Vicente A.V. Girardi

WAS THERE AN EDIACARAN CLYMENE OCEAN IN CENTRAL SOUTH AMERICA

Previous studies have proposed that a major suture resulted from the collision between the Amazonian and São Francisco-Congo cratons during the Cambrian, following the closure of a supposed Clymene Ocean. The proposal tentatively located this ocean along the Araguaia and Paraguay belts at the eastern margin of the Amazonian Craton, and its southern extension reached the Pampean belt in Argentina. In the present study we will argue that the existence of Ediacaran-Cambrian oceanic lithosphere in central South America is highly unlikely. West Gondwana was assembled during the convergence between the Amazonian, West African, São Francisco-Congo and Rio de La Plata cratons as well as the Saharan Metacraton, leading to the closure of the Goiás-Pharusian Ocean during the Neoproterozoic. Final closure and continental collision resulted in the development of the Transbrasiliano-Kandi mega-shear zone that cuts through several mobile belts, but leaves the cratonic areas totally untouched. Consistent results of radiometric dating along the Transbrasiliano (TB) mega-shear in South America and of metamorphic rocks of the Brasília Belt have indicated that the Neoproterozoic collision finished at ca. 620 Ma. After isostatic uplift, cooling, and denudation, between 590 and 500 Ma, emplacement of undeformed K-rich postorogenic granites represented the main tectonic event. At this time or afterwards, a series of small extensional sedimentary basins formed in graben troughs, most of which are within the TB tectonic corridor. They all were of extensional character, contrasting clearly with the convergent tectonics occurring within the coeval Pampean Orogen in Argentina. The main arguments showing that an Ediacaran to Cambrian oceanic closure in central Brazil is untenable include: (i) the assembly of West Gondwana was completed by ca. 600 Ma, when the convergence between the Amazonian, São Francisco and Rio de La Plata cratons had already ended. After this, there is no geological evidence of an oceanic lithosphere (for example, ophiolites, magmatic arcs, et cetera), ruling out the possible existence of an Ediacaran or Cambrian Clymene Ocean in Central Brazil; (ii) the Gurupi and Araguaia belts in Brazil, as well as the Bassaride and Rokelide belts in West Africa, are regarded as aulacogenic-type systems formed within an intraplate tectonic setting. Their tectonic history precedes the collision between the Amazonian and São Francisco-Congo cratons, as demonstrated by the linear structures of the Transbrasiliano megashear which truncate the N-S structural trends of the Araguaia Belt; (iii) there is a close correlation between the Corumbá Group of the Paraguay Belt in Brazil and the Arroyo del Soldado Group in Uruguay. These sedimentary sequences belonged to the same Ediacaran continental shelf and this is a powerful indicator for an Ediacaran connection between the Amazonian and Rio de La Plata cratons, which precludes the existence of a wide ocean (for example, the Clymene) between them. On the other hand, the tentative correlation between the Sierras de Cordoba and the Paraguay Belt cannot be accepted, because these are far apart and there is no similarity in lithology, metamorphism, or structural trends; (iv) the Puga paleopole is the most important evidence for the hypothesis of the Cambrian Clymene Ocean, however the age of about 600 Ma for this paleopole, taken on the basis of Sr and C isotopes, is loosely constrained. In addition this is located at low latitude, not far from the present pole, and therefore could be related to a younger remagnetization; (v) the Pampean Orogen is made up of medium- to high-grade metamorphic rocks constrained between 560 Ma and 520 Ma and therefore was tectonically active during most of the Cambrian. However, at this time, an oceanic lithosphere is not evident in the vicinity of the Paraguay belt, and in central Brazil extensional rather than convergent tectonic processes have been observed.

Calc-alkaline and tholeiitic dyke swarms of Tandilia, Rio de la Plata craton, Argentina: U–Pb, Sm–Nd, and Rb–Sr 40Ar/39Ar data provide new clues for intraplate rifting shortly after the Trans-Amazonian orogeny

Abstract The Tandilia system, Argentina, southernmost part of Rio de la Plata craton (RLPC), is intruded by two Paleoproterozoic unmetamorphosed dyke swarms which are: (i) intermediate (I) and acid (A); and (ii) basic (B1 and B2) in composition. The latter dykes have tholeiitic characteristics, while for both I and A dykes major, minor, and trace elements, including REE (e.g. the higher values of SiO2, K2O and Ba compared to tholeiitic dykes) are characteristic of calc-alkaline suites. The calc-alkaline dykes (I and A) yielded 40Ar/39Ar step-heating plateau ages of emplacement of 2020±24 and 2007±24 Ma. These ages are within error in agreement with a Rb–Sr errorchron of 1956±110 Ma (1σ) [initial 87Sr/86Sr=0.7038±0.0025 (MSWD=19)]. An upper intercept U–Pb age on two baddeleyites from a tholeiitic (B1) dyke places the intrusion at 1588±11 Ma. The tholeiitic dykes (B1 and B2) have K–Ar whole-rock ages from 803±14 to 1193±18 Ma and a 40Ar–39Ar plateau age of 811±36 Ma (2σ) on plagioclase. These discordant apparent ages suggest variable Ar loss of the mineral systems. Calc-alkaline dykes mainly trend E–W, and were emplaced during the transtensional stage of the Trans-Amazonian orogeny during which the plutonic rocks of the Tandilia system were formed. Such a scenario has similarities with the Eburnean evolution of the Richtersveld plutonic arc complex of the southern African subcontinent (Namaqualand) that faces the RLPC in the West Gondwana reconstruction. The significantly younger tholeiitic dykes of Tandilia (1.59 Ga) trend mainly NW-SE. Their presence constrains the time of crustal extension at the Paleo-Mesoproterozoic boundary during which basin-formation tectonics and anorogenic magmatism took place worldwide within a stabilized Paleoproterozoic lithosphere. Such an intraplate regime for the emplacement the youngest dykes of Tandilia is consistent with transcontinental scale, diachronous extensional episodes within the South American continent which initiated shortly after the Trans-Amazonian orogeny, as illustrated by the 1.73 Ga Florida tholeiitic dyke swarm and coeval, anorogenic granitoids scattered across the RLPC (Uruguay and Tandilia). Paleoproterozoic geologic features of the RLPC—namely, the occurrence of plutonic arc rocks (2.14–2.07 Ga) succeeded by emplacement of anorogenic granitoids and mafic dykes (1.73–1.59 Ga)—allow direct correlation with the postulated Gondwana counterpart, mirrored by broadly contemporary plutonic rocks of the Richtersveld and Bushmanland subprovinces of Namaqualand. The broad picture reinforces the idea that the Trans-Amazonian/Eburnean orogenies played an important role for juvenile crustal accretion within the southern South America and its southern Africa counterpart, which was followed by a tendency to dispersion of the stabilized continental fragments during the Mesoproterozoic, preceding the assembly of the Rodinia supercontinent.

Petrology of late proterozoic mafic dikes in the Nico Perez region, central Uruguay

SummaryTholeiitic basaltic and basaltic andesite dikes of Brasiliano (or PanAfrican) age (≈ 600 Ma) intrude the basement of the Nico Perez region, Uruguay. Major and trace element geochemistry of the basalts indicates that they suffered fractionation in shallow magma chambers. The variation in element ratios, which remain virtually unchanged during fractionation (K/Rb, Rb/Ba, Ba/Nb, La/Nb, Zr/Nb and Ti/Zr), indicate that the dikes are not strictly comagmatic. However, they have certain features in common: LILE and LREE enrichment with respect to HFSE and HREE; high Rb/Ba (> 0.9) and Rb/Sr (> 0.08); low K/Rb (< 214); negative Nb and Ti anomalies (Lan/Nbn > 2; Ba/Nb > 22 Ti/Zr < 60).87Sr/86Sr and143Nd/144Nd at 665 Ma are in the range 0.7052 – 0.7119 and 0.51158 – 0.51177, respectively.The lack of correlations between isotope and trace element variations indicate that these characteristics are not controlled by crustal contamination of the melts. They are interpreted as being due to the melting of an enriched mantle (C1) under the influence of a fluid-rich component (C2) which stabilized a Nb-retaining titanate phase in the residuum. Although this process may be related to a subduction environment, it is also possible that it occurred in an ensialic region by the interaction of deep mantle fluids with the lithospheric continental mantle.The parent mantle underlying the early Proterozoic (1.8 Ga) Florida region had isotope and geochemical characteristics which could evolve to the values observed in the adjacent Nico Perez region. It is therefore proposed that mantle enrichment took place in the early Proterozoic and that this mantle melted under the influence of fluids in the late Proterozoic to derive the Nico Perez dikes.ZusammenfassungIn das Grundgebirge der Nico Perez-Region, Uruguay, intrudierten tholeiitische und basaltandesitische Gange panafrikanischen (Brasiliano) Alters (∼ 600 Ma). Die Hauptund Spurenelementgeochemie der Basalte zeigt eine Fraktionierung in seichten Magmakammern an. Die Streuung jener Elementverhältnisse, die während der Fraktionierung scheinbar unverändert bleiben (K/Rb, Rb/Ba, Ba/Nb, La/Nb, Zr/Nb und Ti/Zr), weist darauf hin, daß die Gänge nicht streng komagmatisch sind. Sie haben aber dennoch bestimmte Merkmale gemeinsam: Anreicherung von LILE und LREE in Bezug auf HFSE and HREE; hohes Rb/Ba (> 0, 9) and Rb/Sr (> 0, 08); niedriges K/Rb (< 214); negative Anornalien bei Nb and Ti (Lan/Nbn > 2; Ba/Nb > 22, Ti/Zr < 60). Bei 665 Ma liegen87Sr/86Sr und143Nd/144Nd im Bereich von 0, 7052 bis 0, 7119 beziehungsweise von 0,51158 bis 0,51177.Fehlende Korrelationen zwischen Isotopen- und Spurenelement-Streuungen zeigen, daß these Eigenschaften nicht durch Kontamination der Schmelzen mit Krustenmaterial kontrolliert werden. Sie werden mit der Aufschmelzung von angereichertem Mantel (C1) unter dem Einfluß einer Fluid-reichen Komponente (C2) erklärt. Diese Komponente stabilisierte eine Titanat-Phase im Residuum, die Nb zurückhielt. Obwohl dieser Prozeß einem Subduktions-Milieu zugeordnet werden kann, ist es auch möglich, daß er in einem intrakontinentalen Bereich durch Wechselwirkung von Fluiden aus dem tiefen Mantel mit dem kontinentalen Lithosphärenmantel stattfand.Der Ausgangsmantel, der die frühproterbzoische (1, 8 Ga) Florida-Region unterlagert, weist isotopische and geochemische Merkmale auf, aus denen sich die in der benachbarten Nico Perez-Region beobachteten Werte entwickelt haben könnten. Daher wird angenommen, daß die Anreicherung im Mantel im frühen Proterozoikum stattfand und daß dieser Mantel unter tern Einfluß von Fluiden im späten Proterozoikum Schmelzen bildete, von denen die Gange von Nico Perez abgeleitet werden können.

Petrology of the Proterozoic mafic dyke swarms of Uruguay and constraints on their mantle source composition

Abstract Three dyke swarms in Uruguay range in age from Palaeoproterozoic (1.86 Ga, Florida region) to Neoproterozoic (∼0.7 Ga, Nico Perez and Treinta y Tres regions). The Florida and Nico Perez swarms are basalts, basaltic andesites and andesites with tholeiitic affinity, characterized by LILE and LREE enrichment with respect to HFSE, Nb depletion with respect to K and La, K Rb , Ba Rb , Zr Nb > 13 , (La Yb) n > 4.5 . Both swarms have positive ϵ Sr and negative ϵ Nd , but the Florida isotopic array is dominated by ϵ Nd variations (EM1 type) and that of Nico Perez by ϵ Sr variation (EM2 type). The Treinta y Tres swarm consists of slightly ne -normative transitional or alkaline basalts with geochemical characteristics resembling those of OIB (e.g., Gough Island), but suggestive of a less enriched OIB source, and have slightly positive ϵ Nd and ϵ Sr . Crustal contamination does not appear to be important in the petrogenesis of the swarms. It is proposed that the geochemical characteristics of the Florida swarm derive from melting of lithospheric mantle infiltrated by hydrous fluids released from the thermal breakdown of hydrous phases, either contained in underplated oceanic crust or formed by interaction of asthenospheric fluids with the lithospheric mantle. The EM1-type isotopic features are considered as unrelated with possible fluid addition, but as a pre-existing feature. The geochemical and isotope characteristics of the Nico Perez swarm are attributed to time-integrated enrichment processes, which affected the Nico Perez lithospheric mantle during the Palaeoproterozoic Florida episode. The geochemistry of the Treinta y Tres swarm may be explained by partial melting of a lithospheric mantle which did not suffer Palaeoproterozoic LILE and LREE enrichment but which was isotopically reset in Palaeoproterozoic times. Alternatively, a residual mantle related to the Florida event is required. Mantle melting and dyke intrusion probably occurred in an ensialic environment. The geochemical and isotopic characteristics of the Uruguay dyke swarms are similar to those of the Mesozoic basalts related to the Gondwana break-up, supporting the possibility that the latter derived from a heterogeneous lithospheric mantle source which recorded Proterozoic enrichment processes.

Petrogenesis of the Paleoproterozoic basalt–andesite–rhyolite dyke association in the Carajás region, Amazonian craton

Abstract Paleoproterozoic basaltic, andesitic and rhyolitic dykes crosscut the Archaean Carajas basement. Basalts are distinguished into a high and a low TiO2 group (HTi and LTi), each group consisting of geochemically distinct NE- and NW-trending swarms. The HTi dykes are evolved transitional basalts having essentially EMORB-type geochemistry. The LTi basalts are tholeiites (NE-trending swarm) and high-Al basalts (NW-trending swarm) displaying incompatible trace elements patterns with variably negative Nb anomaly, enrichment in Rb, Ba, K (LILE) and La, Ce and Nd (LREE) and positive Sr anomaly. With respect to orogenic analogues, andesites have lower Al2O3, CaO and Ni, higher FeO, LILE, LREE, Nb, Zr and Ti and negative Sr anomaly. Rhyolites have geochemical characteristics comparable with those of A-type granites. At 1.8 Ga, 87 Sr/ 86 Sr ranges from 0.700 to 0.705 in the HTi basalts and from 0.700 to 0.704 in the LTi group. Andesites define an isochron of 1874±110 Ma (Sro=0.7038±0.0010). Rhyolites from Southern and Northern Carajas define two isochrons of 1802±130 Ma (Sro=0.7062±0.0046) and 1535±82 Ga (Sro=0.7625) respectively, the younger date being interpreted as resetting of the Rb–Sr isotopic system. We propose a petrogenetic model relating LTi basalts with melting of lithospheric mantle metasomatized by acid melts derived from incipient melting of eclogites, representing in turn the subsolidus product of basaltic batches trapped in the mantle. The HTi basalts are explained by melting of the lithospheric mantle containing the complementary residual eclogite. Andesite petrogenesis is consistent with crystal fractionation from a high-Mg andesite parent derived from a mantle source more extensively metasomatized by eclogite-derived melts. Rhyolite composition is consistent with low melting degree of the basement rocks. The basalt–andesite–rhyolite dykes may represent the effects of crustal extension and arching in Carajas, which produced the anorogenic acid to intermediate magmatism (Uatuma group) and affecting a large part of the Amazon craton between 1.85 and 1.7 Ga.

Gravity interpretation and possible regional significance of the Niquelândia layered basic-ultrabasic complex, Goiás, Brazil

Abstract The Niquelândia layered basic-ultrabasic complex is one of several similar Precambrian massifs that constitute a discontinuous, north-trending chain 300 km long through the central Brazilian Shield in the State of Goias. Modeling of the positive gravity anomaly associated with the complex indicates that the complex constitutes a westward-dipping slab that extends no deeper than 6 km. This model is similar to that obtained for the Barro Alto complex, the next massif to the south. The Niquelândia and Barro Alto complexes occur on a hinge line or gradient in the regional gravity field marked by paired Bouguer anomalies: residual highs to the west, lows to the east. The massifs may indicate a suture in the central Brazilian Shield marked by vast layered intrusions emplaced in continental rocks at the onset of an episode of rifting. Subsequent compression thrust remnants of the intrusions eastward, an event that may have accompanied the closing of an ancient ocean.

Petrology and Sr–Nd characteristics of the Nova Lacerda dike swarm, SW Amazonian Craton: new insights regarding its subcontinental mantle source and Mesoproterozoic geodynamics

The NNW-trending Nova Lacerda tholeiitic dike swarm in Mato Grosso State, Central Brazil, intrudes the Nova Lacerda granite (1.46 Ga) and the Jauru granite-greenstone terrain (ca. 1.79–1.77 Ga). The swarm comprises diabases I and II and amphibolites emplaced at ca. 1.38 Ga. Geochemical data indicate that these are evolved tholeiites characterized by high LILE/HSFE and LREE/HSFE ratios. Isotopic modelling yields positive ϵNd(T) values (+0.86 to +2.65), whereas values for ϵSr(T) range from positive to negative (+1.96 to -5.56). Crustal contamination did not play a significant petrogenetic role, as indicated by a comparison of isotopic data (Sr–Nd) from both dikes and country rocks, and by the relationship between isotopic and geochemical parameters (SiO2, K2O, Rb/Sr, and La/Yb) of the dikes. We attribute the origin of these tholeiites to fractional crystallization of evolved melts derived from a heterogeneous mantle source. Comparison of the geochemical and isotopic data of the studied swarm and other tholeiitic Mesoproterozoic mafic intrusions of the SW Amazonian Craton – the Serra da Providência, Colorado, and Nova Brasilândia bimodal suites – indicates that parental melts of the Nova Lacerda swarm were derived from the most enriched mantle source. This enrichment was probably caused by the stronger influence of the EMI component on the DMM end-member. These data, coupled with trace element bulk-rock geochemistry of the country rocks, and comparisons with the Colorado Complex of similar age, suggest a continental-margin arc setting for the emplacement of the Nova Lacerda dikes.

Petrology and geochemistry of the mafic dyke swarm of the Treinta Y Tres region, Northeast Uruguay

Abstract Transitional basalts of the Treinta y Tres region intrude late Proterozoic granites in northeast Uruguay. The age of the swarm is assumed to be close to 750 Ma. Major and trace geochemical diagrams indicate gabbro type fractionation, which is supported by MPR plots. In comparison to the Florida and Nico Perez suites, the Treinta y Tres swarm is enriched in Ti, P, Zr and Sr and depleted in LILE and REE. These features and the Sro and Ndo isotope values are discussed in terms of source heterogeneity and crustal contamination. The geochemical characteristics do not support crustal contamination. The Treinta y Tres suite is better explained by a source affected by two melting episodes. The dikes were originated from the latter melting, which occurred in a an isotopically heterogeneous source. Diagrams related to several incompatible elements ratios indicate that the mantle source of Uruguaian dikes and of continental basalts of Brazil have similar geochemical features since Early Proterozoic up to Mesozoic.

Liquid immiscibility in the Archean Greenstone Belt of Piumhi (Minais Gerais, Brazil)

Abstract Liquid immiscibility has been used for explaining several modern situations either in effusive or in intrusive rocks. Liquid immiscibility is also proposed for explaining the variolitic structures in the Archean rocks of the Abitibi Greenstone belt, Ontario. The volcanic sequence of the lower portion of the Greenstone belt of Piumhi (Minas Gerais, Brazil) shows the occurrence of variolitic structure, that may have been formed by liquid immiscibility. This possibility is supported by the following points: (a) the contact between varioles and matrix is very sharp; (b) spinifex texture occurs both in the varioles and in the matrix and the single crystal may cross the varioles; (c) the varioles and the matrix show contrasting compositions: the latter has the composition of a basaltic komatiite and the former varies form that of an alkali-basalt to andesite; and (d) a test by a thermodynamic model from literature shows that, although with some doubt, the free energy of the possible parent liquid is higher than that of the two unmixed liquids.

U-Pb zircon SHRIMP data from the Cana Brava layered complex: new constraints for the mafic-ultramafic intrusions of Northern Goias, Brazil

Abstract The Cana Brava Complex is the northernmost and least well known layered intrusion of a discontinuous belt of mafic-ultramafic massifs within the Brasilia Belt, which also comprises theNiquelândia and Barro Alto complexes. Available geochronological data from a range of techniques (K/Ar, Ar/Ar, Rb/Sr, Sm/Nd and U/Pb) provide a range of possible ages (time span from 3.9 Ga to 450 Ma), hence a precise and reliable age for the Cana Brava Complex is still lacking. Also, preliminary isotopic and geochemical data of the Cana Brava Complex suggest a significant crustal contamination, which could have affected bulk-rock Sr and Nd systematics resulting in meaningless age determinations. In this paper, we present new U-Pb SHRIMP zircon analyses from four samples of different units of the Cana Brava Complexwhich suggest that the intrusion occurred during the Neoproterozoic, between 800 and 780 Ma, i.e. at the same age ofNiquelândia. Discordant older 206Pb/238U ages are provided by inherited zircons, and match the age of the metamorphism of the embedding Palmeirópolis Sequence.