Arehaean and Lower Proterozoic Units in Central Brazil By Jos6 Caruso M. DANNI, tleinhardt A. FvcK and Othon H. LEONARDOS, Brasilia*) With 8 figures

Zusammenfassung Im zentralen Brasilien wurden folgende Haupteinheiten als Arehaiseher Sockel erkannt: a) hoelnnetamorphes Granit-Gneis-Gebiet mit vorwiegend granodioritischen und tonalitischen Gneisen; diese Gesteine werden yon amphibolitischen G~ngen durchschlagen. Sie bilden einzelne Kerne, die yon Zonen Arehaiseher Griinsteingiirtel umgeben werden, b) Griinsteingiirtel, die basale komatiitisehe ultramafitisehe bis mafischen Laven representieren mik h~iufig auftretenden "Spinifex"-Gefiigen. Diesen folgen Basalte nnd chemisehe Sedimente wie Kieselschiefer, Kalke und graphitisehe Pelite, in die h[iufig saute und intermediiire Vulkanite eingeschaltet sind. Die Metamorphose liegt in der Griinsehiefer-Fazies. e) Mobiler Granulit-Giirtel, der die granulitisierten mafiseh-ultramafisehen Komplexe yon Barro Alto, Niquelgndia und Cana Brava einschliel3t. Das untere Proterozoikum wird dutch folgende Einheiten vertreten: a) Metasedimentfolgen, die die Granit-Griinstein-Serien iiberlageru nnd die einige zinnfiihrende GranitIntrusionen aufweisen, b) Lagige Gabbro-Anorthosit-Plutone mit dazugeh/irenden mafischen Vulkaniten, die von sauren und intermedigren Vulkaniten nnd ozeanischen Sedimenten gefolgt werden, e) Mafiseh-ultramafisehe Intrusionen innerhalb des Granulit-Gebietes.

Abstract The following main units have been recognized as Archaean basement rocks in central Brazil: a) High-grade granite-gneiss terrain composed predominantly by granodioritie and tonalitie gneisses; these rocks are cut by amphibolite dykes and present several nucleii locally surronnded by an array of Arehaean greenstone belts, b) Greenstone belts presenting basal komatiitie ultramafie-mafie lavas with abundant spinifex textures, followed by basalts and chemical sediments such as cherts, limestones and graphitie pelites, that are often interbedded with acid and intermediate volcanics; metamorphism is within greensehist facies, e) Granulite mobile belt, ineluding the granulitized mafie-ultramafie complexes of Barro Alto, Niquelfindia and Cana Brava. The Lower Proterozoie is represented by: a) metasedimentary sequenees overlying the granite-greenstone terrains, with several tin bearing granite intrusions, b) Gabbro-anorthosite layered plutons and their assoeiated marie voleanies which are followed by acid and intermediate volcanies and ocean floor sediments, e) Mafie-ultramafic layered intrusions in the granulite terrain.

Resumo As seguintes unidades geol6gicas principais foram reconhecidas no embasamento da regigo central do Brasil: a) Terrenos granito-gnfiissieos eonstituldos dominantemente pot gnaisses de eomposig~o granodioritica e tonalitiea eortados por diques de anfibolito e mostrando vfirios nfeleos granitieos loealmente envolvidos por seq/i~ncias vuleanosedimentares do tipo greenstone belt. b) CinturSes arqueanos de roehas verdes apresentando na base lavas komatiRicas corn abundantes texturas tipo spinffex, seguidas de basaltos almofadados e de sedimentos quimicos tais como chert, ealefirio e pelitos grafitosos, frequentemente interealados eom vulegnieas intermedifirias e ~eidas; o meta*) Address of the authors: J. C. M. DaNNI, R. A. FUCK and O. H. LEONARDOS,Departamento de Geoci~neias, Universidade de Brasilia, 70.910 Brasilia, Brazil. Band 71, Heft 1, 1982, SeRe 291--317

29"[

J. C. M. DANNIet al. moffismo 6 de facies xisto verde, c) Cinturgo mdvel granulltico incluindo os complexes granulitizados de Barro Alto, Niquelfindia e Cana Brava. O Proteroz6ico Inferior 6 representado per: a) Seqii4ncias sedimentares sobrepostas aos terrenos granite-greenstone, loealmente intrudidas per granites estaniferos; b) PlutSes gabro-anortositicos acamadados; e) Intrus6es mafieo-uhramdtficas alojadas nos terrenos granuliticos. ~paTKoe co~ep~aifite B 14eHTpaYil~HO~ BpaaHJIttH pagJIHqaI-OT e a e ~ y ~ o ~ e OCHOBHt,Ie e~HHHI.IBI a p x e f t c ~ x OT~OmeH~ft: a) m, icoi~oMeTaMopdgHym rpa~TO-rHefmoBym o6aacT5 c npeHMy~ecWBeHHO rpano]IgopHw~-IMH H TOHanHTm,IMH rHe~eaMH; 3TH nopo~,I np0HH3aHI, I 2MqbH6OJIHTHI, IMH }I~HJIaMH H o 6 p a 3 y m r OT~eJIhHI=Ie ~ p a , o~pyz~eHHble 30HaMtt apxefie~oro 3eJIeHoEaMeHHOPO IIOYlCa. 6) 3e$ieHoKaMeHHt, IfI noYic, n p e ~ CTaBJIItIOILIHfI 6asan~H~le I~OMaTI~IITHHecKHe y J ~ T p a M a q b g ~ e c K H e ~ 0 MaqbHHect~HX JIaBI~I C tIaCTt, IM Hp0~IBJIeHtleM ~CIIHHO0/SeECHBIX" CTpOeHH~. 3 a STHMH c/ie~ytOT 6asaJII~TBI I4 XHMHtIeCKHe ce~tIMeHTI~I, EaE IgpeMHHCTBII~ c n a H e ~ , HSBeCTHSIKH H rpa~HTHI, Ie HeJIHTI~I~ B XOTOpI, IX qaCTO BcrpeqaIOT, KaE BKJIIOHeHHYl, KHCJII=Ie H cpe~HHe ByJIKaHIITBI. ]YIeTaM0pqbHSM OTMe~IeH B ~aI~IIII 3eJIeHoI~o CnaHI~a; B) nO~BIDKH]51~ FpaHyJIHTOBI~I~ noac, EOTOpI=Ifl BE/I~0tIaeT FpaHyJnITHSHpoBaHHBIe MaqbHqeeEo-ynBTpaMaqb~qecKI~e KOMHJIeKeBI Barro Alto, Niguelandia H C a n a Brava. B HHX~HeM npoTepo3oe yCTaHOBJIeH0 IIpI4cyTCTBHe c n e ~ y ~ o ~ x cTpaTI4rpaqbIIqecEIIx e/~IIHI4I~:a) CBIITI~IMeTace~IIMeHTOB, IIepeKpI~ITI,IX FpaHIITO-Be/IeHoI~aMeHH~Ii~H c e p ~ s M ~ , H BMeH~a~0L~HMH HeKOTOp~Ie HHTpyBHH rpaHI4Ta, c o ~ e p x < a ~ H e c y p ~ M y ; 5) n p o c n o H r a 5 6 p o - a H o p T o s H r - n n y T O H O E C OTHOCa~HMHe~ E HH~ ~aqbHHecKHMH BynKaHHTaMI4, 3 a KOTOpI~IMH cJIe~ytOT KHCn~Ie H c p e ~ I H e ByJIKaHHTI~I H ~ o p c E H ce~HMeHT~I; B) ~acloH,~ec~o-ya~Tpa~adpH,~ecx~e I4HTpySHH BHyTp~ r p a HynHTOB0fI ~I~aI~HH.

I. Introduction In the last few years, the increasing geologic knowledge of the Preeambrian basement terrains in central Brazil has allowed the recognition and delimitation of typical lithologie associations identical to the ones described in ancient shields of other continents (DANNI ~: FUCK, 1979). Until recently these associations were viewed as a puzzle known under generic denominations like Complexo Basal Goiano (basal complex of Goifis) or Precambriano Indiferenciado (undifferentiated Precambrian) (ALMEIDA, 1968; BA~BOSA et al., 1969). Some were even incorporated within the younger overlying folded sequences. In this paper a geologic synthesis of the main features of the lithostructural associations which form the framework of the so called C o m p 1 e x o B a s a 1 G o i a n o is presented. The basement of the Uruaguano (1800--1100 my) and Brasiliano (700--450 my) fold belts consists of Archaean and Lower Proterozoic units. These units, assembled in the Central Coifis and Cuaxup6 massifs (ALMEIDA e t al., 1976), are exposed within the Tocantins tectonic province (ALMEmA et al., 1981), as shown in fig. 1. The Coi~is massif is mostly underlain by lfigh-grade granite-gneiss terrains, with several associated Archaean greenstone belts (fig. 2). A discontinuous granulite belt extends from Porto Nacional in the north of Goi~s to western Minas GerMs, including the granulitized mafic-nltramafic Barro Alto, Niquelgndia and Carla Brava complexes. The Cuaxup6 massif displays similar features, with a granulite belt bordering the granite-greenstone terrain to the south. 292

Archaean and Lower Proterozoie Units in Central Brazil .

.

.

.

.

o

a ,,o

I

-- ~,o

,;o

Fig. 1. The main structural provinces of central Brazil. The Lower Proterozoic is considered to be represented in the area by a variety of plutonie, volcanic and sedimentary rocks, intruding and/or overlying either the granite-greenstone terrains or the granulite mobile belt. It includes the metasedimentary unit exposed to the north of Goi~s Velho, which conformably overlies the mafic-ultramafic lavas of the greenstone belt, and the Ticunzal formation which rests upon ancient tonalite and granodiorite gneiss in the Cavaleante region, and has been intruded by a number of tin-bearing granites. Within the granulite belt occur mafie-ultramafie layered intrusions of the Americano do Brasil type; gabbro-anorthosite layered plutons and associate marie to acid metavoleanics and ocean floor metasediments are thrusted upon the Barro Alto, Niquelgmdia and Cana Brava granulite complexes. The region was successively involved in at least two post Arehaean tectonic events (Uruaguano and Brasiliano); a third (Transamazonian, 2100--1800 my) is 293

J. C. M. DANNIet al. probably also present. Folding and faulting are very complex, and therefore some of the relationships of the different units are not yet adequately known. More detailed stratigraphic, tectonic, petrologic and geoehronologic work is required in order to clarify the regional geologic evolution. IL The Archaean The main Arehaean units in central Brazil consist of high-grade granite-gneiss terrains, cut by amphibolite dykes and stocks, greenstone belts and a granulite mobile belt.

I - -

e

~_

I~

C

294

Archaean and Lower Proterozoic Units in Central Brazil

1. T h e

granite-gneiss

terrains

High-grade granite-gneiss terrains make up most of the Archaean areas of western Minas Gerais and Goias (fig. 2). They are surrounded by younger Proterozoic metamorphic belts or partially covered by younger deposits. At least in

part, they also constitute the basement upon which the greenstone belts developed. The extensive area of central Brazil underlain by these high-grade gneisses is very poorly known, and investigation is still on the reeonaissanee level. The lack of detailed studies and of systematic geochronologie work renders it very

-

~

~

~-

~

~o

--,~\

i\

~e

7

~

@

g

x t z

<3 0

o N w Q: o_

~j 3~ m m

q

o

~

z

~

5 8

~

-8 ~

295

E

J. C. M. DANNIet al. difficult to establish a logical sequence of events as has been possible in other Archaean areas throughout the world (WlNDLEY, 1977). Although most isotope determinations on gnefsses sampled in the area gave results ranging between 2.000 my and 500 my (HAsuI & ALMEIDA, 1970; CORDANI • HAS~'I, 1975; HASUI et aI., 1980) it has been argued on geological grounds that the high-grade region is of Archaean age (DANNI ~ FUCK, 1979; BERBEItT, 1980). The available radiometric data, most of which K/Ar ages, are therefore evidence of reworking and remobilization of the Archaean rocks during younger Proterozoic tectono-thermal events. Recent results of TASSINARI~ MONTALVAO(1980), HASUI et al. (1980) and J. M. REIS NETO (pets. comm.) who obtained reliable Archaean ages near Crixfis, Dian6polis and Cavalcante, respectively, show that the high-grade region offers good propects for very old ages if Rb/Sr and U/Pb methods are sistematically employed. The lithology of the high-grade region is rather monotonous, comprising chiefly biotite (hornblende) quartzo-feldspathic gneisses which were probably derived from ancient plutonic rocks of tonalitic and granodioritic composition. The most common type is a rock containing biotite and/or hornblende, and plagioclase, potash feldspar and quartz in various proportions. Minor hypersthene, diopside and garnet are sometimes found. The main characteristic of these gneisses is their extensive lateral continuity and remarkable chemical and mineralogical homogeneity. In general, the rocks are medium to coarse grained, locally with large feldspar megacrysts, displaying a weak foliation, which becomes distinct in biotite enriched varieties or in strongly deformed or recrystallized types. Metamorphism is within upper amphibolite facies, although granulite facies mineralogy is sometimes recorded. In some restricted areas, banded gneisses dominate. BARBOSA et al. (1969) believe they have originated from greywackes. Other types of supracrustal rocks appear sporadically as small remains within the granite-gneiss. The most common supracrustals are quartzites, iron formation, aluminous schist and gneiss, and amphibolite (probably metabasalts). The age of the remains has not been established and in many cases, it is not even known whether they are older or younger than the surrounding gneisses. Remnants of mafic-ultramafic complexes occur as folded and recrysta]lized lenses and layers. The ultramafics are pyroxenites and peridotites often altered to serpentinites or tremolite-talc schists. Unlike other Brazilian provinces, as the Tapaj6s or Sgo Francisco provinces (ALMEIDAeta]., 1981), migmatite structures are rather uncommon in the, Goifis high-grade terrains. They have been observed in rather small areas, like in Natividade and Porangatu. Metatexitic and diatexitic structures have been reported, displaying tonalitic-granodioritic or amphibolitic paleosome and granitic neosome (CoR~EIA FILHO & S~,, 1980). In many cases it can be assumed that the migmatites are the result of remobilization associated to Proterozoie events. Several granitic and granodioritic nucleii have been recognized, as in Rubiataba, Hidrolina and Guarinos (DANNI I~= RIBEIRO, 1978; SABdIA, 1979; DANNI & FUCK, 1979) marking an important sialization phase of the Arehaean crust. They constitute large oval structures, partially surrounded by greenstone belts (fig. 8). These structures are in some cases probably older than lavas and sediments of the greenstone sequences, but in others are intrusive domes, intimately related to the tectonic evolution of the greenstones (DANNI N: RIBEI1KO,1978; TASSlNAItI • MON296

Archaean and Lower Proterozoic Units in Central Brazil

TALVAO,1980; BERBERT,1980). Recent Rb/Sr data of granitic rocks from Crix~s indicate an age of 2.929 4 105 my (TA.ssINA.RI& MONTALV_~O,1980); the initial SrSr/Srs6 ratio of 0.701 +- 0.002 leads to the assumption of a limited pre-intrusive crustal history. A similar age has been obtained on the Rubiataba granite (2.800 my, Rb/Sr, HAsur et al., 1980). In Cavalcante, J. M. dos Reis Neto (pers. comm.) obtained Rb/Sr results of 2.600 my and one of 8:200 on tonalitic-granodioritic basement gneiss. Other important features of the granite-gneiss terrains are the extensive shear belts with general N 10--80 E directions, like in Porangatu, Cavalcante, Natividade, or NW-SE directions as in Itafl de Minas. Along these belts, the gneisses have been intensely deformed to cataclastic rocks with low-grade mineralogy. Intrusions of granites and mafic-ultramafic rocks of variable ages are also present (TEIXEIRA, 1978). 2. M e t a b a s i e

Dykes

and Stocks

An important province of amphibolite dykes occurs within the granite-gneiss terrains north of Hidrolina and southeast of Crixas (fig. 8) and north of Goffts Velho (fig. 4). The age of the dykes is not known but there is evidence that they are probably older than the Goifis Velho and Crixgts greenstone belts, within which they have not been found. It is very much possible that with more detailed study the presence or absence ot the dykes will lead to a subdivision of the gneisses and enable to "break through the eonformability barrier", as has been possible elsewhere (McGREcOR, 1978; WINDLEY, 1977; WINGE • DANNI, 1980). The dykes are several kilometres long and their width ranges from a few metres to some hundred metres. Preferential directions are NE-SW and NW-SE. The dyke mineralogy is usually of amphibolite facies (hornblende, plagioclase), although relics of the igneous textures and minerals are very frequently preserved. Greenschist facies minerals (actinolite, epidote, chlorite, albite, calcite, quartz) are also common. The metamorphic texture with hardly any preferential orientation, the continuous and straight character of the dykes, indicate little or no deformation and a remarkable stability of the area during the Proterozoic. Several small stocks of amphibolitized gabbros have also been reported, as in the east :of Hidrolina (RIBEmO FILI~O et al., 1978) and the north of Goifis Velho. Some of them are layered and contain peridotite and pyroxenite facies, metamorphosed to serpentinite and tremolite-talc schist. 8. T h e g r e e n s t o n e

belts

Low grade metasediment and metavolcanic associations identified as Archaean greenstone belts have been described in several areas in Goi~ts and western Minas Gerais (fig. 2). Metamorphic ultramafite lavas with greenschist facies assemblages, unconformably overlying the granite-gneiss basement near Goi~s Velho were first reported by DANNI et al. (1978). COSTA.et al. (1976) interpreted the occurrence of schist belts in the Natividade-Almas area as similar to the Barberton greenstone belt structure of South Africa. TEIXEmA (1978) demonstrated komatiitic chemistry of the uhramafics from the Fortaleza de Minas area, and DANNI ~X:PtIBEIRO(1978) described several sialic blocks surrounded by volcano-sedimentary sequences in 297

,

/

~t

J I I i~'.~ "~' .: .... /,~" iI - )~...:,~.~-'-~..-,,'-~-~.~_

\

L

(

~

~.,,~.

~

o

/

/

I \

\1

\1

I I I

'I I j I

J

t

J

I

.-,,;,... :',,~

I

t

Archaean and Lower Proterozoie Units in Central Brazil

49el7 ' 14o22'

F i g - 3 THE CENTRAL GOIAS GREENSTONE BELTS

~ ' ~ - ~ A R A X J ~ AND SERRA DA MESA ~ROUPS

~

METACHERT, BIF. GRAPHITIC AND cHLORITIC , PHY'LLITE, ACID METAVOLCANICS,METATL~IFFS:ANDDETRITIC METASEDIMENT(QUARTZITED QUARTZ SCHIST, ETC.)

~

BASIC M'ETAVOLCANlOS, METAVOLCAHICLASTICS AND CHEMICAL METASEDIMEHTS

~'['~

ULTRAMAFIC METAVOLCANIC$

~'-~

AMPHIBOLITIZED BASIC INTRUSION!

t

J 6RANITIC- TONALITIC GNEIS NUCLEII

ADAPTED

FROM DANNI 8, RIBEIRO

(19~'8.~) I RIBEIRO FILHO & ql. ( 1 9 7 8 ) AND SAB(~IA 8 TEIXEIRA (1979) 49017'

DNPN/Dn. DE MIRANDA'

IS~

299

0

16~

i

FIG.4

0=21 ,

D*sDE 5Oe 21'

15o~'-

ARAXA

GROUP

~

MIRANDA/DNPM

\

ACiD VOLCANICS,F]NE METASEDIMENT. TO COARSE r ~ GRAINED DETRITIC PELIT ES AND CHEMICAL METASEOIMENTS ........... TIN . . . . . . . . . . . [ ( ME T~AGA$ ALT )

MIOOLE PROTEROZOIC

REGION

GOtAS

%

] GRAN. . . . . . . . . . . . . . . . . .

] AMPHIBOLITIZED

//\l.

"v---v'-~ v v v v v v

~ . ~ h ''m"

DiABASE AND GABBRO

TALC SCHIST, CHLORITE.TALC SCH=ST ( METAULTRAMAF]C VOLCANICS )

GEOLOGIC MAP OF THE GOI~,S VELHO

\

II I I

,|j,

P

i

/

I I

I 16oo 3' 496 54'

C3

Archaean and Lower Proterozoic Units in Central Brazil the Pilar de Gof~s-Hidrolina region, emphasizing their outstanding resemblance to the classical greenstone belts of other continents. Eventual disputes were finally settled when SABOIA(1979) reported the first findings of peridotite flows with well preserved spinifex quench textures at the base of the Crixfis greenstone belt. In the last two years, a rapid succession of papers disclosed new findings and details of the stratigraphy, petrography, chemistry and structure of the central Brazil greenstone belts (SA~olA et al., 1979; SABOIA& TEIXEIRA,1980; TEIXEmA& DANNI, 1979a, b; TEIXEmA et al., 1981; CO~EIA FILr~O & Ss 1980; BIONDI & SCHRANr:, 1980; ScaRANI~& BROUSSE, 1980; DANNIet al., 1980; FRITZSONSIm et al., 1980; MONTALVKOet al., 1981). Geochronology Little progress has so far been done on the radiometrie dating of the central Brazil greenstone belts. Some K/Ar results of HAS~:I & ALMEIDA(1970), ranging between 500--1200 my, point to a resetting of the K/Ar clock during the Proterozie. According to TAssrNARI & MONTALVXO(1980), the greenstones of the Crs de Goifis--Hidrolina region are at least 2.900 my, since some of the granitic reeks they dated contain remains of basic and ultrabasie volcanic rocks, probably related to the volcanic suite of the greenstone belt. General

form

The greenstones of central Brazil occur in NW-SE belts, with several tens of kilometres long and up to 20 km (in general 4--8 kin) wide (fig. 8, 4), looking like elongated troughs set in the ancient polymetamorphie sialie basement. That is probably the case in the Goi~isVelho and western Crix~is areas, where amphibolite and metabasite dykes occur in the granite-gneiss terrains but not within the volcano-sedimentary sequence. Also, to the south of Goi~s Velho, DANm et al. (1978) reported nearly fiat smoothly folded, metamorphosed ultramafic lavas unconformably overlying the granite-gneiss basement. In the Guarinos-Hidrolina area (Fig. 8) the belts conform to granodioritetonalite ovoids, indicating mobility of the sialic nucleii, which is further confirmed by the volcano-sedimentary xenoliths surrounded by the plutonie material adjacent to the greenstone belts (DANm & RmEIRO, 1978; SA~OIAet al., 1979; TASSINA•I & MONTALV~O, 1980). The structure of the belts is synformal, the volcanic and sedimentary rocks being tightly folded in isoclinal synclines with faulted limbs. Antiform structures tend to be deleted by thrusting of the limbs. Axial planes, fold axes and faults strike the general NW-SE direction of the belt. Vergence points to NE. Stratigraphy The stratigraphy of the Goifis and western Minas Gerais greenstone belts is shown in fig. 5. In general the lower part of the sequences consists of ultramafic lava flows (peridotite and subordinate pyroxenite) metamorphosed to serpentinite, talc schist and aetinolite-cblorite-tale schist, which intercalate thin layers of oxide, sulphide 301

J. C. M. DANNIet al. GOIAS V E L H O {DonnE 8~ Dardenne

P I L A R AND HIDROLINA (Donne & R i b e i r o , 197B)

unpub shed))

i1',1:[ GUARINOS

, ,, ,,,,

(DonnE,

[ [ I I LLLL," i ' i

un:oublished )

c RE XA'S (Sabdia,

FORTALEZA

DE MINAS

(Teixefra B

Danni, 1979)

1979)

2:L.7-.:

7.'?.K ~7 .

.

.

.

.

.

.

AU

.........

Fa

.

vvvvv ~vvvv vVvv v

vvv

v

vvvvv

A~

vvvvv

/vvvv\ v v v v v

"L".:'

1: "'.."

r~tl:i,hi/rlill! s A

9' . . - - . . . . . . .

+

+

,.

+

MAC NETITE - CHLORITE -

SERPENT~NITE, TALC SCHIST, SCHtST, V '~ "SPINIFEX TEXTURE ~ ( / ORTHOAMPHIBOL]TE (VV),PILLOW

[EZ] .... , 4 , . . . . . . . . . . . . , . ,

CAR~AT}C ROCKS, MARBLE,

~

.....................

c-,

METACHERT [ FO, Mn, CO~) ~

GRAPHI'TIC PHYLLITE

D

......................... CHLORITE SCHIST, SERICITE--

~

........................ MARBLE WITH BANDED

QUARTZ SERICLTE SCHIST

Fig. 5. Schematic Stratigraphy of Central Brazil greenstone belts. and carbonate faeies metaeherts and graphite phyllite. The occurrences of peridotites near Crix~s, Hidrolfna and Piumhi are especially remarkable for their well preserved spinifex texture and other volcanic features (SA~OlA, 1979; SA~OlA & TEIXEIRA, 1980; DANNI et al., 1980; TEIXEIRA et al., 1981; MONTAVAOet al., 1981). The typical flows display cumulated peridotite at the base, which changes upwards to peridotite with coarse parallel or radiate spinifex texture. At the top of the flow the quench texture is finer and gives way to polyhedral disjunction. Other types consist of massive peridotite flows with polyhedral disjunction at the top, flow breeeias (DANNI et al., 1980), pillowed peridotite flows (FI~ITZSONSJl=t.et al., 1980), and pyroxenite flows (SABoI• et al., 1979; SABOIA & TEIXEIRA, 1980). Individual flows range from a few centimetres up to 80 m in thickness. Associated with the extensive volcanic material are small metamorphosed peridotite, pyroxenite and gabbro intrusions. The thickness of the ultramafic lavas association (C6rrego Alagadinho formation in Crix~s, SABOIAet al., 1979) is of about 1.500 m in Crixas (TEIxEmA et al., 1981), and 1.200 m in Piumhi (FI~ITZSONSIn. et al., 1980).

302

Archaean and Lower Proterozoic Units in Central Brazil The following unit (Rio Vermelho formation in Crix~s, SA~OlA et al., 1979) consists predominantly of basic lavas, metamorphosed to garnet-epidote amphibolite or actinolite-chlorite schist. Sediment intercalations (iron formation, chert, carbonate rock, graphite schist) are common. There are also lens shaped ultramafie flows and intrusions. The basaltic rocks are massive or pillowed, with a variety of variolite, amygdaloid and microlithie textures. Basaltic flow breccias and basic turfs have also been recorded (FRITZSONSJK et al., 1980). Laterally and vertically the metabasalts grade to felsic volcanics including anclesites, daeites, rhyodacites and rhyolites, together with mafic to felsic pyroclastics, represented by biotite-actinolite schist, chlorite-sericite schist, magnetite-chlorite schist, etc. The mafic to felsic suite displays a thickness of 800 m in CrLx~s (SABoZAeta]., 1979) and 2.400 in Piumhi (FRrrzsoNs jI~. et al., 1980). The upper unit of the sequence (RibeirSo das Antas formation in Crixfis, SABOlA et al., 1979) consists predominantly of metasediments with common intercalation of intermediate to acid pyroelastics. Graphite phyllite, iron metachert, sericitequartz schist, chlorite-muscovite schist and quartzite are the main lithologies. Metamorphism is typically of low-grade greenschist facies. Amphibolite facies assemblages have been recorded near the margins of the belts. The presence of chloritoid in some schists (TEIxEIRA& DAXm, 1979 a; SABOIXet al., 1979) possibly indicates a Barrovian type metamorphism. Chemistry Although in small number the available chemical analyses for the central Brazil greenstone belts are sufficient to prove the komatiite nature of the ultramafic flows (Tr,ixEi~& 1978; TEIXEmA & DANNI, 1979b; SAZOIA & TEIXEI~A, 1980; SCHRANK& BaOVSSE, 1980; MONTALV~,Oet al., 1981). Their differentiation is due to early fractionation of olivine and the subsequent enrichment of AleOz and CaO, allowing the derivation of pyroxenite and komatiitie basalt liquids. The upper volcanic suite of Piumhi (BIONDI & SCHRANK,1980) displays the petrochemical character of the calc-alkaline series correlated with the modern island arcs. This is the case of the other greenstone belts as well, as inferred from the mineralogy of the volcanic rocks of their intermediate stage. Mineralization The main types of mineralization recognized within the greenstone belts of Goifis are: a) small pyrrhotite with minor pyrite, pentlandite and chalcopyrite segregations in the peridotite lava flows (Hidrolina, Crix~s); b) pyrite and arsenopyrite and gold (less than 10 ppm) in sulphide facies chert interbedded with graphite phyllite overlying the ukramafie lavas unit (Goi~s Velho, Gnarinos); c) small chromite deposits in the basal serpentinized peridotites; d) gold disseminated in serieite-quartz schist (acid turfs?) and arsenopyritepyrite bearing graphite phyllite, associated with carbonate (ankerite-dolomite)chlorite schist, or remobilized in quartz veins (Pilaf de Goi/~s, Crixfis, Natividade). e) low-grade pyrite-chalcopyrite disseminated ore in biotite schist derived from basic vo]canie and volcanoclastie rocks (Mara Rosa, Natfvidade). 303

J. c. M. DANNIet al. 4. G r a n u l i t e

mobile

belt

The granulite terrains which are part of the Arehaean basement in the median massif of Goi~s make up an extensive but discontinuous belt (Goifis Granulite Belt, WERNmK & ALMEIDA,1979) with lithostruetural features like many of the Arehaean mobile belts recorded in other shields (ANHAEUSSER et al., 1969; SUTTON & WATSON, 1971). According to ALMEIDA (1979) the metamorphic granulite belt traces the western limits of an ancient Archaean craton (Paramirim craton), which contains the S~o Francisco eraton individualized during the later Brasiliano cycle. From a regional point of view, the mobile belt occurs in isolated segments represented by the granulite complexes of Guaxup6, An~polis-Itaucu, Barro Alto, Niquelgmdia, Cana Brava, Gameleira and Porto Naeional (fig. 9.). This longitudinal discontinuity seems to be partly the result of the presence of Proterozoie cover rocks in downfanlted blocks by NW-SE faults. Most of these complexes are thrusted upon sialic Archaean rocks at their eastern limits. In the regions of Cana Brava, Niquelandia and Ceres-Barro Alto (ffg. 6, 7, 8) there is evidence that these thrusts are very old although the last movements are related to the Brasilfano cycle (DANNI& FUCK,1979; FucK & MARINI,1979 a; MAIIINI et al., 1979), as can be inferred from the radiometrie age determinations (644 +_ 27 my, tlb/Sr whole rock isocron, initial ratio 0.786) of GmA~DI et al.

(1978). Investigation of the lithostratigraphic and structural features of the Barro Alto, Niquelandia and Cana Brava complexes allows to point out some important aspects of the evolution of the granulite belt (fig. 6, 7, 8): a) The granulite belt has a width of about 9.0 km, being on both sides limited by high angle thrust faults. The eastern fault system thrust the granulites, mostly derived from gabbro and norite, upon Archaean granite-gneiss basement. The western limit displays features of involvement of the granulite rocks in younger amphibolite facies metamorphic events. There the belt is faulted against younger gabbro-anorthositie complexes or amphibolites derived from gabbros and/or basalts, which according to DANm & LEONAIaDOS(1980) represent oceanic crust formed during the Lower Proterozoie, after the granulitization event. Evidence of this is corroborated by the presence of granulite fragments in gabbro or anorthositie gabbro breecias. b) The stratigraphy of the lower granulite units is well observed in the Serra da Mantiqueira, north of Niquelandia (fig. 7). There the base is made up of metamorphosed spinel peridotites (harzburgite, dunite, and some lherzolite) with textural, mineralogical and chemical features similar to mantle peri0dotites. c) These rocks are overlain by a layered ultramafic sequenee of metapyroxenite (websterite) and metaperidotite (dunite, harzburgite). Deformed thin cromite layers appear near the base. Towards the top there are lenses of metanorites and metatroctolites. Layering and relict cumulus texture result from crystal fractionation and cyclic magmatic sedimentation. d) At the top, the ultramafies grade rapidly to norite and hyperite recrystallized to granulite with annealing texture, metamorphic banding and isoelinal folds with steep axes and axial surfaces. In the Barro Alto complex gabbro-diorite intrusions of the granulite belt contain 304

48 ~ , 12~ '

48oz4' ~'/~-~/~ MYLONITE

BELT

'~

PARANOAGROUP

~

ARAI"GROUP

- ~

SERRADA MESA GROUP

.3,

--~Jl

.&

ALM El ROPOLI S VOLCANO-

SEDIMENTARy SEQUENCE ----F-'~ CATAGLASTIC HORNBLENDE -BIOTtTE GNEIS WITH MYLONITE BELTS//I/ ~

NORITIC AND HYPERITtC GRANULITE

-3.

AMPHIBOLITE ULTRAMAPH~C ROCKS: METAPERI SERPENTINITE AND RARE PYROXENITE

.3.

.5,

.5.

_% ..%

-N

'~ 7

7

/ /

SOS5'

Fig. 6. Geologic map of the Cana Brava complex. 20

Geologischc

P t u n d s c h a u , B d . 71

305

48~

FIG.

14005'

7

-

GEOLOGIC

6

L

I PARANOb*GROUP ~ ; o F r - - ~ B"YTHM'OSLATEOUARTZ,TE--~

MAP

OF

NIQUEL/~NDIA

48~

14~ 48 ~ [7

E G E N DA

~:

~'"L~ Oi ~

UMEBTONE

0 m

o

N

~ .-~

bJ . ~ oc --

CH~OB,T~SO.,GT.P"u =:1 MUSCOVITE SCHIST AND AMPHIBOLITE ( b ) 4, FINE GRAINED AMPHIBOLITE WITH CHERT AND CALCSILI-CATE INTERCALATIONS

O(~ ...I Z ~

ANOBT.OB,TE

LPLYERS AT THE BASE

~: r 2. SERRA DA MESAGROUP(?) I ~ I ~ wig ACID" INTERMEDIATE META" ~ ~O_[]--~/--~F-O VOLCANICS,MUSCOVITE r ' ~

. ~.. ~. .

BIOTITE - HORNBLENDE NORITIC AND GABBROIC GRANULITE

OOABSEORA,NEDA~P.,~-

OI,~IF..S?..I L,T~ *,TH

14~ =

F~'.~ ANORTHOSITE

NOB,T,C ANO OA.BBO,C

GRANULITEI SILEXITE (e) IN STRIKE SLIP ZONE

..........

TROCTOL IT E

<

TONALIT]C GNEIS , CATACLASITE. AND PHYLLJ3NITE

CUMULITIC , METAPERIDODITE AND PYROXENITE, SERPENTtNITE M ETAPERIDODITE SERPENTIN ITE Dos. DE MIRANDA/DNPM

306

Arehaean and Lower Proterozoic Units in Central Brazil 8~7~

BAMBUI"AND PARANOAGROUPS

7 ~

ARAXA GROUP

+

+

G("~'[ VOLCANO- SEDIMENTARYSEQUENCEOF JUSCEL'~NDIA S ~

`4

BANDED AMPHIBOLITE, LOCALLY W[TH ANORTBOSITE LAYERS

`4

4 q

ANORTHOSITE~ ANORTBOSITIG GABBRO AND 4L:z:~,3 OLIVINE GABBRO

`4

4

4

-1

4

l

q

3[~

GRANITIC - TONALITIC - GNEISSES

2

~

NOR[TIG-NYPERITICGRANULITES WITH RARE PYROXENITE; ACID SRANOLITES

~

METAPERIDODITEtSERPENTINITE

I

f 4

/ `4

NUMBERS INDICATE PROBABLE CRUSTAL DIFFERENTIATION ORDER FROM .[ BOTTOM TO TOP

/

"4

4

"4

"4

"4

"4

/

4

"4

"4

"4

ca

"4

"4

< "~

"1

"4

s

,1

"4

=4

"4

"4

"4

4

"4

"4 "4 "4

"4

`4

"4

"4

"4 "1

"4

/

"4

/ /

J /

/'/

.''21P",~ +

T

:.j +/ +

+

T

.,...--j + ~"

r" ,_j

-/'~"r i + + + J ,~' p ;/ "~ + + "~,,r .~-2/ + + ~ "-" /-"

,:-,'_,,-.>,,,_,-x,c, - - " T T - ~" - , ' . , ' - ? , , O L -

+

O ~

~, +

r

+ + +

2,5

+ +

+ +

+ 9

+~. +l

/7/

,~ J

/

,,~ J

/J / '~ . j ,,,-

,--,' ~ r'j

+ j

+ N

~J

GOIANESIA ~

~

/ J ,,j ~J

J~J

Modified after BAETA st o l i i ( 1 9 7 2 ) ,

~km

STACHE (1972), FUCK &01(1981)

FIGUEIREDO ( 1 9 7 6 ) . DI~.DE MIRANDA / D B P ~

49 ~

48040,34,,

Fig. 8. Geologic map of Barro Alto. xenoliths of supracrustal granulitized rocks. Large areas of granulites of supracrustal origin (leptinite, garnet-hypersthene quartzite, fine-grained basic granulite and cale-silicate rocks) overlie the noritic granulites north of GoianGsia. These supracrustal units have been interpreted as roof remnants of the igneous complexes 307

J. C. M. DANNIet al. by BAET~et al. (1972) and STACHE(1976). Judging from the nature and lithological association of these supracrustals, the Arehaean crust at the site of the granulite mobile belt consisted of basic-to-acid voleanics with intercalated chert and impure limestone. To the present writers these materials may possibly represent parts of ancient greenstone belts reworked and recrystallized under high-grade metamorphism, in a similar manner as that reported for the Limpopo, Zambezi and Mozambique belts by KR6Nma (1979 a, b). The Anfipolis-Itaueu complex comprises an extensive region south of the Pirineus inflection. Although not studied in detail it is known by its remarkable lithologieal diversity, liegionally, hornblende-pyroxene bearing gneisses, granodiorite gneisses and leptinites seem to dominate. They sometimes display strong deformation, as well as amphibolitization and migmatization, as in the Itauen region. In the Goianira-Trindade region NILSON • MOTTA (1969) described a differentiated gabbro-pyroxenite body, associated with pyroxene and hornblende bearing gneisses. Metasediments (quartzite, impure marble, tale-silicate rocks) are intercalated with garnet-pyroxene gneisses probably derived from ancient voleanics. These rocks have been intruded by the Serra da Pedra granodiorite. In the Goi~nia-Anfipolis region banded basic granulite is associated with thick leptinite beds. In other areas there are highly deformed anorthosites with ilmenite layers, as well as rocks of the charnoekite-mangerite series and small gabbroperidotite massifs with similar features to Niquelfindia and Calla Brava. The Gameleira complex south of Natividade fs a differentiated mafie-ultramafic complex, partly granulitized, surrounded by migmatites (CoI~REIA FILHO & S~,, 1980). Along the western border of the belt, the Carla Brava, Niquel~ndia and Barro Alto granulites have been retromorphosed to the amphibolite facies, becoming rich in hornblende, garnet and biotite. This indicates that the granulite belt has been submitted to the deformation and recrystallization event of the overlying gabbro-anorthosite-amphibolite sequence. Deformation of the granulite rocks is marked by metamorphic banding, often parallel to the primary igneous layering, and by mineral lineation developed by metamorphic pyroxene and plagioelase. Isoclinal folds with steep axes are frequent as well as transposed fold limbs clue to strike-slip movements (DANNI& LEONAI/DOS, 1980). In Niquelfindia and Cana Brava strong deformation zones are marked by linear trends of silexite and pseudotaehyllite, related to an important transcurrent fault system. In spite of the many controversies, it is believed that the granulite belt is of Archaean age (ALMEIDA, 1979; DANNI & FUCK, 1979; MAltlNI et al., 1979; WERNICK & ALMEIDA,1979; WEBNICK, 1981). Most radiometrie age determinations refer to the rocks of the Cana Brava, Niquel~ndia and Barro Alto complexes (HAsuI & ALMEIDA, 1970; HASUI et al., 1972; Sotrzh, 1978; MATSUI et al., 1976; GllaA~DI et al., 1978). Rb/Sr methods gave unsatisfacto12r results due to the low lib content of the analysed samples (SouzA, 1973). K/Ar determinations gave results scattered between 500 my and 4000 my. The younger ages certainly reveal isotopic rejuvenation of the rocks during the Urucuano and Brasfliano cycles. The very old ages led to the suggestion that the complexes may be slabs of the primitive crust of our planet, formed by differentiation about 4000 my ago (CoRDaNI & HASUI, 1975). However, the possibility of the presence of excess argon in the analysed minerals 308

Archaean and Lower Proterozoic Units in Central Brazil must be taken into account (GIRARDI et al., 1978; HAsuI et al., 1980). We believe that these rocks in fact represent slabs of Archaean lower crust and upper mantle recrystatlized under granulite metamorphism conditions during a Late Archaean event correlatable to the Jequi6 cycle (2700 + 100 my), established by CORDANI (1978) in southern Bahia. From this discussion it is clear that the problem of the age of the granulite is not settled, and that much investigation has to be carried out in order to obtain consistent and reliable data. Important economic laterite type nickel depositsoccur related to the ultramafic rocks of the Niquel~ndia and Barro Alto complexes. The serpentinization of the ultramafic rocks of the Cana Brava complex is responsible for the most important chrysotile asbestos mine of South America (MILEWSKI et al., 1970; MIL~WSKI, 1975). Chrysotile asbestos are also mined in Santo Antonio da Laguna, north of Barro Alto. Besides this, some thin teetonically disrupted chromite layers are associated with the cumulated ultramafie rocks in Niquel~ndia and pyrrhotiteehalcopyrite disseminations have been found in noritie granulites in the Niquelgndia and Barro Alto complexes.

III. The Early Proterozoic The areas immediately west of the granulite terrains of Barro Alto, Niquelandia and Cana Brava display two types of eogenetie magmatic associations: the gabbroanorthosite complexes and the volcano-sedimentary sequences, both metamorphosed to the amphibolite facies. Layered mafie-ultramafie intrusions occur within the An~polis-Itau~u granulite complex. In the domain of the sialie basement Early Proterozoic events were responsible for the platform covers of gneiss, pelitic and graphitie schist of the Ticunzal formation and for the tantalum and tin bearing granites, with associate greisen and pegmatite. The metasedimentary sequence which overlies the marie-ultramarie voleanics of the Goffts Velho greenstone belt probably is also Lower Proterozoic. 1. G a b b r o - a n o r t h o s i t e

plutonic

associations

Gabbrokanorthositie intrusions occur along the western borders of, and overthrusting the Niquelandia an Barro Alto granulite complexes (Fig. 2, 7, 8). The gabbro-anorthosite associations are ancient layered intrusions, deformed into domes and tight isoclines. The core of these structures is made of troetolite and coronitie olivine gabbro, which pass to layered anorthosite, with few amphibolitized marie layers which represent ancient pyroxene rich cumulites (diopside and some hypersthene) transformed to hornblende and garnet. The primitive cumulus minerals responsible for the stratigraphie succession are olivine-labradorite-elynopyroxene-minor orthopyroxene. Anorthosites with more than 90 0/~ plagioclase (An 80--65 ~ are the most typical cumulated rocks. The main metamorphic reerystallization occurred through the cpx 4- plagioclase ~H-20 = garnet + hornblende reaction. In many places, a second generation of hornblende is noted, due to tectonic remobilization of the former. 309

J. C. M. DANNIet al. Towards the top, the mafic layers become progressively more abundant, thicker and more regular in a zone of garnet amphibolite and anorthosite which grades to a thick band of coarse-grained garnet amphibolite. Lenses of ilmenfte occur associated with the anorthosite and amphibolite rocks. Tight isoclinal folding with transposition of limbs and boudins are the most prominent tectonic structures of these rocks. Metamorphism is of the amphibolite facies, the most typical mineral association being Ca--plagioelase, green and brown hornblende and garnet. The age of this plutonie association is not well established. There are a few K/Ar determinations of anorthosite, metagabbro and amphibolite (HAsuI & ALMEIDA,1970; HAsuI et al., 1972; SOUZA, 1978) ranging from 780 to 2450 my, with the clustering of several results around 1700--1900 my, which probably corresponds to the age of the amphibolite metamorphism. Again, the younger ages are probably related to Middle and Upper Proterozoic isotopic rejuvenations. On the other hand, the gabbro-anorthosite intrusives are younger than the granulites, since they have not been granulitized and they do contain granulite fragments in amphiboliti~ed magmatie breceias. Therefore a Lower Proterozoie age is likely. 2. V o l c a n o - s e d i m e n t a r y

sequences

The plutonie gabbro-anorthosite association is sueeeded by metamorphic sequences of basic voleanics (fig. 7, 8) which form the base of the Indaian6polis (DANNI & LEONARDOS,1978, 1980), Juseelandia (FucKet al., 1981) and Palmeir6polls (tllBEIRO FILHO & TEIXEIRA, 1981) volcano-sedimentary sequences. As recognized by DANNI & LEONARDOS (1980) the suites of metabasalts, associated with ferriferous metacherts and talc-silicate rocks are clearly ancient ocean-floor sequences. The sequences start with fine-grained massive generally garnetiferous amphibolite. Mierolithic and amygdaloid textures, as well as the presence of intercalated metaehert give evidence that these rocks are the result of submarine basic volcanism. At the top, the amphibolite is associated with hornblende gneiss and coarse grained metaehert, followed by biotite-muscovite schist with garnet and kyanite and sodie-potassie gneiss with sodic amphiboles of the crossite-riebeekite type (Indaian6polis sequence). DANNI & LEONARDOS(1980) believe that the volcanosedimentary sequence is genetically related with the gabbro-anorthosite complex, the plutonic rocks being the magmatie counterpart consolidated under the ancient oceanic ridge. In the Juscelandia sequence, fine grained amphibolite and metachert intercalations are followed by microeline and oligoelase rich muscovite-biotite gneiss, loeally with relict feldspar phenoerysts. Towards the top, the sequence is dominantly sedimentary, and chiefly formed by garnet-muscovite schist and quartzite. According to t{IBEIItOFILHO & TEIXEIRA (1981), the Palmeir6polis sequence is very similar to the Inda/an6polis and Juscelfindia sequences. The lower unit is composed of fine grained amphibolite (metabasalt) intercalated with gneiss, quartzite and garnet-muscovite-biotite schist; the upper unit consists of micaschist (biotite, muscovite, quartz, chlorite, graphite, garnet, magnetite, and pyrite being 310

Archaean and Lower Proterozoie Units in Central Brazil the usual minerals present) associated with amphibole schist, feldspathie amphibole schist, ealc-silieate rocks, metaehert, and quartzite. There are no geochronological determinations on the volcano-sedimentary sequences. According to the suggestion Of DANNI & LEONM/DOS (1980), they are genetically related to the gabbro-anorthosite layered plutons, and therefore may be of a Lower Proterozoie age. Although the relationship with the Serra da Mesa group, of assigned Middle Proterozoie age, is not yet established (Ft~cK & MAtllNI, 1979 a), there remains the possibility of both units belonging to the same fold belt developed during the Uruaeuano cycle (1400--1100 my) or even during the previous Transamazonian cycle (2000 +_ 200 my). 8. M a f i e - u l t r a m a f f e

layered

complexes

Several mafic-ultramafic layered complexes (Americano do Brasil, Mangabal I and II, Adelfindia, Fronteira do Norte) occur west of Goi~nia, within the An~tpolisItangu block of the granulite belt. The layered intrusions consist of dunite, harzburgite, olivine pyroxenite, olivine-hypersthene hornblendite, eoronitie gabbro, norite and hornblende diorite, which are the result of fractional crystallization of a tholeiite magma and of accumulation of the. olivine, orthopyroxene, elinopyroxene and plagioelase phases. The hornblende reeks seem to have crystallized under high water vapor pressure. The diorites derived from late liquids. There is evidence that these intrusions are older ihan the Middle Proterozoie Arax~ group which regionally overlays the Archaean gneisses and granulites. They are probably of the same Lower Proterozoic age of the previously discussed gabbro-anorthosite layered intrusions near Niquelandia and Barro Alto. Massive Cu, Ni, Co sulphide deposits are related to the mafic-ultramafie rocks of the layered complexes, as in Amerieano do Brasil. 4. M e t a s s e d i m e n t a r y

sequence

of Goidts Velho

i n the Serra de Santa Rita the greenstone belt formations are overlain by a metassedimentary sequence. The sequence consists of basal metaeonglomerate and conglomeratic quartzite grading upwards to muscovite-quartz schist and chloritesericite phyllite, and an uppermost unit of graphite phyllite and metadolomite, with small bands of iron formation and metaehert. No structural unconformity with the underlain ultramafie[ and mafie metavoleanies has been recognized and both units have been affected by the same deformation and the low grade greenschist facies metamorphic event. It is then possible that the sequence represents the upper sedimentary group of the Goi~s Velho greenstone belt. The basal metaeonglomerates and quartzites contain detritie pyrite, gold and uraninite with striking similarities with the Lower Proterozoie Moeda formation of the Quadril~ttero Ferrifero, the Jaeobina group in Bahia, and the Wittwatersrand and Blind River conglomerates. 5. T h e T i e u n z a l

formation

First described by MArtINI et al. (1~78), the Ticunzal formation consists of graphite micaschist, quartz micascbist, and a basal unit of muscovite-biotite gneiss. The formation is about 800 m thick and occurs east of the Tocantins river, ex311

J. c. M. DANNIet al. tending from southwest of Cavalcante as far as Campos Belos and Nova Roma in the eastern part of the State of Coi~s. The Ticunzal formation rests upon the Archaean granite-gneiss basement and is unconformably overlain by the Middle Proterozoic Arai group. Separation of the basal gneiss from the basement gneiss units is difficult and may be only attained through very detailed studies. Preliminary Rb/Sr results (J. M. Reis Neto, pers. inf.) indicate ages around 9,000 my for the Ticunzal gneisses as compared with the Archaean ages of the granodiorite and tonalite basement gneisses (2600--3200 my). In many places, the formation is pierced by pegmatite and granite plutons, some of which bear tin mineralization. The formation of greisen from the micaschist has been recorded at the borders of the Serra Branca granite (ANDRADE, 1978; ANDRADE& I)ANNI,1978; FUCK & MARINI, 1979 a). Metamorphism of the formation is of high temperature greenschist to amphibolite facies, with superimposed retromorphic low temperature greenschist facies assemblages. From the metallogenic point of view, the formation is important for its uranium mineralization. The stratigraphic position, as well as the sedimentary, petrographic and metalogenic characteristics of the formation are strikingly similar to those of the Lower Proter0zoic host formations of the vein type uranium deposits of Australia (Ranger) and Canada (Rabbit Lake). Like those formations it is believed that the Ticunzal formation is representative of euxinic platform sedimentation in localized epicontinental basin, with high salinity and abundant algae activity. 6. T i n b e a r i n g

granites

In the central region of Goi~is, between Mata Azul and Nova ltoma, about twenty granites have been mapped, which constitute a significant tin province, with subordinate Ta, Be, F and tourmaline mfneralizations. The dimensions of the granites are variable, ranging from a few to several ten kilometres, as the Serra Dourada and Serra da Mesa granites. In the eastern part of the province, the granites intruded basement gneiss and migmatite. Contacts are sharp and the intrusion relationship is neatly displayed. Towards the west, the granites intruded the Ticunzal formation, with the local development of andalusite within the contact aureole (MARINI et al., 1978) and the formation of exogreisen tin mineralization (ANDRADE, 1978; ANDnADg& DANNI, 1978). In the western part of the province, the granites are completely surrounded by Middle Proterozoic metasediments of the Arai and Serra da Mesa groups, deformed into large braquianticlines (BAlaBOSAet al., 1969; MAItINI et al., !977). The relationship with these metamorphic rocks is still controversial. In the region between the Preto and Tocantins rivers there is no evidence of intrusion, the basal quartzite of the Arai group resting unconfolznably over the Flor~ncio and Chapada de S~o Roque granites (FucK & MARINI, 1979 b). Further west, at the borders of the Serra da Mesa, Serra Dourada and Serra do Encosto granites, the basal quartzite of the Serra da Mesa group displays sharp contact with fine grained gneissie rocks which may be the deformed border facies of the granites, although according to MAItINI et al. (1976, 1977) there is some evidence of metasedimentary origin. 312

Archaean and Lower Proterozoic Units in Central Brazil tladiometric age determinations also lead to controversial interpretation, due to the high ( > o.780) S§ 86 ratios. Conventional data, as well as ttb/Sr reference isochrons, show several results between 1.100 and 1.800 m.y. for the Serra Dourada and Serra da Mesa granites, and around 1.600 m. y. for the Sueuri and Pedra Branca granites (J. M. do t/eis Neto, pets. comm.). The high 87SrjS6Sr ratios are evidence of probable rejuvenation of the rocks, which is confirmed by K/Ar determinations, with results around 500 m. y. (HAsuI& ALMEmA, 1970). The granitic rocks are usually medium to coarse grained, foliated, sometimes inequigranular, displaying large phenocrysts of K-feldspar and rounded bluish quartz. Grelsen formations are locally abundant. The endogreisen are associated with fractures or with the cupula of the granite (ANDI1ADEI~x:DANNI, 1978). The exogreisen are the product of the transformation of mieashist assigned to the Ticunzal formation. Quartz, yellow Li-mica, albite, beryl, cassiterite, fluorite, topa2 and Fe-Cu sulphides are the most common minerals of the greisen. Cassiterite bearing quartz veins are very common (ANDItADE, 1978).

IV. Concluding remarks The structural and lithological framework of the ancient Precambrian terrains of Goias and western Minas Gerais comprises several teetono-stratigraphie compartments, each one with its own specific evolutionary features, the discussion of which is beyond the scope of this work. The main compartments are: a) high-grade granite-gneiss, cut by amphibolite dykes and stocks, b) low-grade greenstone belts, c) granulite mobile belt, d) metasedimentary sequences overlying granite-greenstone terrains with intrusive tin bearing granites, e) gabbro-anorthosite layered plutons associated with mafie volcanics followed by acid and intermediate volcanics and ocean floor sediments. f) mafic-ultramafic layered intrusions in the granulite belt. These units display regional arrangement similar to other Precambrian shields (see WINDLEY, 1977, for a review). It contains the same lithostructural elements, representing the early evolution stage of the crust of our planet. As a consequence, the same problems related with the understanding of this evolution are also recognized in central Brazil, with many features and relationshisp as yet not well explained. Among these problems concerning the regional geology we stress the following: 1. The recognition of an ancient pre-greenstone belt sialic crust. There is evidence in central Brazil which favors the existence of a pre-greenstone belt sialic crust: a) An unconformity separates the granite-gneiss basement from the overlying nearly horizontal nltramafic lava flows near G0ifis Velho; b) the presence of amphibolite dykes in the granite gneiss basement which are not found within the adjacent greenstone belt; c) the wide extension of the tonaliticgranodioritie gneisses as compared to the narrow greenstone belts, which can be interpreted as ellongate troughs set within the sialie basement. In the Archaean evolution of the area there is an important sialization phase represented by the 2900 my old oval shaped granitic nucleii. Their mobilization has conditioned the 313

J. C. M. DANNIet al. structure of the greeenstone belts, and it is probable that the metamorphism and deformation of the belts are related to this phase of sialization. 2. The geodynamie framework for the build-up, and the metamorphic and structural evolution of the greenstone belts. 8. The definition of the geodynamic conditions necessary to the formation of granulite mobile belts, whether involving the generation of new crust or the reworking of pre-existing materials. The facts established for the Goffts granulite belt point to the reworking of supracrustal, infracrustal and upper mantle material, along deepseated transcurrent zones, as shown in the Niquelfindia, Barro Alto and Cana Brava complexes. 4. The unique association of gabbro-anorthosite layered complexes with volcanic basalt formations, underlying a thick geossynclinal folded pile of volcanic and sedimentary rocks. This situation probably represents a first stage of ocean opening followed by consumption of the oceanic crust under the western continental block, illustrating the possibility of a Wilson-cycle evolution during the Proterozoic in central Brazil.

References

ALMEIDA, F. F. M.: Evolucao TectSnica do Centro-Oeste Brasileiro no Proteroz6ico Superior. Anais Acad. 13rasil. Ci~nc., 40 (supl.): 285--295, Rio de Janeiro, 1968. - - : O Craton do Paramirim e suas rela,c5es corn o do Sao Francisco. I Simp. Geoi. Craton S~o Francisco e suas faixas marginais (resume). Soc. Bras. Geol., Ndcleo Bahia, Salvador, 1979. ALM~IDA, F. F. M., HASUI, Y., & NEVES, 13. 13. t3.: The Upper Precambrian of South America. Bol. Inst. Geol. USP, 7: 45--80, S~o Paulo, 1976. ALMEIDA, F. F. M., HASUI, Y., NEVES, 13. t3. 13., & FUeK, R. A.: Brazilian structural provinces: an introduction. Earth--Science Reviews, 17: 1~29, Amsterdam, 1981. ANDRADE, G. F.: As mineraliza,cSes de estanho, berilio e cobre do granito da Serra 13ranca, Cavalcante - - Go. Univ. Brasilia, unpublished MSc thesis, 1978. ANDRADE, G. F., & DANNI, J. C. M.: As mineraliza,c5es de estanho, berilio e cobre do granito da Serra Branca, Cavalcante, GO. Anais XXX Congr. 13ras. Geol., 6, 2579-2593, Recife, 1978. ANHAUESSER,C. R., MASON, R., VILJOEN,M. J., & VILJOEN, R. 19.: A reappraisal of some aspects of Precambrian Shield Geology. Bull. Geol. Soc. Am., 80: 2175--2200, 1969. 13A~TA JR., J. D. A., FmUEmEDO, A. N., SOUZA, E. P., & MELLO, J. C. R.: Projeto Goian6sia-Barro Alto. Goigmia, DNPM-CPRM, 2 V. (unpublished), 1972. BARSOSA, O., BArTISTA,M. 13., DXER, R. C., BRAUN, O. P. G., & COSTA, J. C.: Geologia e Inven~rio dos Recursos Minerals do Projeto 13rasilia. Rio de Janeiro, DNPMPROSPEC, 225 p. (unpublished), 1969. 13ESBEaT, C. O.: Complexo Basal Goiano. Anais XXXI Congr. 13ras. Geol., 5:2887 2849, Camborifl, 1980. 13IONDI, J., & SC:rtnANK,A.: Sequ8ncia Vulcgnica calco-alcalina em Plum-hi, MG. Considera,cSes para Identificac,go. Anais XXXI Cong. Bras. Geol., 4, 1983--1944, Camborifi, 1980. CORDANI, U. G.: Evolu,cgo geol6gica pr6-cambriana da faixa costeira do Brasil entre Salvador e Vit6ria. Univ. Sgo Paulo, unpublished thesis, 1978. CORDANI, U. G., & HASUI, Y.: Coment~trios sobre os dados geocronol6gicos disponiveis para a Folha Goi~s. In: Brasil, DNPM, Carta Geol6gica do BrasO ao Mili0n6simo. Folha SD-22 Goffts, 85~94, Brasilia, 1975. 314

Archaean and Lower Proterozoic Units in Central Brazil CORREIAFILHO, F. C. L., ~ S~., A. M.: Projeto Natividade. Goffmia, DNPM-CPRM, 6 V. (unpublished), 1980. COSTA, L. A. M., PORTELA,A. C. P., NILSON,A. A., VALE,C. R. O., MARCHETTO,C. L. M., SANTOS, E. L., MENEGUESSO, G., INDA, H. A. V., STERNA, L. H., MARCHETTO,M., BAPTISTA,M. B., FRATIN,O., MOSMANN,R., OLIVEIRA,T. F., & SILVA,W. G.: Projeto Leste Tocantins-Oeste do Silo Francisco., Rio de Janeiro., DNPM-CPRM, 300 p. (unpublished), 1976. DANNI, J. C. M., DARDENNE,M. A., PUCK, R. A., & RIBEIRO,M. J.: Geologia da extremidade sudoeste da Serra Dourada, Goi~s, Brasil. Rev. Bras. Geoc. 3 (3): 160~180, S~o Paulo, 1973. DANNI, J. C. M., ~: FUCK, R. A.: Unidades tcctono-estratigr~ficas do embasamcnto das faixas dobradas da margem ocidental do craton do S~o Francisco. Anais I Simp. sobre a Geol. Craton do S. Francisco e suas faixas marginais. Salvador, Soe. Bras. Geol. (in press), 1979. DANNI, J. C. M., & LEONAI:tDOS,O. H.: Granulite metamorphism and deformation in ancient composite layered intrusions: the Niquelandia Complex, Brazil. XXVI Congr. Geol. Int., V. 1, p. 85, Paris, 1980. - - : The Niquelandia Mafie-Uhramafie Granulites and Gabbro-anorthosite-metavoleanic associations (submitted to Precambrian Research), 1980. DANNI, J. C. M., LEONARDOS,O. H., ~: ANDRADE,G. F.: FeicSes vulc~nicas das rochas ultram~fieas de Hidrolina--Go. Anais XXXI Congr. Brasil. Geol. 4: 1997--2003, Camborifl, 1980. DANNI, J. C. M., ~x: RIBEIRO,C. C.: Caracterizag~o estratigr~fica da sequ~ncia vulcanosedimentar de Pilar de Goi~s e Guarinos, Goi~s. Anais XXX Congr. Bras. Geol., 2: 582--596, Recife, 1978. FRITZSONS Jr., O., BIONDI, J. C., & Cr~ABAN,N.: Geologia da regi~o de Piumhi (MG). Anais XXXI Congr. Bras. Geol. 5: 2906---2917, Camborifl, 1980. FucK, R. A., DANNI, J. C. M., WINGE, M., ANDRADE,G. F., LEONARDOS,O. H., BARREmA, C. F., & KUYUMJIAN,R. M.: Geologia da regi~o de Goian6sia (unpublished data), 1981. FucK, R. A., & MARINI, O. J.: 0 Grupo Araxfi e unidades homotaxiais. I Simp. sobre a Geol. do Craton do S~o Francisco e suas faixas marginais. Salvador, Soc. Bras. Geol. (in press), 1979 a. - - : Projeto S~o F61ix: Geologia da Folha C6rrego S~o Bento. Brasilia, Cony. FUBELETRONORTE, 49 p. (unpublished), 1979 b. GIRARDI,V. A. V., KAWASHITA,K., BASEI, M. A. S., & CORDANI,U. G.: Algumas considera,cSes sobre a evolu,cao geol6gica da regi~o de Cana Brava, a partir de dados geocronol6.gicos. Anais XXX Congr. Bras. Geol. 1: 837--884, Recife, 1978. HASUI, Y., & ALMEIDA, F. F. M.: Geocronologia do Centro Oeste Brasileiro. Bol. Soc. Bras. Geol. 19 (1): 7 26, S~o Paulo, 1970. HASUI, Y., KAWASHITA,K., & IWANU.CH,W.: Geocronologia do maci,co b/tsico-ultrabdtsico de Sao Jos6 do Tocantins, Goffts, pelo m~todo K/Ar. XXVI Congr. Bras. Geol. Bol. Esp. 2 (abstract), Bel6m, 1972. HASUI, Y., TASSlNARI,C. C. G., SIGA Jr., O., TEIXEIRA, W., ALMEIDA,F. F. M., & KAWASHITA, K.: Data,cSes Rb-Sr e K-Ar do centro-norte do Brasil e seu signfficado geol6gico-geotect6nico. Anais XXXI Congr. Bras. Geol., 5: 2659--2676, Camborifl, 1980. KR6NER, A.: Precambrian crustal evolution in the light of plate tectonics and the undation theory. Geologie en Mijnbouw, 58 (2): 281--240, 1979 a. - - : Pan-African mobile belts as evidence for a transitional tectonic regime from intraplate orogeny to plate margin orogeny. In: A. AL-.SHANTI(ed.). Evolution and Mineralization of the Arabian-Nubian shield, vol. I: 21---37. Oxford, Pergamon Press, 1979 b. 315

J. C. M. DANNIet al.

MAltINI, O. J., FucK, R. A., DARDENNE,M. A., & FARIA, A.: Contribui~o ~ Geologia do Precambriano da porc~o central de Goi~s. Rev. Bras. Geoc., 7 (4): 304~324, S~o Paulo, 1977. MARINI, O. J., FUCK, R. A., & FIGUREIREDO,A. N.: Projeto S~o F61ix. Cony. FUBELETRONORTE, 125 p. (unpublished), 1976. MARINI, O. J., FreE, R. A., DANNI, J. C. M., & DARDENNE,M. A.: Geotect6nica da Faixa M6vel Brasilia e do seu embasamento. I Simp. sobre a Geol. Craton do Sgo Francisco e suas faixas marginais. Salvador, Soe. Bras. Geol. (in press), 1979. MARINI, O. J., LIBERAL,G. S., BEIS, L. T., TRINDADE, C. A. H., & SOUZA,S. L.: Nova unidade ]itostratigr~fica do Precambriano do Estado de Goi~s. XXX Congr. Bras. Geol., Bol. 1, 126--127, Reci/e, 1978. MATSUI, K., GIItARDI,V. A. V., BASEI, M. A. S., & HASUI,Y.: Idades K-At no Complexo B~sico-Ultrab~tsico de Cana Brava, Goifis. Anais 29 ~ Congr. Bras. Geol. 4: 279--288, Ouro Preto, 1976. MCGREGOR,V. R.: The Early Precambrian gneisses of the Godthaab district, West Greenland. Phil. Trans. R. Soc. London, A 273: 343--358, 1973. MILEWSI
Archaean and Lower Proterozoic Units in Central Brazil TEIXEmA, N. A.: Geologia, Petrologia e Geoqulmica da Seqti~eia Vulcano-sedimentar do Morro do Ferro, Fortaleza de Minas, MG. Univ. Brasilia, unpublished MSe Thesis, 1978. TEIXEIt/A,N., & DANNI,J. C. M.: Petrologia das lavas ultram~fieas da sequ~neia vulcanosedimentar de Morro do Ferro, Fortaleza de Minas, MG. Rev. Bras. Geoc., 9 (2): p. 151--158, S~o Paulo, 1979 a. - - : Geologia da raiz de um greenstone belt na regi~o de Fortaleza de Minas, MG. tlev. Bras. Geoc. 9. (1) : 17--26, S~o Paulo, 1979 b. TEIXEIIIA, N. A., SABOIA, L. A., FEIIREIRA, C. M. B., TEIXEIIIA, A. S., & CASTRO, J. H. G.: Estruturas e texturas das lavas ultrab~sicas e b~sicas do greenstone belt de Crix~sGoifis, Brasil. Soe. Bras. Geol., Nfieleo Centro-Oeste, Bol. Inf. I0: 33--87, Goi~nia, 1981. THAYE•, T. P.: Some observations and problems concerning the Peridotites and PeridotiteGabbro Complexes of the Goi~s-Belt. Boh Esp. XXVI Congr. Bras. Geol. Bel6m-Pa, 1972. WEnNICK, E.: The Archaean of Brazil. Earth-Science Reviews, 17: 81--48, Amsterdam, 1981.

WERNICK,E., & ALMEIDA,F. F. M.: The geotectonic environments of Early Precambrian granulites in Brazil. Precambrian Res. 8: 1--17, Amsterdam, 1979. WINDLEY,B. F. : The evolving continents. Chichester, Wiley, 885 p., 1977. WlNOE, M., & DANNI, J. C. M.: Compartimentos Geotect6nicos Pr6-Brasilianos entre Caratae~ e Bendeng6, Uau~, Bahia. Anais XXXI Congr. Bras. Geol. 5: 2785--2795, Camborifl, 1980.

317