Keywords

1 Introduction

Gondwana amalgamation was the product of a diachronic convergence of Neoproterozoic continents that culminated in the Early Paleozoic (Kennedy 1964; Boger and Miller 2004; Collins and Pisarevsky 2005; Li et al. 2008; Meert and Lieberman 2008; Merdith et al. 2017a). The convergent margins were a mix of long-lived margins inherited from the Mesoproterozoic, and new convergent margins that succeeded the break-up of the Tonian supercontinent Rodinia. The Neoproterozoic can be summarized as a combinaton of a transition from Rodinia to Gondwana (Hoffman 1991) along with the closure of a vast pre-Rodinian accretionary orogen by the collision of Neoproterozoic India with an East African-Antarctic-West Australian margin (Merdith et al. 2017a). A number of models for this rifting and reconfiguration have been proposed (Dalziel 1991, 1997; Hoffman 1991; Moores 1991; Karlstrom et al. 1999; Burrett and Berry 2000; Wingate et al. 2002; Li et al. 2008, Merdith et al. 2017a). These have recently been kinematically tested by comparing their plate tectonic motion implications (Merdith et al. 2017b).

Table 15.1 Data compilation of Gondwana post-670 Ma orogens with references
Full size table

It has long been proposed that two major tectonic periods were responsible for Gondwana construction (Clifford 1967; Stern 1994; Meert 2003; Collins and Pisarevsky 2005; Cawood and Buchan 2007; Merdith et al. 2017a). The older comprised Late Tonian and Cryogenian (pre-670 Ma) mostly accretionary orogens, preserved within the belts, normally reworked by younger orogenic events. The second and later period included Cryogenian through Cambrian (post-670 Ma) accretionary/collisional settings, responsible for suturing the newly formed supercontinent. These final Gondwana ‘internal orogens’ (e.g., Buzios, Damara, Saldania, Paraguay, Araguaia, Malagasy, Kuunga orogenies) overlapped in time with the initiation of convergent settings on the newly formed margins, certainly influencing the coeval Cambro-Ordovician ‘external’ orogens (e.g., Pampean, Ross, Delamerian, Bhimphedian; Rapela et al. 1998, 2011; Schmitt et al. 2004; Foden et al. 2006; Oriolo et al. 2017). Therefore, by the end of the Cambrian, Gondwana was laced by a series of internal Himalayan-style belts formed by continent–continent collision superimposed on earlier accretionary orogens and ringed by accretionary orogens that although may have been initiated during the final Gondwana formation were to last for the duration of the Phanerozoic. These are dominated by the accretion of oceanic materials along the Gondwana margins (Murphy and Nance 1991; Cawood 2005; Collins et al. 2011).

However, the nature of the ‘internal orogens’ is still a matter of debate. Some authors suggest that they are products of the closure of large oceans, culminating with the collision of faraway paleocontinents (e.g., Clymene Ocean, Tohver et al. 2012; Khomas Ocean, Foster et al. 2015; Adamastor Ocean, Heilbron et al. 2008; Mozambique Ocean, Stern 1994; Meert and Van Der Voo 1997; Collins and Windley 2002; Boger and Miller 2004). Another point of view suggests that these orogens were the product of intracontinental rifting and basin inversion in between blocks that were never far apart (e.g., Araguaia Belt, Cordani et al. 2013a; Damara belt, Nascimento et al. 2017; Ribeira Belt, Meira et al. 2015). These two end members produced distinctive collisional processes during the last convergent phase, providing either a Himalayan or an Alpine-type orogenic style. Differences include the occurrence of a pre-collisional arc in the former, and the development of a hyperextended margin with exhumation of subcontinental lithospheric mantle in the second type (Manatschal and Müntener 2009).

Here we present a review of the main orogenic belts that consolidated Gondwana, using the geological database of the new geological map of Gondwana (Schmitt et al. 2016a), focusing on the post-670 Ma orogenic pulses. The geological data confirms that the Cryogenian-Cambrian orogenic events are widespread in all belts that sew Gondwana together, in both eastern and western major fragments. In addition, we discuss these events in SW Gondwana, mostly preserved on the actual coastal regions of South America and Africa with an inland branch (Damara belt). This implies that the basement of the Atlantic continental margins comprises Ediacaran-Cambrian belts (Schmitt et al. 2016b). Our data also shows that during the last period of Gondwana formation, there was an extensional stage (c. 610–570 Ma) that pre-dated the final collisional events. The ‘South Atlantic orogenic system’ contains geological units of oceanic nature from this stage. The cause of closing these Ediacaran basins could have been the initiation of the marginal Gondwana convergent settings or the far-field effects of the Himalayan-scale collision caused by Neoproterozoic India colliding with both Australia and Africa at this time (Merdith et al. 2017a).

2 Gondwana Amalgamation

Although not including all of the Earth’s major continental fragments (e.g., Laurentia, Baltica), Gondwana is commonly referred to as a supercontinent (Fig. 15.1). Its landmass represents around 64% of today’s continental crust (Torsvik and Cocks 2013), composed of the continents of Africa, South America, India, Australia and Antarctica, including several smaller fragments now incorporated into Asia, Europe and North America. The supercontinent title is not only related to the amount of participating continental crust but also to its endurance, rarely registered in the evolution of the Earth’s history. Gondwana lasted as a merged landmass c. 320 myr, from the Cambrian (c. 500 Ma) until the Jurassic (c. 180 Ma), comparing only to the lifetime of Archean cratons (Condie et al. 2015). Gondwana’s lifetime expands to a value between 800 and 600 myr when taking into account the amalgamation and break-up processes. Supercontinental cycles are assumed to vary from 250 to 1000 myr, including both assembly and dispersal (Condie et al. 2015). Gondwana’s break-up interval started at c. 183 Ma, with Madagascar splitting away from East Africa, and ended at c. 85 Ma, with Antarctica and Australia’s separation (White et al. 2013; Reeves 2014), lasting c. 100 myr. Nevertheless, the amalgamation period was longer, from 670 to 480 Ma (c. 190 myr), considering all the orogenic belts that sutured the Neoproterozoic cratons (Fig. 15.2). However, if the Late Tonian-Cryogenian tectonic events (850–670 Ma) are included in this timeframe, then the amalgamation period lasted for more than 370 myr.

Fig. 15.1
figure 1

Gondwana configuration at c. 500 Ma with cratons (pre-Gondwana continents) sutured by Gondwana mobile belts. This figure is mostly compiled from the new Gondwana geological map at 1:5 M scale (Schmitt et al. 2016a). It is also compiled from Collins (2003), Kröner and Cordani (2003), Heilbron et al. (2004), Jacobs et al. (1998, 2008), Schmitt et al. (2008), Ramos (2010), Frimmel et al. (2011), Offler et al. (2011), Fritz et al. (2013), Torsvik and Cocks (2013), Ramos and Naipauer (2014). Reconstruction by Richetti et al. (2016). The letters and numbers represent Gondwanan cratons and mobile belts, respectively. They are AA Arequipa/Antofalla; AM Amazonia; RA Río Apa; SL São Luís; PR, Parnaíba; SF, São Francisco; PP, Paranapanema; LA Luís Alves; RP Río de La Plata; KH, Kalahari; B Bangweulu Block; TZ Tanzania; CC Congo; WA West Africa; SH Sahara; DW Dharwar; GH Grunehogna; EA East Antarctica; WAU West Australia; SAU South Australia; NAU North Australia. 1—Pampeana; 2—Caapucú High; 3—Paraguai; 4—Araguaia; 5—Gurupi; 6—Borborema (North and Central); 7—Borborema (South); 8—Rio Preto; 9—Brasília; 10—Araçuaí (West), Ribeira (Paraíba/Embu); 10b—Apiaí; 11—Araçuaí (East), Ribeira (Oriental Terrane) and Costeiro Domain; 12—Cabo Frio Tectonic Domain; 13—Dom Feliciano, Kaoko (Coastal); 14—Cuchilla Dionísio; 15—Nico Pérez; 16—Saldania; 17—Gariep; 18—Damara; 19—Kaoko (Central-East); 20—Angolan Belt; 21—West Congo; 22—Oubanguides; 23—Dahomey; 24—Rockelides; 25—Anti-Atlas; 26—Hoggar; 27—Ad Dawadimi and Ar Rayn; 28—Arabian/Nubian Shield (North); 29—Arabian/Nubian Shield (South); 30—Galana (Azania); 31—Western Granulite; 32—Eastern Granulite; 33—Zambesi; 34—Lufilian; 35—Nampula Block; 36—Dronning Maud Land (West); 37—Dronning Maud Land (Sor Rondane Mountains and Yamato-Belgica Complex); 38—Prince Olaf Coast/Kemp Land, (Lützow-Holm Complex); 39—Sri Lanka; 40—Southern Granulites; 41—Madagascar (Vohibori); 42—Madagascar (Antananarivo, Androyen and Bemarivo); 43—Seychelles; 44—Reworked border of the Napier Complex; 45—Eastern Ghat; 46—Princess Elizabeth Land; 47—Meghalaya Plateau; 48—Pinjarra; 49—Petermann; 50—Delamerian; 51—Ross (Northern and Southern Victoria Land); 52—Ross (Pensacola Mountains). TKL stands for Transbrasiliano-Kandi Lineament

Full size image
Fig. 15.2
figure 2

a Time-space chart with all recorded eastern and western Gondwana orogens from the Ediacaran and Cambrian. The data was compiled from references listed in Table 15.1. b At the top right there is an insert of Gondwana with the areas corresponding to the two main amalgamation phases. Note that the proto-Gondwana core is centred on Africa

Full size image

The earlier Tonian-Cryogenian events are dispersed within some belts. In terms of the volume of continental crust reworked or generated, these are minor in comparison with the equivalent Ediacaran–Cambrian regions. It is noteworthy that the Tonian-Cryogenian domains have a large percentage of juvenile material (Johnson et al. 2011; Oriolo et al. 2017), suggesting that much of the continental crust was generated at this time, in contrast to the Ediacaran-Cambrian, which involved more tectonothermal reworking of existing crust.

Here we consider the 670–480 Ma tectonic events to be directly responsible for Gondwana amalgamation. According to the compiled data, within this timeframe, two stages of convergent tectonics are identified, overlapping partially with a transitional extensional period (c. 610–570 Ma) (Figs. 15.2 and 15.3).

Fig. 15.3
figure 3

Diagram with distribution of 670–480 Ma orogenic peaks in Gondwana. Based on Fig. 15.2

Full size image

The birth of Gondwana is best represented by the map of Neoproterozoic cratons and Neoproterozoic-Cambrian belts (Trompette 1994; Collins 2006; Gray et al. 2008; Fig. 15.1). We present a new version of the Gondwana cratons-belts map, based on the new geological map of Gondwana (Schmitt et al. 2016a). Our compilation shows 55 post-670 Ma belts which register Gondwana’s final amalgamation, suturing c. 16 cratonic blocks (Figs. 15.1 and 15.2).

Many authors suggest that the approximation and collision between the Neoproterozoic paleocontinents was orchestrated by the Rodinia supercontinent break-up (Hoffman 1991; Cawood and Buchan 2007). This common sense is based mostly on the similarities between the cratons’ ‘barcodes’ (age pattern of crustal growth) and paleomagnetic data. However, some western Gondwana cratons (and possibly India; Merdith et al. 2017a) might not have been Rodinian participants (Cordani et al. 2003; Oriolo et al. 2017). Over recent decades it has become clear that the Neoproterozoic was an Era of several individual continents, inconsistent with the earlier ideas of only two single continents, named East and West Gondwana (cf. Shackleton 1996).There are large differences between the eastern and western orogens, but we will show that at the end of the amalgamation process (late Ediacaran-Cambrian), these orogens started to share a similar evolution.

2.1 Eastern Gondwana Orogens (Including the EAO)

Rocks deformed and metamorphosed in the East African Orogen (EAO; Stern 1994) extend, in a reconstructed Gondwana, from the eastern Mediterranean in the north (e.g., Candan et al. 2016), through Arabia (including NW India/Pakistan/Afghanistan), eastern Africa, Madagascar, southern India and Sri Lanka to East Antarctica (Fig. 15.1). The orogen is likely to follow the subglacial East Antarctica Mountain Range (An et al. 2015) to the Gambutschev suture (Ferraccioli et al. 2011), where it meets the Kuunga Orogen (Meert et al. 1995; also known as the Pinjarra-Prydz-Denman Orogen; Fitzsimons 2003a, b; Fig. 15.1). Together these orogens delineate the western, southern and eastern margins of Neoproterozoic India.

The northern East African Orogen is part of the Arabian-Nubian Shield (Johnson et al. 2011) and is characterized by voluminous juvenile Neoproterozoic crust that formed as a series of volcanic arcs (Robinson et al. 2014, 2015a, b; Blades et al. 2015, 2017; Fig. 15.1). Less extensive continental terranes exist in the region, particularly in the Sinai (Be’eri-Shlevin et al. 2012; Eyal et al. 2014), in the Khida and Afif Terranes of Saudi Arabia (Stoeser et al. 2001; Whitehouse et al. 2001) and in Yemen (Windley et al. 1996, 2001; Whitehouse et al. 1998, 2001) with corollaries along the southern Gulf of Aden escarpment (Sassi et al. 1993; Whitehouse et al. 2001; Collins and Windley 2002). The eastern margin of the East African Orogen in Arabia is often left at the margin of the exposed Neoproterozoic in Saudi Arabia, but similar magnetic anomalies to the easternmost exposed Saudi terrane (the Ar-Rayn terrane; Doebrich et al. 2007; Cox et al. 2012) occur beneath the Ediacaran Rub Al-Khali Basin of Saudi Arabia (Johnson and Stewart 1995). Where Precambrian basement is exposed in the east of the Arabian Peninsula in Oman, it is again Neoproterozoic juvenile crust that formed in volcanic arc tectonic environments (Bowring et al. 2007; Whitehouse et al. 2016; Alessio et al. 2017). This led Cozzi et al. (2012) and Merdith et al. (2017a) to extend the East African Orogen to regions that now form the basement of southern Afghanistan, Pakistan and NW India.

The Mozambique Belt is the common name for the southern East African Orogen. Here, Tonian and pre-Tonian terranes came together in two main orogenic events (Fig. 15.1). Many of these Tonian and pre-Tonian terranes have a long history of pre-Cryogenian arc-related subduction-arc magmatism (Handke et al. 1999; Blades et al. 2015; Elburg et al. 2015; Jacobs et al. 1998, 2015; Archibald et al. 2016, 2017a, b). The earlier one occurred at c. 650–640 Ma as indicated by the time of peak metamorphism in Uganda, Kenya, Tanzania and northern Mozambique (Appel et al. 1998; Hauzenberger et al. 2004; 2007; Fritz et al. 2013; Tenczer et al. 2013). This orogenic event was interpreted as being the result of intra-arc extension (Appel et al. 1998), based on its counterclockwise P-T-t path, but on a regional scale it correlates with the amalgamation of the main Arabian-Nubian Shield along the Keraf Suture to the north, and it is focused along the suture of Azania (an extensive pre-Neoproterozoic terrane including central Madagascar, parts of Somalia, Ethiopia, Yemen and the Madurai Block of southern India) with the Congo-São Francisco continent. This is particularly apparent in Madagascar, where c. 650–640 Ma metamorphic ages were reported from the west of the country (Jöns and Schenk 2011), whereas in the eastern part, metamorphic ages of c. 570–540 Ma dominate (Tucker et al. 1999, 2014; Collins et al. 2003; Jöns and Schenk 2011; Fig. 15.1). This younger, eastern orogenesis correlates with the Ediacaran arc accretion recorded in the far east of the Saudi Arabian Shield that separates the exposed Saudi Shield from the basement of Oman (Fig. 15.1). These observations led Collins and Pisarevsky (2005) to propose that the western c. 650–640 Ma orogenesis was due to late Cryogenian collision of Azania with the Congo-São Francisco continent (the East African Orogeny sensu stricto; e.g., Stern 1994; Meert and Van der Voo 1997), while the younger c. 570–540 Ma orogenesis was the result of the final collision of Neoproterozoic India with the then amalgamated Azania/Congo-São Francisco continent, closing the Mozambique Ocean (and forming the Malagasy Orogeny; Collins and Pisarevsky 2005). Studies from southern India support this hypothesis because orogenesis in the Southern Granulite Belt is restricted to 570–520 Ma and forms a part of the Malagasy Orogeny (Collins et al. 2007a, b, 2014; Plavsa et al. 2012, 2014, 2015; Clark et al. 2015; Johnson et al. 2015; Taylor et al. 2015; Richard et al. 2015; Vijaya Kumar et al. 2017).

The Pinjarra-Prydz-Denman (Kuunga) orogeny led to the final major amalgamation of continental crust in eastern Gondwana, with the suturing of Australia-East Antarctica against India and Kalahari. The Pinjarra orogeny refers to the entire orogen, but here we separate the three to discretely treat varying tectonic events. The Pinjarra orogeny preserves the suture along the west coast of Australia (Fig. 15.1). Further south, in Antarctica, ice covers most exposures, but the suture crops out in the Denman glacier area, and, further south, in the Prince Charles Mountains-Prydz Bay area, where India and the Rayner province collided with the main crustal part of Antarctica (e.g., Boger 2011).

Exposure of the Pinjarra orogeny in Australia is limited to small inliers along the western coastline of the continent, such as the Leeuwin, Northampton and Mullingara Complexes (Fig. 15.1). The Leeuwin Complex in the southwest best preserves the orogeny (e.g., Collins and Fitzsimons 2001; Collins 2003). Here, pink granitic gneisses had their protoliths emplaced at c. 750 Ma and they exhibit upper amphibolite-granulite metamorphism dated at c. 522 Ma (Collins 2003). This is broadly coeval with the end of tectonism, as c. 520 Ma dykes that intrude the Leeuwin Complex exhibit no deformation (e.g., Fitzsimons 2003a, b). The tectonic environment of emplacement was inferred to be a rift, related to Rodinia break-up (Collins 2003), since at the time it was postulated that Kalahari was attached to this margin of Australia (e.g., Powell and Pisarevsky 2002). Sinistral shearing is preserved in the Northampton Complex (Embleton and Schmidt 1985), and alkali granitoids in the Leeuwin Complex, originally inferred to be rift related, are now thought to have been emplaced in a sinistral transpressive environment (e.g., Harris 1994; Fitzsimons 2003a, b).

Further south, the Antarctica Denman Glacier area fits tightly against the Leeuwin Complex in a reconstructed Gondwana (Fig. 15.1). Here, U–Pb dating of zircon from syenite yielded an age of c. 516 Ma, and orthogneisses with a protolith age of c. 3 Ga record a metamorphic overprint age of between 550 and 520 Ma (e.g., Halpin et al. 2008). Some data shows substantial lead loss between 600 and 520 Ma (Black et al. 1992), indicating an Ediacaran history similar to that of rocks further north in Australia. The suture between India-Antarctica and Australia-Antarctica is typically traced south of this area, towards Prydz Bay and the Prince Charles Mountains (e.g., Boger et al. 2001).

The Prydz Bay area is further south and east in Antarctica and is also strongly affected by the Gondwana-forming orogeny between India-Antarctica and Australia-Antarctica. Owing to the similarity of protoliths, neodymium model ages and metamorphic events, the Prydz Bay area is inferred to be part of the Indo-Antarctica plate (e.g., Zhao et al. 1995; Kelsey et al. 2007; Wang et al. 2008; Liu et al. 2009; Boger 2011). Here, too, late Ediacaran-early Cambrian metamorphism up to granulite facies (Liu et al. 2003; Kelsey et al. 2007) is evident, with 540–500 Ma charnockite and granite plutons intruding gneisses (Liu et al. 2006, 2009; Mikhalsky and Sheraton 2011). Further inland from Prydz Bay, zircon from gneiss in the Grove Mountains suggests magmatic emplacement at c. 900 Ma, with a high-grade metamorphic overprint between 530 and 520 Ma (Zhao et al. 2000; Liu et al. 2003). Younger c. 500 Ma granitic dykes exhibit no metamorphism, suggesting that deformation had finished by this time (Zhao et al. 2000).

A broad region of Ediacaran-Cambrian deformation and metamorphism occurs within Australia. This stretches from the eastern Pilbara, where it is called the Patterson Orogen (Martin et al. 2017), and then passes through central Australia, where it is known as the Petermann Orogen. This intra-Australian orogen involves significant dextral transpressional deformation (Raimondo et al. 2009, 2010) and has been linked to a suggested Neoproterozoic 40° anticlockwise rotation between the North Australian Craton and the South Australian Craton interpreted from paleomagnetic data (Li and Evans 2011). Cryogenian sedimentary rocks of the Centralian Superbasin are found on both sides of the Petermann Orogen, supporting the hypothesis that the orogen is an intracontinental orogen and, although there is significant crustal shortening (Raimondo et al. 2010), it doesn’t represent an oceanic suture (Close et al. 2003).

2.2 Western Gondwana Orogens (Pan-African–Brasiliano Events)

Amazonia, West Africa, São-Francisco-Congo, Kalahari and Rio de La Plata are the major outcropping paleocontinents of western Gondwana (Fig. 15.1). The Saharan block, considered to be a metacraton, is poorly exposed but also may have played an important role in the amalgamation process (Abdelsalam et al. 2002). These Neoproterozoic paleocontinents are partially covered by Phanerozoic deposits, also comprising Gondwana intracratonic basins. Two cratonic blocks are inferred below the Paraná (Paranapanema paleocontinent) and Parnaíba (Parnaíba block) basins in Brazil, based on geophysical and geological data from the basement (Mantovani and Brito Neves 2009; Daly et al. 2014; Fig. 15.1). Smaller blocks are São Luis (which links with the West Africa Craton), Luis Alves (which might link with Paranapanema) and Rio Apa (an inlier within the Cambrian Paraguay-Pampean orogens; respectively, Basei et al. 2008; Klein and Moura 2008; Dragone et al. 2017; Fig. 15.1).

One of the major crustal scale structures within western Gondwana, the Transbrasiliano-Kandi shear zone (Cordani et al. 2013b; Fuck et al. 2014), is considered by Ganade de Araujo et al. (2014a) to have developed in a c. 3000 km collisional orogen (the West Gondwana Orogen). According to Ganade de Araujo, this orogen, which includes at least the Brasilia, Borborema, Dahomey and Hogar belts, is comparable to the East African Orogen because it represents the closure of a large oceanic basin and evolved from a long-lasting subduction environment (Fig. 15.1). The West Gondwana Orogen was formed as a result of the convergence of two groups of blocks: Amazonia-West African cratons and Central African blocks (Saharan, São Francisco-Congo, Kalahari and Rio de La Plata). The consequent closure of the Goais-Pharusian Ocean is registered by high-pressure subduction-related rocks along the orogen and also some mafic oceanic units. These collisional events occurred between 630 and 580 Ma (Ganade de Araujo et al. 2014b).

At first glimpse it might seem logical to visualize the Gondwana amalgamating framework as three blocks and two major north–south running orogens. However, when one looks in detail, the framework is more complex, crosscut by roughly east–west-trending orogens. These interference zones, or transverse orogenic triple junctions (Passchier et al. 2016; Goscombe et al. 2017), demonstrate that Gondwana amalgamated through the collision of multiple blocks. However, the various belts do not eliminate the possibility that there was a previous connection between larger blocks or linked smaller fragments.

The oldest western Gondwana orogens, grouped in the 670–575 Ma interval, include mostly the West Gondwana Orogen (Hoggar, Dahomey, Oubanguides, Borborema north and Brasilia belts; Ganade de Araujo et al. 2014a, 2016) and the Dom Feliciano belt (in southern South America; Philipp et al. 2016; Fig. 15.1). Some authors suggest that these connect in one superorogen (Oriolo et al. 2016, 2017). The Hoggar–Dahomey belts are products of convergence among the West African and Saharan cratons (Fig. 15.1). In the northern Borborema province (the South American counterpart), it is not clear which are the colliding blocks. To the east, the Borborema province is composed of Paleoproterozoic blocks reworked in the Brasiliano events, the São Francisco-Congo Craton representing one of these blocks. To the west, underneath the Paleo-Mesozoic Parnaiba basin, the mostly inferred Parnaíba cratonic block possibly represents the other colliding block (Daly et al. 2014; Fig. 15.1). Continuing to the south, the northern sector of the Brasilia belt represents the collision between the São Francisco Craton western margin with the Goiás block (Pimentel 2016; Fig. 15.1). Further south the Paranapanema Craton (mostly covered by the Paraná Paleo-Mesozoic basin) is the counterpart for the agglutination and generation of the southern Brasilia belt (Fig. 15.1). The Dom Feliciano belt is a product of the intereaction between Paranapanema, Rio de La Plata and Kalahari blocks, and seems to have an important period of magmatism c. 650–590 Ma, attributed to collision, as discussed in item 3.

A smaller group of c. 600–550 Ma western Gondwana orogenic belts are characterized, some on the fringes of the 670–590 belts. The Sergipano-Oubanguides collisional belt accomplished the suturing of the northern São Francisco Craton margin and the Saharan Metacraton at this time (Fig. 15.1). On the southern São Francisco margin, the Araçuai-West Congo Orogen, plus the Central and South Ribeira belts, have a more enigmatic evolution. It is agreed that there is a c. 600–550 Ma metamorphic peak event, but the cause is still controversial (Pedrosa-Soares et al. 2008; Degler et al. 2017; Bento dos Santos et al. 2010; Richter et al. 2016). Some propose that the metamorphism is related to arc emplacement in a subduction environment (Duffles et al. 2016), while others propose a collisional setting (Heilbron et al. 2004; Vinagre et al. 2014). An alternative model suggests an intracontinental setting and a major extension that produced the large granitic batholiths and hence the regional metamorphism (Meira et al. 2015).

The period 560–510 Ma comprises the largest Neoproterozoic-Cambrian peak of collisional metamorphism and tectonic activity (Fig. 15.3). In western Gondwana these belts are related to the approximation of two major blocks: the Kalahari in the south and the Amazonia in the north (Fig. 15.1). The Rockelides-Araguaia belts register the collision of the Amazonia Craton suturing western Gondwana. Tohver et al. (2012) and Trindade et al. (2006) attribute this collision to the closure of the Clymene Ocean, after a long-lasting subduction zone, with a hidden magmatic arc below the cover of the Parnaiba Basin (Fig. 15.1). Cordani et al. (2013a) advocate that this belt is the result of the closure of an intracontinental basin. Towards the south the Araguaia belt merges into the Paraguai belt that runs until the Rio Apa cratonic inlier (Fig. 15.1). The metamorphic ages here are younger, coeval with the Pampean belt and the Puncoviscana belt, possibly related to the collision of the Pampia and Arequipa-Antofalla blocks (Rapela et al. 2002; Ramos 2008; Ramos et al. 2010; Escayola et al. 2011; Fig. 15.1) or the final closure of the Clymene Ocean separating the Paranapanema from Rio Apa (and Amazonia; McGee et al. 2015a, b). These younger belts of c. 530–480 Ma represent the transition between the internal suturing Gondwana belts and the external marginal belts.

In the central part of western Gondwana there is also a generation of Cambrian belts (530–480 Ma) fringing the Ediacaran-Cambrian domains (560–510 Ma; Brito Neves et al. 2014). There are two major ramifications of orogenic systems: the Damara-Lufilian-Zambesi and the Ribeira-Kaoko-Dom Feliciano-Gariep-Cuchilla Dionisio (named here the South Atlantic orogenic system, detailed on item Sect. 15.3). The former has recently received extensive attention in the literature with metamorphic and geochronological data (Goscombe et al. 2017). The Damara Belt is a classical product of convergence between the Kalahari Craton and the southern Congo Craton, named the Angola block (some authors even suggest that this block had independent kinematics, not linked with Congo cratonic blocks; Heilbron et al. 2008; Fig. 15.1). Evidence suggests at least two hypotheses for the origin of the Damara Belt: (1) an intracontinental orogeny with a small ocean formed during an extensional period (Porada 1979; Nascimento et al. 2017); and (2) a large ocean separating both cratons (de Waele et al. 2003; Johnson et al. 2005; Schmitt et al. 2012). The conclusion is key to the understanding of eastern and western Gondwana amalgamation (see Sect. 15.4).

3 Ediacaran-Cambrian Orogens in SW Gondwana—the South Atlantic Orogenic System

SW Gondwana was constructed as a result of the interaction between the Kalahari, Rio de La Plata, Southern São-Francisco-Congo, Paranapanema and Luis Alves paleocontinents (Figs. 15.1 and 15.4). Our review focuses on the orogens that are younging towards today’s South Atlantic continental margins, registering Ediacaran-Cambrian tectonic events (Fig. 15.4). Although controversial at some points (Will and Frimmel 2018; see Sect. 15.4), it is certain that the South Atlantic break-up followed the youngest sutures of Gondwana’s amalgamation (Schmitt et al. 2016b). Therefore we propose the name South Atlantic Orogenic System for the following belts in conjunction: Gariep-Saldania-Cuchilla Dionisio-West-Kaoko-Dom Feliciano-South and Central Ribeira and Angola.

Fig. 15.4
figure 4

SW Gondwana with cratons, and post-670 Ma orogens, major crustal-scale structures and oceanic-derived units. The letters and numbers on the map stand for the Gondwana cratons and mobile belts, respectively. They are AM Amazonia; PR Parnaíba; SF São Francisco; PP Paranapanema; LA Luís Alves; RA Río Apa; RP Río de La Plata; KH Kalahari; CC Congo. 1—Pampeana; 2—Caapucú High; 3—Paraguay; 4—Araguaia; 5—Borborema (South); 6—Rio Preto; 7—Brasília; 8a—Araçuaí (West), 8b—Ribeira (Paraíba/Embu) and 8c—Apiaí to East Araçuai and Occidental Terrane; 9—Ribeira (Oriental Terrane); 10—Cabo Frio Tectonic Domain; 11—Curitiba Terrane; 12—Paranaguá Terrane; 13—Dom Feliciano; 14—Cuchilla Dionísio; 15—Nico Pérez; 16—Mar del Plata Terrane; 17—Saldania; 18—Gariep; 19—Damara; 20—Kaoko (Coastal); 21—Kaoko (Central-East); 22—Angola; 23—West Congo; 24—Lufilian; 25—Zambesi. TKL stands for Transbrasiliano-Kandi Lineament

Full size image

Even though separated today by more than 6000 km of ocean, these belts share some important features that hide the final amalgamation of Gondwana: (1) the tectonometamorphic evolution occurred between 670 and 480 Ma; (2) most of the belts present mafic-ultramafic lithostratigraphic units as tectonic slivers; (3) all belts show contraction structures and evidence for crustal thickening; and (4) all belts present extensional Cambrian structures and late Ediacaran-Cambrian basins.

Below we briefly describe the evolution of this orogenic system through the Ediacaran-Cambrian.

3.1 670–575 Ma Orogens

The Dom Feliciano is the only belt developed during this interval in SW Gondwana that corresponds to the Cryogenian to early Ediacaran period. It is mostly a granitic domain that extends from Uruguay (Aigua Batholith) to south Brazil (Pelotas-Florianópolis batholiths; Fig. 15.4; Bitencourt and Nardi 2000; Florisbal et al. 2012; Philipp et al. 2016). It is predominantly composed of calcalkaline magmatic rocks, with shoshonitic and alkaline terms (Lara et al. 2017). Xenoliths of Paleoproterozoic and Tonian gneisses are present. This igneous unit also intrudes Tonian metavolcanosedimentary sequences (Gruber et al. 2011). The emplacement is controlled by low-angle structures with west vergence that evolve to steep northeast–southwest shear zones segmenting the belt after c. 610 Ma (Martil et al. 2017).

The generation of the batholiths is considered to be related to a syn to post-collisional setting at c. 650–550 Ma (Oyhantçabal et al. 2011; Philipp et al. 2016). This collision between the Kalahari and Rio de La Plata cratons would be the consequence of the Adamastor ocean closure along a southeast dipping subduction zone (Basei et al. 2008). Its correspondent in Africa is interpreted by some to be the 660–620 Ma Coastal Terrane, which is considered to be an exotic domain in the Kaoko belt thrust over the Congo-Angola Craton margin at c. 590 Ma (Goscombe and Gray 2007; Figs. 15.1and 15.4). In agreement with the Brazilian counterpart it is also interpreted as a product of southeast subduction (Goscombe et al. 2017).

In Figs. 15.1 and 15.4, the Dom Feliciano belt is represented as a domain with peak metamorphism at c. 640–590 Ma, coeval with the Brasilia Belt. However, its surrounding terranes register the influence of a younger metamorphic tectonic event, well represented in the Cuchilla Dionisio, Eastern Kaoko, Damara, Gariep and Saldania belts (Figs. 15.1 and 15.4). The 585–485 Ma orogenic events that fringe the eastern sector of the Dom Feliciano belt are largely overlooked when authors interpret this domain as the product of the collision between the Kalahari and Rio de La Plata Craton (see Sect. 15.4).

Further north, the Ribeira belt contains a series of 640–590 Ma arc-related batholiths (Campanha et al. 2008; Faleiros et al. 2011; Tupinambá et al. 2012; Vinagre et al. 2014; Fig. 15.4). The southwestern segment is interpreted as having developed due to a northwest dipping subduction zone (Campanha et al. 2015), while to the northeast a southeast dipping subduction is proposed (Heilbron et al. 2004; Trouw et al. 2013). The work of Heilbron et al. (2004) and Trouw et al. (2013) considers a collisional phase, regarding the docking of the magmatic arc domain onto the São Francisco Craton, and the end of this subduction at c. 580–560 Ma. A similar period is attributed to the final stage of subduction in the southwestern sector (Apiaí Terrane) due to collision between the arc terrane and the Curitiba terrane (Campanha et al. 2008; Faleiros et al. 2016; Fig. 4).

3.2 575–480 Ma Orogens

The best representative of the Ediacaran-Cambrian orogens is the Damara belt (Fig. 15.4). The convergence between the Kalahari Craton and the southern Congo Craton (Angola Block) resulted in a collisional event starting at c. 550 Ma with late collisional plutons at c. 530–500 Ma (Schmitt et al. 2012). Lehmann et al. (2016) and Passchier et al. (2016) suggested that precollisional convergence started as early as c. 590 Ma, based on Ar–Ar data in biotite, with the metamorphic climax in the central part of the orogen at c. 530 Ma related to the final collision (Schmitt et al. 2012; Goscombe et al. 2017).

Today the Damara belt is an inland branch of the Gondwana-forming orogens (Miller 2008; Nascimento et al. 2016). The South Atlantic continental margins also show younger metamorphic/collisional ages, represented by the Gariep Belt, Saldania Belt, Cuchilla Dionisio Terrane, Kaoko Belt, Paranaguá Block, Oriental Terrane (Ribeira Belt), Cabo Frio Tectonic Domain (Buzios Orogen) and Angolan belt (Figs. 15.1 and 15.4).

These Ediacaran-Cambrian domains have several common features: (1) medium- to high-grade metamorphic rocks; (2) mafic-ultramafic lenses tectonically interleaved with supracrustals; (3) ductile compressional structures; and (4) late extensional structures with syntectonic Cambrian to Ordovician magmatic intrusions.

The Cabo Frio Tectonic Domain is the only one that shows high P-T metamorphism dated at c. 530 Ga, interpreted as being related to the collision between the Angolan Block and the southern São Francisco Craton (Fig. 15.4; Schmitt et al. 2016b). These metamorphic conditions were attained during the collisional-exhumation phase by supracrustal sequences from a basin developed on transitional crust between the continent and the ocean (Schmitt et al. 2008; Fernandes et al. 2015; Capistrano et al. 2017).

The mafic-ultramafic rocks of the Forte de São Mateus unit are considered to be relics of an oceanic crust, dated at c. 610 Ma (Schmitt et al. 2008, 2016b). Other mafic-ultramafic occurrences are (Fig. 15.4) the Araçuai belt (Pedrosa-Soares et al. 1998), Pien Suite (in between the Luis Alves Craton and Curitiba Terrane; Cury 2009), Paso del Dragon Complex and Arroio Grande Complex (in the northern Cuchilla Dionisio Terrane; Bossi and Gaucher 2004; Ramos and Koester 2014), and Marmora Terrane (in the Gariep Belt; Frimmel and Frank 1998). The Damara Belt contains the Matchless Amphibolite, interleaved with accretionary prism sedimentary and volcanoclastic deposits (Kukla and Stanistreet 1991; Meneghini et al. 2014). These are interpreted as representing oceanic crust formed in a precollisional stage, and most of them are attributed to a back-arc basin setting (Will and Frimmel 2018). Some ultramafic occurrences are alternatively interpreted as exhumed subcontinental lithospheric mantle (e.g., Paso del Dragon Complex in Uruguay; Benedek et al. 2017). Regardless of the origin, this 610–570 Ma extensional phase is also well recorded in the older domains, such as the Dom Feliciano belt (Camaqua Basin), the Luis Alves Craton (Itajai Basin), the Paranapanema Block (Castro Basin) and the Oriental Terrane of the Ribeira Belt (Pico do Itapeva Basin; Riccomini et al. 1996; Almeida et al. 2012; de Oliveira et al. 2014), and even in eastern Gondwana sectors, as in Madagascar (Collins 2006).

The tectonic inversion of these basins started with convergence at c. 580 Ma and culminated with continental collision between 560 and 520 Ma (Fig. 15.3). The oceanic crust from the Buzios basin (in the Cabo Frio Tectonic Domain) is interpreted to be subducted towards the northwest-west, based on the high P-T conditions of the sequence (Schmitt et al. 2016b). A 570–540 Ma igneous domain in the Oriental Terrane of the Ribeira belt (to the northwest) is considered to be representative of a magmatic arc related to this Cambrian subduction (Martins et al. 2016). On the other hand, to the south, the Marmora terrane, obducted at c. 545 Ma onto the Kalahari Craton, is considered to be an oceanic unit of a back-arc basin developed in a post-collsional setting (Frimmel and Frank 1998; Will and Frimmel 2017).

Widespread ductile deformational phases are related to convergence and collision in all of these South Atlantic belts from 580 Ma until 530 Ma, coeval with high- to low-grade metamorphism, which varies according to the exposed crustal level and proximity to the suture zones. Low-angle structures predominate in the Cabo Frio Tectonic Domain, Angolan Belt, Apiaí and Curitiba terranes, Gariep belt and the external parts of the Damara and Kaoko belts.

High-angle structures related to steep shear zones are common in the Dom Feliciano Belt, Central Ribeira Belt and Western Kaoko Belt, mostly attributed to late collisional strain zones, developed during the final amalgamation of the cratons (Oyhantçabal et al. 2010; Konopásek et al. 2016; Oriolo et al. 2016; Passchier et al. 2016; Philipp et al. 2016). The age of this last kinematic activity of these mylonitic zones is constrained to the period 580–480 Ma. These transpressional and transtensional kinematics are probably the result of lateral adjustments of the multiple collisions. Some authors try to make transatlantic correlations between the shear zones (de Wit et al. 2008; Konopásek et al. 2016). Although they are subvertical structures, the connection is very speculative since the opening of the South Atlantic produced wide continental margins and probably a rifted microcontinent (Rio Grande Rise; Santos et al. 2014; Szatmari and Milani 2016). Approximately 300–500 km of continental crust is estimated on both conjugate margins submerged in this Atlantic sector (Schmitt et al. 2016b).

Magmatism at c. 580–550 Ma is related to these shear zone systems in the Dom Feliciano Belt, Cuchilla Dionisio and the Coastal Terrane of Kaoko Belt (e.g., Oyhantçabal et al. 2011; Konopásek et al. 2016). In the northern sector of the South Atlantic belt, a pre-eminent 540–440 Ma magmatic event intrudes into these high strain zones (Martins et al. 2016). During their reactivation in a transtentional phase, 510–440 Ma plutons intrude into this shear zone system (Bongiolo et al. 2016).

4 Discussion

The main contribution of this chapter is to present a compilation of all the collisional orogens/metamorphic belts that sutured Gondwana in the late Ediacaran Early Cambrian (Fig. 15.1). The timeframe attributed to orogenic activity in each of the 55 belts comprises the metamorphic peak and deformational structures representing the continental collision phase, which includes the extrusion of nappes and development of steep shear zones (Figs. 15.1 and 15.2). Two main periods of orogenic activity related directly to the amalgamation processes can be recognized (Fig. 15.3). The first period, starting c. 670 Ma, includes relatively few orogens (c. 15) but involves larger areas in Gondwana than subsequent orogenesis. These orogens are shown in orange and dark green in Fig. 15.1. This period amalgamated the Saharan, West African, São Francisco-Congo and Paranapanema paleocontinents, and also the juvenile Tonian terranes of the Arabian Nubbian shield and some microcontinents of the East African Orogen. This c. 575 Ma proto-Gondwana was the core of the supercontinent, centred on Africa (Fig. 15.2b).

The second and last period of Gondwana amalgamation includes a vast number of orogens, c. 40 (Fig. 15.3). These, shown in light- and very light-green, were responsible for suturing the Amazonia, Rio de La Plata, Kalahari, Dhawar, East Antarctica and Australian paleocontinents (Fig. 15.1). The external (marginal) orogens overlapped in time with this second amalgamation stage. They are the Pampean, Ross and Delamerian belts (also plotted in Fig. 15.3).

4.1 Correlating Ediacaran-Cambrian Orogens Throughout Gondwana

The distribution of the collisional orogens throughout western and eastern Gondwana is similar (Fig. 15.2), with coeval and interleaving events in the period 670–480 Ma (Figs. 15.1 and 15.3). This pattern indicates that although Gondwana was constructed by the convergence of several paleocontinents and microcontinents (Neoproterozoic Cratons) with different geological evolutions, their approximation might be orchestrated by global geodynamics (Hoffman 1991; Condie et al. 2015; Merdith et al. 2017a). The two stages of Gondwana amalgamation interpreted here, 670–575 Ma and 575–480 Ma, represent periods with abundant collisions in all sectors of Gondwana (Fig. 15.2). The younger stage has many more orogens (c. 40 orogens) than the first stage (c. 15 orogens; Fig. 15.3).

The separation was arbitrarily drawn at c. 580–570 Ma (Fig. 15.3). These 10 myr of low orogenic activity match with the opening of extensional/transtensional basins (e.g., SW Gondwana; Riccomini et al. 1996; Almeida et al. 2012; de Oliveira et al. 2014). These Ediacaran rifting events, sometimes forming oceanic crust (e.g., Buzios, Cuchilla Dionisio, Gariep; Fig. 15.4), probably indicate a period when the convergence rate slowed down. These basins were rapidly formed and even more quickly inverted in the second and final stage of Gondwana amalgamation. One important global tectonic event that coincides with the beginning of this second stage is the opening of the Iapetus ocean (Nance et al. 2014).

The 575–480 Ma collisional orogens that sutured Gondwana are recognized in the east–west Kuunga and north–south South Atlantic belts (Fig. 15.1). The intersection zone between these composite belts is located in the central part of SW Gondwana, today in the South Atlantic (Fig. 15.4). In this zone, the final amalgamation was completed by the Kalahari and Rio de La Plata collision with the São Francisco-Congo Craton. This was coeval with the development of the marginal orogens of the Pampean-Ross-Delamarian events (Fig. 15.1).

4.2 Do the Internal Western Gondwana Ediacaran-Cambrian Orogens Represent Closure of Oceanic Realms?

The Ediacaran-Cambrian South Atlantic rogenic system is here considered to be the record of the final suturing of western Gondwana. These orogens share certain features that might indicate that there was mantle exhumation and/or generation of oceanic crust at c. 610–590 Ma (Fig. 15.4). The sediments deposited in these basins present a major Ediacaran provenance (Basei et al. 2005; Foster et al. 2015; Fernandes et al. 2016) mostly from volcanic and plutonic sources. These igneous domains are interpreted as continental magmatic arcs (Basei et al. 2008; Heilbron et al. 2008; Gaucher et al. 2009; Goscombe et al. 2017) ovelapping partially in age with syn- to post-collisional intrusions (Oyhantçabal et al. 2011; Florisbal et al. 2012) and extensional intraplate magmatism (Meira et al. 2015). Most authors agree that these basins were developed in a back-arc environment at the end of southeast-directed subduction (Basei et al. 2008; Heilbron et al. 2008; Gradim et al. 2014).

The map shows that the reality is more complex, exposing flaws in this model that should be reviewed (Fig. 15.4). The back-arc basin model apparently works well for the Marmora basin-Cuchilla Dionisio Terrane, but, northwards, in the Apiaí Terrane and the Cabo Frio Tectonic Domain, evidence suggests that the major metamorphic peak, probably related to continental collision, is much yonger, c. 570–520 Ma (Schmitt et al. 2004, 2008, 2016b; Cury 2009; Faleiros et al. 2011). High- to medium-pressure rock units, identified in both domains, are not compatible with a back-arc setting.

Transform limits are usually not considered in Neoproterozoic-Cambrian tectonic models. The opening of the Khomas sea (Damara belt), Adamastor Ocean (Dom Feliciano belt) and Buzios Ocean (Cabo Frio Tectonic Domain) is geometrically plausible including transform limits, which are intrinsic to oceanic crust formation. However, the closure of all these oceans/seas within a short period implicates a complex array that would need to be resolved with triple junction migration along subduction zones (e.g., Passchier et al. 2016).

In addition, the evolution of the Damara belt is well constrained to the period 570–500 Ma with a major metamorphic peak and magmatism around 540–530 Ma (Miller 2008; Goscombe et al. 2017). It shows that the Kalahari-southern Congo main frontal collision was coeval approximately with the Buzios orogen (Cabo Frio Tectonic Domain; Schmitt et al. 2004), the Apiai-Embu orogenic domains (Cury 2009; Faleiros et al. 2011), and the deformation and metamorphism in the Cuchilla Dionisio-Gariep-Saldania domains/belts (Bossi and Gaucher 2004; Gaucher et al. 2009; Frimmel et al. 2013).

A major 620–600 Ma collision in the Dom Feliciano belt is not compatible with the evolution of the fringing post-585 Ma belts (Fig. 15.4). The reconstruction shown in the SW Gondwana cratons-belts maps suggests a connection between the Cuchilla Dionisio and the Damara-Kaoko orogenic junction (Fig. 15.4). This 570–500 Ma orogenic system strongly reworks the Coastal Terrane of the Kaoko Belt (Foster et al. 2015; Goscombe et al. 2017; Nascimento et al. 2017). On the South American side, the Florianopolis batholith does not offer published data relating to this younger orogenic activity, though the Paranagua Terrane (still with very limited geochronological data) has Cambrian syntectonic magmatism, and the Curitiba-Apiai terranes also show this younger metamorphism (Cury 2009).

We propose that the collision on the Damara belt developed coevally with the Cuchilla Dionisio-Saldania-Gariep-Dom Feliciano-Kaoko-Ribeira-Cabo Frio orogenic system between 570 and 500 Ma, including the Mar del Plata Terrane (Rapela et al. 2011; Fig. 15.4). This South Atlantic orogenic system contains the main SW Gondwana suture that was reactivated in the extensional setting 350 myr later. Indeed, most of the the South Atlantic opening follows the rule of suture inheritance, with old orogenic high-pressure and oceanic-derived rocks preserved on its actual continental margins (Beltrando et al. 2010, 2014). In disagreement with the interpretation of Will and Frimmel (2018), we consider that back-arc basins are not likely to favour continental break-up since they usually do not represent major lithospheric boundaries.

In addition, the recent discovery of offshore continental units, 1000 km off the Brazilian coast at the Rio Grande Rise (Santos et al. 2014), reinforces the idea that a large volume of Gondwana lithosphere is hidden in the continental margins and even as remnants encrusted on the Atlantic oceanic floor. Therefore the c. 630–600 Ma Dom Feliciano Belt in southern South America and its probable counterpart Coastal Terrane in Namibia might represent the external limits of a belt that only terminated its evolution in the Cambrian.

5 Conclusion

The compilation of 55 orogenic belts that sutured Gondwana in the period 670–480 Ma provides the following conclusions:

  • Two main orogenic stages are recognized. The first, 670–575 Ma, includes few orogens (c. 15) but larger areas. During this stage, Saharan, West African, São Francisco-Congo and Paranapanema paleocontinents, Arabian Nubian shield juvenile Tonian terranes and some East African Orogen microcontinents were amalgamated, forming the proto-Gondwana core.

  • The second stage, from 575 to 480 Ma, incorporates more orogens, c. 40, suturing the Amazonia, Rio de La Plata, Kalahari, Dhawar, East Antarctica and Australian paleocontinents.

  • The collisional orogen pattern throughout Gondwana is similar, indicating that Gondwana was built up by the convergence of distinct paleocontinents. This comparable pattern regarding metamorphic and deformational peaks suggests that Gondwana amalgamation processes, although complex, were partially orchestrated by global geodynamic forces.

  • In SW Gondwana the South Atlantic system of Ediacaran-Cambrian orogens is here considered to register the final suturing of western Gondwana.

  • The belts share certain features, such as the opening of basins, coinciding with the interval between the two main orogenic stages. Mafic-ultramafic sequences might indicate that there was mantle exhumation and/or generation of oceanic crust at c. 610–570 Ma.

  • These basins were rapidly formed and even more quickly inverted during the 575–480 collisional stage, represented by the major east–west Kuunga and north–south South Atlantic belts.

  • We propose that the 570–500 Ma convergence and collision of the Damara belt was coeval with the Cuchilla Dionisio-Saldania-Gariep-Dom Feliciano-Kaoko-Ribeira-Cabo Frio orogenic system. This South Atlantic orogenic system holds the key to the SW Gondwana suture, reactivated 350 myr later, evidenced by old orogenic high-pressure and oceanic-derived rocks preserved in the present conjugate continental margins.