro-Ordovician platformoccurs. As discussed above, these carbonates rep-resent deposition on an exotic fragment of Lauren-tia that was welded to Gondwana during Middle toLate Ordovician times. In the Precordillera, theseexotic carbonates are overlain by a Silurian toDevonian siliciclastic package bearing faunas andfloras typically endemic to Gondwana. Inboard ofthe Precordillera, Cambrian–Ordovician andSilurian–Devonian packages constitute a set ofautochthonous siliciclastic sequences derived fromerosion of the Gondwana shield to the north.

The Sauce Grande Basin adjacent to the Sierrade la Ventana (or Sierras Australes) Fold Belt ofArgentina and the Cape–Karoo Basin in the CapeFold Belt of southern Africa share a close Gondwa-nide geological history (Keidel 1916; du Toit 1927,1937; Lopez-Gamundı & Rossello 1998; Rapelaet al. 2003). Both regions are underlain by a thickpackage of shield-derived and relatively matureclastic sediments, including thick quartzite sectionsof Early Palaeozoic age (uncertain mid-Cambrian toDevonian, Andreis et al. 1989; Johnson 1991;Broquet 1992; Armstrong et al. 1998). An angularunconformity separates these deposits from Meso-to Neo-Proterozoic basement intruded by CambrianA-type granites. U–Pb dating and geochemistry hasshown that the granites and associated rhyoliticextrusive rocks below the famous Cape unconfor-mity, which separates them from the overlyingsiliciclastic rocks of the Cambrian–OrdovicianTable Mountain Group, overlap in age and chemicalcomposition with similar rock types below theunconformity of the thick clastic sediments (Cura-malal Group) in the Sierra de la Ventana (Rapelaet al. 2003). This is the most dramatic follow-upcorrelation work yet vindicating the earliest lithos-tratigraphic correlations between Africa and SouthAmerica made almost a century ago by Keidel(1916). The absence of body fossils in bothsequences on these opposite sides of the Atlanticis conspicuous; yet the same trace fossils (Cruzi-ana) have been described in both sequences(Broquet 1992; Rapela et al. 2003). In both thewestern Cape and in the Sierra de la Ventana,basal conglomerates with locally derived felsicigneous clasts occur sporadically (Barnett et al.1997; Rapela et al. 2003). Throughout the CapeBasin, the thick siliciclastic rocks of the lowerTable Mountain Group are abruptly overlain, andlocally scoured into, by a thin sequence of diamic-tites (the Pakhuis Formation) that represent theshort-lived latest Ordovician glaciation. The


related palaeo-pole and extensive polar ice cap werecentred near what is today southern Chad in centralAfrica. There is no record of related glacial depositsin the Sierra de la Ventana, but since this is a glob-ally recognised glacial event, and because theseglacial deposits are present in the internal basal sec-tions of the Parana Basin (see below), we mayassume its non-preservation in the Sierra de laVentana.

The subsidence history of the southwesternGondwana Palaeozoic foreland basin (Fig. 7) issummarized by using data from the referenceareas above. A subsidence plot was calculated forthe Bolivian case using the back-stripping methodof Steckler & Watts (1978). Previously publishedcurves were used for the other areas. All plots

show the existence of periods of acceleratedsubsidence and suggest a lithospheric flexuralloading mechanism for the subsidence (e.g.,Williams 1995). In all regions there appears to bea strong correlation between the time when thesebasins first experienced accelerated subsidenceand the age of onset of the classical orogenic epi-sodes recognized in the geology of the Gondwa-nides, as discussed above.

The internal domain: correlating the

Parana and Karoo Basin fills

Six supersequences, each one comprising a geologi-cal record of some tens of millions of years,

500 300400 2000

1

2

34

5

6

CuevoVillamontes

ChuquisacaTacsara

Ma

(2)

Bolivia

TectonicSubsidence

550 500 400450 350Ma

0

km 3

2

1

4

5

6

TectonicSubsidence

(1)

(3)

(4)

Precordillera

TotalSubsidence

TotalSubsidence

350 300 250Ma

2

0

1

3

4

5

6

Paganzo

(6)

(7)

TectonicSubsidence

km

Cape-Karoo400 300 200

0

2

4

6

8

10

km

(5)

(8)

TectonicSubsidence

TotalSubsidence

Ma

km

Fig. 7. Subsidence plots for some areas along the Gondwanides, as shown in Figure 2. The curve for Bolivia wascalculated according to stratigraphic data from Gohrbandt (1993); Precordillera–Paganzo—curve for the Famatiniancycle by Ramos (1993), and that for the Gondwanic cycle by Fernandez-Seveso et al. (1993). The stratigraphy of theSauce Grande basin is poorly constrained in age and does not allow the calculation of a reliable subsidence curve.Cape–Karoo curve after Cloetingh et al. (1992). Relevant features are: (1) the passive margin response of theCambrian–Early Ordovician carbonate platform of the Precordillera; (2) the accelerated period of subsidence in theBolivia curve starting at c. 450 Ma, which can be correlated with the collision and docking of the Precordillera terraneagainst Gondwana; (3) the resultant Ocloyic orogeny and foreland subsidence. A second period of acceleratedsubsidence in the Precordillera curve was during Early Devonian, with its highest rates at c. 400–390 Ma (4); a similarsignal was also detected in the Cape diagram (5), but, as yet, no cause-and-effect relationship can be established. Thehigh rates of subsidence observed in the Paganzo curve during the interval 350–320 Ma (6) were interpreted byFernandez-Seveso (1993) as related to transtensional subsidence, most likely derived from the stresses of the Chanicorogeny. The Late Permian period of accelerated subsidence in the Paganzo curve (7) and in the Karoo curve (8) isthe classic flexural response for the Sanrafaelic (Cape–Sierra de la Ventana) orogeny.


constitute the stratigraphic framework of theinterior Parana Basin (Milani 1997). Much of thetime is represented by a series of lacunae that separ-ate the supersequences (Figs 5 and 8). The Rio Ivaı(Caradoc–Llandovery), Parana (Lochkovian–Fras-nian) and Gondwana I (Pennsylvanian–Scythian)supersequences represent major Palaeozoic trans-gressive–regressive cycles, whereas Gondwana II(Anisian–Norian), Gondwana III (Upper Juras-sic–Berriasian) and Bauru (Senonian) are fully con-tinental packages, the latter being an up to 250 mthick post-volcanic section accumulated in theflexural depression created by the load of theGondwana III lava pile.

The Rio Ivaı Supersequence comprises the oldestsedimentary rocks of the Parana Basin. This Ordovi-cian–Silurian package is widespread, but its thick-ness varies considerably with some thick elongateddepocentres striking SW–NE. There is also ageneral trend of thickening to the west, with thepackage reaching about 1000 m in the Paraguayanportion of the basin. A regional, fluvial palaeocur-rent pattern from NE to SW is evident (Milaniet al. 1996) in the lowermost section of the RioIvaı package. Seismic data (Marques et al. 1993 inMilani 1997 (unpublished Petrobras internalreport) show that the thickest Rio Ivaı occurrenceis confined to a 600 km long, SW–NE orientatedrift system that runs from Paraguay to the northeast-ern portion of the basin. The sequence includes basalconglomerates and sandstones (Alto Garcas For-mation) with a section of basalts (the Tres Lagoasbasalt) that suggests significant extension andrifting during the inception of the formation of theParana Basin. The Balcarce Formation (Rapelaet al. 2003; Zimmermann & Spaletti 2005) of theTandilia System, adjacent to the Sierra de laVentana, is the most likely correlative unit of theRio Ivaı Supersequence in that area. The latest Ordo-vician glacial diamictites (Iapo Formation) are alsopreserved here and, together with overlying fossili-ferous shales and siltstones (Vila Maria Formation),span Caradoc–Llandovery times. The shales recordmaximum flooding of the Ordovician–Siluriancycle, following rapid de-glaciation, as they do inthe Cape Basin (where the Cedarberg Shalescontain ‘giant’, cold-water conodonts). The top ofthe Rio Ivaı Supersequence is cut by a peneplaincovered by the sheet-like Devonian Parana Superse-quence. The latter represents a complete transgres-sive–regressive cycle of sedimentation startingwith continental to neritic Lower Devonian sandyrocks (Furnas Formation) followed by marineshaly sediments (Ponta Grossa Formation)that span the Pragian to Frasnian stages. ThePragian shales have sedimentological and strati-graphic characteristics indicating maximum Devo-nian flooding and rapid drowning of the shallow

Furnas platform. A second basin-scale unconfor-mity surface marks the upper limit of the Devonianpackage. This sub-Pennsylvanian unconformity is abenchmark surface in the Parana Basin, and rep-resents a lacuna of almost 55 Ma. This unconformityis distinctly angular in the Gondwanides affected bythe Chanic orogeny, such as in the western Argen-tine basins and the Sauce Grande foreland basin(Lopez-Gamundı & Rossello 1993).

The subsequent accumulation of the Gondwana ISupersequence of the Parana Basin occurred duringde-glaciation episodes and reflected increased sedi-mentary influx from those areas freed from icecover. Sediments were dominated by mass flowsand re-sedimentation, defining a singular deposi-tional style for the Pennsylvanian to LowerPermian interval throughout the Parana Basin. The1500 m thick de-glaciation related section (namedItarare Group in the southern portion of theParana Basin and the Aquidauana Formation inthe north) is composed of diamictites intercalatedwith sandstone and shale packages (Franca &Potter 1988), of both glacio-terrestrial (minor) andglacio-marine environments. The interpretation ofmarine affinities of the Itarare sedimentation is sup-ported by the overall presence of Tasmanites andacritarchs (Daemon & Quadros 1970). The glacialdeposits onlap the sub-Pennsylvanian unconformityfrom north to south and extend over progressivelywider areas. In the Early Permian, onlapping sedi-mentation reached the southernmost portion of thebasin, presently located in Uruguay. Basin-widecorrelations of the glacial record in the ParanaBasin (and for the rest of the sedimentary sectionas well) is almost exclusively constructed fromextensive biostratigraphy (Daemon & Quadros1970; several other references for local sections).Calibrated to European palinozones, the resultingchronostratigraphy (particularly for the Carbonifer-ous–Permian package) is imprecise and uncertain.This has recently been improved through U–Pbdating of zircons from intercalated bentonites(ash-fall tuff beds, as mentioned above, Coutinhoet al. 1991; Guerra-Sommer et al. 2005; Santoset al. 2006). Through this work, as in the KarooBasin, pre-existing stratigraphic schemes are nowknown to be substantially incorrect by up to tensof millions of years, highlighting the necessity forintensive geochronology-based stratigraphy that isonly just starting. In any case, the glacial recordin the Parana Basin has been traditionally ascribedto the range Westphalian–Artinskian, a time-spanof about 35 Ma. A Mississippian record is lackingin the Parana Basin, but is present in western depo-centres like the Paganzo Basin of Argentina, wheredepositional space was created in a time when theParana Basin domain was experiencing peakglacial conditions.


cprado

Highlight

cprado

Highlight

PA

RA

NÁ

Paraná & other South American domains

II

KM

ES

OZ

OIC

J

P

Tr

C

D

Mis

Pen

S

O

C

100

Ma

200

300

400

500

PA

LAE

OZ

OIC

Cape–Karoo & Cape Fold Belt

vvv

GOND

WAN

AGO

NDW

ANA

III

IRI

OIV

AÍ

lowerPonta Grossa

Vila Maria

Iapó

glac

iatio

n

++++++Palermo

Periglacialsed.

Pea

k

Car

boni

fero

usgl

acia

tion

Cho

iyoi

mag

mat

ism

mag

mat

ism

Localized rift basin

Lacustrine / fluvial

Ser

raG

eral

WG

ondw

ana

brea

kup

Aeo

lian

sand

yde

sert

sTe

rres

tria

lto

suba

quou

sM

arin

e

Ope

n-t

o-P

anth

alas

saph

ase

(Nat

ure

ofse

dim

enta

tion)

(Bas

inco

nfig

urat

ion)

Ocl

oyic

Pre

cord

iller

anC

hani

cS

anra

fae l

ic(

Oro

geni

es)

(G

eody

nam

icep

isod

es)

Gon

dwan

i cF

amat

inia

n

CA

PE

Cedarberg

Rift

KA

NG

O

lowerBokkeveld

Pak

huis

glac

iatio

n

PrinceAlbert

Dw

yka

glac

iatio

nC

FB

defo

rmat

ion

Mai

npu

lses

Aeo

lian

sand

yde

sert

s

Dra

kens

berg

volc

anis

mE

tend

eka

mag

mat

ism

vvv

vvv

KA

RO

O

GO

ND

WA

NA

Intr

acra

toni

cph

ase

Fig. 8. Major stratigraphic units of the Parana and Cape–Karoo basins, and correlation in time with some importanttectonic, magmatic and climatic events that took place in SW Gondwana during the Phanerozoic. Arrows indicatemajor events of flooding. Note the conspicuous trend of drying in these basins, which were marine in their origin(open towards Panthalassa) and became intracratonic, trapped inside Gondwana, by Permian times. Main source ofdata for Parana: Milani (1997) and Milani & Ramos (1998); Cape–Karoo: Cole (1992) and Broquet (1992); otherSouth American domains: Milani (1997) and references therein. Time scale after Gradstein et al. (2004).


By contrast, in southern Africa there is no signof a major hiatus before the onset of the mid-Palaeozoic glaciation (Figs 5 and 8). In the KarooBasin the glacial period is represented by sedimentsof the Dwyka Group (Visser 1997). Sedimentationin the Cape Basin terminated during the Lower Car-boniferous period (Visean) after deposition of thetidal flat to lacustrine sandstones of the WaaipoortFormation of the Witteberg Group. The age of theoverlying Dwyka deposits has been a matter oflong debate. It was previously believed that ahiatus of approximately 30 Ma duration may haveoccurred before deposition of the Karoo Sequencecommenced (Visser 1990, 1997). However,palaeontology has shown that the top of the Waai-poort Formation is Visean in age (Streel & Theron1999) and lacustrine (Evans 1999). Dropstonesand soft sediment deformation features believed tobe glacial in origin are present in the uppermostbeds of the Witteberg Group (Streel & Theron1999). Therefore, glaciation was in progressduring deposition of the uppermost Witteberg sedi-ments, and it appears that there is in fact a minimalhiatus before glaciers advanced across theWitteberg surface.

Glaciogenic sedimentation and erosion domi-nated until the earliest Permian (upper Asselian–lowermost Sakmarian) accumulating up to 800 mof thick diamictites and rhythmites in the southernpart of the Karoo Basin, but much less in the vicin-ity of the classical striated pavements along thenorthern fringes of the basin (Visser 1989, 1997;Cole 1992). Ice-flow directions indicate sources tothe north (Cargonian Highlands), east (EasternHighlands, now in East Antarctica), and southwest(Southern Highlands, now in West Antarctica),which may represent major ice-spreading centresat that time (Visser 1989, 1997). Visser (1997) esti-mated an ice sheet cover of the basin floor of morethan 4 km, similar to that in much of present-dayEast Antarctica. In the Carboniferous, at the timeof maximum glaciation of Gondwana, the southpole was located on the Antarctic shield (Opdykeet al. 2001) then situated less than 1000 km to theeast of the Karoo Basin (e.g., Rakotosolofo et al.1999; Scotese 2000; Reeves et al. 2002). InAfrica, the northern margins of the ice sheetsreached at least as far as southern Madagascar(Rakotosolofo et al. 1999), the Central AfricanRepublic, northern Angola and southern Sudan.

Sections of up to 800 m of sediment in the formof lodgement till, rain-out and sub-aqueous and sub-glacial melt-water sands, suspended mud, andturbidity-current sands and silts, were deposited inat least seven upward-fining cycles related toadvance and retreat of the ice cap, each startingwith thick coarse tillites/diamictites and terminat-ing in thin shaly horizons and rhythmites/varves.

Although some investigators infer a marine settingfor the deposition of part of these diamictites andrelated varves (Cole 1992; Johnson 1997; Visser1997; Catuneanu et al. 1998; Rubidge et al.2000), there is no unequivocal evidence for this. Itis possible therefore that the bulk of the Dwykasequence of the Karoo Basin was deposited in a ter-restrial setting (du Toit 1926). Marine fossils occurat the top of the Dwyka along the western marginsof the basin, particularly in southern Namibia andalong the deep glacial valley through theVryburg–Prieska region along the northernextents of the Karoo Basin. These are witnesses ofa short marine transgression related to eustatic sea-level rise following rapid global de-glaciation (duToit 1954; McLachlan & Anderson 1973; Visser1987, 1989). Elsewhere, these sediments indicatenon-marine conditions, including those of the over-lying carbonaceous mudstones of the Prince Albert,Whitehill and Collingham formations, when theKaroo Basin had become a gigantic inland lake(Faure & Cole 1999). These observations implythat the water-lain diamictites, previously modelledas deposited in a wide marine-shelf environment(Visser 1989, 1997), may represent glaciogeniclake sediments deposited beneath and peripheralto the major continental ice sheet that coveredmuch of Gondwana at that time.

Two occurrences of rhyolitic–andesitic volcanictuff are present in the Dwyka Group of southernAfrica (Bangert et al. 1999). These were mostlikely derived from a magmatic arc to the SW ofthe Parana Basin (Cole 1992; Bangert et al. 1999).U–Pb dating of zircons from tuffs near Laingsburgabout 400 m above the base of the Dwyka yielded adate of 297 + 1.8 Ma (Bangert et al. 1999). The ageof the top of the Dwyka has been similarly derivedusing zircon dating from tuffs in the lowermost bedsof the overlying Prince Albert Formation(288 + 3 Ma and 289 + 3.8 Ma). Thus, theDwyka Group in southern Africa spans about50 Ma, with the upper 200 m of the Dwyka Grouprepresenting less than 10 Ma or at most 20% ofthe glacial period. Although the mid-Palaeozoicglacial deposits of the Dwyka and the ItarareGroups have been correlated in general termssince the time of du Toit (1927), it is clear thatthere are some differences between them in termsof environmental conditions.

The Karoo Basin evolved further as a flexuralforeland basin around the Permo-Triassic boundary,linked to the emerging Cape Fold Belt (Halbich1992; Cloetingh et al. 1992; Catuneanu et al.1998, 2002). Many workers have related theUpper Palaeozoic Karoo sequences to tectonic pro-cesses associated with subduction of the palaeo-Pacific plate beneath the Gondwana plate duringthe Permo-Carboniferous (Lock 1980; Cole 1992;


Smith 1995; Johnson 1997; Catuneanu et al. 1998,2002). However, the above-mentioned models arebased on lithostratigraphic analysis complementedby only limited biostratigraphy due to a scarcityof zonal fossils with precise time resolution. Thishas hampered detailed tectonic modelling andbasin analysis, and is the cause of significant contro-versy about the age and position of lithostrati-graphic boundaries (e.g., Zawada & Cadle 1988;Rubidge et al. 2000) and sequence stratigraphy(e.g., Catuneanu et al. 1998; Turner 1999). Thereis simply not enough reliable chronostratigraphyto verify detailed correlations, but available datashow that some of the basin evolution models areextrapolations based on low resolution biostratigra-phy. For example, Catuneanu et al. (1998) place thelithostratigraphic top of the Dwyka sequence atc. 268 Ma and 265 Ma in the southern and northernparts of the Karoo Basin, respectively, 30–35 Mayounger than the lithostratigraphy based on morerecent U–Pb zircon chronostratigraphic data. Simi-larly, their dating of the Collingham Formation is atleast 12 Ma younger than the U–Pb dates derivedfrom zircons in the widespread air-fall tuffs fromthis formation, as will be mentioned below.

Considerable debate remains also whether theKaroo Basin at this time is entirely terrestrial or ifthere are substantial marine sequences preserved(e.g., Faure & Cole 1999; Visser 1992a). Rapidmelting of the ice sheets in the Early Permian wasfollowed by slow deposition of black suspendedmuds (Prince Albert Formation, lower EccaGroup; Sakmarian–Kungurian in age). Geochemis-try (Zawada & Cadle 1988; Faure & Cole 1999)indicates that most of these deposits are fresh- tobrackish-water lake deposits and not marine as isoften inferred (e.g., Cole 1992; Visser 1997;Catuneanu et al. 1998, 2002; Rubidge et al.2000). Thereafter, post-glacial isostatic rebound ofthe northern and eastern provenance areas resultedin an influx of fluvial deltaic sands and onset ofextensive coal deposition in the northeastern partof the Karoo Basin.

At the end of the Early Permian, a distinctiveblack pyrite-rich mud horizon, the Whitehill For-mation, was deposited throughout a gigantic andhighly reducing basin, which by then extendedacross into the Parana Basin of South America,where the coeval Irati Formation was deposited.Both these formations have a characteristic faunaof the fresh-water Mesosaurus and dragonflies (duToit 1927, 1937; Milani 1992; Visser 1992b;Araujo et al. 2001). This distinctive geochemicalmarker is a true ‘time line’ for correlation; and ithas recently been shown that it can be easilytraced in subsurface using relatively fast andsimple magnetotelluric sounding (Branch et al.2007). Both the Whitehill and Irati shales have a

very high total organic carbon content (up to 24%in condensed sections of the latter), a sulphurcontent as high as 8%, and elevated uraniumconcentrations. In South Africa, the WhitehillFormation is abruptly overlain by the upward-coarsening sequence of turbidites of the Collinghamand Laingsburg formations.

In turn, distal turbidites of the Collingham For-mation are overlain by coarser turbidites of theVischkuil and Laingsburg formations and then thedeltaic sequences of the Fort Brown Formation.These deposits are succeeded by a package ofshoreline and braided-river deposits of the BeaufortGroup (Upper Permian, Rubidge et al. 2000); thesecontain the terrestrial mammal-like reptile Dicyno-don, described in detail by King (1990). A tuffhorizon in the lowermost Beaufort, just below theCistecephalus–Dicynodon boundary (Rubidge,pers. comm. 2001) yields a U–Pb zircon date of258 + 1 Ma (uppermost Guadalupian). Thesedeposits were abruptly followed by meanderingriver deposits of the upper Beaufort (Triassic)across the Permo-Triassic boundary at c. 251 Ma(Smith 1995; MacLeod et al. 2000; Ward et al.2000). Thereafter, a new Mesozoic fauna, charac-terized by Lystrosaurus, assumed the landscape,and terrestrial sedimentary processes continued todominate with a progressive desertification,interrupted at times by wet–dry cycles straddlingthe Triassic–Jurassic boundary (Decker & deWit 2005). This environment was terminatedabruptly by the Drakensberg flood basalts of theKaroo Large Igneous Province, at 182 + 1 Ma(Duncan et al. 1997; Turner 1999). At the contact,basalt flows can be seen to fill a landscape ofdune deposits.

Comparative subsidence histories

In South America, an important phase of structuralrearrangement of the Parana Basin geometry alsofollowed the Carboniferous–Permian glaciation.Accompanying the Late Permian deformationalong southwestern Gondwana (the evolvingSierra de la Ventana Fold Belt, von Gosen et al.1991; Cobbold et al. 1992), subsidence and sedi-ment accumulation in the Parana Basin followed aperiod of accelerated flexural subsidence until ear-liest Triassic times, accommodating about 1400 mof mostly terrestrial sedimentary rocks of the Tere-sina and Rio do Rasto formations. Lopez-Gamundıet al. (1995) pointed out the syntectonic character ofsedimentation in the foreland basin adjacent to theSierra de la Ventana Fold Belt during the latePermian. In the Parana these are basin-centred dis-continuous aeolian fields and shallow lake depositsframed by huge sandy deserts (Piramboia For-mation in the northern portion; Sanga do Cabral


cprado

Highlight

and Buena Vista formations in the southern area;Cabacua Formation in Paraguay). A progressiveand irreversible continentalization of the deposi-tional systems of the Parana Basin occurred fromthe Late Permian to the late Triassic and can beseen from the sedimentary record of the upperportion of the Gondwana I Supersequence (Rio doRasto Formation) onwards. By then, sandy desertscovered the entire basin until the latest Jurassic(Botucatu Formation), followed by the Early Cre-taceous basaltic lavas and intrusives of the SerraGeral Formation, and its equivalents in the Eten-deka region of Namibia, whose lava sequenceshave been correlated both geochemically and isoto-pically (Marsh et al. 2001).

The presence of thick and extensive EccaPermian turbidites in the Karoo, therefore, definesan important difference between the two basins.This appears to signify a greater degree of flexuralresponse of the foreland basin to the Capeorogeny than is the case in South America duringthe Sanrafaelic orogeny. In South America, theregional sense of onlap of sedimentary beds wassuddenly inverted (Fig. 9), as shown by thenorthward-wedging retrograding sequence of theGuata Group (Milani & Ramos 1998). Maximumpalaeo-bathymetric conditions for the Gondwana ISupersequence (external neritic at most) are docu-mented in the Palermo Formation of earliest LatePermian age (Daemon & Quadros 1970). The over-lying package, accommodated by a renewed cycleof flexural subsidence of the basement, is a regres-sive section up to 1400 m thick (Passa Dois Group)that culminates in Early Triassic aeolian sandstones

(Sanga do Cabral, Piramboia and Buena Vistaformations). Subsequently, the two interior basinsof Gondwana merged into a giant sand desert.This was first interrupted in the Cape–KarooBasin only by basalt flows of the Karoo igneousprovince during the late Jurassic; then in bothbasins by outpourings of basalts during the secondlarge (Parana) igneous province, focussed predomi-nantly along the west coast of southern Africa butbest preserved in South America.

To summarise, the subsidence history of theinterior domain of the Parana and Cape–Karoobasins are compared using their respective subsi-dence curves (Fig. 10). In South America, the Ordo-vician–Silurian phase cannot be adequatelyevaluated quantitatively due to poor chronostrati-graphic resolution within the dominantly sandysection of the Rio Ivaı Supersequence. The geo-metry of this unit, defining narrow, elongateddepocentres along SW–NE weakness zones of thebasement, suggests some kind of rifting or trans-tension as the initial tectonic mechanism respon-sible for inception of the basin. This is also seenin the subsidence curves of the Cape Basin ofsouthern Africa, where extensional processes wereresponsible for beta factors ranging between 1.2and 2.2 (Cloetingh et al. 1992).

In the Parana Basin, Devonian subsidencestarted with low rates related to a flat and stable sub-stratum, in accordance with the overall character-istics of its basal section, the Furnas Formation.This sandy unit exhibits a blanket-like geometrywith remarkably constant thickness and sedimento-logical characteristics across the basin. From

Fig. 9. Gondwana I Supersequence (Carboniferous to Lower Triassic) basin-scale correlation of stratigraphic data(Milani 1997) using information of 40 deep boreholes (not shown). The stratigraphic record of this supersequence ofthe Parana Basin documents an abrupt change in the sense of onlap: accompanying the southward-retreating icecap it developed from north to south during the accumulation of the basal, glacially-influenced package of the ItarareGroup and Aquidauana Formation, whereas the overlying post-glacial transgressive package of the Guata Grouponlaps to the north, with subaerial exposure and the development of a regional unconformity in the northern portion ofthe basin. Such an important change in basin configuration is attributed to intracratonic response to plate margintectonics. The end of Palaeozoic history of the Parana Basin was marked by the advance of continental depositionalsystems (Piramboia, Sanga do Cabral and Buena Vista formations) towards the remnant central water body (Rio doRasto Formation). For location see Figure 2.


cprado

Highlight

Pragian time onwards, a pattern of increasing subsi-dence rates was established. This period of acceler-ated subsidence led to the maximum flooding by theDevonian sea, recorded by the laminated blackshales of the basal Ponta Grossa Formation. Shalysedimentation under highstand marine conditions,punctuated by deltaic prograding sandy bodies,proceeded up to Frasnian time. Similar acceleratedsubsidence is observed at c. 400 Ma in the Capesubsidence plot (Fig. 10). Both areas presumablyresponded rapidly to lithosphere flexures generatedfar away in the Gondwanides. The general responseof sandy shallow marine to coastal deposits (Furnasand Nardouw) changing to shaly neritic environ-ments with second-order maximum flooding con-ditions (basal Ponta Grossa and basal Ceres) andthen a highstand setting with Devonian deltas(middle/upper Ponta Grossa and upper Bokke-veld/Witteberg) are the sedimentological responsesthat we infer to trace a linked subsidence history ofthese basins.

The Late Permian–Early Triassic subsidenceevolution in the history of the Parana and Karoobasins is also remarkably similar (Fig. 10), inspite of very different lithostratigraphy at thosetimes: the Parana Basin was definitively trappedinside Gondwana, experiencing the final drying-upof its sea and the accumulation of extensive conti-nental deposits, surprisingly thick (1.4 km),whereas in the Karoo a c. 1-km thick pile of distalturbidites, deposited during low subsidence rates,was succeeded by several kilometres of morerapidly accumulated coarse sediments derivedfrom the emerging Cape mountains during latestPermian, in a true foreland domain. Despite thesedifferences, the subsidence plots delineate distinctaccelerated subsidence at around 250 Ma from thestratigraphic records of both basins. The ‘classic’response of foreland development in the ParanaBasin is an unusual and still poorly understoodmechanism of intracratonic flexure. Nevertheless,it appears that the general model of a LatePermian Gondwanide ‘foredeep’, which includedthe Karoo and Sauce Grande basins, and a contem-poraneous Gondwanide ‘foreland’, where theParana Basin developed, was an outstandinginsight of du Toit’s genius.

Concluding remarks

The data set at hand today allows a more confidentcorrelation between the Gondwana basins of SouthAmerica and Southern Africa than was possible indu Toit’s time. First, it is now more evident that aclose relationship existed between the developmentof the Parana Basin in the continental interior andthe Palaeozoic tectonic regime along the flankingGondwanides. By treating the Cape and Karoobasins as part of a continuum, such a relationshipalso emerges more clearly. The lithosphere ofsouthwestern Gondwana reacted by flexure underthe stresses generated along the Gondwanides, andthis provided an effective mechanism to create‘intracratonic’ depositional space well into theGondwana shield. In both South America andsouthern Africa, this mechanism of subsidenceseems to reflect the craton-ward propagation ofthe flexural bending of the lithosphere that charac-terizes the foreland domain, so that the ParanaBasin experienced phases of accelerated subsidencethat correlate well with those in the adjacent fore-land. Yet, in detail the lithostratigraphies of theforeland basin sequences do not correlate wellbetween the Cape–Karoo and Parana basinsbecause the Palaeozoic collisional tectonic historyalong the Gondwanides was different and/or dia-chronous (Fig. 11). Only by Permian–Early Trias-sic times did plate boundary processes operate in

450 400 300350 250 200

3

11

4

D CSO P

Ma

1

2

22

(1)

(2) Cape–Karoo

Paraná(2)

(1)

km

km

Fig. 10. Backstripped, tectonic subsidence of basementof Cape–Karoo (Cloetingh et al. 1992) and Paranabasins (Milani 1997). Note (1) the important Devoniancycle of accelerated subsidence; (2) the second cycle ofaccelerated subsidence during Late Permian–EarlyTriassic. The similarity of tectonic histories implicit inthese diagrams, in spite of the attenuated rates of Paranarelative to Karoo, is very significant considering theposition of each basin and the differential distance toPalaeozoic plate margins. This suggests a coupledforeland–intracratonic flexural mechanismof subsidence.


unison to yield similar coupled foreland basinsequences. Even then distinct stratigraphic succe-ssions are evident, related to differences inlocal subsidence rates, and presumably to spatialand temporal variations in degrees of tectonicloading. These differences still need to be fullyunderstood. During the early Mesozoic the respect-ive foreland basins transformed and merged into atruly trans-Gondwana terrestrial basin. But even atthis stage the Karoo Basin history differed in thatthe terrestrial desert conditions were interruptedby the widespread rapid outpourings of the short-lived (,1 Ma) Karoo LIP. However, at the veryend of their common history, towards the timewhen the two continents finally separated, theSerra Geral–Etendeka basalts and their associatedseaward-dipping volcanic sequences were eruptedall along both margins of the South Atlantic.

The evolution of both the Parana and Cape–Karoo basins as described herein is an integratedproduct of a particular local history of subsidence

(and the variable elastic thickness of their base-ment), and the variable rates of transformationinto intracratonic basins during the concomitant,but diachronous Palaeozoic tectonics within theGondwanides. Thickness variations across bothbasins are therefore controlled to a much greaterdegree by loading effects during the evolution ofthe Gondwanides than by global eustasy, and thisprovides a fundamental key towards improvingunderstanding of the geological history of central,continental Gondwana.

We both acknowledge discussions with many Gondwanacolleagues over a long time, especially at the Gondwanaconferences. E.J.M acknowledges Petrobras for providingthe means for his research into southern South Americangeology. M.d.W particularly thanks Arthur Fuller, JohnRogers, Ingo Halbich and Peter Booth for introducinghim to the lure of the Cape Fold Belt and the Cape–Karoo Basin. His research over the years was supportedthrough the National Research Foundation of SouthAfrica. Pat Eriksson and Stephen Flint provided greatinputs towards the improvement of the manuscript. Thisis AEON contribution no. 29.

References

ANDREIS, R. R., INIGUEZ, A. M., LLUCH, J. L. &RODRIGUEZ, S. 1989. Cuenca paleozoica de Ventania,Sierras Australes, Provincia de Buenos Aires. In:CHEBLI, G. A. & SPALLETTI, L. A. (eds) Cuencassedimentarias argentinas. Universidad Nacional,Tucuman, Geological Correlation Series, 6, 265–298.

ARAUJO, L. M., RODRIGUES, R. & SCHERER, C. M. S.2001. Expressao estratigrafica do carbono organico,resıduo insoluvel, enxofre, parametros de pirolise eelementos-traco nas sequencias deposicionais Irati,Bacia do Parana. In: 7 Congresso Brasileiro de Geo-quımica, Curitiba, abstracts 207.

ARMSTRONG, RA., DE WIT, M. J., REID, D., YORK, D. &ZARTMAN, R. 1998. Cape Town’s Table Mountainreveals rapid Pan-African uplift of its basement rock.Journal of African Earth Sciences, 27–1A, 10–11.

ASTINI, R. A. 1996. Las fases diastroficas del PaleozoicoMedio en la Precordillera del oeste argentino—evidencias estratigraficas. In: 13 Congreso GeologicoArgentino y 3 Congreso de Exploracion de Hidrocar-buros, Buenos Aires, Actas, 5, 509–526.

ASTINI, R. A., BENEDETTO, J. L. & VACCARI, N. E.1995. The early Paleozoic evolution of the ArgentinePrecordillera as a Laurentian rifted, drifted, and col-lided terrane—a geodynamic model. GeologicalSociety of America Bulletin, 107, 253–273.

ASTINI, R. A., RAMOS, V. A., BENEDETTO, J. L.,VACCARI, N. E. & CANAS, F. L. 1996. La Precordil-lera: un terreno exotico a Gondwana. In: 13 CongresoGeologico Argentino y 3 Congreso de Exploracion deHidrocarburos, Buenos Aires, Actas, 5, 293–324.

ASTINI, R. A., THOMAS, W. A. & MCCLELLAND, W. C.2005. Provenance of Middle Ordovician detritus in theArgentine Precordillera. In: PANKHURST, R. J. &VEIGA, G. D. (eds) Gondwana 12: Geological and

Fig. 11. Schematic tectonic setting of SW Gondwanamargin during the two main Palaeozoic geodynamiccycles that affected the region. The main area ofcollision and orogenic deformation migrated from NW(during the Famatinian cycle) to SE (during theGondwanic cycle). Parana (P) and Cape–Karoo (K)basins were affected to different degrees by the varioustectonic episodes active along the margin of thepalaeo-continent.


cprado

Highlight

Biological Heritage of Gondwana, Abstracts. Acade-mia Nacional de Ciencias, Cordoba, tArgentina, 51.

BAHLBURG, H. & BREITKREUZ, C. 1991. Paleozoic evol-ution of active margin basins in the Southern CentralAndes (Northwestern Argentina and Northern Chile).Journal of South American Earth Sciences, 4,171–188.

BANGERT, B., STOLLHOFEN, H., LORENZ, V. & ARM-

STRONG, R. 1999. The geochronology and signifi-cance of ashfall tuffs in the glaciogenicCarboniferous–Permian Dwyka Group of Namibiaand South Africa. Journal of African Earth Sciences,29, 33–50.

BARNETT, W., ARMSTRONG, R. & DE WIT, M. 1997.Stratigraphy of the upper Neoproterozoic Kango andlower Paleozoic Table Mountain Groups of the CapeFold Belt, revisited. South African Journal ofGeology, 100, 237–250.

BARRETT, S. F. & ISAACSON, P. E. 1988. Devonianpaleogeography of South America. In: MCMILLAN,N. J., EMBRY, A. F. & GLASS, D. J. (eds) Devonianof the world. Canadian Society of Petroleum Geol-ogists, Memoirs, 14, 655–667.

BASEI, M. A. S., SIGA, O., JR., MASQUELIN, H.,HARARA, O. M., REIS NETO, J. M. & PRECIOZZI,F. P. 2000. The Dom Feliciano belt of Brazil andUruguay and its foreland domain, the Rio de la Platacraton: framework, tectonic evolution and correlationwith similar provinces of southwestern Africa. In:CORDANI, U. G., MILANI, E. J., THOMAZ FILHO,A. & CAMPOS, D. A. (eds) Tectonic evolution ofSouth America. 31st International Geological Con-gress, Rio de Janeiro, 311–334.

BORRELLO, A. V. 1965. Sobre la presencia del Cambricoinferior olenellidiano en la Sierra de Zonda, Precordil-lera de San Juan. Ameghiniana, 10, 313–318.

BRANCH, T., WECKMANN, U. & RITTER, O. 2007. TheWhitehill Formation—a highly conductive marker inthe Karoo Basin. South African Journal of Geology,110, 465–476.

BROQUET, C. A. M. 1992. The sedimentary record of theCape Supergroup: a review. In: DE WIT, M. J. &RANSOME, I. D. (eds) Inversion tectonics of theCape Fold Belt, Karoo and Cretaceous basins ofSouthern Africa. A. A. Balkema, Rotterdam, 159–184.

CATUNEANU, O., HANCOX, P. J. & RUBIDGE, B. S. 1998.Reciprocal flexural behaviour and contrasting strati-graphies: a new basin development model for theKaroo retroarc foreland system, South Africa. BasinResearch, 10, 417–439.

CATUNEANU, O., HANCOX, P. J., CAIRNCROSS, B. &RUBIDGE, B. S. 2002. Foredeep submarine fans andforebulge deltas: orogenic off-loading in the under-filled Karoo Basin. Journal of South African EarthSciences, 35, 489–502.

CLOETINGH, S., LANKREIJER, A., DE WIT, M. J. &MARTINEZ, I. 1992. Subsidence history analysesand forward modelling of the Cape and KarooSupergroups. In: DE WIT, M. J. & RANSOME, I. D.(eds) Inversion tectonics of the Cape Fold Belt,Karoo and Cretaceous basins of Southern Africa.A. A. Balkema, Rotterdam, 239–248.

COBBOLD, P. R., GAPAIS, D., ROSSELLO, E. A.,MILANI, E. J. & SZATMARI, P. 1992. Permo-Triassic

intracontinental deformation in SW Gondwana. In:DE WIT, M. J. & RANSOME, I. D. (eds) Inversion tec-tonics of the Cape Fold Belt, Karoo and Cretaceousbasins of Southern Africa. A. A. Balkema, Rotterdam,23–26.

COLE, D. I. 1992, Evolution and development of theKaroo Basin. In: DE WIT, M. J. & RANSOME, I. D.(eds) Inversion tectonics of the Cape Fold Belt,Karoo and Cretaceous basins of Southern Africa.A. A. Balkema, Rotterdam, 87–100.

CORDANI, U. G., BRITO-NEVES, B. B., FUCK, R. A.,PORTO, R., THOMAZ FILHO, A. & CUNHA,F. M. B. 1984. Estudo preliminar de integracao doPre-Cambriano com os eventos tectonicos das baciassedimentares brasileiras. Petrobras, Rio de Janeiro,Ciencia-Tecnica-Petroleo Series, 15.

COUTINHO, J. M. V., HACHIRO, J., COIMBRA, A. M. &SANTOS, P. R. 1991. Ash-fall derived vitroclastic tuf-faceous sediments in the Permian of the Parana Basinand their provenance. In: ROCHA-CAMPOS, A. C.(ed.) Gondwana Seven Proceedings. University ofSao Paulo, Brazil, 147–160.

DAEMON, R. F. & QUADROS, L. P. 1970. Bioestratigrafiado Neopaleozoico da Bacia do Parana. Brasılia. In:Congresso Brasileiro de Geologia 24, Anais. Socie-dade Brasileira de Geoloiga, 359–412.

DALLA SALDA, L. A. 2005. The Appalachian-Famatinianorogenic belt. In: PANKHURST, R. J. & VEIGA, G. D.(eds) Gondwana 12: Geological and Biological Heri-tage of Gondwana, Abstracts. Academia Nacional deCiencias, Cordoba, Argentina, 127.

DALY, M. C., LAWRENCE, S. P., KIMUN’A, D. & BINGA,M. 1991. Late Palaeozoic deformation in centralAfrica: a result of distant collision? Nature, 350,605–607.

DECKER, JE. & DE WIT, M. J. 2005. Carbon isotopeevidence for Crassulacean Acid Metabolism in theMesozoic. Terra Nova, 18, 9–17.

DE WIT, M. J. & RANSOME, I. D. 1992. Regional inver-sion tectonics along the southern margin of Gondwana.In: DE WIT, M. J. & RANSOME, I. D. (eds) Inversiontectonics of the Cape Fold Belt, Karoo and Cretaceousbasins of Southern Africa. A. A. Balkema, Rotterdam,15–22.

DE WIT, M. J., JEFFERY, M., BERGH, H. & NICOLAY-

SEN, L. 1988. Geological map of sectors of Gond-wana, reconstructed to their disposition c. 150 Ma.1:10,000,000. American Association of PetroleumGeologists/University of Witwatersrand.

DIAZ-MARTINEZ, E. 1995. Devonico Superior y Carbo-nifero del Altiplano de Bolivia—estratigrafia, sedi-mentologia y evolucion palaeogeografica. Informesde ORSTOM Bolıvia, La Paz, 46.

DUANE, M. J. & BROWN, R. 1992. Geochemical open-system behaviour related to fluid-flow and metamor-phism in the Karoo Basin. In: DE WIT, M. J. &RANSOME, I. D. (eds) Inversion tectonics of theCape Fold Belt, Karoo and Cretaceous basins ofSouthern Africa. A. A. Balkema, Rotterdam, 127–140.

DU TOIT, A. L. 1926. The Geology of South Africa. Oliverand Boyd, Edinburgh.

DU TOIT, A. L. 1927. A geological comparison of SouthAmerica with South Africa. The Carnegie Institution,Washington, Publication 381.


DU TOIT, A. L. 1937. Our wandering continents. Oliverand Boyd, Edinburgh.

DU TOIT, A. L. 1954. The Geology of South Africa.Haughton and Boyd, London.

DUNCAN, R. A, HOOPER, P. R., REHACEK, J., MARSH,J. S. & DUNCAN, A. R. 1997. The timing and durationof the Karoo igneous event, southern Gondwana.Journal of Geophysical Research, 102, 18127–18138.

EGLINTON, B. M. & ARMSTRONG, R. A. 2001. Geochro-nological and isotope constraints on the Mesoprotero-zoic Natal-Namaqua Belt: evidence from deepborehole intersections in South Africa. PrecambrianResearch, 125, 179–189.

EVANS, F. J. 1999. Paleobiology of Early Carboniferousbiota of the Waaipoort Formation (Witteberg Group),South Africa. Palaeontologia Africanus, 35, 1–6.

FAURE, K. & COLE, D. 1999. Geochemical evidence forlacustrine microbial blooms in the vast PermianMain Karoo, Parana, Falkland Islands and Haubbasins of southwestern Gondwana. PalaeogeographyPalaeoclimatology Palaeoecology, 152, 189–213.

FERNANDEZ-SEVESO, F. & TANKARD, A. J. 1995.Tectonics and stratigraphy of the Late PalaeozoicPaganzo Basin of Western Argentina and its regionalimplications. In: TANKARD, A. J., SUAREZ

SORUCO, R. & WELSINK, H. J. (eds) Petroleumbasins of South America. American Association ofPetroleum Geologists, Memoirs, 62, 285–301.

FERNANDEZ-SEVESO, F., PEREZ, M. A., BRISSON, I. E.& ALVAREZ, L. 1993. Sequence stratigraphy andtectonic analysis of the Paganzo Basin, WestArgentina. In: International Conference on Carbon-iferous and Permian, Comptes Rendus, Buenos Aires,2, 223–260.

FOUCH, M. J., JAMES, D. E., VANDECAR, J. C. & VAN

DER LEE, S. KAAPVAAL SEISMIC GROUP, 2004.Mantle seismic structure beneath the Kaapvaal andZimbabwe cratons. South African Journal ofGeology, 107, 33–44.

FRANCA, A. B. & POTTER, P. E. 1988. Estratigrafia,ambiente deposicional e analise de reservatorio doGrupo Itarare (Permocarbonıfero), Bacia do Parana(Parte 1). Boletim de Geociencias da Petrobras, 2,147–191.

FRANCA, A. B., MILANI, E. J. ET AL. 1995. Phanerozoic cor-relation in Southern South America. In: TANKARD, A. J.,SUAREZ SORUCO, R. & WELSINK, H. J. (eds)Petroleum basins of South America. American Asso-ciation of Petroleum Geologists, Memoirs, 62, 129–161.

FRIMMEL, H. E. & FRANK, W. 1998. Neoproterozoictectono-thermal evolution of the Gariep Belt and itsbasement, Namibia and South Africa. PrecambrianResearch, 90, 1–28.

FURQUE, G. 1965. Geologia de la region del Cerro laBolsa (Provincia de La Rioja). In: Segundas JornadasGeologicas Argentinas, Buenos Aires, Actas,181–215.

GOHRBANDT, K. H. A. 1993. Palaeozoic palaeogeo-graphic and depositional developments on the centralproto-Pacific margin of Gondwana: their importanceto hydrocarbon accumulation. Journal of South Amer-ican Earth Sciences, 6, 267–287.

GRADSTEIN, F., OGG, J. & SMITH, A. 2004. A GeologicTime Scale 2004. Cambridge University Press.

GUERRA-SOMMER, M., CAZZULO-KLEPZIG, M.,FORMOSO, M. L., MENEGAT, R. & BASEI, M. A. S.2005. New radiometric data from ash fall rocks inCandiota coal-bearing strata and the palynostrati-graphic framework in southern Parana Basin(Brazil). In: PANKHURST, R. J. & VEIGA, G. D.(eds) Gondwana 12: Geological and Biological Heri-tage of Gondwana. Abstracts, Academia Nacional deCiencias, Cordoba, Argentina, 189.

HALBICH, I. W. 1992. The Cape Fold Belt orogeny: stateof the art 1970s–1980s. In: DE WIT, M. J. &RANSOME, I. D. (eds) Inversion tectonics of theCape Fold Belt, Karoo and Cretaceous basins ofSouthern Africa. A. A. Balkema, Rotterdam, 141–158.

HARVEY, J. D., DE WIT, M. J., STANKIEWICZ, J. &DOUCOURE, C. M. 2001. Structural variations of thecrust in the southwestern Cape deduced from seismicreceiver functions. South African Journal ofGeology, 104, 231–242.

INIGUEZ, A. M., VALLE, A., POIRE, D. G., SPALLETTI,L. A. & ZALBA, P. E. 1989. Cuenca precambrica/palaeozoica inferior de Tandilia, Provincia de BuenosAires. In: CHEBLI, G. A. & SPALLETTI, L. A. (eds)Cuencas sedimentarias argentinas. UniversidadNacional, Tucuman, Geological Correlation Series, 6,245–264.

JOHNSON, M. R. 1991. Sandstone petrography, prove-nance and plate tectonic setting in Gondwana contextof the southeastern Cape-Karoo Basin. South AfricanJournal of Geology, 94, 137–154.

JOHNSON, M. R. 1997. The foreland Karoo Basin, SouthAfrica. In: SELLEY, R. C. (ed.) African basins. Sedi-mentary basins of the World. Elsevier, Amsterdam,269–317.

JOHNSTON, S. T. 2000. The Cape Fold Belt and Syntaxis,and the rotated Falkland Island: dextral transpressiontectonics along the southwest margin of Gondwana.Journal of African Earth Sciences, 31, 51–63.

KEIDEL, J. 1916. La geologıa de las sierras de la Provınciade Buenos Aires y sus relaciones con las montanas deSud Africa y los Andes. Anales del Ministerio de Agri-cultura de la Nacion, Seccion Geologıa, Mineralogıa yMinerıa, Buenos Aires, 3, 1–78.

KING, G. 1990. The dicynodonts. Chapman and Hall,London.

KOKOGIAN, D. A., FERNANDEZ-SEVESO, F. &MOSQUERA, A. 1993. Las secuencias sedimentariastriasicas. In: RAMOS, V. A. (ed.) Geologia y recursosnaturales de Mendoza. Asociacion GeologicaArgentina—Instituto Argentino de Petroleo, BuenosAires, 65–78.

LINDEQUE, A. S., RYBERG, T., STANKIEWICZ, J.,WEBER, M. H. & DE WIT, M. J. 2007. Seismicimaging of the crust and Moho in a section acrossthe Beattie Magnetic Anomaly, South Africa. SouthAfrican Journal of Geology, 110, 419–438.

LOCK, B. E. 1980. Flat-plate subduction and the CapeFold Belt of South Africa. Geology, 8, 35–39.

LOPEZ-GAMUNDI, O. & ROSSELLO, E. A. 1993.Devonian-Carboniferous unconformity in Argentinaand its relation to Eo-Hercynian orogeny in southernSouth America. Geologische Rundschau, 82, 136–147.

LOPEZ-GAMUNDI, O. & ROSSELLO, E. A. 1998.Basin fill evolution and palaeotectonic patterns


along the Samfrau geosyncline: the Sauce Grandebasin–Ventana foldbelt (Argentina) and Karoobasin–Cape foldbelt (South Africa). GeologischeRundschau, 86, 819–834.

LOPEZ-GAMUNDI, O. R., ESPEJO, I. S., CONAGHAN,P. J. & POWELL, C. M. A. 1994. Southern SouthAmerica. In: VEEVERS, J. J. & POWELL, C. M. A.(eds) Permian-Triassic Pangean basins and foldbeltsalong the Panthalassan margin of Gondwanaland.Geological Society of America, Memoirs, 184,281–329.

LOPEZ-GAMUNDI, O. R., CONAGHAN, P. J., ROSSELLO,E. A. & COBBOLD, P. R. 1995. The Tunas Formation(Permian) in the Sierras Australes Foldbelt, eastcentral Argentina—evidence for syntectonic sedimen-tation in a foreland basin. Journal of South AmericanEarth Sciences, 8, 129–142.

MACLEOD, K. G., SMITH, R. M. H., KOCH, P. I. &WARD, P. 2000. Timing of mammal-like reptileextinction across the Permian-Triassic boundary inSouth Africa. Geology, 28, 227–230.

MARSH, J. S., EWART, A., MILNER, S. C., DUNCAN,A. R. & MILLER, R. MCG. 2001. The EtendekaIgneous Province: magma types and their stratigrapicdistribution with implications for the evolution of theParana- Etendeka flood basalt province. Bulletin ofVolcanology, 62, 464–486.

MCLACHLAN, I. R. & ANDERSON, A. 1973. A review ofthe evidence for marine conditions in southern Africaduring Dwyka times. Palaeontologia Africanus, 15,37–64.

MELO, J. H. G. 1988. The Malvinokaffric realm in theDevonian of Brazil. In: MCMILLAN, N. J., EMBRY,A. F. & GLASS, D. J. (eds) Devonian of the world.Canadian Society of Petroleum Geologists, Memoirs,14, 669–704.

MILANI, E. J. 1992. Intraplate tectonics and the evolutionof the Parana Basin, Southern Brazil. In: DE WIT, M. J.& RANSOME, I. D. (eds) Inversion tectonics ofthe Cape Fold Belt, Karoo and Cretaceous basinsof Southern Africa. A. A. Balkema, Rotterdam,101–108.

MILANI, E. J. 1997. Evolucao tectono-estratigrafica daBacia do Parana e seu relacionamento com a geodina-mica fanerozoica do Gondwana sul-ocidental. DScthesis, Universidade Federal do Rio Grande do Sul.

MILANI, E. J. 2004. Comentarios sobre a origem e evolu-cao tectonica da Bacia do Parana. In: MANTESSO-NETO, V., BARTORELLI, A., CARNEIRO, C. D. R. &BRITO-NEVES, B. B. (eds) Geologia do continenteSul-Americano: evolucao da obra de FernandoFlavio Marques de Almeida. Beca Editora, SaoPaulo, 265–279.

MILANI, E. J. & RAMOS, V. A. 1998. Orogenias paleozoi-cas no domınio sul-ocidental do Gondwana e os ciclosde subsidencia da Bacia do Parana. Revista Brasileirade Geociencias, 28, 473–484.

MILANI, E. J., ASSINE, M. L., SOARES, P. C. &DAEMON, R. F. 1996. A Sequencia Ordovıcio-Siluriana da Bacia do Parana. Boletim de Geocienciasda Petrobras, 9, 301–320.

OPDYKE, N. D., MUSHAYANDEBUN, M. & DE WIT, M. J.2001. A new palaeomagnetic pole for the DwykaSystem and correlative sediments in sub-Saharan

Africa. Journal of African Earth Sciences, 33,143–154.

PEREIRA, E., RODRIGUES, R. & BERGAMASCHI, S.2005. Stratigraphical, bioestratigraphical and geo-chemical characterization of Silurian and Devonianworld-wide transgressive events in the Parana Basin:a high-resolution approach. In: PANKHURST, R. J. &VEIGA, G. D. (eds) Gondwana 12: Geological andBiological Heritage of Gondwana, Abstracts. Acade-mia Nacional de Ciencias, Cordoba, Argentina, 290.

PANKHURST, R. J. & RAPELA, C. W. 1998. Introduction.In: PANKHURST, R. J. & RAPELA, C. W. (eds) TheProto-Andean Margin of Gondwana. GeologicalSociety, London, Special Publications, 142, 1–9.

PANKHURST, R. J., RAPELA, C. W. Fanning C. M. &MARQUEZ, M. 2006. Gondwanide continental col-lision and the origin of Patagonia. Earth-ScienceReviews, 76, 235–257.

PITTS, B. E., MAHER, M. J., DE BEER, J. H. & GOUGH,D. I. 1992. Interpretation of magnetic, gravity andmagnetotelluric data across the Cape Fold Belt andKaroo basin. In: DE WIT, M. J. & RANSOME, I. D.(eds) Inversion tectonics of the Cape Fold Belt,Karoo and Cretaceous basins of Southern Africa.A. A. Balkema, Rotterdam, 27–32.

POWELL, C. M. A. 1993. Assembly of Gondwanaland—open forum. In: Proceedings of Gondwana Eight.A. A. Balkema, Rotterdam, 219–237.

RAKOTOSOLOFO, N. A., TORSVIK, T. H., ASHWAL,L. D., EIDE, E. A. & DE WIT, M. J. 1999. TheKaroo supergroup revisited and Madagascar-Africafits. Journal of African Earth Sciences, 29, 135–151.

RAMOS, V. A. 1988. Late Proterozoic — Early Palaeozoicof South America—a collisional history. Episodes, 11,168–174.

RAMOS, V. A. 1990. Field guide to geology of the CentralAndes (318–338 SL). Buenos Aires, Universidad deBuenos Aires, Central Andes Field Seminar.

RAMOS, V. A. 1993. Interpretacion tectonica. In: RAMOS,V. A. (ed.) Geologia y recursos naturales de Mendoza.Asociacion Geologica Argentina—Instituto Argentinode Petroleo, Buenos Aires, 257–268.

RAMOS, V. A. 2005. The Proterozoic-Early Palaeozoicmargin of western Gondwana. In: PANKHURST, R. J.& VEIGA, G. D. (eds) Gondwana 12: Geologicaland Biological Heritage of Gondwana, Abstracts.Academia Nacional de Ciencias, Cordoba,Argentina, 304.

RAMOS, V. A., JORDAN, T. E., ALLMENDINGER, R. W.,KAY, S. M., CORTES, J. M. & PALMA, M. A. 1984.Chilenia: un terreno aloctono en la evolucion palaeo-zoica de los Andes centrales. In: IX Congreso Geolo-gico Argentino, Buenos Aires, Actas, 2, 84–106.

RAMOS, V. A., JORDAN, T. E. ET AL. 1986. Palaeozoicterranes of the central Argentine-Chilean Andes.Tectonics, 5, 855–880.

RAPELA, C. W. 2000. The Sierras Pampeanas of Argen-tina: Paleozoic building of the southern proto-Andes.In: CORDANI, U. G., MILANI, E. J., THOMAZ

FILHO, A. & CAMPOS, D. A. (eds) Tectonic evolutionof South America. 31st International GeologicalCongress, Rio de Janeiro, 381–387.

RAPELA, C. W., PANKHURST, R. J., FANNING, C. M. &GRECCO, L. E. 2003. Basement evolution of the


Sierra de la Ventana Fold Belt: new evidence forCambrian continental rifting along the southernMargin of Gondwana. Journal of the GeologicalSociety, London, 160, 613–628.

RAPELA, C. W., CASQUET, C. ET AL. 2008. The Rıo de laPlata craton and the assembly of SW Gondwana.Earth-Science Reviews, 83, 49–82.

REEVES, C. V., SAHU, B. K. & DE WIT, M. J. 2002. Are-examination of the palaeo-positions of Africa’seastern neighbours in Gondwana. Journal of AfricanEarth Sciences, 34, 101–108.

RUBIDGE, B. S., HANCOX, P. J. & CATUNEANU, O. 2000.Sequence analysis of the Ecca-Beaufort contact in thesouthern Karoo of South Africa. South African Journalof Geology, 103, 81–96.

RUST, I. C. 1975. Tectonic and sedimentary framework ofGondwana basins in southern Africa. In: Third Gond-wana Symposium, Australia, Actas, 554–564.

SANTOS, R. V., SOUZA, P. A. ET AL. 2006. Shrimp U–Pbzircon dating and palynology of bentonitic layers fromthe Permian Irati Formation, Parana Basin, Brazil.Gondwana Research, 9, 456–463.

SCOTESE, C. R. 2000. Paloemap Project: Earth History(palaeogeographic maps). Department of Geology,University of Texas, Arlington.

SEMPERE, T. 1995. Phanerozoic evolution of Bolivia andadjacent regions. In: TANKARD, A. J., SUAREZ

SORUCO, R. & WELSINK, H. J. (eds) Petroleumbasins of South America. American Association ofPetroleum Geologists, Memoirs, 62, 207–230.

SLOSS, L. L. 1963. Sequences in the cratonic interior ofNorth America. Geological Society of America Bulle-tin, 74, 93–114.

SMITH, R. M. H. 1995. Fluvial environments across thePermian-Triassic boundary in the Karoo Basin, SouthAfrica and possible causes of the tetrapod extinctions.Palaeogeography Palaeoclimatology Palaeoecology,117, 84–104.

STANKIEWICZ, J., RYBERG, T., SCHULZE, A.,LINDEQUE, A. S., WEBER, M. H. & DE WIT, M. J.2007. Results from the wide-angle seismic refractionlines in the Southern Cape. South African Journal ofGeology, 110, 407–418.

STECKLER, M. & WATTS, A. B. 1978. Subsidence of theAtlantic-type continental margin off New York. Earthand Planetary Science Letters, 41, 1–13.

STERN, C. R. & DE WIT, M. J. 2004. Rocas Verdes ophio-lite, southernmost South America: remnants ofprogressive stages of development of oceanic typecrust in a continental back arc basin. In: DILEK, Y.& ROBINSON, P. T. (eds) Ophiolites in EarthHistory. Geological Society, London, Special Publi-cations, 218, 665–683.

STREEL, M. & THERON, J. N. 1999. The Devonian-Carboniferous boundary in South Africa and the ageof the earliest episode of the Dwyka glaciation: Newpalynological result. Episodes, 22, 41–44.

TANKARD, A. J., JACKSON, M. P. A., ERIKSSON, K. A.,HOBDAY, D. K., HUNTER, D. R. & MINTER, W. E. L.1982. Crustal evolution of southern Africa. Springer-Verlag, Berlin.

THOMAS, W. A. & ASTINI, R. A. 2005. Vestiges of anOrdovician Ocloyic west-vergent, thin-skinned thrustbelt in the Argentine Precordillera, southern central

Andes. In: PANKHURST, R. J. & VEIGA, G. D. (eds)Gondwana 12: Geological and Biological Heritageof Gondwana, Abstracts. Academia Nacional deCiencias, Cordoba, Argentina, 350.

TROUW, R. A. J. & DE WIT, M. J. 1999. Relation betweenthe Gondwanide Orogen and contemporaneous intra-cratonic deformation. Journal of African EarthSciences, 28, 203–213.

TURNER, B. R. 1999. Tectonostratigraphical developmentof the upper Karoo foreland basin: orogenic unloadingversus thermally-induced Gondwana rifting. Journalof African Earth Sciences, 28, 215–238.

TURNER, S., REGELOUS, M., KELLEY, S., HAWKES-

WORTH, C. & MANTOVANI, M. 1994. Magmatismand continental break-up in the South Atlantic: highprecision Ar–Ar geochronology. Earth and PlanetaryScience Letters, 121, 333–348.

VAIL, P. R., MITCHUM, R. M. & THOMPSON, S. 1977.Seismic stratigraphy and global changes of sea level,part 3: relative changes of sea level from coastalonlap. In: PAYTON, C. E. (ed.) Seismicstratigraphy—applications to hydrocarbon explora-tion. American Association of Petroleum Geologists,Memoirs, 26, 63–81.

VAUGHAN, A. P. M. & LIVERMORE, R. A. 2005. Episo-dicity of Mesozoic terrane accretion along the Pacificmargin of Gondwana: implications fro superplume–plate interactions. In: VAUGHAN, A. P. M., LEAT,P. T. & PANKHURST, R. J. (eds) Terrane processesat the margins of Gondwana. Geological Society,London, Special Publications, 246, 143–178.

VAUGHAN, A. P. M., LEAT, P. T. & PANKHURST, R. J.(eds) 2005. Terrane processes at the margins ofGondwana. Geological Society, London, SpecialPublications, 246.

VEEVERS, J. J., COLE, D. I. & COWAN, E. J. 1994.Southern Africa: Karoo Basin and Cape Fold Belt.In: VEEVERS, J. J. & POWELL, C. M. A. (eds)Permian-Triassic pangean basins and foldbeltsalong the panthalassan margin of Gondwanaland.Geological Society of America, Memoirs, 184,223–280.

VILLAR, L. M., CHERNICOFF, C. J. & ZAPPETTINI, E.2005. Evidence of a Famatinian continental magmaticarc at Paso del Bote, La Pampa province, Argentina.In: PANKHURST, R. J. & VEIGA, G. D. (eds)Gondwana 12: Geological and Biological Heritageof Gondwana, Abstracts. Academia Nacional deCiencias, Cordoba, Argentina, 365.

VISSER, J. N. J. 1987. The palaeogeography of part ofsouthwestern Gondwana during the Permian-Carboniferous glaciation. Palaeogeography Palaeo-climatology Palaeoecology, 61, 205–219.

VISSER, J. N. J. 1989. The Permo-carboniferous DwykaFormation of southern Africa: deposition by a predo-minantly marine ice-sheet. Palaeogeography Palaeo-climatology Palaeoecology, 70, 377–391.

VISSER, J. N. J. 1990. The age of the late Palaeozoic gla-ciogene deposits in southern Africa. South AfricanJournal of Geology, 93, 366–375.

VISSER, J. N. J. 1992a. Sea-level changes in aback-arc-foreland transition: the late Carboniferous-Permian Karoo Basin of South Africa. SedimentaryGeology, 83, 115–131.


VISSER, J. N. J. 1992b. Deposition of the early to latePermian Whitehill Formation during a sealevel highstand in a juvenile foreland basin. South AfricanJournal of Geology, 95, 181–193.

VISSER, J. N. J. 1997. Deglaciation sequences in thePermo-Carboniferous Karoo and Kalahari basins ofsouthern Africa: a tool in the analysis of cyclic glacio-marine basin fills. Sedimentology, 44, 507–521.

VON GOSEN, W., BUGGISCH, W. & KRUMM, S. 1991.Metamorphism and deformation mechanisms in theSierras Australes fold and thrust belt (Buenos AiresProvince, Argentina). Tectonophysics, 185, 335–356.

WARD, P. D., MONTGOMERY, D. R. & SMITH, R. 2000.Altered river morphology in South Africa related tothe Permian-Traissic extinction. Science, 289,1740–1743.

WECKMANN, U., RITTER, O., JUNG, A., BRANCH, T. &DE WIT, M. J. 2007. Magnetotelluric measurementsacross the Beattie magnetic anomaly and the SouthernCape Conductive Belt, South Africa. Journal of Geo-physical Research, 112, B5, B05416.

WEGENER, A. 1912. Die Entstelhung der Konitinente.Geologische Rundschau, 3, 267–292.

WICKENS, H. D. V. 1992. Submarine fans of the PermianEcca Group in the SW Karoo Basin: Their origin andreflection on the tectonic evolution of the basin and

its source areas. In: DE WIT, M. J. & RANSOME,I. D. (eds) Inversion tectonics of the Cape Fold Belt,Karoo and Cretaceous basins of Southern Africa.A. A. Balkema, Rotterdam, 117–126.

WILLIAMS, K. E. 1995. Tectonic subsidence analysisand Palaeozoic palaeogeography of Gondwana. In:TANKARD, A. J., SUAREZ SORUCO, R. &WELSINK, H. J. (eds) Petroleum basins of SouthAmerica. American Association of Petroleum Geol-ogists, Memoirs, 62, 79–100.

ZALAN, P. V., WOLFF, S. ET AL. 1990. The Parana Basin,Brazil. In: LEIGHTON, M. W., KOLATA, D. R., OLTZ,D. F. & EIDEL, J. J. (eds) Interior cratonic basins.American Association of Petroleum Geologists,Memoirs, 51, 681–708.

ZAWADA, P. K. & CADLE, A. B. 1988. Position of theEcca-Beaufort boundary contact in the southwesternFree State: an evaluation of four possible alternatives.South African Journal of Geology, 91, 49–56.

ZIMMERMANN, U. & SPALLETTI, L. A. 2005. Prove-nance of the Balcarce Formation: an indicator forEarly Palaeozoic volcanism in Eastern Argentina. In:PANKHURST, R. J. & VEIGA, G. D. (eds) Gondwana12: Geological and Biological Heritage of Gondwana,Abstracts. Academia Nacional de Ciencias,Cordoba, Argentina, 377.


Milani&de Wit Karoo-parana

Documents