The Brasiliano collage in South America: a review

Manuscrito ID 30148. Recebido em: 05/08/2014. Aprovado em: 29/08/2014. ABSTRACT:Analysis of recent geological and geochronological data from the basement of the South American platform indicates that the Brasiliano orogenic collage took place in four distinct pulses: a) Early Cryogenian (ca. 800 – 740Ma); b) Late Cryogenian-Early Ediacaran (ca. 660 – 610 Ma); c) Early-Middle Ediacaran (c. 590 – 560 Ma); and d) Late Cambrian (520 – 500 Ma). e rst three pulses are well represented in most Neoproterozoic structural provinces in West Gondwana. e youngest orogenic phase/pulse, however, is only seen in Argentina (Pampean Orogeny) and Brazil, in eastern Rio de Janeiro State (Búzios Orogeny). e period between ca. 750 and 500 Ma is comparable to that reported for the amalgamation of various continental fragments in East (Arabian-Nubian, Mozambique, Kuunga) and North Gondwana (Cadomian). However, important di erences in the nature and ages of events are recognized, which can be expected in view of the magnitude of Gondwana agglutination and the diversity of paleogeographic and tectonic scenarios. West Gondwana shows an interesting peculiarity: lithologically and tectonically diversi ed Tonian terranes underlie Brasiliano orogenic buildups. ey were strongly reworked during most of the orogenic pulses. e Tonian terranes (1000 – 900 Ma) and their relation with Rodinia or with the processes of Gondwana fusion remains an open question. Indications of their presence in East Gondwana are still poorly documented.


INTRODUCTION
Over the last ve decades the word "Brasiliano" has been loosely used to refer to many aspects of the Neoproterozoic geology of Brazil (e.g. cycle, orogeny, tectonic evolution, metamorphism, mobile belts, cratons, etc.).
When the rst formal de nitions were proposed (Cordani et al. 1968, Almeida et al. 1973Almeida e Hasui 1984), the term appeared as "Brasiliano cycle" ("Ciclo Brasiliano"), comprising the period from 700 to 450 Ma. ese authors separated it from the pre-Neoproterozoic cycles, also de ned in the early 70s (e.g. the "Uruaçuano", "Espinhaço", and "Transamazonian" cycles). ese seminal concepts were heavily based on Rb-Sr and K-Ar geochronology, the only methods available at that time. e poor dating was made worse by the non-uniform availability of the data over the South American territory. ese de nitions were also strongly biased by poor geological and geotectonic knowledge, associated to the need to establish the timing of geological/geotectonic processes in South America (see a synthesis of original de nitions in Brito Neves et al. 1990, and in Tab. 1).
With the progress of systematic geological mapping, strongly supported by recent aerogeophysical surveys, and introduction of new isotopic and geochronological methods, especially Sm-Nd and U-Pb (Pb-Pb evaporation, TIMS, LA-ICPMS, SHRIMP), those concepts adopted during the last decades of the past century need to be reviewed and updated. As new data is being rapidly produced, new reviews may be necessary in the near future The misuse of one of the older designations, the "Transamazonian" cycle (ca. 2,100 -1,800 Ma), is persistent in Brazilian literature. With the improvement of geological knowledge in the last decades, it has been recognized that all systems of the Paleoproterozoic era are well represented in the continent (Brito Neves, 2011). erefore, choosing a single unifying rack (the "Transamazonian") to include all the di erent systems is both undesirable and inconsistent, although understandable in the past. Even for Amazonia, in northern South America, the accumulated geological knowledge -according to Santos 2003;Tassinari and Macambira 2004;Rosa Costa et al. 2006;Cordani and Teixeira 2007, among many others -is sufcient to place the initial concept of "Transamazonian" in a untenable position, since isotopic determinations originally labeled as "Transamazonian" (ca. 2,000 Ma) are restricted to a small area in the northeastern part of the Amazonian craton. Likewise, using the same designation in other basement blocks (e.g. São Francisco Craton, Rio de La Plata Craton, São Luis Craton, basement of Neoproterozoic provinces, etc.), is inadequate, given that these blocks, or plates, were -probably -distant from each other, and unlikely to have shared a single cycle of plate interaction. e "Uruaçuano cycle" (ca. 1,300-1,100 Ma) proposal was consolidated in the same publication of Almeida et al. (1973), although it was present in several earlier studies, in other structural provinces of South America. is was an interpretive mistake of geological and isotopic data (mostly Rb-Sr and K-Ar), which gradually disappeared from the scienti c archive of the continent, due to the growth in the geological knowledge of the region.
At that time, it was a common practice to take a few groups of age and "validate" them as a tectonic "cycle".
is was done without the necessary commitment to paleogeographic and paleotectonic scenarios. Concepts such as "cycle", "phase", "orogeny", and "tectonic event" were used loosely, without commiting to a ruling principle. Additionally, there was the widespread unduly idea of a continent (South America) always acting as a large monolithic continental mass, which is not adequate.
As far as the Brasiliano "cycle" and the Neoproterozoic structural provinces are concerned, new information is continuously being produced. A few subdivisions have been proposed (e.g. Schobbenhaus & Brito Neves 2003;Silva et al. 2005), with very good scienti c support, providing valuable aid to the approach proposed in the present paper.
Scienti c progress is instrumental to challenge inadequate concepts and subdivisions. In addition to that, advances in geological knowledge have gradually demonstrated the existence of other major tectonic events associated to the concept of interaction of several paleocontinents (e.g. those from the Siderian, Tonian, Cambrian, among others), barely hypothesized or imagined during the last quarter of the past century. e interesting question posed by Collins et al. (2011), in the XIV Gondwana Symposium, about the nal assembly of Gondwana ("Where and when did Gondwana nally amalgamate?") seems to have a good answer based on the most remarkable record in South America. A review of the di erent main phases (pulses) of convergent plate interactions during the Neoproterozoic and the nal fusion of West Gondwana in late Cambrian times are the main objectives of this work.
platform. In the 1970s, several classic papers were published (Almeida 1971(Almeida , 1978Almeida et al. 1973;Cordani 1973), in which the Neoproterozoic orogenic activity was encompassed in the so-called "Brasiliano" (between 700 and 450 Ma, Almeida et al. 1973). is is a sequence of relevant contributions, each successively taking advantage of the progress of the knowledge (Tab. 1).
Contributions by Schobbenhaus and Campos (1984) and Almeida and Hasui (1984), although sticking to the rules of international committees, and taking hold of the continent as a whole, brought about advances to the proposals of the previous decade. As mentioned above, Brito Neves et al. (1990) presented a synthesis of these classi cations and systematizations (Tab. 1). Subsequent works -by Pimentel et al. 1997;, among others -, brought to light new data from mapping and geochronological data using new isotopic methods, and made it clear that previous schemes could not prevail. From then on, the number of relevant contributions grew and led to rapid progress. e new data represented the starting point for new schemes of geological time subdivision in the continent (Schobbenhaus & Brito Neves 2003;Delgado et al. 2003). In both cases, Neoproterozoic tectonic events were considered with particular care, including the previous proposals to subdivide the "Brasiliano" sensu lato. Silva et al. (2005), considering the development of the Mantiqueira Province, attempted to improve and update the scheme proposed by Delgado et al. (2003). ese contributions provide excellent synthesis of geochronological data available for the Proterozoic of the continent. In addition, they provided the rst attempts to organize the "Brasiliano" geological record in space and time.
De ning orogenic processes, taking into account geological time subdivision and main stages supported by tectonic events, is a complex and demanding exercise, and it needs to be carried out permanently. Timely revisions are necessary to keep track of knowledge advancement.

PLATE INTERACTIONS, OROGENY, OROGENIC COLLAGE, FUSION (ASSEMBLY)
ere is no consensus in the international literature concerning the de nition of the concepts above. ey are simply overlooked, or in some cases, treated frivolously. However, in our view and experience, attempts to understand the geotectonic development of continents, following the initial assumptions of Mitchell and Reading (1969), Dewey and Bird (1970) and Dewey and Burke (1973), as suggested by Helwig (1974), and again by Sengör (1990), is very useful in the systematization of tectonic processes. According to Helwig (1974) and Sengör (1990), orogenesis is de ned as a plate interaction (development of a Wilson cycle), orogenic collage is a concatenated sequence of orogenies, and fusion is the sum of these processes, leading to the formation of a large continental mass or a supercontinent.
As mentioned above, several geoscientists, including authors of textbooks, do not follow any robust systematics, and the consolidation of these concepts is usually missingsee Moores and Twiss 1995;der Pluijm and Marshak 2004;Kearey et al. 2009, among other more recent textbooks -, which leaves room for many inferences and counterintuitive "concepts". For instance, in the case of the several orogenies of the Appalachian Mountains involving distinct terranes, arcs and plates, no discrimination is proposed, and the term orogeny persists until the end of a long and complex evolution. Orogenic collage should have been used, since the assembly of Pangaea was the nal result of the processes. Indeed this oversimpli cation ("orogenesis") has been common practice, especially when some linearity of orogenic processes persists. In cases where co-linearity and/or contiguity of successive orogenic processes are absent, and where interacting participants (basement inliers, terranes, "structural highs", shear belts, etc.) are abundant, the simplistic treatment has not been possible or is usually confusing. is is the case of the branching systems of orogens or "mosaic-like folded regions" present not only in Proterozoic terrains (e.g. Neoproterozoic structural provinces in Brazil, Baikalian provinces of Asia, etc.), but also in Phanerozoic terrains (e.g. European Hercynian, Ural-Mongolia, Alps). In every case there is then the necessity to investigate each of the belts, their paleogeographic framework, as well as their lithostructural and isotopic records, and perform comparison exercises in order to achieve meaningful syntheses.
Even recognizing the disregard of formal nomenclature and de nitions in the Geotectonic literature, for which the necessity of formalization equals that of the stratigraphic nomenclature code, we decided to follow the "Helwig (1974)- Sengör (1990)" proposal, aiming to understand the problems presently posed for South America. We hope to establish a basis for the better appreciation of the problem and for its future solution.

Archean
Eoarchean rocks have not been found in South America up to now, although there are some indirect records from Sm-Nd model ages and from rare U-Pb ages of detrital zircon grains.
Exposures of Paleoarchean high-and low-grade rocks are recognized within the main cratonic nuclei (São Francisco, Amazonian), in basement inliers within and between mobile belts, and eventually in detrital zircon grains from Proterozoic metasedimentary rocks. e oldest rocks (> 3.4 Ga) were identi ed in the central and northern domains of the São Francisco craton (Nutman & Cordani 1993), and in the São José do Campestre block (Dantas et al. , 2010(Dantas et al. , 2013, in the Borborema province. Up to this date, there is no su cient data to try to establish workable tectonic subdivisions for this era in South America. Mesoarchean records are far more abundant in cratons, basement of Proterozoic mobile belts, as well as in detrital material of Proterozoic metasedimentary rocks. However, in spite of the available data, it seems still early to suggest subdivisions. Occasionally, some denominations of orogenic nature are usede.g. "Guriense cycle (Schobbenhaus & Campos, 1984) -to distinguish between di erent tectonic/ metamorphic peaks. Although the increasing volume of geological and geochronological data is recognized, formal classi cations should wait until additional data are available. In addition to that, many Mesoarchean terranes were variably reworked during Proterozoic orogenies, requiring caution when making new propositions.
In general, among the Archean eras, Neoarchean records are the most frequent in the basement of the cratons and the Proterozoic mobile belts. e wealth of data has lead researchers to propose some local designations, such as the "Jequié cycle" (Cordani 1973), "Jequié-Aroense" (Schobbenhaus & Campos, 1984), and "Rio das Velhas" (Noce et al. 1998). Such designations express the necessity of knowledge systematization, and are understandable, acceptable, and should be stimulated. However, caution should be exercised when such designations are extended from a speci c geological domain to the whole continent.
With the necessary caution, possible Neoarchean continental nuclei in Brazil should be confronted with what is known about Neoarchean rocks worldwide. Indeed, several neoarchean continental reconstruction scenarios have been suggested in the literature (e.g. Bleeker 2003): ■ a single continent (Kenoran); ■ two or three large continental masses (Sclavia, Superia, Vaalbara); or ■ many di erent masses, plates and nuclei, which appears to be the less probable solution, according to Bleeker (2003).
In our point of view, this xcenario considered as the "less probable solution" (Bleeker, 2003), involving many di erent neoarchean masses, appears to be the most realistic model (taking into consideration the data of the South American Platform).

PALEOPROTEROZOIC
ere are good lithological and structural records of all Paleoproterozoic periods recognized in South America, both from orogenic and anorogenic processes, also including basin-forming tectonics (Brito Neves 2011).
Siderian rock units (2,500 -2,300 Ma) were recently identi ed in the cratonic basement of the Amazonian and Luis Alves cratons, and in basement inliers of Proterozoic belts, including volcano-sedimentary formations (Brito Neves 2011).
e Rhyacian (2,300 -2,050 Ma), mainly in its middle part (ca. 2,200 Ma), is the Paleoproterozoic time interval with the largest number of tectonic, magmatic and sedimentary events recorded in the diverse components of the basement in South America, particularly in its central and eastern sectors.
e Orosirian (2,050 -1,800 Ma) runs second place in terms of record wealth, comprising independent orogenies, all well constrained in their time intervals, but also representing the nal stages of orogenies that started in the Rhyacian. Aside from the orogenic processes, important granite plutonism and abundant volcanic and sedimentary sequences have been identi ed. In eastern Amazonia, these processes were extensive, representing a large igneous province (LIP), covering an area that exceeds 1 million km 2 , in ve di erent countries (Brito Neves 2011; Klein et al. 2013).
During the Statherian (1,800 -1,600 Ma), there is a single orogen in the central sector of the Amazonian craton, known as the Rio Negro-Juruena belt, exposed to the west of an Orosirian belt. Nd-Sr isotopic evidence points to juvenile accretion processes in the origin of this orogen (Tassinari & Macambira 2004;Cordani & Teixeira 2007).
is orogen is believed to continue into North America (Transcontinental-Labrador) and northern Europe (Gothian-Kongsbergian), according to continental reconstructions put forward by several authors (Brito Neves 2005). However, the most outstanding feature of the Statherian in South America is taphrogenesis, with many examples in all sectors of the continent. is is represented by plutonism, volcanism, and sedimentation, including precursors of large intracratonic sedimentary basins ("steer's head basins" or koilogens). In part, these extensive covers continued to form well into the Mesoproterozic, for example the Espinhaço Supergroup in the São Francisco craton. Some of the basins were formed in large rift zones, deformed later on during the Late Mesoproterozoic and Neoproterozoic orogenic processes (e.g Espinhaço, Orós-Jaguaribe, Araí).

MESOPROTEROZOIC
ere are two large geological domains in South America, the Amazonian (or pre-Brasiliano, or north-west domain) and the "Brasiliano" (or central-east domain) ones, according to Brito Neves and Fuck (2013. eir distinction is easily recognized when analyzing Mesoproterozoic and younger rock assemblages (Table 2).
In the Amazonian domain, the Mesoproterozoic (1,600 -1,000 Ma) is represented in several NNW-ESE mobile belts that developed between 1,590 and 1,000 Ma (Tab. 2), organized from northeast to southwest in a chelogenic pattern (Tassinari & Macambira 2004;Bettencourt et al. 2010, among others). e continuity of these belts into North America is a frequent matter of debate (Brito Neves 2005;Bettencourt et al. 2010;Brito Neves & Fuck 2014, among others), especially among researchers investigating the history of supercontinents.
Mesoproterozoic magmatic and sedimentary processes have been recorded within the Brasiliano domain. However, so far, no true mobile belt has been recognized, except, perhaps, for the Ceres-Rialma belt, exposed along the high course of the Tocantins River, in central Brazil (Moraes et al. 2004(Moraes et al. , 2006. Most units generated during the Mesoproterozoic were strongly a ected by Neoproterozoic orogenic events. Some restricted cratonic covers escaped this, for instance the central part of the Chapada Diamantina Group, in the São Francisco craton (Süssenberger et al. 2014). Several examples of Mesoproterozoic rift deposits were involved in younger Brasiliano mobile belts, many hundreds of million years later (e.g. part of Espinhaço, Orós-Jaguaribe, Araí, etc.).
Available geological and geochronological information indicate that the Mesoproterozoic time scale for the successive orogenic events in Amazonia is distinct from that recorded in the extra-Amazonian portion of the continent (Tab. 2). It is not an overstatement to say that the end of the Mesoproterozoic marks the end of the history of Precambrian orogenies in Amazonia, since Neoproterozoic events are scarce (Brito Neves & Fuck 2013). On the other hand, Neoproterozoic systems are of utmost importance outside Amazonia, in the so-called Brasiliano domain.

NEOPROTEROZOIC
Tonian Terranes/Orogenies e progress of the geological knowledge arising from 1:100.000 and larger scale mapping of almost the whole area of the Brasiliano domain and also from abundant isotopic studies, led to the establishment of a much more complete scenario of crustal evolution than that described by the pioneering authors mentioned above.
Tonian terranes, in the sense of Howell (1995), have been identi ed in all Neoproterozoic structural provinces of South America. ey usually represent fractions of Early-Middle Tonian accretionary -some juvenile in nature -and collisional orogens, including volcanic-sedimentary associations metamorphosed under low-to high-grade conditions. Rock assemblages, typical of active continental margins, magmatic island arcs and remnants of ocean oor have been recognized in these units (e.g. Pimentel & Fuck 1992;Pimentel et al. 1997Pimentel et al. , 2000Pimentel et al. , 2004Pankhurst et al. 2008;Santos et al. 2010;Siga Jr. et al. 2009Van Schmus et al. 2011). It is not always possible to reconstruct the paleogeography of these terranes, as they were disrupted during subsequent Neoproterozoic orogenies. In all cases, it has been hard to discriminate them from younger Neoproterozoic lithostructural and tectonic units, and they have frequently been interpreted as basement inliers. In general, time lapses of hundreds of million years have been recognized between these Tonian terrains and the younger Neoproterozoic processes. e plate con gurations and duration of these periods still need to be accurately established.   Table 2. Generalized scheme of evolution and main events of the Amazonian (Pre-Brasiliano, north and northwest parts) and Brasiliano (central and eastern parts) domains of the basement of the South American platform. Based on Saes and Leite (1993), Geraldes et al. (2000), Ruiz et al. (2004), Tassinari and Macambira (2004), Bettencourt et al. (2010), Rizzotto and Hartmann (2012), Rizzotto et al. (2013), Brito Neves and Fuck (2013)

Borborema province (NE South America)
Tonian terranes, apparently the largest in the continent, occupy wide areas within the Borborema Province ( Fig. 1). ey extend along ca. 1,000 km from the north of João Pessoa to SW Piauí, forming a belt up to ca. 100 km wide (the so-called Alto Pajeú terrane + Riacho Gravatá "subterrane") and in some other minor occurrences in the Transversal (central) and southern domains of the province. Volcanic-sedimentary associations, including felsic, ma c and restricted ultrama c rock associations, as well as orogenic and collisional granites, are dominant. In some cases, arc magmatism displays juvenile geochemical and isotopic signatures (Accioly et al. 2010;Santos et al. 2010;Van Schmus et al. 2011). ese rocks formed largely between 1,050 and 930 Ma. Tonian rocks are also exposed in the north (Belém do São Francisco, Cabrobó) and south (Marrancó, Poço Redondo) of the Pernambuco-Alagoas massif, in the northern portion of the Sergipano belt, between the post-Cryogenian supracrustals and the Paleoproterozoic basement of the massif (Oliveira et al. 2010). A wealth of data is available for the Borborema Tonian rock assemblages Van Schmus et al. 2011;Caxito et al. 2014aCaxito et al. , 2014b. According to De Wit et al. (2005, 2008, it is likely that these Tonian units of the Borborema Province have also been recognized in the Pan African domains, from Cameroon to the south of Sudan.

Tocantins province
e Goiás Magmatic arc is one of the most important tectonic features of the Brasília Belt, in central Brazil. It is exposed for approximately 1,000 km in the NNE-SSW direction along the western part of the belt. It comprises two sectors, separated by the Archean terranes of the Goiás massif (see Pimentel et al. 2004 and Figs. 2 and 3).
e southern sector is known as the Arenópolis arc, and the northern sector, as the Mara Rosa arc (Pimentel et al. 2000;Laux et al. 2004Laux et al. , 2005. In both sectors, early evolution of the arc took place in Tonian intra-oceanic island arcs, between ca. 930 and 810 Ma, with the crystallization of very primitive tholeiitic to calc-alkaline volcanics and associated tonalites/granodiorites (Fig. 3). e rocks display positive ε Nd(T) values and Nd T DM model ages mostly between 0.8 and 1.1 Ga (Pimentel & Fuck 1992;Pimentel et al. 1997Pimentel et al. , 2000. Geochemical and isotopic data suggest that some of the original tonalite/andesite magmas were similar to modern adakites (Pimentel et al., 1997).
A younger cycle of magmatic activity took place between ca. 670 and 600 Ma (Laux et al. 2005;Fuck et al. 2006;Oliveira et al. 2006). is second cycle also comprises volcanic-sedimentary sequences, calc-alkaline tonalite-granodiorite intrusions, as well as bi-modal intrusions comprised of gabbro/diorite/granite. Sm-Nd data produced T DM values varying in a much wider range, from ca. 0.9 to 2.2 Ga, suggesting a higher contribution/contamination with older continental crust and indicating that this second event took place in a continental arc. Metamorphism under greenschist to amphibolite facies took place at ca. 630-620 Ma. is was followed by the emplacement of a large volume of post-tectonic bimodal intrusions formed by gabbros and granites dated between ca. 600 and 540 Ma.
Although these two main phases of arc magmatism are recognized on the basis of geochronological and, locally, geophysical data, their spatial distribution is still not clear for most of this geotectonic unit. In a large part, both sectors of the Goiás Magmatic Arc have been strongly a ected by dextral strike-slip shear zones of the Transbrasiliano-Kandi lineament (Schobbenhaus et al. 1975;Cordani et al. 2013b).
is is a striking geologic feature that corresponds to a wide (approximately 200-100 km wide) tectonic corridor of continental dimensions, cutting through most of South America and extending into Africa. e strike-slip faults have disrupted the arc rock units, causing important displacements in the NNE-SSW direction, obliterating the original spatial, structural, and stratigraphic relationships between the two di erent phases of arc magmatism. In addition, younger and older tonalite and granodiorite intrusions, preserved or not from strong deformation, are virtually identical petrographically, hampering their distinction in the eld. erefore, one has to rely on tools such as geochronology and airborne geophysics to distinguish between them. is was partially achieved in the Mara Rosa arc, where an appropriate amount of Sm-Nd isotopic data allied with U-Pb zircon ages seem to indicate that the Tonian rocks (930-800 Ma) dominate the eastern part of the arc (here informally called the Amaralina In the Arenópolis arc (the southern part of the Goiás magmatic arc, Figs. 2 and 3) strike slip faults separate regions where the Tonian juvenile rocks dominate from regions dominated by the Ediacaran phase. However, the available geological and geochronological data do not allow their representation in geological maps at an appropriate scale. Small ultrama c bodies consisting of serpentinite, metaperidotite, and amphibolite are not rare and have been interpreted as remnants of oceanic basins, separating di erent island arc terranes. Similar rocks have also been recognized in the Mara Rosa arc. ey may represent former suture zones between colliding magmatic arc systems, which have been obliterated by later deformational events (Pimentel et al. 2000).

Mantiqueira province a) Italva Klippe/Terrane
Rock assemblages of the Italva terrane, including tonalite and granodiorite gneiss, marble, amphibolite, intercalated with rocks of the Rio Negro arc, occur in a large synform (a probable klippe, Peixoto & Heilbron 2010), within the "Oriental terrane", in the central northern part of Rio de Janeiro (Fig. 4)  superposed trends of the Neoproterozoic orogenic association (Rio Negro arc and surroundings).
Geochemical data of the Italva rocks indicate extensional magmatism in a back arc-type environment, lled up with detritus derived from a neighboring basement (Peixoto and Heilbron 2010). Geochronological data of amphibolites, intercalated with marbles, indicate the crystallization age of 848 ± 11 Ma for the original ma c magmatism (Heilbron and Machado 2003). is is ca. 50 Ma older than the oldest age recorded so far in the Rio Negro arc, and more than 200 Ma older than the main magmatic activity within the magmatic arc. Metamorphism of the Italva rocks took place at higher P-T conditions than that seen in the younger rocks, and their contacts are tectonic in nature.

b) Itaiacoca Group
e Itaiacoca Group occurs in the westernmost area of the Mantiqueira Province, in Paraná and São Paulo. It is a ca. 180 km long and > 15 km wide metavolcanic-sedimentary sequence, exposed between the ca. 620 Ma old Cunhaporanga and Três Córregos magmatic arcs (Fig. 5).
e metamorphic belt is covered in the north and south by Phanerozoic deposits of the Paraná Basin. It is composed of a QPC-type (quartzite-pelite-carbonate) metasedimentary association. Stromatolitic dolomite marble, frequently rich in talc, and phyllite make up the lower part of the sequence, followed by quartzite and metavolcanic rocks, with phyllite at the top (Siga Jr. et al. 2009. e U-Pb zircon age of the Itaiacoca metavolcanics is set at 934 ± 36 Ma, and detrital zircon ages vary from late Stenian to Early Tonian (Siga Jr. et al. 2009. e sequence is thought to represent continental passive margin deposits related to the Paranapanema microplate (Mantovani & Brito Neves 2009). Cryogenian orogenic processes later truncated it, and it is covered locally by metaconglomerate, metarkose, metagraywake and metavolcanics of the Cryogenian Abapã Formation, dated at 628±18 Ma.

INTERCONTINENTAL INTEGRATION
As previously mentioned, a recent contribution, by Brito Neves and  showed that the basement of the South American Platform is divided into two distinct major geological and geotectonic domains, the Amazonian or pre-Brasiliano domain and the Brasiliano domain, which are separated by the transcontinental Transbrasiliano lineament (Tab. 2). ese domains are separated from each other by the Transcontinental Kandi-Transbrasiliano lineament. e distinction between the two domains is underlined by the striking di erences in stratigraphic organization in both domains, especially when the tectonic events in the Brasiliano domain are considered. From the Mesoproterozoic onwards the two domains display very distinct geological features and tectonic evolution. In Table 3, the di erent tectonic events comprising the Brasiliano collage are listed and detailed.
Tonian Terranes e presence of Tonian terranes within the Brasiliano collage is well documented. However, based on the available data, it has been di cult to recognize them as representative of an independent (post-Stenian, pre-Brasiliano) phase of crustal evolution. Although there is evidence of structural discontinuity between the Tonian and the Brasiliano rock assemblages in some areas, the precise meaning of this break is hard to understand due to the strong overprint of Brasiliano orogenic events.
As far as geological knowledge and lithostructural context of each province go, there is no evidence to link the Tonian occurrences in Brazil to those of the classic Meso-Neoproterozoic Grenvillian orogenies of the northern and southern hemispheres. Also, there are fundamental di erences in age and composition, and it has been truly dicult to provide paleographic reconstruction, even following several di erent models.
In relation to Africa, based on previous reconstitutions and age data, the Meso-Neoproterozoic Namaqua-Natal belt is possibly the most interesting orogen (candidate) to be correlated to South American Tonian terrains. Some similarities have been pointed out during the analysis of the literature. However, many di culties are also evident as, for instance, the lack of paleomagnetic support. Meso-Neoproterozoic geological units in Nigeria, Chad and southern Sudan, comparable with those in Brazil, have been reported (de Wit et al. 2005Wit et al. , 2008), but they have been strongly reworked by the Pan-African orogeny. If this supposed correlation will be con rmed, the Tonian terranes in West Gondwana may extend along strike for several thousands of kilometers. Actually, the presence of Tonian terranes in Rodinia, Western (and Eastern) Gondwana has been neglected, and this is persisting as a challenging open question.
Tonian processes and orogenies were absent and have not been considered in the Rodinia concept; this supercontinent was formed at ca. 900 Ma ago (Li et al. 2008). It is still an open question whether the Tonian processes precede (or follow) Rodinia amalgamation, and this has important implications for the con guration of this supercontinent. It appears that, in general, awareness of this problem is absent in the international scienti c community. B ra z il U ru g u a y P a r a n a p a n e m a C r a t o n   e rst events of convergent plate interaction of the Brasiliano collage are believed to have taken place in the Early Cryogenian (≥750 Ma), as recorded in most of the structural provinces, except for the Pampean Orogen. Investigation of the initial events may be hampered by the superposition of younger magmatic or tectonic events, or by the lack of outcrops, access problems or even by the lack of geochronological data.
Furthermore, the initial convergent processes are coeval with taphrogenic events (elsewhere) linked to the breakup of Rodinia, which appear to have more evidence due to a wealth of geological and isotopic data (850 -740 Ma, Silva    Lena et al. 2014).
Recently, Philipp et al. (2014) reported some older ages in the southern part of the São Gabriel terrane. Tonalite orthogneis from the Imbicuí rock unit was dated at 875 ± 5 Ma. Intrusive diorite was dated at 745 ± 1.2 Ga. According to the authors, this may represent an older arc, possibly related with the Tonian terrains discussed above.

Cerro Bori orthogneisses
Early Cryogenian rocks are exposed in Cerro Bori, Uruguay, east of the southernmost tip of the Pelotas-Aiguá batholith. Tonalite and granodiorite gneisses, ma c and felsic granulites and amphibolite are the dominant rock types (Lenz et al. 2011). Petrological, geochemical and isotopic data con rm magmatic arc a liation.
U-Pb SHRIMP data of mafic granulites indicate ages between 770 and 796 Ma. Gneisses samples were dated between 770 and 802 Ma. Regional metamorphism was dated at ca. 676 Ma, implying in a complete accretionary and collisional history before the emplacement of the Pelotas-Aiguá batholith, preceding, therefore, all other accretionary processes related with the climax of the Brasiliano orogeny. Lenz et al. (2011) suggested an early Brasiliano event (or Brasiliano I), following the earlier proposal by Silva et al. (2005). Despite the small size of this area (ca. 300 km 2 ), this is an important occurrence that should be remarked. Relations with the almost coeval São Gabriel arc exposed to the west side of the Pelotas-Aiguá batholith still have to be investigated.

The "Embu terrane"
e "Embu terrane" is exposed in the central Mantiqueira Province, to the northwest and north of the Curitiba terrane up to the northern part of the province (Figs. 4 and 5), striking NE-SW for more than 250 km and up to 40 km wide (Heilbron et al. 2004;Campos Neto et al. 2007;Alves et al. 2013;Trouw et al. 2013). e Embu terrane is recognized as a continental fragment docked at the Ribeira belt during the Ediacaran, when the Paranapanema and São Francisco plates collided. It comprises Paleoproterozoic gneiss (shown as a basement inlier in Figs. 4 and 5), medium-to highgrade supracrustal rocks, as well as Cryogenian gneiss, and granite intrusions dated at 600-590 Ma.
Early Cryogenian tonalite and granodiorite gneisses with magmatic arc a nities exposed to the south of São Paulo were dated at 811 ± 13 Ma (Cordani et al. 2002) while muscovite leucogranites from the western "Embu terrane" were dated at 760 -740 Ma (Passarelli et al. 2004). Xenotime and monazite from the Mogi das Cruzes granite yielded ages of ca. 790 Ma (Vlach 2001), similar to the age of regional metamorphism.
Orogenic and anorogenic granite magmatism took place between 650 and 500 Ma, with the main peak between 590 and 540 Ma (Alves et al. 2013). e available data suggest that the "Embu" is a Brasiliano terrane, although early Cryogenian, as well as Meso-and Paleoproterozoic events have also been identi ed (Alves et al. 2013). However, connections with coeval events in neighboring terranes and belts have yet to be investigated. ere are some unpublished data pointing out the occurrence of some pre-Proterozoic rock units (at its northernmost part).

The Ceres-Rialma Metamorphic Belt
A high-grade metamorphic belt is exposed along the headwaters of the Tocantins River in central Brazil, east of the Mara Rosa arc and Goiás massif. It extends for more than 350 km, striking NNE-SSW, and was formed in the Early Cryogenian (Moraes & Fuck 1994;Moraes et al. 2004Moraes et al. , 2006. e western part of the belt comprises sedimentary and volcanic rocks of the Juscelândia, Indaianópolis and Palmeirópolis sequences and the Serra da Malacacheta and Serra dos Borges gabbro-anorthosite complexes, formed in a rift environment that evolved to ocean oor at ca. 1,270 Ma. ese rocks overlie tectonically the eastern part of the belt that is made of the Barro Alto, Niquelândia and Cana Brava ma c-ultrama c layered complexes (Moraes et al. 2004;Ferreira Filho et al. 2010, among others). It has been suggested that the layered complexes are originally related to the taphrogenesis processes that led to the Rodinia breakup (Brito Neves & Fuck 2013), at about 800 Ma Della Giustina et al. 2011).
Although the original igneous stratigraphy and mineralogy of the complexes were disturbed by late Eocryogenian tectonic-metamorphic processes (750 -770 Ma, Ferreira Filho 1994;Moraes et al. 2006), it has been possible to retrieve petrological and geochemical data. ese suggest that they may have been a single large body, representing an Eocryogenian LIP (Moraes et al. 2004), an inference which appears to be supported by geophysical data (Soares 2005;Soares et al. 2010). e two parts of the belts are singularly di erent in origin and age, but were juxtaposed tectonically, deformed and metamorphosed together under high-grade conditions of high-T amphibolite and granulite facies. ey form an unusual high-grade belt between the Goiás massif and the Paleoproterozoic Natividade-Cavalcante block, basement to the external zone of the Brasília belt. Another notable feature of the high-grade belt is that its metamorphic gradient is opposite to that of the Brasília belt: it increases from west to east.
Despite the numerous investigations carried out on the belt, several questions are still pending. e age of closure and inversion of the western Ectasian rift is still debatable, although Moraes et al. (2006) present preliminary data suggesting metamorphism between ca. 0.75 -0.76 Ga, as indicated by U-Pb analysis of metamorphic zircon and Sm-Nd garnet-whole rock isochrons on rocks of the Juscelândia sequence. e causes of orogeny and high-grade metamorphism are still poorly understood. Subsequent Neocryogenian tectonics and magmatism are rather complex, and hamper the understanding of previous events. It seems likely that the orogenic process may have been caused by compressive far-eld stresses resulting from the inversion of the large rift that led to the intrusion of the layered ma c-ultrama c complexes. So far, no obvious connection could be established with the other Cryogenian terrains discussed above.

Neocryogenian-Early Ediacaran (650/600 -600/590 Ma)-"Pulse" 2 of the Brasiliano collage
e Neocryogenian-Early Ediacaran time interval is marked globally by post-Rodinia and pre-Gondwana events, and by diachronic accretionary and collisional processes in all of the Neoproterozoic structural provinces in South America. Large orogenic chains with important crustal accretion events were then developed. ey also include important recycling of Mesoproterozoic to Paleoarchean crustal material. Narrow and wide oceans Schobbenhaus & Brito Neves 2003;Basei et al. 2008) were consumed in a varied and complex scenario of plate interaction and continent formation. Collisional processes were followed soon after (late Cryogenian-early Ediacaran) by the intrusion of typically collisional syenite and granite bodies emplaced in thickened crust. In other cases -e.g., SE of Brasilia belt, north of Mantiqueira province -, the collision was entirely con ned within the early Ediacaran.

Tocantins Province
In the Tocantins Province (Figs. 2 and 3), the Brasiliano part of the Goiás magmatic arc (Pimentel et al. 2000Fuck et al. 2006;Oliveira et al. 2006;Dantas et al. 2006) and of the Socorro-Guaxupé nappe, further south (Campos Neto & Caby 2000), are relatively well studied, with reference to interacting plates, closed ocean realms, as well as to their accretion history (650 -630 Ma) and subsequent collision (ca. 625 Ma). Metamorphic processes and post-collisional events are relatively well constrained.
e Araguaia (Rokelide) orogenic system further north within the Tocantins Province displays good geological and geochronological records of ocean closure (770 -550 Ma), as attested by the Quatipuru, Serra do Tapa and Tucuruí ophiolites (Paixão et al. 2008). So far, there is no clear evidence of related arc magmatism. However, based on positive gravity Bouguer anomalies observed in the basement, it has been suggested that the Goiás magmatic arc might branch out northwards below the Parnaíba basin (Castro et al. 2014).

Borborema Province
An important accretionary event has been recorded in the western Borborema Province (Fig. 1). e Santa Quitéria-Tamboril system is ca. 380-km-long, striking north-south, and up to 110-km-wide (Fetter et al. 2003;Santos et al. 2008Santos et al. , 2009Araújo et al. 2012;Costa et al. 2013;Vasconcellos et al. 2013). Several aspects of arc formation and evolution are still unknown, due to the lack of data and to the superimposed extrusion tectonics. Recent magnetotelluric data display evidence of double convergent subduction below the Santa Quitéria arc (Padilha et al. 2014), corroborated by remnants of eclogite/retroeclogite and high-pressure granulite on both sides of the arc (Santos et al. 2009;Amaral et al. 2011Amaral et al. , 2012. In the "Zona Transversal" of the Borborema Province, and in several other areas, there are references of granites formed in magmatic arcs (e.g. "Conceição", "Brejinho", "Riacho do Icó", "Timbaúba", "Caboclo-Nova Olinda", "Curral de Cima", etc. Ferreira et al. 2004;Sial et al. 2008), the majority of which were dated between 650 and 620 Ma. ese granites of the "Zona Transversal" domain (central-western part) are mostly piercing rock assemblages of the Piancó-Alto Brígida (SPAB), Riacho Gravatá (RG) and Alto Pajeú terranes (AP) (See Fig. 1). is group of granites deserves further investigation, and, informally, they have been grouped as part of the "Alto Sertão Magmatic Arc".
Accretionary orogenic evolution has also been proposed for the northern part of the Sergipano belt, on the southern border of the Pernambuco-Alagoas massif (Oliveira et al. 2010), where ages are well constrained between 630 and 617 Ma.
In the African Dahomeyan-Trans-Saharan counterpart of the Borborema Province, Caby (2011) reports that the main calk-alkaline plutonic phase of active continental margin took place at 640 -630 Ma, followed by regional metamorphism at 620 -600 Ma.

Mantiqueira Province
Accretionary processes have also been recognized in the Mantiqueira Province, from eastern Minas Gerais to eastern Uruguay (Figs. 4 and 5), forming di erent branches for more than 1,500 km along the eastern margin of the continent, between 18ºS and 34ºS (Cordani et al. 2000;Bizzi et al. 2003;Silva et al. 2005, among others). e accretionary systems are not continuous along the strike of the province, but are supported by a wealth of geological, geochemical and geochronological data. ese include: ■ Rio Doce, in Minas Gerais, Espírito Santo and Rio de Janeiro (Heilbron et al. 2004;Pedrosa Soares et al. 2006, e scheme of plate interactions is not fully understood. In the northern part of the Mantiqueira Province, the western "São Francisco plate" represents the lower plate. e upper plate to the east was possibly the "Congo plate", but this is still an open question (also, there are some objections). Other authors envisage collision in a broader scenario, such as the "nutcracker model" proposed by Alkmim et al. (2006).
In the central part of the province, several lithospheric fragments have been mentioned as interacting plates/microplates/microcontinents; these include the Paranapanema, Curitiba, and Luis Alves blocks (Figs. 4 and 5). However, the proposed models need additional tectonic and geophysical support in order to be validated (Basei et al. 2008). e model developed for the southern part of the province suggests that the "Rio de La Plata" represents the lower plate (to the west) and the Kalahari (to the east), would be the upper plate (Basei et al. 2008). Many of the di culties in identifying the plates involved are related to the post-collisional shear systems resulting from Neoediacaran-early Ordovician escape tectonic processes, which hamper paleogeographic reconstructions. Additionally, evolution of the South Atlantic continental margin and the Cenozoic uplift and other brittle tectonic events also mask some of the (former) Brasiliano features.

Some local and precursor events of the Neocryogenian-early Ediacaran processes and/or inheritance -Preceding "pulse" 2
In some cases, initial accretionary events have been discriminated from the dominant Neocryogenian ones, due to reliable isotopic data, although the lack of appropriate eld data may hamper their interpretation.
In the Santa Quitéria magmatic arc, W Ceará (Borborema Province), the large majority of age determinations are in the 640-610 Ma interval. However, in the eastern part of the arc, granodiorite, metatexite and diatexite have crystallization ages between 876 and 831 Ma (Araujo et al. 2012), this means ca. 200 Ma older than the main phase of arc building.
In the Rio Negro arc, Rio de Janeiro (Mantiqueira Province), the main rock assemblages were formed mainly between 630 and 590 Ma. A granodiorite gneiss sample was recently dated at 792 ± 12 Ma and a granite sample at ca. 845 Ma, interpreted as the product of initial arc magmatism (Heilbron et al. 2004). Although additional isotopic data are needed, the new ages are duly noted.
As mentioned previously, rocks representing "pulse" 3, aside from those of the so-called "Rio Doce arc" in Minas Gerais and Espírito Santo, according to Campos Neto and Figueiredo (1995) and Campos , are exposed in the coastal zone of Rio de Janeiro and São Paulo, along the Serra do Mar ("Serra do Mar terrane", . According to this peculiar concept of the "Rio Doce arc" (590 -570 Ma), its rock forming events overlap those of the preceding orogenic pulse represented by the Rio Negro arc, exposed eastwards. According to Campos Neto and Figueiredo (1995) and Campos , these events have (some type of ) connection with those of the Rio Doce orogeny in the Araçuaí belt, further north in the Mantiqueira Province. So, a precise separation of the orogenic "pulses" (arc magmatism) 2 and 3 along the coastal zone of the Mantiqueira Province is still uncertain .
In addition to the successive accretionary events, there are several indications of intraplate orogenic processes without any evidence of previous subduction. Collisional/metamorphic events (ca. 590 -560 Ma) have been recorded in some particular cases of the Brasiliano structural provinces (e.g. part of the northern Mantiqueira Province, to the north of the Patos lineament in the Borborema province, etc.). e identi cation of truly orogenic events occurring during the Early-Middle Ediacaran has been di cult, due to concurrent post-orogenic tectonic and magmatic events, as well as anorogenic magmatism, which overlap in time and space with the orogenic rocks.
According to Tupinambá et al. (2007Tupinambá et al. ( , 2010, the accretionary phase of the Rio Negro arc ended around 600 Ma ago, and the collisional phase started soon after, including several syn-collisional magmatic events until 560 Ma. ese authors admitted also that another magmatic arc may have been accreted at the beginning of the collisional phase, reiterating previous observations by Campos . Post-collision processes have been recorded in many places, such as granite and syenite magmatism (e.g. Itu and Graciosa suites, Teixeira-Terra Nova tectonic zone, late granites of the Rio Doce and Rio Negro arcs, Araguaia and Brasilia belts granites, granite magmatism piercing through basement inliers in the Borborema Province, etc.). Given the widespread occurrence of processes and their record, this is clearly a time interval composed of a diversi ed plethora of tectonic and magmatic activities. e distinction between the events of this time interval, late and post-collisional processes, from the previous ones, arc magmatism and metamorphism, still needs to be completed with new and better geological eld data and isotopic information. It is also useful to bear in mind that escape-(extrusional) tectonic processes and foreland tectonics a ecting the South American Platform as a whole (in fact a large part of West Gondwana) started soon after 580 Ma. ese processes obliterated some of the original geological relationships.

Cambrian Orogenies (530 -490 Ma) -"Pulse" 4 of the Brasiliano collage
From the late Ediacaran to the early Ordovician, the most remarkable tectonic events in South America are related to post-collisional processes, marking the transition from unstable to stable tectonic conditions. ey mainly include escape tectonics and associated sedimentary processes, foreland tectonics, extension-related intrusions and isostatic adjustments.
While such processes were underway, two local minor orogenic developments were in process in the eastern and southwestern parts (Búzios and Pampean orogenies, respectively) of the continental area that soon after became Western Gondwana.

Búzios orogeny
e Cambrian Búzios orogeny, the youngest event of the Ribeira belt in the Mantiqueira Province, is recognized in the coastal zone of Rio de Janeiro (Região dos Lagos, Schmitt et al. 2008;Schmitt & Armstrong 2014).
e tectonic evolution started with the history of a small (residual?) ocean basin from 610 Ma onwards, where volcanic-sedimentary sequences accumulated. ese rocks were deformed and metamorphosed under high-grade conditions, and transported to the northwest at ca. 520 Ma. e hypothesis is a process of subduction of the "Angola plate" (to the east) under the "Terreno Oriental" (to the west). is is estimated to have taken place between 590 and 530 Ma, which was followed by collision at ca. 530 -520 Ma. Late igneous and tectonic activity took place between 520 and 490 Ma (Schmitt et al. 2008). e authors compare this complete Wilson cycle with the products of Pan-African orogeny in the Kaoko belt in Angola and Namibia. Schmitt and Armstrong (2014) admit that the orogeny extended until ca. 490 Ma in a stepwise process that progressed to a poorly understood Ordovician orogenic collapse, until middle Ordovician times.

Pampean orogeny
e Pampean orogenic belt, east of Cordoba, Argentina (Fig. 6), developed in response to the subduction of the lower "Pampia microplate" to the west, made of high-grade "Grenvillian" rocks, under the Rio de La Plata upper plate in the east (Rapela et al. 1998;Escayola et al. 2007;Schwartz et al. 2008). e orogen comprises two parallel belts, the eastern one comprising mostly calc-alkaline magmatic rocks, and the western made of metasedimentary rocks and peraluminous granites. Calc-alkaline magmatism was active between 555 and 525 Ma, whereas peraluminous intrusions associated with regional metamorphism took place in the Early Cambrian (525 -515Ma). Ophiolite remnants were dated by the Sm-Nd method at 647 ± 70 Ma (Escayola et al. 2007). Schwartz et al. (2008) proposed a possible connection with the "Terra Australis orogen", along the closing paleo-Paci c Ocean. In a regional perspective, there is the possibility of a northward connection with the Paraguay belt (south-western border of the Amazonian plate), although there is no evidence for subduction of ocean lithosphere in the latter. Additional geological and geophysical data are necessary to test this model.

PRELIMINARY CONSIDERATIONS ABOUT GONDWANA AMALGAMATION
Processes leading to amalgamation of supercontinental masses are complex, long lasting, and accomplished according to amalgamation fronts of distinct paleogeography. Eventually, amalgamation processes in one province or front concur in time with taphrogenesis or breakup processes taking place in other areas. ese are fundamentally diachronic processes from one amalgamation front to another and from one province to another. erefore, understanding supercontinent formation requires a large amount of geological, tectonic, paleomagnetic and isotopic data. Additionally, it should be remembered that supercontinents are not the nal product of con guration, therefore improvement of tectonic models are necessary.
Several distinct models for the assembly and evolution of Gondwana have been proposed, but the problem is far from resolved. Many questions still remain, demanding  di erent levels of geological and geophysical investigations in di erent continents. According to the recent proposal by Collins et al. (2011), processes took place along three di erent fronts of present day southern continents. Should this be the case, two of these fronts are represented in the South America basement, in addition to Mediterranean-Middle East-Arabia, East Africa-Madagascar-India-Sri Lanka-Antarctica, with rami cations to Zambezi and Damara, and Rokelides-Araguaia-Paraguay-Pampean.
ere are other models/versions of Gondwana formation, e.g. Stern (1994), Meert (2003) (2008), Linnemann et al. (2008Linnemann et al. ( , 2014 (North Africa, Cadomian orogeny), etc. See discussions in the next item of this paper. Consensus seems to be hard to achieve, and it will certainly take time to better understand amalgamation of Gondwana.
Independent of academic dissent, and aware of the need for new data, it can be stated that the Brasiliano collage in South America (and in part of westernmost Africa) comprises four distinct phases (main "pulses"), from the early Cryogenian (Neoproterozoic) to the early Ordovician (in some special cases, up to the upper Ordovician, ca. 440 Ma). e latest processes of collage fronts that ended up in Western Gondwana amalgamation are recorded in South America, and some similar and coeval events have been recorded in westernmost Africa (e.g. Schmitt et al. 2008;Schmitt & Armstrong 2014).
From Bahia to Uruguay, the Mantiqueira Province displays the most complete record of the main tectonic-magmatic events referred to above, including the Tonian terranes and the four Neoproterozoic pulses. None of the other provinces present such a complete record. However, as mentioned previously, there are some unsolved questions related to the adequate and precise discrimination of the pulses discussed above (pre-740 Ma; 650 -630 Ma; 590 -560 Ma; 530 -490 Ma). Also, it will be necessary to re ne time interval limits, as well as to search for additional magmatic and metamorphic records. It should be noted that in the Araçuaí belt, northern portion of the Mantiqueira Province, late-collisional granites were intruded from 590 to 530 Ma. Similar intrusions were reported from the Borborema Province and other areas. However, robust isotopic data are lacking.
Where and when did Gondwana finally amalgamate?
In general, the orogenic processes that led to Gondwana amalgamation (Collins et al. 2011) are somewhat similar and coeval ( ~ 750Ma ± Δx -500 Ma ± Δx), in its western as well as in its eastern portion. Early phases, nevertheless, display some important time di erences. On the other hand, lateto post-orogenic events tend to overlap in the late Ediacaran, or in the middle Cambrian, occasionally extending into the Ordovician. is most important set of geotectonic processes coincides with one of the most dynamic periods of Earth evolution, involving all its geospheres and their components. e overall time lapse comprised in this long, complex and polyphase assembly process (ca. 280 -250 Ma) is comparable to that of the assembly of Pangea, between 500 and 230 Ma ago (270 Ma), for which scienti c foundation (Stratigraphy, Paleontology, Paleomagnetism) is much more reliable.
Similarities notwithstanding, detailed analysis of each portion of the continent shows that the orogenic collage pulses are quite distinct in time and importance. It should be noted that the polyphase development of Gondwana (Tab. 4) bears many similarities with the formation of Pangaea, where several phases (e.g. Caledonian, Alleghenian, etc.), are constantly highlighted. However it is well known that they are not exclusive.

Northern part of Gondwana
In northern Gondwana, three main orogenic phases have been distinguished (Linnemann et al. 2008(Linnemann et al. , 2014Pereira et al. 2012, among

Eastern Gondwana
Several distinct, although not necessarily opposite, digressions apply to eastern Gondwana evolution, from the Arabian-Nubian shield, to India, western Australia and Antarctica. a) Collins and Pisarevsky (2005) provided an excellent synthesis based on rich paleogeographic documentation and paleomagnetic data of the whole evolutionary process, starting at 750 Ma. e authors display the continental masses involved, as well as the intervening oceans (Mozambique, "Adola", Adamastor, "Braziliano", "Paci c"). ree orogenic phases of the collage are recognized: ■ e rst orogenic phase (c. 630 Ma) is recorded between east Africa ("Azania") and the "Congo/Tanzania/ Bangweulu" ("CTB") masses, and between these and the continental masses of the "Saharan metacraton", including closure of several minor ocean basins between them; ■ Early Ediacaran (ca. 570 Ma), when "CTB" collided with the São Franciscano block. ere is no record of this collision in South America, or of its continuity into the collision with "Azânia"; ■ Early Cambrian (540 -530 Ma), encompassing a series of collision processes east and west of India with "Austrália-Mawson", and "Azânia", and of "CTB" with Kalahari (including northern Mantiqueira Province?) and Rio de La Plata (Don Feliciano belt?). Associated with these processes, high-grade metamorphism took place in several of the resulting mobile belts.
b) Based on geochronological and other databases, Meert (2003) and Meert and Lieberman (2008) present di erent concepts and syntheses, displaying several noticeable and well-documented phases for the 750 -530 Ma time interval: East African orogeny: ■ ca. 750 Ma (between 780 and 720 Ma): important accretion phase, arc magmatism from northern Africa (Arabia-Nubian shield) to Antarctica. ■ ca. 640 Ma (between 700 and 610 Ma): important phase of collisions, high-grade metamorphism, and in some cases initiation of escape tectonics (Arabia-Nubian shield, Somalia, Eritrea). ■ ca. 550/540 Ma (between 590-520 Ma) -Kuunga orogeny -: collision, high-grade metamorphism and post-orogenic magmatism in several belts. Structures related to the Kuunga orogeny strike orthogonally to those of the East African orogen, between the southwest coast of Africa (Damara), southern India and western Australia. Meert (2003) has suggested that the 500 Ma old Ross-Delamerian belt (Australia-Antarctica-South Africa) should be ascribed to a later event that took place after nal suturing of Gondwana, adding to the controversy about the belt.
is suggestion is a matter of dispute.

West Gondwana
As discussed above, and previously highlighted by Brito Neves and Fuck (2013), four main orogenic pulses are recorded

Buzios Orogeny
In the central-northern portion of the Mantiqueira Province, another Cambrian orogen is recognized, as discussed above. e orogenic belt appears to correspond to the closing of a small oceanic basin or back-arc, which remained between the "Terreno Oriental", the upper plate of the Ribeira orogenic belt, and a colliding continental fraction (Cabo Frio terrane), in the east, interpreted by some authors as a slice from the Congo craton/plate. e collisional process probably closed a remnant fraction of the Adamastor Ocean, and led to intense low-angle deformation and high-T amphibolite to granulite facies metamorphism. Geological and geochronological studies (Schmitt et al. 2008;Schmitt & Armstrong 2014) identi ed Cambrian ages of 530 -510 Ma for the orogeny, which was followed by collapse and extrusion between 510 and 480 Ma. Geological and geochronological equivalents were found in western Africa, within the Kaoko belt (Schmitt et al. 2008). Recently, Schmitt and Armstrong (2014) presented new data, reiterating the Middle to Late Cambrian age of 515 -510 Ma, possibly extending up to 490 Ma, for the Búzios orogeny. Still not well-understood collapse processes occurred until the Mid Ordovician (c. 440 Ma). Escayola et al. (2007) reported the Sm-Nd isochron age 647 ± 77 Ma for ophiolite remnants and interpreted it as the best age estimate for the development of oceanic lithosphere in the belt. Early Cambrian ages of 530 -514 Ma were recorded for the calk-alkaline arc magmatism and subsequent metamorphism (collision). Schwartz et al. (2008) concluded that subduction started at 555 Ma, in the Ediacaran, establishing ages of 555 -525 Ma for the arc magmatism, partially overlapping with the data of Escayola et al. (2007). ey pointed out 525 Ma as the age of collision and metamorphism, revealing the record of Early Cambrian amalgamation of Gondwana in southwest South America.

Pampean
Recent data (Rapela et al. 2014) added new information to the age of the Pampean orogeny. Metasedimentary rocks a ected by the orogeny include a 525 Ma old detrital zircon peak among Neo-and Mesoproterozoic and older zircon grains. erefore, ages younger than 525 Ma should be expected for the orogeny, allowing the possibility of ages from the Middle to Late Cambrian (< 530 Ma).
It is also possible that these orogenic processes reached farther north, encompassing the southern branch of the Paraguay belt (MacGee et al. 2012). is proposition has been challenged by Cordani et al. (2013a).
Although remote, there is the possibility of a connection between the Pampean and the Ross-Delamerian orogens (positioned to the south of Gondwana, south of Antartica and Australia). Lack of data hampers discussion at this time.