Age, composition, and sougranites of the NeoproterProvince, Brazil

Elson P. Oliveira a, *, Juliana F. BuRosemery S. Nascimento d, Josea Department of Geology and Natural Resources, InstituSP, Brazilb John de Laeter Centre of Mass Spectrometry, School ofc Department of Geology, Federal University of Pernambd Faculty of Geology, Geosciences Institute, Federal Univ

e the Poo Redondobelt are the mac

edimentary domain.inho and Queimadae that their parentalmost probably from90e570 Ma graniteanites do not showical and Sr-Nd datahe Macurure micas-by the Belo Monte-el ow for magmael ow model pro-

. All rights reserved.

1. Introduction

Granites are of prime importance in studies of crustal evolutionbecause they are one of the main components of continental areas,are related in space and time with orogenic belts, and isolate

* Corresponding author.

Contents lists availab

Journal of South Ame

Journal of South American Earth Sciences 58 (2015) 257e280E-mail address: elson@ige.unicamp.br (E.P. Oliveira).by mixing between a juvenile mac source and a crustal component that could bmigmatites or the Macurure metasediments. Other 630e618 Ma granites in theenclave-rich Coronel Jo~ao Sa granodiorite and the Camara tonalite in the Macurure sThese granites have similar geochemical and isotopic characteristics as the LajedGrande granitoids. We infer for the Camara tonalite and Coronel Jo~ao Sa granodioritmagmas have had contributions from mac lower crust and felsic upper crust,underthrust S~ao Francisco Craton, or Pernambuco-Alagoas Domain. The younger 5group is conned to the Macurure metasedimentary domain. Although these grtypical features of S-type granites, their UePb age, eld relationships, geochemsuggest that their parental magmas have originated from high degree melting of tchists. Field observations support a model in which the Macurure domain, limitedJeremoabo and S~ao Miguel do Aleixo shear zones, behaved as a ductile channmigration and emplacement during the Neoproterozoic, very much like the channposed for emplacement of leucogranites in the Himalayas.

2014 Elsevier Ltdgranites are product of partial melting of Poo Redondo migmatites. Sr-Nd isotopes of the QueimadaGrande granodiorite and Lajedinho monzodiorite suggest that their parental magma may have originateda r t i c l e i n f o

Article history:Received 30 March 2014Accepted 6 August 2014Available online 21 August 2014

Keywords:Sergipano beltGranitesGeochemistryUePb geochronologySreNd isotopesWest Gondwanahttp://dx.doi.org/10.1016/j.jsames.2014.08.0030895-9811/ 2014 Elsevier Ltd. All rights reserved.rce of continental arc- and syn-collisionozoic Sergipano Belt, Southern Borborema

eno a, Neal J. McNaughton b, Adejardo F. Silva Filho c,P. Donatti-Filho a

te of Geosciences, P.O. Box 6152, State University of Campinas e UNICAMP, 13083-970 Campinas,

Applied Physics, Curtin University of Technology, Perth, WA 6845, Australiauco, 50740-530 Recife, PE, Brazilersity of Para, Belem 66075-110, Brazil

a b s t r a c t

The Sergipano belt is the outcome of collision between the Pernambuco-Alagoas Domain (Massif) andthe S~ao Francisco Craton during Neoproterozoic assembly of West Gondwana. Although the under-standing of the Sergipano belt evolution has improved signicantly, the timing of emplacement,geochemistry and tectonic setting of granitic bodies in the belt is poorly known. We recognized twogranite age groups: 630e618 Ma granites in the Caninde, Poo Redondo and Macurure domains, and 590e570 Ma granites in the Macurure metasedimentary domain. UePb SHRIMP zircon ages for granites ofrst age group indicated ages of 631 4 Ma for the Stios Novos granite, 623 7 Ma for the Poo Redondogranite, 619 3.3 Ma for the Lajedinho monzodiorite, and 618 3 Ma for the Queimada Grandegranodiorite. These granitoids are dominantly high-K calc-alkaline, magnesian, metaluminous, macenclave-rich (Queimada Grande and Lajedinho), or with abundant inherited zircon grains (Poo Redondoand Sitios Novos). Geochemical and isotope data allow us to propose that Stios Novos and Poo Redondojournal homepage: www.elsevier .com/locate/ jsamesle at ScienceDirect

rican Earth Sciences

snapshots of the superimposed deformations as they freeze part oftheir structural evolution. During ascent through the lithosphere,granitic magmas crosscut an expressive crustal thickness andentrain xenoliths from both their sources and the country rocks. Forthis reason, granites are of particular geological interest for directand indirect investigation of how the continental crust evolves.

Granites are a common component of the Sergipano belt buttheir ages and petrogenesis are only locally known (e.g. Silva Filhoet al., 1997; Guimar~aes and Silva Filho, 1995; McReath et al., 1998;Long et al., 2005). More uncertain is their tectonic signicance(e.g. Bueno et al., 2009), melt source(s) and evolution. The belt isone of the most important Precambrian orogenic belts of North-eastern Brazil, not only because it was considered as evidence forcontinental drift (e.g. Allard and Hurst, 1969), but also because itcontains several structural and lithologic domains that allow it tobe compared with Phanerozoic orogens (Oliveira et al., 2006, 2010).The Sergipano belt is located in the southernmost part of the Bor-borema Province (Fig. 1) and originated through collision betweenthe Congo-S~ao Francisco Craton and the Pernambuco-AlagoasDomain during the Neoproterozoic Brasiliano/Pan-AfricanOrogeny (e.g. Brito Neves et al., 1977; Van Schmus et al., 1995;Brito Neves and Fuck, 2013). It is a key belt for reconstructing part of

The Macurure, Vaza Barris and Esta^ncia are dominated by meta-morphic to non-metamorphic sedimentary rocks, whereas theother domains are more diverse and composed of igneous, meta-morphic and sedimentary rocks. Granites are abundant in theMacurure, Caninde, and Poo Redondo-Maranco domains (Fig. 2).

Threemain events of regional deformation are recognized in thesedimentary domains of the belt (Del-Rey Silva, 1995; Oliveiraet al., 2010; and references therein). The rst event is character-ized by south-verging D1 nappes and thrust zones, which probablydisplaced the metasedimentary rocks of the Macurure and VazaBarris domains for large distances over the edge of the S~ao Fran-cisco Craton in the south; a few granitic bodies were emplaced intothe Macurure Domain during or shortly after D1. The D2 event ismarked by reactivation of D1 and has a transpressive characterassociated with signicant vertical movements; most granite plu-tons were emplaced during this event. The D3 event is the lastductile deformation event in the Sergipano Belt and it took placeduring uplift of the belt in response to compression in a brittle toductile-brittle regime.

Part of the studied granites (Fig. 2) occurs in the Macururedomain, which is mostly composed of garnet micaschists withminor marble and quartzites. The Macurure domain was meta-

) in

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e280258the history of West Gondwana assembly.In this paper we present eld relationships, new UePb zircon

ages, whole-rock geochemichal and Sr-Nd isotope data for granitesof the Sergipano Belt as a contribution to understanding theirsource(s) and evolution. Our results indicated that crust- andmantle-derived magmas conributed to granite genesis, and thatboth arc-like and syn-collision granites are present in domains ofthe Sergipano belt.

2. The Sergipano Belt, NE-Brazil

The Sergipano Belt is a triangular shape orogenic belt withWNW-ESE direction (Fig. 1), located in the southern part of Bor-borema Province, NE-Brazil. It comprises ve lithostructural do-mains: Caninde, Poo Redondo-Maranco, Macurure, Vaza Barrisand Esta^ncia (Davison and Santos, 1989; Del-Rey Silva, 1995;Oliveira et al., 2006, 2010) limited each from the other by thefollowing major shear zones from north to south: Macurure, BeloMonte-Jeremoabo, S~ao Miguel do Aleixo and Itaporanga (Fig. 1).

Fig. 1. Symplied geology of the Sergipano Belt. A. Location of the Sergipano Belt (square

domains (modied after Oliveira et al., 2010). MSZ, BMJSZ, SMASZ and ISZ stand, respectivelzones.morphosed under amphibolite facies conditions and is separatedfrom the Vaza Barris Domain in the south by the S~ao Miguel doAleixo shear zone, and from the Poo Redondo-Maranco Domain inthe north by the Belo Monte-Jeremoabo shear zone. The originalsedimentary basin and its depositional settings are no longer easyto reconstruct owing to deformation and erosion. However, in lessmetamorphic or deformed portions of the Macurure domain,Davison and Santos (1989) recognized centimetre-thick rhythmiclayers of micaschists, with plane-parallel structures and abruptcontact indicative of deposition in deep water settings such asturbidites. Also, Oliveira et al. (2010) report on a sequence of chaoticblocks of mica-schist, phyllite, meta-rhythmite and rare graniteembedded in a meta-sandstone matrix; the entire rock packagewas subsequently deformed by D2. The authors interpret thissequence as an ancient alluvial fan and suggest a signicant timegap between the D1 and D2 deformation events.

The granites occupy large portion of the Macurure domain(Fig. 2) and are of two types: (1) pre-collisional granites (pre-toearly-D2 granites) including tonalite-granodiorite of restrict

Brazil. SFC-S~ao Francisco Craton, BP-Borborema Province. B. The Sergipano Belt and its

y for the Macurure, Belo Monte-Jeremoabo, S~ao Miguel do Aleixo and Itaporanga shear

umb- Pe- LajVBD

ericoccurrence with numerous micaschist xenoliths; these graniteswere later deformed during D2 and D3; (2) syn-collisional granites(syn-to late-D2 granites), which comprise mostly pink granites andless often grey granodiorites. The syn-collisional granites are sheet-like bodies that preserve magmatic structures such as mineralbanding, abundant schlieren and mac enclaves paralleling thehost rocks S2 foliation. These granites were injected as sheets alongthe F2 fold hinges and the axial plane foliation of micaschists, ul-timately forming plutons of large size. In some places, the granitescrosscut the schists S2 foliation (Bueno et al., 2009). Bueno et al.(2009) have obtained UePb (zircon, SHRIMP) age of 628 12 Mafor the pre-to early-D Camara tonalite and UePb (titanite, TIMS)

Fig. 2. Geological map of part of the Sergipano Belt showing the main granite plutons. N4- Santa Helena, 5- Canudos, 6- Formosa, 7- Itabi, 8- Gloria, 9- Capivara, 10- Carabas, 11and Caninde domains: 15- Stios Novos, 16- Queimada Grande, 17- Poo Redondo, and 18Caninde Domain; PRMD e Poo Redondo-Maranco Domain; MRD e Macurure Domain;Monte-Jeremoabo and S~ao Miguel do Aleixo shear zones.

E.P. Oliveira et al. / Journal of South Am2ages of 584 10 Ma and 571 9 Ma for the syn-to tardi-D2 Angicoand Pedra Furada granites, respectively. Using the age of theCamara tonalite as a maximum age for onset of the main collisionalevent (D2) in the belt and the age of the Pedra Furada granite as thewaning stage of D2 event, the authors have suggested that the maincollision, and its related magmatism, may have lasted at least 57million years.

The other studied granites occur in the Poo Redondo-Marancodomain and Caninde domain (Fig. 2). The Poo Redondo-Marancodomain was further divided into two sub-domains, namely PooRedondo and Maranco (Oliveira et al., 2010). The former iscomposed of migmatites and granites, and the latter by pelitic topsammitic metasedimentary rocks, rhythmites interleaved withcalc-alkaline andesite to dacite, intercalations of basalt, andesite,gabbro, and serpentinites. Granodiorite palaeosomes of the PooRedondomigmatites yielded twoUePb SHRIMP ages of 980Ma and961 Ma; they also show slightly negative to positive Nd(T) values,and dominant calcic to calc-alkaline geochemistry. On the otherhand, the swarm of dacite-andesite sills or lavas in slates andphyllites of the Maranco sub-domain are ca. 603 Ma old, show calk-alkaline to alkali-calcic major element signature, and negative Nd(T)values. Oliveira et al. (2010) suggested that the andesite and dacitesof the Maranco sub-domain and the protoliths of the migmatites ofthe Poo Redondo sub-domain formed in continental arcs.

The granitoids of the Poo Redondo-Maranco domain includethe Sitios Novos, Queimada Grande and Poo Redondo granitoids(Fig. 2). The Sitios Novos and Queimada Grande granitoids aretypical I-type granites to monzogranites, occasionally withnumerous mac enclaves, whereas the Poo Redondo granodioriteis more homogeneous. All of these granites are emplaced into thePoo Redondo migmatites.

The Lajedinho monzodiorite (Fig. 2) is another granite bodyselected for this study. It contains oriented mac enclaves, andoccurs in the Caninde domain. The Caninde Domain contains thefollowing lithodemic units: (i) The Novo Gosto-Mulungu unit ismade up of ne-grained amphibolites intercalated with phyllites,metasiltstones, metacherts, graphite schists, calc-silicate rocks andmarbles, cross-cut by mac and felsic dykes, granites and FeeTi-rich gabbros (Nascimento et al., 2005; Oliveira and Tarney, 1990;

ers refer to the following granitoids: Macurure domain: 1- Angico, 2- Areias, 3- Lagoas,dra Furada, 12- Monte Alegre, 13- Camara, 14- Coronel Jo~ao Sa; Poo Redondo-Marancoedinho. Insert is a zoom of the Garar region. PEAL e Pernambuco-Alagoas Massif; CD ee Vaza Barris Domain. MSZ, BMJSZ and SMASZ stand; respectively for Macurure, Belo

an Earth Sciences 58 (2015) 257e280 259Oliveira et al., 2010); (ii) The Garrote unit is a continuous, up to2 km-wide, strongly deformed granite sheet intrusive into rocks ofthe Novo Gosto-Mulungu unit; (iii) The Gentileza unit is made up ofamphibolites and diorites intercalated with porphyritic quartz-monzonite, and minor dolerite and gabbroic bodies; (iv) The Can-inde gabbroic complex comprises massive and layered olivine-gabbronorite, leucogabbro, anorthosite, troctolite, and minorpegmatitic gabbro, norite and peridotite. Granites, granodiorites,and rapakivi granites cross-cut these units. Oliveira et al. (2010)interpreted the Caninde Domain as a rift sequence that was laterdeformed and accreted to the Poo Redondo-Maranco Domain.

3. Granites: eld characteristics and petrography

The granites studied here can be separated into two age groups:630e618 Ma and 590e570 Ma. The rst group occurs in the Mac-urure, Poo Redondo-Maranco, and Caninde domains, whereas thesecond group is conned to the Macurure domain. Because in thelatter domain the deformation events are well established, Buenoet al. (2009) named the two granite groups as pre-collisionalgranites (pre-to early-D2 granites; 630e618 Ma) and syn-collisional granites (syn-to tardi-D2 granites; 590e570 Ma). In thePoo Redondo-Maranco domain the 630e618 Ma granite groupincludes the Stios Novos, Queimada Grande and Poo Redondogranitoids, whereas in the Caninde domain it includes only theLajedinhomonzodiorite (Fig. 2). Details of all of these granitoids arepresented below.

3.1. Granites in the Macurure domain

The Macurure domain pre-collisional granitoids include grano-diorites to tonalites. They are composed of quartz, biotite, horn-blende, plagioclase and epidote, and the accessories titanite, apatiteand allanite with epidote core. Representatives of the group are theCoronel Jo~ao Sa granodiorite (Long et al., 2005) and the Camaratonalite (Bueno et al., 2009). These granitoids contain mac en-claves (Fig. 3A) and/or xenoliths of garnet-biotite schists (Fig. 3B,Camara tonalite), and they are variably deformed. The Camaratonalite shows a penetrative foliation paralleling the micaschistfoliation, and contains quartz with undulose extinction and sub-grain boundaries, as well as quartz ribbons along the S2 foliation.The plagioclase shows mechanical twinning, undulose extinctionand subgrain boundaries; sometimes it is more deformed andshows recrystallized tail and pressure shadows with quartz. TheCoronel Jo~ao Sa granodiorite shows no record of penetrativedeformation in its central portion; the granite shows igneous foli-ation marked by attened mac enclaves and schlieren of elon-gated hornblende (Fig. 3C). On the other hand, macroscopicdeformation is common at its margin (Fig. 3D). The Coronel Jo~ao Sagranodiorite shows quartz and biotitewith undulose extinction andplagioclase with mechanical twinning.

The syn-collisional granites are comprised mostly by pinkgranites and less often by grey granodiorites. The granitoids arene-to medium grained, with equigranular texture and interlobategrain boundaries. They are composed mostly of quartz, microcline,

schists S2 foliation. Some of the granites are massive, whereasothers are strongly deformed with penetrative foliation, but all ofthem contain microstructures indicative of solid-state deformationsuch as undulose extinction of quartz and biotite, quartz and feld-spar recrystallization by subgrain rotation and boundary migration,and static recrystallization of quartz grains (Bueno et al., 2009).Where the granite plutons are intrusive into micaschists, the con-tact is knife sharp and rare contact metamorphism is observed.These granites contain large rafts of metasedimentary rocks(Fig. 4C) and abundant biotite-rich schlieren aligned parallel to theS2 schist foliation (Fig. 4D).

3.2. Granites in the Poo Redondo-Maranco domain

The main granite plutons in the Poo Redondo-Maranco domainare the Queimada Grande granodiorite, Stios Novos granite andPoo Redondo granite. These granitoids form E-Welongated bodies(Fig. 2), and are representatives of the 630e618Ma age group in theSergipano belt.

The Queimada Grande granodiorite (#16 in Fig. 2) is a largeintrusion in the domain, and contains many mac enclaves(Fig. 5A), as well as biotite-rich schlieren. It shows different struc-tures in the centre and at the margins. In the intrusion centre, thegranodiorite is porphyritic with centimetre-long K-feldspar phe-nocrysts; it does not show macroscopic deformational structures,but presents microstructures indicative of solid-state deformation

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e280260plagioclase, biotite, muscovite, and epidote. The accessory mineralsare allanite, zircon, apatite and titanite.

The syn-collisional granites in the Macurure domain are sheet-like bodies with preserved magmatic structures such as mineralbanding, abundant schlieren and mica-rich enclaves paralleling thehost rocks S2 foliation (Fig. 4A). These granites were emplaced assheets along the F2 fold hinges and axial plane foliation of themicaschists (Fig. 4B), eventually forming large plutons, such as theItabi granite (#7 in Fig. 2). In some places, the granites crosscut theFig. 3. Field aspects of Macurure domain 630e618 Ma granitoids. A) Camara Tonalite with mCoronel Jo~ao Sa granodiorite with aligned enclaves in its central portion; and D) deformedsuch as undulose extinction of quartz and biotite and quartzrecrystallization by boundary migration. On its northern andsouthern margins, the pluton is more deformed; it is ne-grainedwith biotite-rich schlieren and oriented mac enclaves, whichdene a high angle magmatic foliation. This part of the granodioriteshows undulose extinction of quartz and biotite, quartz and feld-spar recrystallization by subgrain rotation and boundary migration,static recrystallization of quartz grains, and plagioclase with me-chanical twinning. In the northern contact with the Poo Redondomigmatite the Queimada Grande granodiorite foliation parallelsac enclaves; B) and deformed xenoliths of garnet-biotite schist; C) Isotropic feature ofby D2 in the northern contact with the Macurure micaschist.

that of the migmatite and in some places offshoots of the grano-diorite are conformable with the migmatite's foliation indicating

boundary migration. A few mac enclaves occurs occasionaly(Fig. 5B). The contact with the Queimada Grande granodiorite is

Fig. 4. Field aspects of the syn-to late-D2 granites in the Macurure domain. A) Santa Helena granite (#4 in Fig. 2) emplaced and parallel to the S2 foliation of host micaschists; B)Angico granite (#1 in Fig. 2) emplaced along the axial plane of F2 fold; C) Angico granite with rafts of metasedimentary rocks; D) Areias granite (#2 in Fig. 2) with biotite-richschlieren paralleling the micaschist S2 foliation. Insert shows the structural interpretation.

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e280 261that the pluton was emplaced during the last migmatization event.The Stios Novos granite (#15 in Fig. 2) is a pink, ne-to medium

grained granite with equigranular texture. The granite is composedof quartz, plagioclase, K-feldspar and biotite. At mesoscopic scalethis granite does not show deformation, but presents microstruc-ture representative of solid-state deformation such as unduloseextinction of quartz and biotite, and quartz recrystallization byFig. 5. Field aspects of granites in the PooRedondo-Maranco and Caninde domains. A) Queimshowing a few mac enclaves; C) Poo Redondo granite with off-shoots into the host migmasharp.The Poo Redondo granite (#17 in Fig. 2) is dominantly grey,

ne-to medium grained, with equigranular texture; it is composedof quartz, plagioclase, K-feldspar, muscovite, andminor biotite. Likethe Sitios Novos granite, the Poo Redondo granite does not showmacroscopic deformation structures but presents microstructuresindicative of solid-state deformation; however the Macurure shearadaGrandeGranodioritewith numerousmac enclaves; B) isotropic Sitios Novos granitetites (dark grey); D) Lajedinho monzodiorite showing oriented mac enclaves.

zone at north deformed it. The contact between the Poo Redondogranite and the migmatites is intrusive (Fig. 5C), and in severalplaces the granite contains migmatite xenoliths.

3.3. Granite in the Caninde domain

The Lajedinho monzodiorite (#18 in Fig. 2) in the Canindedomain is also representative of the 630e618Ma granite age group.It was emplaced into metadiorite and amphibolites of the Gentilezaunit with zircon UePb age of ca. 688 Ma (Oliveira et al., 2010). Themonzodiorite entrains elongated mac enclaves (Fig. 5D) and iscomposed of hornblende, plagioclase, quartz, and minor K-felsdpar,apatite and zircon.

4. UePb SHRIMP zircon dating

Here we present new UePb ages for three granite plutons fromthe Poo Redondo-Maranco domain and one from the Canindedomain. These granites alongwhith the 625Ma-old Coronel Jo~ao Sagranodiorite (Long et al., 2005) and the 628Ma-old Camara tonalite(Bueno et al., 2009) make up the 630e618 Ma granite age group inthe Sergipano Belt. The geographic coordinates of each datedsample are given in Fig. 6.

Zircons were dated with the Sensitive High Resolution IonMicroprobe using the Perth Consortium SHRIMP II at the CurtinUniversity of Technology,Western Australia, based on the operationprocedure described by Compston et al. (1984) and operationconditions described by Smith et al. (1998). After separation withconventional gravimetry and magnetic techniques, the zircongrains were mounted in epoxy resin along with ships of BR266zircon standard (U 550 ppm; 206Pb/238U 0.0914), and polished

observations selected grains were imaged on the Scanning ElectronMicroscope (SEM) for qualitatively analysis of morphology andinternal structure. The UePb SHRIMP analysis followed the oper-ational procedures described by Compston et al. (1984) with cyclesof 7-scan for granite, incident O2- ray of 2 nA andmass resolution of5000 ca. The data were reduced using the SQUID software (Ludwig,1999a) and ISOPLOT (Ludwig,1999b). The ages reported here are for206Pb/238U with between 95% and 105% concordance. Pooled agesare quoted with 95% condence level errors. The age uncertaintiesin relation to the concordia intercept are around 1s.

Zircon grains from the Stios Novos granite gave the age of631 4 Ma; those from the Poo Redondo granite the age of623 7 Ma, those from the Queimada Grande granodiorite the ageof 618 4 Ma, and the Lajedinho monzodiorite the age of619 3 Ma (Table A1, Fig. 6). The Poo Redondo and Sitios Novosgranites contain numerous Early Neoproterozoic inherited zircongrains (Table A1).

5. Major- and trace-element geochemistry

Geochemical data were acquired for 82 samples from 18 gran-ites in the Macurure, Poo Redondo-Maranco, and Caninde do-mains. The samples are from random localities in each pluton(Fig. 2).

Major elements were analyzed on fusion beads and trace ele-ments on pressed powder pellets using a Philips PW-2404 X-rayspectrometer at the Geochemistry Laboratory of Campinas Uni-versity. The fusion beads were made with a mixture of lithiummetaborate and tetraborate (80/20 p/p e Spectroux 100B JonhsonMattey/USA) in the 5:1 proportion (melter/sample) in a Fluxy 300melting equipment. The powder pellets were prepared by mixing

ain

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e280262to half of mean grain thickness for further imaging. After optical

Fig. 6. Concordia diagrams for granites in the Poo Redondo-Maranco and Caninde dom (sample CRN-11 e W37.69 ; S9.80 ); C. Queimada Grande granodiorite (sample JUD-91 e

Table A1 show all analyzed grains.9 g of the sample with 1.5 g of wax and pressed in a hydraulic press.

s: A. Stios Novos granite (sample JUD-96 eW37.62; S9.92); B. Poo Redondo granite W37.66 ; S9.96 ); D. Lajedinho monzodiorite (sample CRN-109B e W37.79 ; S9.63 ).

The quality control has been done by comparison with interna-tional standard samples AC-E, WS-E and RGM-1, and quadrupli-cating of two studied samples. The precision is 0.1% for SiO2 andAl2O3, 0.01% for the other major elements and less than 2 ppm forthe trace elements. The data for La, Ce, Nd and U are only indicative.Representative whole-rock analyses are given in Table A2. Addi-tional rare earth elements and other trace elements of selectedsamples were analysed on a Thermo (Xseries2) quadrupole ICP-MSfollowing the in-house adapted analytical procedures of Egginset al. (1997) and Liang et al. (2000), and instrument conditions ofCotta and Enzweiler (2009); the results have less than a 10% de-viation from the recommended values for the international stan-dards BRP-1, RGM-1 and GSP-2.

Fig. 7 shows the chemical classication of Debon and Le Fort(1983) for the studied granitoids. Most granitoids of the590e570Ma age group are relatively homogeneous in compositionvarying from granodiorite to granite, with only two samples plot-ting in the syenite or quartz-syenite elds. On the other hand,granitoids of the older age group show larger compositional vari-

plutons that show iron enrichment (Fig. 8C). In this diagram, the590e570Ma granite group spreads over the ferroan andmagnesianelds (Fig. 8C). The CaO, Na2O and K2O relationships (Frost et al.,2001) for the studied granitoids indicate that the 590e570 Magranite group is mainly alkali-calcic to alkalic, whereas the oldergranite group is calc-alkalic to alkali-calcic (Fig. 8D).

Themost relevant trace element characteristics of the SergipanoBelt granites are illustrated in Figs. 9 and 10. Fig 9 shows theprimitive mantle-normalized multi-element diagrams (spider-gram) for representative samples of each granite group, along withaverages for oceanic and continental arcs. As shown in Fig. 9A, the630e618 Ma granite group has little similarities with felsic rocksfrom oceanic arcs, at least for the elements Sr to Th; the group hastrace element signature very much like the felsic rocks of conti-nental arcs. On the other hand, representatives of the younger,590e570 Ma granite group are more scattered in the diagram andtheir patterns are not similar to either the oceanic or continentalarc felsic rocks. This is particularly true for the very right end of thediagram where the elements Gd to Yb show low abundances and

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e280 263ation. For instance, the Lajedinho pluton of the Caninde domain iscomposed dominantly of quartz-monzodiorite, whereas the Quei-mada Grande granitoid of the Poo Redondo-Maranco domainvaries from quartz-monzodiorite through quartz-monzonite andgranodiorite to monzogranite (adamellite). Similarly, the SitiosNovos and Poo Redondo granitoids show signicant variation andare represented by granodiorite to granite. On the other hand, theCamara and Coronel Jo~ao Sa plutons are more homogenous, withgranodiorite composition.

Other chemical characteristics of the Sergipano Belt granitoidsare shown in Fig. 8. In the Al2O3/(Na2O K2O) molar vs. Al2O3/(CaO Na2O K2O) molar diagram (Fig. 8A) most 630e618 Magranitoids are metaluminous with some samples of the PooRedondo granite and also some from the Macurure domainshowing a tendency to peraluminous. The 590e570 Ma granites ofthe Macurure domain straddle the elds of metaluminous to per-aluminous granites. Considering 1.1 as a limit between I- and S-typegranites, the older granite group is composed mainly of I-typegranites (Fig. 8A) and the younger group of I-type with a fewsamples of the S-type; in this gure the Poo Redondo plutonshares similar geochemical characteristics with the 590e570 Magranites. The majority of samples are high-K calc-alkaline granites(Fig. 8B) with only one sample of the Sitios Novos pluton plotting inthe shoshonite eld (Fig. 8B). Most 630e618 Ma granitoids aremagnesian; the exceptions are the Lajedinho and Poo RedondoFig. 7. Chemical-mineralogical classication of the two granites age groups of the SergipanFields after Debon and Le Fort (1983).steep patterns. These characteristics can be more clearly seen inFig. 10, a plot of Y vs. Sr and (Gd/Yb)n vs. (La/Yb)n. As shown, theyounger granite group has considerably lower Y values, and higher(Gd/Yb)n and (La/Yb)n ratios than the older granite group, a featurethat may be associated with garnet left in the residue duringgranitic magma production by partial melting of a garnet-richsource, such as the garnet-bearing Macurure micaschists.

In the tectonic setting discrimination diagram of Pearce et al.(1984) samples of the older granite group plot in the eld of arcgranites with a few samples of the Sitios Novos granite plotting alsoin the syn-collision eld; all the Lajedinho monzodiorite samplesplot in the within-plate eld (Fig. 11A). Similar behaviour isobserved for samples of the younger granite group but in this case agreater number of samples plot in the syn-collision granite eld(Fig. 11B).

6. Nd and Sr isotope geochemistry

Sr and Nd isotope analyses were performed in the Geochro-nology Laboratory of the University of Braslia following the tech-niques of Gioia and Pimentel (2000). Approximately 60 mg ofpowdered rock samples were dissolved for Sr, Sm, and Nd extrac-tion in successive acid attacks with concentrated HF, HNO3, andHCl. A mixed 149Sme150Nd spike was added to the solution beforethe rst acid attack. Sr and the REE group were separated from theo Belt. A) 630e618 Ma plutons; B) 590e570 Ma plutons. Q quartz, P plagioclase.

K2ndt et

ericFig. 8. Geochemical characteristics of granites from the Sergipano Belt. A) Al2O3/(Na2O 1989), dashed line represents the boundary between I- and S-type granites (Chappell aitalics, and Rickwood (1989) in parentheses; C) SiO2 vs FeOt/(FeOt MgO) diagram (Fros

E.P. Oliveira et al. / Journal of South Am264whole-rock solutions using a conventional ion exchange. Subse-quently, Sm and Nd were extracted by reverse-phase chromatog-raphy in columns packed with HDEHP (diethylhexyl phosphoricacid) supported on PTFE powder. Sr, Sm, and Nd aliquots wereloaded onto double Re evaporation laments, and the isotopicmeasurements were carried out on a multicollector Finnigan MAT-262 mass spectrometer in static mode. Mass fractionation correc-tions were made using a 88Sr/86Sr ratio value of 8.3752. 1s uncer-tainty on the measured 87Sr/86Sr ratio was better than 0.01%. ForSm/Nd and 143Nd/144Nd ratios, the uncertainties are better than0.1% (2s) and 0.003% (2s), respectively, after repeated analyses ofinternational rock standards BCR-1 and BHVO-1. The 143Nd/144Ndratios were normalized to a 143Nd/144Nd ratio of 0.7219. Nd and Srprocedure blanks were less than 150 and 300 pg, respectively. TheTDM values were calculated using the model of DePaolo (1981).

Sr and Nd isotopic analyses were carried out for 35 samples fromthe Macurure and Poo Redondo-Maranco rocks (Tables A3 andA4). Initial Nd and 87Sr/86Sr ratios for 630e618 Ma granites andPoo Redondo migmatite were calculated to 625 Ma, whereasinitial Nd and 87Sr/86Sr values for 590e570 Ma granites and Mac-urure schists the age of 580Mawas chosen on the basis of the UePbzircon and titanite ages.

In the Macurure domain, the Camara tonalite is one of the630e618Ma granites and has a (87Sr/86Sr)i ratio of 0.70916, Nd(t) of-7.45 and TDM of 1.71 Ga (Tables A3 and A4, Fig. 12A). The other630e618 Ma granite group is the Coronel Jo~ao Sa granodiorite thathas been previously studied by Silva Filho et al. (1997), McReathet al. (1998) and Long et al. (2005). These authors obtained(87Sr/86Sr)i ratios for the Coronel Jo~ao Sa granodiorite ranging from0.7123 to 0.7167, Nd(t) values of -4.8 to -6.9 (Fig. 12A), and TDMfrom 1.50 to 1.70 Ga. The 590e570 Ma granites in the Macururedomain show a large range in (87Sr/86Sr)i ratios (0.70782 toO) molar vs. Al2O3/(CaO Na2O K2O) molar diagram (modied by Maniar and Piccoli,White, 1992); B) SiO2eK2O diagram with nomenclature after Le Maitre et al. (1989) inal., 2001); D) SiO2 vs Na2O K2OeCaO diagram (Frost et al., 2001). Symbols as in Fig. 7.

an Earth Sciences 58 (2015) 257e2800.71219) and Nd(t) values (-1.63 to -11.79) (Fig. 12A), but present arange in TDM, varying from 1.22 to 1.86 Ga. The Macurure micas-chists show a very large range in (87Sr/86Sr)i ratios from 0.70515 to0.76379, Nd(t) ranging from -1.89 to -7.49 (Fig. 12B), and TDMranging from 1.37 to 1.78 Ga.

The 630e618 Ma granite group in the Poo Redondo-Marancodomain gave the following results (Fig. 12A): (i) Queimada Grandegranodiorite: (87Sr/86Sr)i ratios ranging from 0.70656 to 0.70789,Nd(t) slightly negative ranging from -1.15 to -2.55 and TDM varyingof 1.18 to 1.32 Ga; (ii) Stios Novos granite: (87Sr/86Sr)iratio 0.71164, Nd(t) -5.47 and TDM 1.51 Ga and (iii) PooRedondo granite: (87Sr/86Sr)i ratios ranging from 0.71353 to0.71417, Nd(t) ranging from -4.23 to -5.50 and TDM varying of 1.40to 1.57 Ga. Two analyses were obtained for the Poo Redondomigmatites and the data are: (87Sr/86Sr)i ratios (0.71066e0.71832),Nd(t) values (-1.47 to -5.65) and TDM 1.48 Ga.

The Lajedinho monzodiorite is a representative of the 630e618Ma granites in the Caninde domain; its Nd isotope analyses(Table A3) are from Nascimento (2005). Initial Nd values for Laje-dinho monzodiorite were calculated to 625 Ma. This granite hasNd(t) values ranging from -1.10 to -0.08 and TDM from 1.14 to 1.22Ga (Table A3).

7. Discussion

7.1. Sources for the 630e618 Ma granites in the Sergipano belt

7.1.1. 630e618 Ma granites in the Poo Redondo-Maranco domainThe oldest recognized Neoproterozoic granites of the Poo

Redondo-Maranco domain are the Stios Novos (631 4 Ma) andthe Poo Redondo granites (623 7 Ma). The Stios Novos is a high-K calc-alkaline to alkaline granite, metaluminous, magnesian, with

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e280 26587Sr/86Sr(i) ratio of 0.71163, Nd(t) of -5.47, and with depletedmantle Nd model age (TDM) of 1.51 Ga. Poo Redondo granite is ahigh-K calc-alkaline to alkali-calcic granite, peraluminous, ferroan,with 87Sr/86Sr(i) ratio ranging from 0.71352 to 0.71417, Nd(t) from-4.23 to -5.50, and TDM varying from 1.40 to 1.57 Ga. In the tectonicsetting discrimination diagram (Pearce, 1996), the Poco Redondogranite falls in the arc eld and the Sitios Novos in both the arc andsyn-collision elds (Fig. 11A). Given that the Poo Redondo andSitios Novos granites contain several early Neoproterozic inheritedzircons, potential sources for their parental magmas are the ca.980e960 Ma Poo Redondo migmatites (Oliveira et al., 2010) andthe Maranco and Macurure metasediments (with ca 900e1000 Madetrital zircons; Carvalho, 2005).

Fig. 10. Trace element characteristics of the granites from the Sergipano Belt. A) SreY di590e570 Ma granite group (X-ray uorescence data); B) Chondrite-normalized (Evensen ethigh ratios of the younger group (ICP-MS data).

Fig. 9. Primitive mantle normalized multi-element diagram of representative samples ofContinental and oceanic arcs after Condie and Kroner (2013).In the Nd isotope evolution diagram (Fig. 13A), the data for thegranites plot in the elds of Poo-Redondo migmatite, Marancometasediments, Macurure micaschists, and Pernambuco-AlagoasMassif rocks. On the Nd(625) versus (87Sr/86Sr)i diagram(Fig. 13B), data for Stios Novos granite plot in the eld of the PooRedondo migmatite samples. Similar conclusion holds for the PooRedondo granite, which plot partially in the same eld and in theMacurure micaschist eld. Based on the very close crystallizationages of the two granites (631 4 Ma and 623 7 Ma), their similarhigh-K calc-alkaline, continental arc geochemical signatures,intrusion into migmatites, and Nd-Sr isotope characteristics akin tothe Poo Redondo migmatites, we suggest that partial melting of asource similar to the Poo Redondo migmatites was the most likely

agram showing the higher Y abundances of the 630e618 Ma granite group than theal., 1978) Gd/Yb-La/Yb diagram for the two granite age groups showing the distinctive

the two granite age groups. Normalizing values from Sun and McDonough (1989).

petrogenetic model for genesis of Poo Redondo and Stios Novosgranites. According to Oliveira et al. (2010), the 980e960 Ma-oldmigmatite paleosomes of the Poo Redondo-Maranco domain arecalc-alkaline to calcic and have positive to slightly negative Nd(t)values suggesting similarity with continental arc granites; the PooRedondo and Stios Novos granites may have inherited their arc-like major and trace element geochemical signature from theirpossible source, i.e. the Poo Redondo migmatites.

The Queimada Grande granodiorite is another pluton in thePoo Redondo-Maranco domain (Fig. 2). Rocks from this plutonpresent less radiogenic 87Sr/86Sr(i) ratios in the range 0.70656 to0.70789, Nd(625) slightly negative (-1.15 to -2.55), and TDM varyingfrom 1.18 to 1.32 Ga (Tables A3 and A4, Figs. 10 and 11). Thecalculated Nd(625) for the Queimada Grande granodiorite isamongst the most juvenile so far found in granites of the Macurureand Poo Redondo-Maranco domains, suggesting contributionfrom a juvenile source in its genesis. In the Nd(t) evolution diagram(Fig. 13A), the data for the granodiorite plot in the elds of Marancometasediments, Pernambuco-Alagoas Massif, and partially in theeld of Macurure micaschists and Poo-Redondo migmatite. In theNd(625) versus (87Sr/86Sr)i diagram (Fig. 11B), samples of the

7.1.2. 630e618 Ma granites in the Macurure domainThe 630e618 Ma granite group in the Macurure domain is

represented by the Camara granodiorite and Coronel Jo~ao Sagranodiorite (Fig. 2). These granitoids are deformed in differentscales by the tectonic events that affected the Macurure domain.The two plutons have similar UePb zircon ages, i.e. 628 12 Ma forthe Camara granodiorite (Bueno et al., 2009) and 625 2Ma for theCoronel Jo~ao Sa granodiorite (Long et al., 2005).

Long et al. (2005) commented on the petrogenesis of the CoronelJo~ao S~ao granodiorite. On the basis of eld relationships and Sr andNd isotope these authors suggested that the granodiorite originatedby partial melting of a basaltic lower crust, represented byamphibolite xenoliths entrained in the granodiorite, and an un-known crustal source represented by zircons with inherited cores.These authors proposed that the likely source candidates of appro-priate age and Sm-Nd isotope characteristics could be the Archaean/Palaeoproterozoic S~ao Francisco craton and early Neoproterozoic(Cariris Velhos) material. Long et al. (2005) discussed that the dataavailable in their study were not sufcient to draw quantitativeconclusions about magma sources. As a qualitative conclusion theauthors proposed that the granodiorite magma was the product of

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e280266Queimada Grande granodiorite are very distinct of the othergranites in the Poo-Redondo Maranco domain, suggesting limited,or no contribution from the Poo-Redondomigmatite to its genesis.

The Queimada Grande granodiorite shows trace element abun-dances characteristic of volcanic arc granitoids such as depletion inNb and Ti relative to Rb, Ba and K (Fig. 9). Enrichment of these ele-ments is generally assigned to uids from subducted sediments orsubducted ocean crust (Briqueu et al., 1984; Pearce et al., 1984;Pearce, 1996). In the diagram proposed by Pearce (1996) the grano-diorite plots in the volcanic arc eld (Fig. 11A). This pluton showscharacteristics of I-type granite such as mol Al2O3/(CaO Na2O K2O) < 1.1 (Fig. 6A). The granodiorite showsnumerous aligned mac xenoliths that are evidence of magmamixing andmingling. The combined eld relations,major- and trace-elements, and Nd-Sr isotopic data suggest that the Queimada Grandegranodiorite has a strong subduction signature, having been formedpossibly in a continental arc by magma mixing between a mantle-derived mac crustal source that has previously experienced sub-duction zone element depletion and a crustal component that couldbe represented by the Maranco and Macurure metassediments, aswell as Pernambuco-Alagoas rocks (Figs. 16A, B).Fig. 11. Trace element tectonic setting discrimination diagram for granites of the Sergipartial melting of local heterogeneous crustal sources.The plot of Nd(625) versus crystallization age shows that the

data of Long et al. (2005) for Coronel Jo~ao Sa granodiorite fall in theelds of the Poo Redondo migmatite, Maranco metasediments,Macurure micaschists and Pernambuco-Alagoas Massif rocks(Fig. 13C). In the Nd(625) versus (87Sr/86Sr)i diagram (Fig. 13D), thesamples of the Coronel Jo~ao Sa granodiorite plot partially in theeld of Macurure micaschists. Representatives of Cariris Velhosrocks in the Sergipano belt, the Poo Redondo migmatite, are alsoshown in the Nd(625) versus (87Sr/86Sr)i diagram in order to test forits likelihood as source for the Coronel Jo~ao Sa granodiorite magma.Accordingly, this suggest that the Poo Redondo migmatites do notappear to have contributed with material to magma source of thegranodiorite as proposed by Long et al. (2005).

Another example of 630e618 Ma granitoids in the Macururedomain is the Camara tonalite (Fig. 2). This pluton also containsamphibolite xenoliths (Fig. 3A) but unlike the Coronel Jo~ao Sagranodiorite it entrains xenoliths of deformedMacurure micashists(Fig. 3B). The Camara tonalite has (87Sr/86Sr)i ratio of 0.70916, Nd(t)of -7.45 and TDM of 1.71 Ga (Tables A3 and A4, Fig. 12A). In the Ndevolution diagram (Fig. 13C), the tonalite plot into the elds forpano Belt (Pearce et al., 1984). A) 630e618 Ma granites; B) 590e570 Ma granites.

Marancometasediments, Pernambuco-AlagoasMassif and partiallyin the eld of Macurure micaschists. In the Nd(625) versus(87Sr/86Sr)i diagram (Fig. 13D), the tonalite plots below the eld forthe Macurure micaschists.

The Camara tonalite shows isotope and geochemical charac-teristics very similar to the Coronel Jo~ao Sa granodiorite. Bothgranitoids were emplaced pre-to early-D2 (Bueno et al., 2009) intoMacurure micaschists, have mac enclaves and similar crystalli-zation age. As such, they may have had similar petrogenetic evo-lution, especially regarding the sources of melting. According to

Long et al. (2005) simple fractional crystallization is not appli-cable in the petrogenesis of Coronel Jo~ao Sa granodiorite, becausethis process cannot account for the range of variation in initial Srand Nd isotopic compositions. Our data together with the dataobtained by Long et al. (2005) suggest that the source for bothCoronel Jo~ao Sa granodiorite and Camara tonalite may be a mixturebetween at least two end-members represented by basaltic lowercrust (amphibolite enclaves) and an upper continental crustcomponent. The magma must have been contaminated with theMacurure micaschists because the granites show many zircon

Fig. 12. Neodimium and Strontium isotope diagrams for rocks of Macurure and Poo Redondo-Maranco domains. A) Plot of Poo Redondo migmatite and 630e618 Ma granitoids:Queimada Grande granodiorite, Stios Novos granite, Poo Redondo granite, Camara tonalite, and Coronel Jo~ao Sa granodiorite. Data for Coronel Jo~ao Sa granite are from Long et al.(2005); B) Plot of 590e570 Ma granites and Macurure micaschists.

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e280 267Fig. 13. Sr-Nd isotope characteristics of 630e618 Ma granitoids in the Sergipano Belt. A) Egranite and Stios Novos granite and their possible source components. Data for Poo RePernambuco-Alagoas Massif (PEAL) are from Silva Filho et al. (2002); B) Nd(t) vs (87Sr/86Scomponents; C) evolution of Nd for the Camara tonalite and Coronel Jo~ao Sa granodiorite, anSa tonalitic enclave and Coronel Jo~ao Sa amphibolite xenolith are from Long et al. (2005); D)possible source components.volution of Nd(625) with time for the Queimada Grande granodiorite, Poo Redondodondo migmatites and Maranco metasediments are from Carvalho (2005). Data forr)i plot for granites in the Poo Redondo-Maranco domain and their possible sourced their possible source components. Data for Coronel Jo~ao Sa granodiorite, Coronel Jo~aoNd(t) vs (87Sr/86Sr)i plot for the 630e620 Ma granites in the Macurure domain and their

grains with inherited core, and this contamination may have beenmore intense in the Camara tonalite because it shows more nega-tive Nd(625) values (and micaschist xenoliths) than the Coronel

natively, that the micaschists may have contributed material to the

samples of the Formosa granite that presents one Nd(t) value of-4.39 and another of -11.79. Fig. 14B shows a plot of Nd(t) vs(87Sr/86Sr)i for the Macurure syn-collisional granites and associated

Fig. 14. Sr-Nd isotope characteristics of 590e570 Ma granitoids in the Sergipano Belt. A) Evolution of Nd(580) for the granites and their possible source components. Data for PooRedondo migmatites and Maranco metassediments are from Carvalho (2005). Data for Pernambuco-Alagoas Massif (PEAL) are from Silva Filho et al. (2002); B) Nd(t) vs (87Sr/86Sr)ifor the granites and Macurure micaschists.

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e280268bulk geochemistry. There is only one exception observed inJo~ao Sa granodiorite.

7.2. Source(s) for the 590e570 Ma granites in the Macururedomain

Our Sm-Nd and RbeSr isotope data provide insights intopossible magma sources for the 590e570 Ma Macurure granites.The granites Nd(580 Ma) variation from -1.63 to -11.79 (Fig. 14) isobserved also in the Macurure micaschists, Maranco metasedi-ments, Poo Redondo migmatites, and in rocks of the Pernambuco-Alagoas Massif (PEAL) (Fig. 14A); these rock units can be potentialmelting sources for the Macurure granites. The majority of theNd(t) values shown by the 590e570 Ma granites are in the rangefound for theMacurure schists (Table A3 Fig 14), suggesting that theMacurure micaschists were sources for granite magma or, alter-Fig. 15. Geochemical modelling for the 590e570 Ma syn-collisional granite group. A) Nd(58The grey rectangles represent Macurure micaschists and S~ao Francisco Craton rocks. A and Bas end-members of compositional modelling. The ornamented square represents the calcula100% of each component. Note that the majority of samples for the 590e570 Ma granitesMacurure micaschists eld. B) Nd(580) vs Nd for the 590e570 Ma granites plotted as inplagioclase fractionation in granititc magmas.micaschists. The majority of granites plot into the eld representedby the micaschists, suggesting that they might have been partialmelts of the latter. Another piece of evidence that supports thishypothesis is the granites TDM ages in the range 1.26-1.78 Ga(Table A3), which are very similar to TDM values for Macururemicaschists, from 1.36 to 1.76 Ga. The more negative Nd(t) value(-11.16) of the Formosa granite, which plots well below the eld forMacurure micaschists in Fig. 14B, can be explained by source mix-ing, or contamination with less radiogenic crust such as the S~aoFrancisco craton, which is likely to have imbricated in depthbeneath the Macurure metasedimentary domain according to theevolution model of Oliveira et al. (2010) for the Sergipano Belt.

In order to test the hypothesis of source mixing in genesis of the590e570 Ma granites, we performed quantitative modelling usinga simplemixture equation for Nd and Nd(t) datawith theMacururemicaschists and rocks of the S~ao Francisco Craton as end-members.The obtained results (Fig. 15A) match very well with the Macurure0) vs Nd source mixing for the 590e570 Ma granite group represented as black ellipses.represents samples of Macurure micaschist and S~ao Francisco Craton, respectively, usedted composition for a mixture between the two end-members in a 0, 20, 40, 60, 80 andplot in between the two compositional calculated lines for mixture and mostly in thedividual plutons. The arrow indicates the theoretical trend of quartz, K-feldspar, and

raniOlie Ca

ericmicaschists as the major component in genesis of the syn-collisional granites. The majority of 590e570 Ma granites plot inthe micaschists eld and in between the two compositionalcalculated lines for the mixture between the Macurure micaschistsand the S~ao Francisco Craton rocks. In Fig. 15B, the possibility offractional crystallization as a major process in genesis of thegranites is less likely because this process cannot account for theobserved large variation of Nd(t) values. Fractional crystallizationmay have been signicant only in the Monte Alegre and Areiasgranites.

The Macurure domain is a typical sedimentary domain of theSergipano belt bound by two regional-scale shear zones, namely theBeloMonte-Jeremoabo to the north and the S~aoMiguel do Aleixo tothe south (Fig. 2). This domain has undergone a substantial crustalcompression during the main deformation event D2 that probablyfacilitated high degrees of partial melting of the metasedimentary

Fig. 16. A) Evolution of Nd(625) vs TDM for the Poo Redondo-Maranco and Macurure gCarvalho (2005), Macurure schists and Coronel Jo~ao Sa granodiorite are respectively fromvs. geographic location (from North to South) of the possible arc-type granites from th

E.P. Oliveira et al. / Journal of South Ampile to form the parentalmagmas of the 590e570Ma granites. In theeasternmost part of the Macurure domain the michaschists showmigmatization features such as quartz-feldspar segregation. Thecontact between the 590e570 Ma granites and the micaschists areknife sharp, and intrusion of the granite barely caused contactmetamorphism in the metasediments (Santos et al., 1988). Thesegranites occasionally contain large rafts of metasedimentary rocks(Fig. 4C), apparently split apart during granite intrusion.

According to Bueno et al. (2009), eld observations support thesuggestion that the granitic magmas migrated/crystallized alongthe S2 axial plane foliation and were collected at the hinge zones ofF2 folds. In this scenario, the axial plane foliation probably acted as achannel for magma migration and collection to form large-scalebatholiths. The space necessary for granite emplacement alongthe country rock's axial plane foliation may have been generated byhydraulic fracturing in a scenario similar to that suggested forleucogranites in the Himalayas (Searle et al., 2003).

The 590e570Ma granites show characteristics of I-type granites(Fig. 8B, D), such as decrease of P2O5 with increasing SiO2, positivecorrelation between Pb and SiO2, Al2O3/(CaO Na2O K2O) < 1.1and titanite. According to the original characteristics of S-typegranites (Chappell and White, 1974, 2001) these features conictwith derivation of the 590e570 Ma granites by partial melting oftheMacururemicaschists. However, the S-type characteristicsweredened in granites originated by melting of pelites of the LachlanFold Belt. The Macurure Sequence is more Ca-rich and melting of itcan generate magmas with geochemical signatures similar to thoserecorded in the 590e570 Ma granites.

Oliveira et al. (2005) have obtained UePb SHRIMP detrital zircondata for Macurure domain quartzite and micaschist. The maincluster of zircon ages is around 980 Ma with some zircon grainsshowing Archaean ages (2.8 Ga and 3.1 Ga). According to theseauthors the protoliths of the Macurure metasediments resulteddominantly from erosion of sources with ages between 1.0 and 2.0Ga, with a few grains coming from Archaean sources. More signif-icantly, no zircon grain younger than 900 Ma was observed, whichindicate a maximum deposition age for the Macurure sedimentswell before the Neoproterozoic Brasiliano orogeny, possibly shortlyafter the ca. 1.0 Ga Cariris Velhos orogeny (Brito Neves et al., 1995;Santos et al., 2010; Van Schmus et al., 2011). This afrmative isconrmed by the Sm-Nd isotope data of Oliveira et al. (2005) that

tes and their possible source components. Data for Maranco metassediments are fromveira et al. (2010) and Long et al. (2005); depleted mantle from DePaolo (1981); B) Nd(t)ninde domain through the Poo Redondo-Maranco domain to the Macurure domain.

an Earth Sciences 58 (2015) 257e280 269indicate Nd model ages (TDM) for metasediments varying from 1.2to 1.8 Ga. These TDM are very similar to those found for the PooRedondo-Maranco domain and Pernambuco-Alagoas Massif rocks.According to detrital zircon ages and the TDM, the protoliths ofMacurure domain clastic metasediments may have originated byerosion of rocks from the Poo Redondo-Maranco domain and thePernambuco-Alagoas Massif.

The Poo Redondo-Maranco Domain comprises metasedi-mentary rocks, volcanic and plutonic rocks within migmatiticbasement (Santos et al., 1988; Carvalho, 2005). The Pernambuco-Alagoas Massif is composed mostly of orthogneisses and granitesin its southern part (Silva Filho et al., 2002). If the Macururemicaschists resulted from erosion of the Poo Redondo-MarancoDomain and the Pernambuco-Alagoas Massif, they will havegeochemical characteristics similar to the source rocks. Followingthis reasoning, if the Macurure micaschists have undergone highdegrees of partialmelting to form the 590e570Ma granites parentalmagmas, then these granites will have isotope geochemistry veryclose to that of the micaschists. This was demonstrated in Fig. 14B.

Crawford and Searle (1993) proposed that the collision-relatedleucogranites in North Pakistan are partial melts of biotite-richsedimentary protoliths, even though the granites do not showRb/Sr isotopic characteristics compatible with derivation fromcrustal protoliths. These authors concluded that the sedimentswere immature and leucogranites have inherited their isotopic

7.3.1. Hot asthenosphere upwelling by ~630e618 Ma?Sr-Nd isotope ratios, UePb crystallization ages, and major and

erictrace elements data indicate that the Stios Novos and PooRedondo granites could have originated by partial melting of localPoo Redondo-Maranco domainmigmatites, and that the Lajedinhoand Queimada Grande granitoids require contribution from morejuvenile sources.

Oliveira et al. (2010) proposed that rifting/extension in theCaninde domain, which is the domain immediately north of thePoo Redondo-Maranco domain, was intermittent from ca. 715 Mawith emplacement/extrusion of bimodal magmas of the Garroteand Novo Gosto unit until approximately 640 Mawhen intrusion ofthe A-type Boa Esperana rapakivi granite took place. According tothe model put forward by these authors, compression in the Can-inde and Macurure domains was more intense when the S~aoFrancisco craton (plate) began to underthrust the Pernambuco-Alagoas massif, forming the Macurure shear zone that limits theCaninde and Poo Redondo-Maranco domains, and probablymarking the onset of D1 and D2 deformations in sedimentary do-mains of the Sergipano Belt.

The beginning of collision between the S~ao Francisco craton/plate and the Pernambuco-Alagoas massif is unknown but takinginto account that the 630e618Ma granite age group is possibly arc-related (see next section), we suggest that the younger granite ofthis group (618 Ma) denes a minimum age for the onset of colli-sion. Subduction of the S~ao Francisco craton/plate might have beenfollowed by slab break-off or slab tearing, which allows theasthenosphere to uplift. Uplift of the asthenosphere may haveprovided the necessary heat for partial melting of the PooRedondo migmatites to originate the Stios Novos and PooRedondo granites, and to contribute with mantle material to formcharacteristics. In our case, the 590e570 Ma granites showgeochemical and isotopic characteristics of the likely source rocks,particularly the Macurure micaschists, and lesser of the PooRedondo-Maranco domain and the Pernambuco-Alagoas massif.

For many geologists it is a consensus that the geochemicalcomposition of granites depends on the sources and crystallizationhistory of the melt (Pearce, 1996; Forster et al., 1997; Barbarin,1999; Frost et al., 2001; Clemens, 2003; Sun et al., 2010). For thisreason it is too difcult to consider classication schemes that canenclose all types of granites present in the crust. Chappell andWhite (1974) proposed the I- and S-type classication on the ba-sis of observations made in low-temperature granites, in south-eastern Australia. There the granites show isotopic evidencedemanding contrasting source reservoirs for S- and I-type graniticmagmas. For some reasons, granites of the Lachlan Fold Belt pre-serve these differences, whereas in some other regions granite ty-pology can be somewhat less distinct. Crawford and Searle (1993)reported on collision granitoids in North Pakistanwith compositionvarying from biotite-bearing aplitic granodiorites and mon-zogranites through two-mica granites to pegmatitic, garnet-muscovite leucogranites. This means that in the Himalayas thereare collisional granites that resulted from partial melting of meta-sedimentary rocks that do not show the common S-type charac-teristics. In conclusion, our data support the hypothesis that590e570Ma granites are product of high degrees of partial meltingof the Macurure micaschists but they cannot be classied strictly asS-type granites because they do not show the typical features ofthis granite type.

7.3. Tectonic implications

E.P. Oliveira et al. / Journal of South Am270the more juvenile Queimada Grande and Lajedinho granitoids.7.3.2. 630e618 Ma-old Neoproterozoic arc development in theSergipano belt?

The 630e618 Ma-old granitoids of the Sergipano Belt belong tothe high-K calc-alkaline series and some of them entrain macenclaves. High-K calc-alkaline granitoids are very common inorogenic belts (continental arcs), in post-collisional (Caledonian-type) tectonic setting (Bonin, 1990; Roberts and Clemens, 1993;Wang et al., 2004; Karsli et al., 2007), and in intracontinentalshear zones (Neves and Mariano, 1997; Njanko et al., 2006). Thepresence of mac enclaves in the granites implies magma minglingor capture of pre-existing mac crust by the granitic magma.

Two tectonic scenarios can be envisaged for the origin of the630e618 Ma high-K calc-alkalic igneous activity in the SergipanoBelt. One model is based on mantle plume activity, and was pro-posed by Neves and Mariano (1997) for the origin of high-K calc-alkalic plutons in the Borborema Province. The authors suggestedthis model to explain the association between mac to interme-diate (diorite to granodiorite) and felsic rocks (coarse-grained toporphyritic quartz monzonites to granites) in the BorboremaProvince. These authors concluded that the main petrogeneticprocess responsible for origin of these rocks was magma mixingand that the source for the granitoid magmas was the lower crustand for the diorites was the metasomatized subcontinental litho-spheric mantle. The granitoids studied by Neves and Mariano(1997) emplaced along major strike-slip shear zones of the Bor-borema Province. However, according to the authors, granitoidemplacement was not controlled by a tectonic event because thescale of magmatism is too large to be assigned to transcurrentfaulting and the increase of temperature promoted by the shearzones would not be high enough to trigger melting under uid-absent conditions that prevailed in the deep crust. Another argu-ment used by the authors in support of the mantle plumemodel forgranitoid genesis in the Borborema Province is the absence ofcommon features found in a tectonic environment related to sub-duction, such as ophiolites, suture zones, or high-pressure meta-morphic rocks in the internal portion of the province.

The other model that we prefer is for origin of the QueimadaGrande and Lajedinho high-K calc-alkaline plutons in a continentalarc. Oliveira et al. (2010) proposed a complete plate tectonic cyclefor the Sergipano belt and had already proposed that arc-type rockswere generated in the time spam 630e617 Ma for the PooRedondo-Maranco and Macurure domains. Another argument thatbuttress the hypothesis that the Queimada Grande and Lajedinhoplutons formed in a Neoproterozoic arc in the Sergipano belt is thepresence of volcanic rocks with arc geochemical signature in thePoo Redondo-Maranco domain with UePb SHRIMP zircon agesabout 603 Ma (Carvalho 2005; Oliveira et al., 2010).

According to Neves and Mariano (1997) in the BorboremaProvince there is no evidence of collision tectonics. In the Borbor-ema Province there are granites with similar crystallization ages,such as the 591 Ma-old Teixeira batholith and the 576 Ma-old SerraRedonda granite-diorite pluton (Archanjo et al., 2008) all of thememplaced along regional shear zones. In the Sergipano Belt, the 584Ma-old Angico granite and 571 Ma-old Pedra Furada granite hadtheir emplacement controlled by collision (Bueno et al., 2009). TheTeixeira and Serra Redonda granitoids are located in the northernpart of the Borborema province, whereas the Angico and PedraFurada granites are located in the southernmost part (SergipanoBelt). On a regional scale, what geological scenario can account forthe simultaneous emplacement of syn-collision high-K calc-alkaline granites in the Sergipano Belt (approximately between590e570 Ma) and strike-slip-related granitoids (590e520 Ma) indomains at north in the Borborema Province? Bueno et al. (2009)suggested that this scenario is possible during continent-

an Earth Sciences 58 (2015) 257e280continent collision, such as the Himalayas, when syn-collision

ericgranites formed in the collision zone (e.g. Sergipano Belt) andstrike-slip controlled granites formed in the passive indentor (e.g.mainland Borborema Province) owing to far eld stress duringcoeval extrusion tectonics. Fetter et al. (2003) have recognized acontinental arc in the northwestern portion of the BorboremaProvince represented by the Santa Quiteria batholith dated at ca.665 and 591 Ma; the arc was formed during collision of the Bor-borema Province with the West African-S~ao Lus craton (Santos etal., 2008). All of these evidences support the hypothesis that theBorborema Province has endured multiple Neoproterozoic colli-sional events, at least at its margins.

Despite the high-K calc-alkaline granite series being generallyassociated with extensional tectonics (Whalen et al., 2004;Guimar~aes et al., 2004; Njanko et al., 2006; Karsli et al., 2007), thisrock series is themost abundant one in continental arcs, such as theAndes (Winter, 2001) and in post-collision settings in the Alpine-Himalayan belt in Turkey (e.g. Karsli et al., 2007). Silva Filho et al.(2000) suggested that the high-K calc-alkaline granitoids alongthe boundary of Pernambuco-Alagoasmassif and the Sergipano beltcould be remnants of an arc during the Brasiliano orogeny. There areother high-K calc-alkaline granitoids ascribed to the development ofarc-type setting, like the St Peter Suite in Australia (Swain et al.,2008), I-type high-K calc-alkaline and S-type granitoids fromsoutheastern Roraima, Brazil (Almeida et al., 2007), the SaghroMassif in Morocco (El Baghdadil et al., 2003), and the Sierra deMacon I-type high-K calc-alkaline granitoid in the Argentina, whichshows negative Nd(t) values (Poma et al., 2004). In conclusion, ourintegrated eld, geochemistry and isotope data suggest that theQueimada Grande and Lajedinho monzodiorite plutons belong to aca. 618 Ma-old Neoproterozoic magmatic arc in the Poo Redondo-Maranco and Caninde domains of the Sergipano belt.

We propose that the Camara tonalite and the Coronel Jo~ao Sagranodiorite represent approximately coeval arc-rocks in the Mac-urure domain. There are many similarities between these granitesand the Queimada Grande granodiorite. All of the granites showsimilar crystallization age, between 618Ma and 628Ma (Long et al.,2005; Bueno et al., 2009). From the geochemical point of view, thegranitoids are also similar; they are magnesian, metaluminous, I-type, high-K calc-alkaline granites with trace element signatures ofvolcanic arc granites. Additionally, the granitoids show numerousmac xenoliths of amphibolitic composition. The differenceamongst these granites lies in their Nd(t) values. The Camaratonalite and Coronel Jo~ao Sa granodiorite showmore negative Nd(t)values, i.e. -7.45 and -4.83 to -6.86 respectively, when comparedwith the Queimada grande Granodiorite (-1.15 to -2.55). This dif-ference can be explained by greater amounts of crustal assimilationduring magma ascent and emplacement in the continental crust.

The probable Neoproterozoic arc granites in the Sergipano beltshow increasing contamination (or interaction) with upper crustalcomponents from north to south (Fig. 16). The granites TDM valuesincrease and Nd(t) values generally decrease from north to south inthe belt. The Queimada Grande granodiorite and the Lajedinhomonzodiorite have the least negative Nd(t) values, whereas theCamara tonalite and some parts of the Coronel Jo~ao Sa granodioritehave the most negative Nd(t) values.

The only possibility for generation of arc-type rocks in the PooRedondo-Maranco and inMacurure domains simultaneously, is thattheMacurure domainwas connected to the Poo Redondo-Marancodomain before the beginning of Brasiliano orogeny. Oliveira et al.(2010) had already proposed this connection based on TDM anddetrital zircon data for Macurure domain quartzite and micaschist.Rocks of Macurure domain show Ndmodel ages (TDM) varying from1.2 to 1.8 Ga (Oliveira et al., 2005) and cluster of zircon ages is around980Ma indicating these rocks as product of erosion of Cariris Velhos

E.P. Oliveira et al. / Journal of South Amsources around 1.0 Ga found in Poo Redondo-Maranco domain.7.3.3. Generation of 590e570 Ma granites within a channel ow?After generation of a continental arc in the Sergipano belt rep-

resented by ~ 625 Ma-old granitoids, the Macurure crust started tothicken until partial melt took place to originate the ca. 580 Ma-oldcollisional granites. During collision of two or more plates thereshould be an interval of time between the rst contact of thecolliding blocks and the onset of granitic magmas. This period oftime is required for the crust to be sufciently thickened and itslowest levels reach the pressure-temperature conditions forgranitic magma generation. For the Himalayas, which is the mostrecent example and still active continent-continent collision zone(Nelson et al., 1996; Klemperer, 2006), the incubation period be-tween the beginning of continental collision and the production ofgranites is approximately 25 million years, a number derived fromthe initial contact between India and Asia at 57 Ma and the oldestleucogranite at 32 Ma (Leech et al., 2005). In ancient orogenic beltsthe beginning of collision is more difcult to infer. Nevertheless,Ferre et al. (2002) estimated in 60 million years the span of timebetween collision and S-type granite generation in the ProterozoicThans-Sahara belt. In the Sergipano belt, if the ~625 Ma-old gran-ites are taken as the minimum age for onset of the main collisionalevent (D2) in the belt and the age of the Angico granite (584 10Ma) as the rst syn-D2 granites, we then have a minimum timespan of about 41 million years since the beginning of collision andgeneration of the rst syn-collisional granite.

It is possible to drawn an analogy between theHimalayas and theSergipano Belt because the Sergipano Belt contains several struc-tural and lithologic domains that render it comparable to Phanero-zoic orogens. Searle and Szulc (2005) suggested that the HighHimalayametamorphic sequence operated as a ductile channel owapproximately 15-20 km thick, extruding southwards, and boundbymajor shear zones above and below where the leucogranites weregenerated. According toSearle et al. (2003) amid-crustal layerwas athigh temperature, deforming in a ductile manner with a combina-tionof both pure and simple shear, andwas partiallymelted in situ toproduce leucogranite sheets, whichmigrated horizontally followingthe planes of anisotropy dened by the metamorphic foliation.

Field observations are consistent with the interpretation of theMacurure domain as a Neoproterozoic analogous of the ductilechannel owmodel proposed for the High Himalaya. The Macururedomain is a metasedimentary domain and is located between shearzones, i.e. the S~aoMiguel do Aleixo, to the south, and the BeloMonteJeremoabo, to the north. During collision between the Pernambuco-Alagoas massif and the S~ao Francisco Craton, the Macurure domainwas compressed, possibly between two regional shear zones,thereby generating a great crustal shortening. Partial melting of themetasedimentary rocksmayhave takenplaceduring this shorteningevent to form in situ granitic magmas. The magma then migratedalong the S2 axial plane foliation towards the region of less tensionand was collect in the hinge of F2 folds. From the structural point ofview, the Macurure domain operated as a ductile channel owbound by two shear zones, between which the metasedimentarypile was exhumed and eroded, exposing side by side contrastinglithotectonic domains such as the low-grade metamorphicVazaBarris domain and the higher-grade Macurure domain.

Acknowledgements

The authors acknowledge the nancial support of the Brazilianagencies FAPESP (05/60119-5, 04/05054-2; 02/03085-2; 02/07536-9), CNPq (308424/2011-5), Millenium Project (42.0222/2005-7), andINCT project (573713/2008-1). Barbara Lima and Jeane Chaves arethanked for whole-rock Sm-Nd and Sr laboratory facilities at Uni-versity of Brasilia. We also thank Ignez de Pinho Guimar~aes and an

an Earth Sciences 58 (2015) 257e280 271anonymous reviewer for their valuable comments of themanuscript.

Appendix A. Data tables

Table A1UePb geochronologic data for granitic rocks of the Poo Redondo-Maranco, Caninde, and Macurure domains.

Spot U ppm Th ppm 232Th % 206common

Isotope ratios Age (Ma) Concordance%238U 207 Pb 1s 207 Pb 1s 206 Pb 1s 206 Pb 1s

206 Pb 235U 238U 238U

Stios Novos granite15-1* 384 210 0.56 0.06 0.1133 0.0012 4.16 0.07 0.2664 0.0032 1852 19 824-1 118 44 0.38 0.2 0.0842 0.0036 2.63 0.12 0.2264 0.0023 1297 84 10114-1 295 187 0.66 0.09 0.0726 0.0008 1.66 0.02 0.1664 0.0013 1002 22 9917-1 918 14 0.02 0.07 0.0718 0.0006 1.61 0.02 0.1626 0.0013 981 18 995-1 45 36 0.82 0.21 0.0706 0.0019 1.58 0.05 0.1625 0.0023 944 56 10321-1 344 235 0.71 0.27 0.0714 0.0010 1.59 0.02 0.1612 0.0012 969 28 9926-1 219 247 1.17 0.02 0.0717 0.0009 1.57 0.02 0.1593 0.0014 978 27 973-1 251 139 0.57 0.06 0.0726 0.0008 1.59 0.02 0.1586 0.0012 1002 23 9516-1 526 286 0.56 0.03 0.0708 0.0006 1.51 0.02 0.1543 0.0012 950 17 978-1 228 190 0.86 0.64 0.0699 0.0015 1.44 0.03 0.1492 0.0013 925 43 9712-1 422 381 0.93 0.4 0.0712 0.0018 1.42 0.04 0.1445 0.0011 964 51 909-1 120 70 0.6 0.25 0.0658 0.0014 1.28 0.03 0.1415 0.0014 799 43 10718-1 290 174 0.62 0.73 0.0693 0.0016 1.35 0.03 0.1410 0.0012 906 47 9425-1 858 108 0.13 0.26 0.0611 0.0007 0.90 0.01 0.1068 0.0007 654 4 10211-1 46 45 1.02 0.01 0.0592 0.0036 0.86 0.05 0.1056 0.0017 647 10 11320-1 187 69 0.38 0.12 0.0586 0.0022 0.85 0.03 0.1049 0.0011 643 6 11619-1 411 189 0.47 0.24 0.0591 0.0010 0.85 0.02 0.1039 0.0008 637 5 11224-1 457 42 0.09 0.13 0.0613 0.0008 0.87 0.01 0.1031 0.0007 633 4 986-1 345 147 0.44 0.11 0.0600 0.0008 0.85 0.01 0.1030 0.0008 632 5 1052-1 76 35 0.48 0.59 0.0578 0.0024 0.82 0.03 0.1027 0.0014 630 8 12123-1 188 237 1.3 0.35 0.0586 0.0014 0.82 0.02 0.1022 0.0009 627 5 11410-1 77 75 1 0.32 0.0573 0.0017 0.80 0.03 0.1012 0.0012 622 7 1237-1* 51 37 0.75 1.3 0.0508 0.0021 0.69 0.03 0.0979 0.0014 602 8 2591-1* 451 40 0.09 2.94 0.0653 0.0028 0.83 0.04 0.0927 0.0007 572 4 7322-1* 750 281 0.39 1.66 0.0620 0.0047 0.78 0.06 0.0910 0.0007 561 4 83Queimada Grande granodiorite1-1* 107 62 0.60 0.62 0.0546 0.0023 0.75 0.03 0.0990 0.0011 608 6 1541-2 140 111 0.82 0.21 0.0616 0.0018 0.86 0.03 0.1016 0.0010 624 6 952-1@ 319 365 1.18 0.15 0.0602 0.0008 0.82 0.01 0.0985 0.0008 606 4 992-2 354 245 0.71 0.08 0.0597 0.0009 0.84 0.01 0.1016 0.0008 624 4 1053-1 400 108 0.28 0.13 0.0594 0.0009 0.82 0.01 0.1002 0.0007 616 4 1064-1* 450 329 0.76 1.23 0.0588 0.0017 0.76 0.02 0.0938 0.0007 578 4 1035-1 264 179 0.70 0.18 0.0591 0.0010 0.82 0.01 0.1008 0.0008 619 5 1095-2* 302 128 0.44 0.54 0.0576 0.0011 0.79 0.02 0.0999 0.0008 614 5 1196-1 301 229 0.79 0.18 0.0594 0.0009 0.82 0.01 0.0997 0.0008 613 5 1056-2 203 147 0.75 0.00 0.0610 0.0010 0.86 0.02 0.1022 0.0009 627 5 987-1@ 212 146 0.71 0.24 0.0695 0.0012 1.38 0.03 0.1442 0.0012 914 35 958-1 322 115 0.37 0.11 0.0604 0.0008 0.84 0.01 0.1005 0.0008 617 5 1009-1 240 164 0.71 0.24 0.0598 0.0010 0.82 0.02 0.0991 0.0008 609 5 10210-1 313 248 0.82 0.56 0.0597 0.0014 0.83 0.02 0.1005 0.0008 617 5 10410-2* 324 219 0.70 0.12 0.0587 0.0007 0.83 0.01 0.1022 0.0008 627 5 11311-1* 229 179 0.81 0.18 0.0574 0.0013 0.79 0.02 0.1004 0.0008 617 5 12212-1 196 201 1.06 0.40 0.0578 0.0015 0.80 0.02 0.0998 0.0009 613 5 11713-1 196 168 0.89 0.16 0.0589 0.0011 0.82 0.02 0.1005 0.0009 617 5 11014-1@ 333 64 0.20 0.13 0.0601 0.0009 0.86 0.01 0.1034 0.0008 634 5 10415-1 91 56 0.63 0.03 0.0591 0.0018 0.83 0.03 0.1021 0.0012 627 7 11016-1 215 190 0.91 0.08 0.0615 0.0010 0.84 0.02 0.0992 0.0009 610 5 9317-1 156 163 1.08 0.29 0.0572 0.0016 0.80 0.02 0.1013 0.0010 622 6 12518-1 253 204 0.83 0.55 0.0581 0.0016 0.81 0.02 0.1016 0.0009 624 5 11719-1 84 66 0.82 0.56 0.0567 0.0022 0.78 0.03 0.0996 0.0012 612 7 12720-1 152 114 0.78 0.16 0.0594 0.0016 0.84 0.02 0.1025 0.0010 629 6 10821-1 245 171 0.72 0.19 0.0583 0.0010 0.81 0.01 0.1002 0.0008 615 5 11322-1 218 241 1.14 0.24 0.0602 0.0013 0.83 0.02 0.1000 0.0009 615 5 101

Spot U ppm Th ppm 232Th % 206common

Isotope ratios Age (Ma) Concordance %238U 207 Pb 1s 207 Pb 1s 206 Pb 1s 206 Pb 1s 207 Pb 1s

206 Pb 235U 238U 238U 206 Pb

Poo Redondo granite

.33

.07

.58

.61

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e28027204-116B1B.4-3* 110 47 0.44 0.75 0.1422 0.0025 41B.2-2* 410 84 0.21 0.02 0.1167 0.0025 31B.14-2 146 75 0.53 0.23 0.0707 0.0019 11B.6-1 387 180 0.48 0.05 0.0723 0.0007 10.09 0.2209 0.0024 1201 12 2254 30 1880.07 0.1908 0.0011 1075 7 1906 38 1770.04 0.1624 0.0014 971 8 949 54 980.02 0.1611 0.0009 962 5 994 19 103

Table A1 (continued )

Spot U ppm Th ppm 232Th % 206common

Isotope ratios Age (Ma) Concordance %238U 207 Pb 1s 207 Pb 1s 206 Pb 1s 206 Pb 1s 207 Pb 1s

206 Pb 235U 238U 238U 206 Pb

1B.10-1 128 71 0.57 0.11 0.0702 0.0012 1.54 0.03 0.1590 0.0015 952 9 933 35 981B.2-1 117 70 0.62 0.31 0.0688 0.0015 1.49 0.04 0.1569 0.0016 941 9 893 46 951B.8-2 191 119 0.64 0.00 0.0725 0.0008 1.55 0.02 0.1552 0.0014 927 8 1001 22 1081B.11-1 49 14 0.30 0.27 0.0664 0.0030 1.26 0.06 0.1381 0.0021 834 12 820 95 981B.11-4 614 93 0.16 0.45 0.0672 0.0008 1.27 0.02 0.1375 0.0007 830 4 843 26 1021B.11-2 187 93 0.51 0.11 0.0651 0.0011 1.23 0.02 0.1368 0.0010 828 6 776 37 941B.15-5 508 87 0.18 0.13 0.0673 0.0008 1.25 0.02 0.1344 0.0007 812 4 846 24 1041B.11-3* 539 360 0.69 1.56 0.0727 0.0016 1.13 0.03 0.1122 0.0008 677 4 1007 44 1491B.15-2 261 87 0.35 0.47 0.0615 0.0015 0.93 0.02 0.1092 0.0008 668 5 656 52 981B.8-1 159 33 0.22 -0.03 0.0606 0.0009 0.88 0.02 0.1055 0.0009 647 6 626 33 971B.14-1* 321 41 0.13 0.15 0.0590 0.0009 0.85 0.01 0.1039 0.0007 639 4 568 35 893B.2-4 791 251 0.33 0.11 0.0613 0.0006 0.88 0.01 0.1039 0.0004 637 3 651 21 1021B.8-3 293 41 0.15 0.06 0.0622 0.0009 0.89 0.01 0.1036 0.0007 635 4 681 31 1073B.4-4# 245 35 0.15 0.71 0.0607 0.0026 0.86 0.04 0.1024 0.0008 628 5 628 92 1003B.4-3# 371 45 0.13 0.96 0.0590 0.0021 0.83 0.03 0.1018 0.0006 626 4 568 79 911B.15-4# 435 21 0.05 0.24 0.0599 0.0011 0.84 0.02 0.1018 0.0006 625 3 600 39 961B.4-2# 341 46 0.14 0.19 0.0607 0.0009 0.85 0.02 0.1017 0.0010 624 6 628 33 1011B.2-3# 521 144 0.29 0.07 0.0609 0.0007 0.85 0.01 0.1007 0.0005 618 3 637 24 1033B.2-5* 819 319 0.40 0.37 0.0622 0.0010 0.86 0.01 0.1001 0.0004 613 2 682 34 1111B.12-1* 70 50 0.74 0.73 0.0718 0.0044 0.99 0.06 0.0996 0.0015 604 8 979 126 1621B.1-1* 440 37 0.09 -0.04 0.0630 0.0006 0.86 0.01 0.0985 0.0006 603 3 708 21 1171B.12-2* 676 109 0.17 1.77 0.0608 0.0017 0.80 0.02 0.0961 0.0005 591 3 630 62 1071B.8-6* 470 132 0.29 0.49 0.0604 0.0013 0.74 0.02 0.0889 0.0005 548 3 616 45 1131B.8-4* 468 108 0.24 0.62 0.0598 0.0014 0.72 0.02 0.0871 0.0019 537 11 595 49 1111B.7-2* 516 309 0.62 1.07 0.0736 0.0015 0.87 0.02 0.0858 0.0005 520 3 1032 42 1981B.10-2* 270 59 0.23 0.85 0.0604 0.0024 0.69 0.03 0.0824 0.0006 508 4 617 86 1211B.13-1* 506 174 0.35 0.43 0.0625 0.0013 0.61 0.01 0.0704 0.0004 435 2 690 44 15909-28C2C.50-2 299 204 0.70 0.22 0.0699 0.0009 1.58 0.03 0.1644 0.0015 983 9 924 28 942C.50-1 97 64 0.68 -0.04 0.0712 0.0013 1.50 0.03 0.1522 0.0019 912 11 965 39 1062C.49-1* 60 38 0.65 0.89 0.0620 0.0030 1.25 0.06 0.1463 0.0022 887 12 675 105 762C.46-1 133 24 0.19 0.23 0.0650 0.0013 1.28 0.03 0.1424 0.0016 861 9 774 40 902C.49-2* 122 24 0.20 -0.20 0.0706 0.0032 1.29 0.06 0.1329 0.0017 800 9 946 94 1182C.46-2* 80 64 0.84 1.02 0.0663 0.0035 1.20 0.07 0.1311 0.0018 793 10 814 112 1032C.51-1 119 13 0.12 0.05 0.0613 0.0018 0.99 0.03 0.1166 0.0014 712 8 650 62 912C.48-1 136 69 0.52 -0.04 0.0621 0.0017 0.93 0.03 0.1086 0.0012 664 8 679 59 1022C.41-2* 177 47 0.28 0.13 0.0584 0.0014 0.86 0.02 0.1070 0.0011 658 7 545 54 832C.48-2* 226 83 0.38 0.22 0.0587 0.0016 0.86 0.02 0.1060 0.0011 652 6 557 59 852C.42-1* 334 54 0.17 0.20 0.0591 0.0011 0.86 0.02 0.1057 0.0010 650 6 569 40 882C.43-1 322 164 0.52 -0.17 0.0619 0.0010 0.90 0.02 0.1057 0.0010 647 6 669 35 1032C.52-1 514 105 0.21 -0.06 0.0601 0.0007 0.87 0.01 0.1049 0.0009 644 6 606 25 942C.41-1* 188 48 0.27 -0.14 0.0634 0.0011 0.92 0.02 0.1054 0.0011 644 7 722 35 1123C.43-3 511 224 0.45 0.12 0.0602 0.0010 0.87 0.02 0.1047 0.0005 643 3 609 37 952C.45-1* 263 98 0.39 -0.29 0.0630 0.0014 0.91 0.02 0.1048 0.0010 641 6 709 48 1112C.45-2 273 79 0.30 -0.17 0.0604 0.0011 0.87 0.02 0.1039 0.0010 637 6 619 40 972C.42-2* 611 97 0.16 0.73 0.0585 0.0013 0.83 0.02 0.1034 0.0009 636 5 547 47 863C.42-4 398 81 0.21 0.81 0.0609 0.0018 0.87 0.03 0.1033 0.0007 634 4 636 64 1002C.47-1* 439 123 0.29 0.54 0.0581 0.0012 0.82 0.02 0.1018 0.0009 627 6 533 45 853C.42-3# 375 74 0.20 0.61 0.0624 0.0014 0.88 0.02 0.1022 0.0006 626 4 686 49 1103C.45-4# 447 199 0.46 0.00 0.0618 0.0009 0.87 0.01 0.1015 0.0006 622 3 668 33 1073C.45-3* 195 37 0.20 0.37 0.0548 0.0026 0.76 0.04 0.1006 0.0009 622 5 403 106 653C.43-2# 424 212 0.52 0.35 0.0602 0.0013 0.84 0.02 0.1009 0.0006 620 3 611 47 992C.44-1* 346 83 0.25 1.47 0.0600 0.0022 0.80 0.03 0.0965 0.0009 594 5 605 78 1022C.44-2* 413 172 0.43 2.95 0.0540 0.0035 0.48 0.03 0.0642 0.0007 402 4 370 145 92

Spot Isotope ratios Age (Ma) Concordance%

U ppm Th ppm 232Th238U

% 206common

207 Pb206 Pb

1s 207 Pb235U

1s 206 Pb238U

1s 206 Pb238U

1s 207 Pb206 Pb

1s

Lajedinho monzodioriteL-11.1 425 281 0.68 0.22 0.0599 1.6 0.83 1.68 0.1002 0.5366 615.4 3.1 601 35 99L11-1.2 357 158 0.46 0.24 0.0607 1.9 0.84 1.93 0.1 0.5103 614.4 3 630 40 103L11-2* 225 184 0.84 -0.03 0.0625 1.4 0.89 1.51 0.1027 0.6237 630.3 3.7 691 29 110L11-2.2* 677 435 0.66 -0.08 0.062 0.9 0.87 0.94 0.1014 0.3712 622.8 2.2 676 19 108L11-3* 223 345 1.6 0.11 0.0617 2.9 0.84 2.94 0.0985 0.647 605.7 3.7 665 61 110L10-1 224 165 0.76 0.25 0.0607 2.7 0.84 2.74 0.0998 0.6433 613 3.8 629 57 103L10-2* 205 161 0.81 0.05 0.063 2.1 0.89 2.16 0.1021 0.6547 626.4 3.9 708 44 113L10-3 343 237 0.71 0.25 0.0601 1.8 0.83 1.92 0.1004 0.5198 616.9 3.1 606 40 98L10-4** 322 208 0.67 0.06 0.0608 2.2 0.87 2.22 0.1042 0.5175 639.2 3.1 631 47 99

(continued on next page)

E.P. Oliveira et al. / Journal of South American Earth Sciences 58 (2015) 257e280 273

Table A1 (continued )

Spot Isotope ratios Age (Ma) Concordance%

U ppm Th ppm 232Th238U

% 206common

207 Pb206 Pb

1s 207 Pb235U

1s 206 Pb238U

1s 206 Pb238U

1s 207 Pb206 Pb

1s

L2-1* 602 592 1.02 -0.04 0.0613 1 0.87 1.31 0.1027 0.7959 630 4.8 650 22 103L2-2** 447 223 0.52 -0.02 0.0611 1.5 0.89 1.56 0.1056 0.4813 647.4 3 642 32 99L3-1 268 171 0.66 0.06 0.0608 1.5 0.85 1.62 0.1012 0.5637 621.5 3.3 631 33 102L3-1.2 494 477 1 0.03 0.061 1 0.85 1.08 0.101 0.4247 620.1 2.5 639 21 103L8-2* 217 161 0.77 0.33 0.0586 3.1 0.82 3.13 0.1019 0.6749 625.3 4 553 67 88L8-3* 190 17 0.09 -0.55 0.0676 3.1 0.89 3.32 0.0952 1.2983 586.3 7.3 856 63 146L8-3.2* 411 286 0.72 0.21 0.0584 1.8 0.82 1.93 0.1015 0.7719 623.4 4.6 545 39 87L6-2* 519 239 0.48 -0.18 0.063 1.3 0.89 1.4 0.1026 0.4349 629.9 2.6 709 28 113L6-3** 348 300 0.89 -0.05 0.0611 1.2 0.89 1.31 0.1051 0.5298 644.5 3.2 642 26 100L7-1* 415 288 0.72 0.2 0.0569 3.1 0.82 3.15 0.1048 0.4999 642.5 3.1 487 69 76L7-2 186 147 0.82 -0.02 0.0615 1.6 0.86 1.8 0.1016 0.7504 623.5 4.5 656 35 105L7-3 261 207 0.82 0.33 0.0598 2.6 0.85 2.74 0.1025 0.7068 629.2 4.2 598 57 95L7-3.2* 316 189 0.62 -0.2 0.063 2.9 0.87 2.94 0.0999 0.6442 613.7 3.8 709 61 116L14-1* 350 288 0.85 0.26 0.0594 2.4 0.82 2.47 0.1005 0.6506 617.3 3.8 583 52 94L17-1** 290 162 0.58 0.15 0.0603 1.6 0.88 1.66 0.1058 0.5827 648.5 3.6 616 34 95L22-1* 322 218 0.7 0.05 0.0622 1.4 0.87 1.49 0.1013 0.5389 622.1 3.2 680 30 109L18-1* 651 300 0.48 0.11 0.0606 1.4 0.89 1.41 0.1063 0.3831 651.4 2.4 626 29 96

Obs: (1) Reproducibility of Pb/U for BR266 zircon standard was 1.15% (2s; n 14). (2) Pb isotope ratios corrected for common Pb. (3) Ranked age shown is 206 Pb/238U ageif < 800Ma and 207Pb/206 Pb age for others. (4) * Data with common Pb correction >1% and/or discordant: i.e. 206 Pb/238U and 207Pb/206 Pb ages not overlapping as 2s; datanot considered in age discussion. (5) @ statistical outlier; data not considered in age discussion.OBS: (1) Notation: data collected during three analytical sessions (1, 2 and 3); sample B corresponds to 04-116B, and C to 09-28C. (2) Pb isotope ratios corrected for commonPb. (3) Reproducibility of Pb/U for BR266 zircon standard was: session 1 (1.88%; 2s; n 7); session 2 (1.54%; 2s; n 12); session 3 (0.88%; 2s; n 6); assigned error tocombined data sets is 2.00% (2s). (4) * Data with common Pb correction >1% and/or >10% discordan; these data not considered in age discussion (see text). (5) # Youngestconcordant population; interpreted as the emplacement event at 623 7 Ma (n 8; MSWD 0.89).Obs: (1) Reproducibility of Pb/U for Temora zircon standard was 0.44% (2s; n 13).* Data with U>1% or >5% discordant were not considered in age discussion.**- statisticaloutlier; data not considered in age discussion.

Table A2Major and Trace element data for granitic rocks of the Poo Redondo-Maranco, Caninde, and Macurure domains.

Granite Angico Areias Lagoas Santa Helena Canudos Capivara

Sample Ponto JUTC JUTC JTC JUMS JUMS JTC JUMS JUMS JUD JUD JUD JUD JUDNumber 13 11 33 112 27 33 35 20 22B 72 86 88 5 33

Major elements (wt%)SiO2 71.59 71.39 71.31 70.86 68.73 71.93 70.74 69.54 72.70 70.14 68.99 70.47 69.45 64.95TiO2 0.291 0.275 0.295 0.315 0.424 0.246 0.328 0.368 0.188 0.414 0.446 0.396 0.410 0.627Al2O3 15.10 15.08 15.18 14.77 15.83 15.16 15.26 15.70 15.04 15.57 16.06 15.61 16.00 14.80Fe2O3t 1.21 1.47 1.55 1.31 1.70 1.11 1.73 1.90 0.92 1.57 1.78 1.56 2.06 4.21MnO 0.02 0.02 0.02 0.02 0.02 0.02 0.03 0.03 0.01 0.01 0.02 0.01 0.02 0.08MgO 0.33 0.39 0.43 0.68 0.50 0.33 0.67 0.71 0.26 0.54 0.61 0.57 0.69 2.76CaO 1.42 1.30 1.26 1.12 1.25 0.97 2.55 2.78 0.89 1.92 1.96 1.87 1.44 3.90Na2O 4.42 4.35 4.27 4.41 4.67 4.55 4.33 4.60 4.10 4.89 4.74 4.89 4.72 3.31K2O 4.77 4.86 4.91 5.02 5.29 5.28 3.37 3.33 5.35 3.92 4.34 3.80 4.41 4.28P2O5 0.101 0.097 0.135 0.176 0.247 0.090 0.095 0.111 0.074 0.127 0.141 0.131 0.170 0.331LOI 0.37 0.39 0.40 1.14 1.04 0.40 0.75 0.70 0.43 0.60 0.37 0.50 0.82 0.39Total 99.60 99.60 99.80 99.80 99.70 100.10 99.90 99.80 99.90 99.70 99.50 99.80 100.20 99.60Trace elements by XRF (mg/g)V 19.7 21.1 19.1 21.2 23.9 17.8 32 35 10.3 21.9 29.1 22.8 33 78Cr 22.6 20.5 11.6 21.5 20.2 26.8 15.7 72 5.2 93 159 132 48 203Ni

Table A2 (continued )

Granite Formosa Itabi Gloria Carabas

Sample JUD JUD JUD JUTC JUMS JUMS JUD JUD JUD JUD JUD JUD JUD JUDNumber 13C 15B 18 138 03C 9 34 35B 37A 37B 80A 80B 10A 10B

CaO 0.53 0.19 2.84 3.29 3.55 3.54 0.67 0.69 0.97 0.78 3.81 0.81 2.18 1.35Na2O 4.51 4.32 4.52 3.92 3.85 4.51 4.66 4.36 4.83 5.23 3.81 4.91 4.42 4.91K2O 4.49 5.43 2.65 3.47 3.44 2.22 4.79 4.63 4.79 4.73 2.08 4.55 3.44 4.00P2O5 0.056 0.112 0.203 0.189 0.189 0.227 0.029 0.023 0.023 0.027 0.386 0.030 0.103 0.085LOI 0.65 0.66 2.07 0.63 0.68 0.47 0.40 0.48 0.23 0.54 0.61 0.40 0.80 0.52Total 99.60 99.10 99.00 100.00 100.20 99.90 99.40 99.40 100.30 100.10 99.50 100.30 99.80 99.80Trace elements by XRF (mg/g)V 21.8 15.4 17.5 50 52 47 7.8 6.3 4.7 11.1 91 8,7 35 24Cr 246 64 250 38 49 98 168 46 23 95 122 20.1 143 53Ni

Table A2 (continued )

Granite Lajedinho Coronel Jo~ao Sa Stios Novos

Sample FS FS CRN JUD JUD JUD JUD JUD JUD JUD JUD JUD JUD JUDNumber 169 170 109A 56 66 69 70 104 106 107 108 203 204 205

Ga 27.4 23.8 24 25.3 22.5 24 22.6 25.4 25.3 23.3 25 22.2 22.1 22.3Rb 68 82 78.7 109 103 102 103 168 206 235 191 131 125 121Sr 577 471 599 472 615 529 667 490 476 762 561 325 350 342Y 42 46 37.8 5.5 12.9 6.4 19 4.9 4.1 5 11.7 10.6 10.6 9.7Zr 363 398 553 224 214 179 301 204 252 447 174 180 181 168Nb 16.8 19.1 17.4 12.4 11.4 8.5 12.3 4.9 4.7 4.5 9 9.9 9.9 8.6Ba 1472 1364 1862 903 1010 912 1283 1245 1367 2214 1116 690 765 648Pb 16.1 18 19.5 23.8 31 26.5 22.4 59 55 44 53 54 47 48

Granite S. Novos Queimada Grande

Sample JUD JUD JUD JUD JUD JUD JUD JUD JUD JUD JUD JUD JUDNumber 206 94A 99 102 109 116 118 119 128A 128B 133A 133B 177

Major elements (wt%)SiO2 67.52 65.81 64.64 65.18 70.89 66.63 64.04 59.54 61.77 61.94 57.60 64.55 68.03TiO2 0.548 0.805 0.703 0.655 0.407 0.681 0.758 1.034 0.760 0.980 1.099 0.731 0.582Al2O3 14.82 15.14 15.21 14.75 14.14 15.87 15.26 15.34 15.35 15.23 16.21 15.07 16.09Fe2O3t 3.46 3.93 4.25 4.20 2.27 3.36 4.44 6.33 5.47 5.33 7.31 4.58 2.84MnO 0.05 0.06 0.07 0.07 0.04 0.05 0.07 0.10 0.09 0.08 0.12 0.07 0.04MgO 2.27 1.88 2.16 2.56 0.94 1.04 2.28 3.74 3.19 3.01 4.52 2.38 0.87CaO 3.46 3.39 3.68 4.01 1.91 2.93 3.76 4.87 3.94 4.14 5.08 3.57 2.57Na2O 3.74 4.24 3.93 3.69 3.80 4.14 3.95 4.13 4.13 3.77 4.38 3.65 4.09K2O 3.20 3.33 3.71 3.54 4.61 3.95 4.02 3.44 3.25 3.79 2.27 3.90 4.15P2O5 0.148 0.351 0.298 0.221 0.176 0.233 0.293 0.422 0.331 0.396 0.364 0.286 0.186LOI 0.51 0.54 0.48 0.53 0.78 0.41 0.78 0.65 0.84 0.72 1.05 0.79 0.44Total 99.70 99.50 99.10 99.40 100.00 99.30 99.60 99.60 99.10 99.40 100.00 99.60 99.90Trace elements by XRF (mg/g)V 59 83 74 81 39 63 78 126 101 97 140 79 47Cr 393 129 127 284 81 96 140 159 162 257 352 327 33Ni 31 19.1 20.9 33 8.3 2.2 21.5 40 34 34 59 25.4 1.6Zn 64 77 69 58 48 80 72 103 99 89 118 71 69Ga 21.1 22.2 21.3 20.3 23.5 24.5 22.7 24.2 24.1 24.4 26 21.2 24.3Rb 107 107 126 124 163 141 123 144 151 131 131 127 151Sr 371 723 581 481 352 615 589 663 634 688 528 563 571Y 9.7 21.4 16.9 20.3 9.2 14 15.1 14.4 12.6 17.3 14.7 15.7 10.6Zr 156 226 237 233 158 241 217 211 206 271 259 232 200Nb 8.4 11.5 12.3 12.7 9.7 13.9 11.6 10.4 7.4 12.3 10.3 11.6 12.2Ba 693 1047 1135 840 616 1420 1118 1008 896 1101 362 952 1362Pb 45 38 37 37 48 35 41 38 44 43 27.3 45 39

Granite Queimada Grande Poo Redondo

Sample JUD JUD JUD JUD JUD JUD JUD JUD JUD JUD JUD JUD JUDNumber 179 188 157 159 163 182 183 184A 184B 184C 185 186 191

Major elements (wt%)SiO2 69.73 64.05 75.41 72.09 72.24 74.46 72.03 70.31 72.85 66.17 73.68 74.50 71.85TiO2 0.322 0.893 0.08