paleoproterozoic late-orogenic and anorogenic … · paleoproterozoic late-orogenic and anorogenic...
TRANSCRIPT
Precambrian Research 104 (2000) 47–75
Paleoproterozoic late-orogenic and anorogenic alkalinegranitic magmatism from northeast Brazil
J. Pla Cid a,*, M.F. Bitencourt a, L.V.S. Nardi a, H. Conceicao b, B. Bonin c,J.M. Lafon d
a Curso de Pos-Graduacao em in Geociencias, Campus de Agronomia, Instituto de Geociencias,Uni6ersity Federal do Rio Grande do Sul (UFRGS), A6. Bento Goncal6es, 9500, CEP-91509-900 Porto Alegre, RS, Brazil
b CPGG-PPPG/UFBA, Rua Caetano Moura, 123, Instituto de Geociencias, UFBA, CEP-40210-350, Sal6ador, BA, Brazilc Departement des Sciences de la Terre, Laboratoire de Petrographie et Volcanologie, Uni6ersite Paris, Sud. Centre d’Orsay,
Bat. 504, F-91504 Paris, Franced Centro de Geociencias, Laboratorio de Geologia Isotopica, Uni6ersidade Federal do Para (UFPA), Para, Brazil
Received 21 January 1999; accepted 4 May 2000
Abstract
During the Transamazonian cycle (2.090.2 Ga), two silica-saturated alkaline granite suites were intruded alongthe border of the Sao Francisco craton, northeastern Brazil. The Couro de Onca intrusive suite (COIS) is late tectonicrelative to major deformational events, and the Serra do Meio suite (SMS) postdates this event, being interpreted asanorogenic type, deformed during the Brasiliano cycle (0.65–0.52 Ga). The COIS encompasses alkali-feldspargranites, syenogranites, monzogranites, with dykes of similar composition and age (2.15795 Ma). They aremetaluminous rocks of alkaline affinity, whose evolution was probably controlled by mineral fractionation processesinvolving mainly plagioclase, amphibole, and Fe–Ti oxide. The SMS, whose probable age is 2.01 Ga, is composedof metaluminous and peralkaline granites, with associated quartz syenites, with Ti-aegirine transformed during theBrasiliano deformation to riebeckite+ titanite+aenigmatite. The temperatures of two stages of metamorphictransformation, overprinted during the Brasiliano event, are estimated at about 550 and 400–450°C. Both magmaticsuites are derived from enriched mantle sources. The metaluminous and peralkaline trends observed in the SMS,characterized by high concentrations of HFS and RE elements controlled mainly by F-activity, reflect the evolutionof different primary magmas produced by successive melting of the same mantle sources, in an anorogenic setting.© 2000 Elsevier Science B.V. All rights reserved.
Keywords: Alkaline granites; Geochemistry; Anorogenic; Late-orogenic
www.elsevier.com/locate/precamres
1. Introduction
The studied region is situated near the borderof the Bahia state, northeast Brazil (Fig. 1A).Two alkaline suites, which intrude Archean to
* Corresponding author.E-mail addresses: [email protected] (J. Pla Cid), lnardi@
if.ufrgs.br (L.V.S. Nardi), [email protected] (H. Con-ceicao), [email protected] (B. Bonin), [email protected](J.M. Lafon).
0301-9268/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.
PII: S 0301 -9268 (00 )00081 -4
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7548
Paleoproterozoic sequences, have been investi-gated for their geological, petrographical, geo-chemical and geochronological aspects. TheCouro de Onca intrusive suite (COIS) corre-sponds to the Couro de Onca granites, as pro-posed by Pla Cid et al. (1997a) to encompass fourgranitic bodies which intrude Archean TTGgneisses. The Serra do Meio suite (SMS) corre-sponds partly to the term previously employed by
Leite (1997) with the exception of the peralumi-nous granites, as pointed out by Pla Cid (1994).Thus, the term refers to the granitic magmatismwhich is part of the Campo Alegre de Lourdesalkaline province (Conceicao, 1990).
The SMS has been studied by Leite (1987),Conceicao (1990), Pla Cid (1994), Leite (1997),who presented consistent petrographic, electronmicroprobe, and geochemical data. In this paper,
Fig. 1. Bahia state in northeast Brazil, with location of studied areas-A. Main geological units of the Sao Francisco Craton, afterAlmeida et al. (1977), 1, Phanerozoic cover; 2, Neoproterozoic cover; 3, Neoproterozoic fold belts, and 4, Craton terrains. Alkalinegranites in rectangles. Brasiliano Riacho do Pontal and Rio Preto (RPFB); Aracuaı (AFB); Sergipano (SFB) and Brasılia fold belts(BFB)-(C). Couro de Onca intrusive suite (1) and Serra do Meio suite (2)-B. Simplified geological map of the COIS, after Figueroaand Silva Filho (1990) (modified). White circles indicate the visited outcrops-C.
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 49
additional petrographic data and new electronmicroprobe analyses are shown, as well as a re-interpretation of the geochemical data, with em-phasis on magmatic derivation and possiblesources. Isotopic determinations at Campo Ale-gre de Lourdes alkaline province point to em-placement related with the Transamazonianevent (2.090.2 Ga, Wernick, 1981), as evidencedby U–Pb data from zircon/baddeleyite in theAngico dos Dias Carbonatite (2.01 Ga, Silva etal., 1988).
Field relations, geochemical and isotopic dataof the COIS and other granitoids from this areawere studied by Figueroa and Silva Filho (1990).Geochronological data from high-K calc–alka-line granitoids indicate a Transamazonian mag-matism (2.0 Ga) related to Paleoproterozoictranscurrent deformation. According to these au-thors, the COIS is probably late to post-strike-slif. Pb–Pb in zircon isotopic determinations, aswell as field structures, petrographic and geo-chemical data from COIS, are presented and dis-cussed in this paper, together with some electronmicroprobe analyses of amphibole and plagio-clase.
2. Geological setting
The studied magmatic suites are intrusive intwo main geotectonic units in northeast Brazil,the Sao Francisco craton (SFC), and the Riachodo Pontal fold belt (RPFB) (Fig. 1B). TheRPFB, which represents the northwestern borderof the cratonic terrain, is one of the severalBrasiliano-age (0.65–0.52 Ga Van Schmus et al.,1995; Barbosa and Dominguez, 1996) fold beltssurrounding the SFC. According to Brito Neves(1975), major ENE–WSW trending structures inthe RPFB were developed during theNeoproterozoic. Jardim de Sa (1994) concludedthat the Brasiliano cycle in most fold belts fromnortheast Brazil is characterized essentially byintracontinental reworking, showing no evidenceof subduction-related settings. On the otherhand, the same author, based on structural pat-terns, geochemical and geochronological data,
suggested a Paleoproterozoic subduction settingin these terrains, which has promoted extensivemantle metasomatism. In fact, the geochemicalpatterns of Brasiliano-age ultrapotassic magma-tism (Ferreira et al., 1994; Pla Cid, 1994),confirm a subduction signature, attributed to theTransamazonian (Paleoproterozoic) cycle.
The Sao Francisco craton (SFC) has been astable crustal terrain since 0.52 Ga (Van Schmuset al., 1995), which corresponds to the last stagesof the Brasiliano cycle. This cratonic structure isroughly represented by an Archean nuclei cross-cut by NS-trending Paleoproterozoic belts.Within these older nuclei, TTG associations, aswell as charnockitic, enderbitic, and metasedi-mentary rocks of 2.6–3.3 Ga (Barbosa andDominguez, 1996) are found. The Paleoprotero-zoic belts are characterized by widespread occur-rence of syenitic and granitic rocks of alkaline,shoshonitic and ultrapotassic signatures (Con-ceicao, 1990). The syenitic rocks are late-Transa-mazonian, about 2.0 Ga in age (Conceicao, 1990,1994; Rosa, 1994).
According to Leite et al. (1993), Pla Cid(1994), Leite (1997), the Brasiliano event in thisarea is responsible for the collision of crustalblocks, which has resulted in extensive thrust tec-tonics along ENE–WSW striking surfaces (Fig.2). Transport direction along these planes is top-to-NW. NS-striking, km-wide transcurrent zoneslocated to the east of thrust structures have beeninterpreted by Leite (1997) as lateral ramps, ac-tive during the same event. Ductile shearingstructures found in the SMS granitoids are at-tributed to this event, and occur both alongmain thrust planes and lateral ramps.
A pre-Brasiliano, possibly Transamazonian,extensional setting is argued for this area (PlaCid, 1994; Leite, 1997), which is succeeded byintensive sedimentation events during Meso-proterozoic times (Leite, 1997). Such tectonicframework is compatible with the assumptionmade by Pla Cid (1994) that lithological contactsin the SMS, as observed now, do not representoriginal igneous geometry, but result from strongsuperposition of tectonic events during theBrasiliano cycle.
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7550
Fig
.2.
Geo
logi
cal
map
ofth
eC
ampo
Ale
gre
deL
ourd
esre
gion
,sh
owin
gth
eSM
S,af
ter
Lei
te(1
997)
(mod
ified
).
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 51
3. Local geology
The studied granitic suites are represented by,(i) SMS, part of the Campo Alegre de LourdesAlkaline Province (Leite, 1987; Conceicao, 1990;Pla Cid, 1994); and (ii) Couro-de-Onca Intrusivesuite (COIS- Figueroa and Silva Filho, 1990; PlaCid et al., 1997a,b). These suites are intrusive inthe cratonic rocks (COIS, Fig. 1B), next to thelimit with RPFB, and within the fold belt terrains(SMS, Fig. 1b).
3.1. Serra do Meio suite
This suite is intrusive in the RPFB terrains (Fig.2), where Leite (1997) identified six different units,(i)-the gneissic–migmatitic complex, of Archeanage, as indicated by a Rb–Sr determination of 2.6Ga (Dalton de Souza et al., 1979), composed oftwo mica-bearing ortogneisses of granitic compo-sition, migmatitic gneisses of trondhjemite–tonalite composition, and leucocratic gneisses ofgranitic composition. According to Dalton deSouza et al. (1979), this complex is metamor-phosed in the amphibolite facies; (ii)-the Serra daBoa Esperanca unit, encompassing metasedimen-tary rocks of greenschist facies, is composed ofquartz-mica schist and calcareous schist, with in-ferred Paleo to Mesoproterozoic age; (iii)-theSanto Onofre group, composed of strongly recrys-tallized metapsamitic and metapelitic rocks ofMesoproterozoic age; (iv)-the plutonic rocks,comprising tholeiitic to transitional basalts(Couto, 1989), a carbonatitic complex, and alka-line granites (SMS), in addition to mafic andultramafic plutons. The basaltic rocks are charac-terized by significant Fe–Ti-(V) deposits (Couto,1989). The plutonic rocks were deformed andmetamorphosed during the Brasiliano cycle,which confirms its pre-Brasiliano emplacementage (Couto, 1989; Pla Cid, 1994). Whole-rockRb–Sr determinations show that these graniteshave undergone strong isotopic system opening,with ages varying from 500 to 850 Ma, whichrepresent a mixing age between the end of Brasil-iano cycle and crystallization age (Pla Cid, 1994),(v)-the Parnaıba basin, formed by sedimentaryrocks with late-Ordovician to early-Carboniferousage, and (vi)-Cenozoic cover sediments.
The SMS constitutes several intrusions of ca400 km2 total area within a major ENE–WSWalignment (Fig. 2). They intrude the Paleoprotero-zoic Serra da Boa Esperanca unit (Silva et al.,1988; Leite and Froes, 1989; Leite et al., 1993), aswell as the Archean gneissic–migmatitic complex.cm To m-sized xenoliths of the former unit areabundant within these intrusions. Granites havegneissic, locally mylonitic structure, with amarked foliation, attributed to the Brasiliano col-lisional cycle (Jardim de Sa, 1994), trendingENE–WSW (Fig. 2) and dipping less than 25°towards SE or NW. In the northeastern portionof this granitic belt, NS-trending transcurrentzones are dominant, and imprint on the granites afoliation which dips at an angle usually higherthan 45°.
The alkaline granites are leucocratic, with fine-grained and inequigranular texture marked byalkali-feldspar phenocrysts, as well as riebeckiteand magnetite porphyroblasts. The gneissic struc-ture evolves locally into a mylonitic one (Pla Cid,1994), although in some outcrops the granitesshow pod-like portions of coarse-grained andisotropic rocks, surrounded by deformed por-tions. In spite of the Brasiliano-age deformationin this area, the alkaline granites sometimes pre-serve igneous structures, such as biotitic schlieren,found near the village of Peixe (Fig. 2), as well ascoarse-grained, isotropic textures, as previouslydescribed. Enclaves (autoliths) and dykes are notobserved.
Alkaline granites are not commonly deformed.However, deformational features are described inalkaline suites of similar tectonic environment,such as the Willyama supergroup, Olary Block(South Australia, Ashley et al., 1996) and Nogoyagranites (South Africa, Scogings, 1986).
3.2. Couro-de-Onca intrusi6e suite
These intrusions are located in the northwesternborder of the SFC (Fig. 1), near the town ofPetrolina (Fig. 1C). The following Archean unitshave been identified by Figueroa and Silva Filho(1990) in this portion of the SFC, (i) the Caraıbacomplex, formed by gneisses and migmatitic or-togneisses, both with trondhjemite–tonalite–gra-
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7552
nodiorite composition; (ii) volcano-sedimentarysequences, represented by the Rio Salitre andTanque Novo complexes, the former includingschists, paragneisses, quartzites, and bandediron-formations, and the latter containing mafic-ultramafic and felsic volcanic rocks. (iii) Archeangneisses to orthogneisses of TTG compositionwhich are intrusive in these metamorphic com-plexes, as suggested by the presence of xenolithsof the Rio Salitre and Tanque Novo rocks.
The magmatic rocks are represented by syn- tolate-tectonic granites, relaated to the Transama-zonian cycle, represented by, (a) Sobrado-typegranitoids of calc–alkaline affinity (Figueroa andSilva Filho, 1990), where Rb–Sr isotopic deter-minations indicate an age of 20049107 Ma, andSri ratio of 0.737; (b) biotite monzogranites ofalkaline affinity and age of 19289409 Ma, ob-tained by Rb–Sr isotopic method (Figueroa andSilva Filho, 1990) and, (c) COIS, with isotopicdata obtained by Pb–Pb in zircon suggestingcrystallization age of 215795 Ma (Pla Cid,1994). Seven crystals were analyzed, and the ob-tained age varies from 2142 Ma, consideringthree grains, to 2157 Ma, four grains. The dataconfirm the late-Transamazonian age of this in-trusive suite.
The COIS is composed of at least four plutonsof irregular to slightly circular shape, comprisinga total area of about 35 km2 (Fig. 1C). Thesegranites are intrusive in Archean trondhjemite–tonalite–granodiorite rocks (Figueroa and SilvaFilho, 1990), found as centimeter to meter-sizedxenoliths. The rocks are medium-grained, leuco-to mesocratic, medium-gray or pink in color, andhave a pronounced foliation, marked by thealignment of amphibole crystals. Banded (gneis-sic) structure is sometimes observed, parallel tothe main foliation. Massive structure is locallypresent. The dominant orientation trends 010–050°, dipping 45–70° NW, locally to SE. Cen-timeter-sized shear zones are parallel to thegranite foliation, and imprint a locally myloniticstructure on these granites, suggestive of theirsyn- to late-tectonic emplacement. Mafic mineralsoccur in two principal forms, (i) irregular to cir-cular clots containing amphibole and biotite of
medium grainsize, and (ii) smaller-sized, wide-spread amphibole crystals.
The banding is characterized by alternating,10–15 cm thick bands of mesocratic and leuco-cratic granitic compositions. Near to outcrops 12and 13 (Fig. 1C), the COIS foliation is folded,with axial planes parallel to the banded struc-ture. m-Sized xenoliths of gneissic basementrocks are also found here, which are partiallyassimilated by the host granites.
A common feature is the presence of synplu-tonic-dyke swarms, either with coarse-gained orpegmatitic texture, the latter ones showing dm-sized lenses of blue quartz. Both have centimetricto metric thickness, and display diffuse contactswith granitic host rock. Two preferential orienta-tions have been observed for these dykes, anearlier one, striking 080°, and dipping 60° to-wards NW, represented by coarse-grained dykes,and a later one, which crosscuts the former,trending 025° and dipping 30° NW, representedby coarse-grained and pegmatitic types.
Their composition varies from alkali-feldspargranites to monzogranites, both having amphi-bole as the dominant mafic phase. Pegmatiticdykes are of subordinate occurrence and containblue-quartz, alkali-feldspar and amphibole. Com-monly, one of the contacts between dyke andgranitic host rock is irregular and diffuse, whilethe other is sharp. The sharp, upper contact isalso marked by coarsening of minerals in thedyke, which shows progressive grainsize decreasetowards the lower, diffuse contact. The orienta-tion of mafic minerals within folded dykes issimilar to the one in the host rocks. These char-acteristics are suggestive of the synplutonic char-acter of these dykes. The occurrence of coarsergrainsize at the upper contact is probably relatedto an original volatile gradient established duringemplacement.
Local assimilation of dyke rocks by the gran-ites is also observed. Near these intrusions, anumber of fragments are found within the hostrock, which shows progressive textural modifica-tion due to partial assimilation processes. Thecoarser margins, containing amphibole-plagio-clase concentrations, remain virtually intact.
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 53
4. Petrographic features
4.1. Serra do Meio suite
These rocks are alkali-feldspar granites, withsubordinate quartz–alkali-feldspar syenites,showing strong mineral alignment and recrystal-lization features. In spite of their strong solid-state deformation, probably of Brasiliano age,original magmatic minerals, such as alkali-feldspar phenocrysts and some mafic phases, arepreserved. The SMS granites are formed by alkali-feldspar, quartz, albite, microcline, aegirine, ae-girine–augite, hedenbergite, riebeckite, magnetiteand biotite. Accessory minerals are titanite, allan-ite, aenigmatite, fluorite, apatite and zircon, andcalcite occurs as a secondary mineral phase.
Three different facies were identified. Metalu-minous types, with biotite as the only mafic min-eral, slightly peralkaline granites with aegirinecrystals, and strongly peralkaline ones, presentingriebeckite–aegirine–aenigmatite paragenesis. Themain petrographic evidence for preserved igneoustextures, as well as metamorphic assemblages, areshown in the Tables 1 and 2. Specific features ofeach type are discussed below.
4.1.1. Metaluminous granitesThese are alkali-feldspar granites with intersti-
tial, subhedral, brown to greenish-brown biotite.The foliation is marked by alignment of intersti-tial crystals, and also by elongate, millimetric,biotite concentrations. Colorless and purple fluor-ite grains are interstitial. In the northeastern plu-tons, these characteristics are dominant, but in thecentral portion of the belt (Fig. 2), biotite issometimes transformed into muscovite, associatedto magnetite blasts and interstitial calcite, suggest-ing the presence of Fe–CO2-rich fluids. Mi-croprobe data indicate that these are anniticbiotites (Conceicao, 1990; Pla Cid, 1994; Pla Cidet al., 1995).
4.1.2. Slightly peralkaline granitesThese are represented by alkali-feldspar gran-
ites, which predominate over quartz–alkali-feldspar syenites. The mafic mineralogy iscomposed of green and colorless pyroxene, biotite
and occasionally magnetite. The green to pale-green crystals are aegirine (Ac 72–94%), and ae-girine–augite (Ac 60–72%); colorless pyroxenewas optically identified as hedenbergite. Heden-bergite and aegirine–augite grains are interpretedas magmatic mafic minerals, since metamorphictemperature estimations, as discussed later, arebelow their stability field. Acicular, greenish crys-tals are included in the cores of alkali-feldsparphenocrysts, not observed in the newly-formed,clear borders, attesting their magmatic origin.Sometimes, as shown by Conceicao (1990), Leiteet al. (1991), these included crystals are aegirine–augite in composition. Aegirine constitutes eitherphenocryst, enclosing alkali-feldspar, albite, andquartz, or randomly distributed, interstitialcrystals.
4.1.3. Strongly peralkaline granitesThese are represented by alkali-feldspar gran-
ites, with local occurrence of the quartz–alkali-feldspar syenites, showing strong variation ofmafic mineral contents, which are riebeckite, ae-girine, aegirine–augite, colorless pyroxene, aenig-matite, biotite, and magnetite. The stronglyperalkaline granites exhibit preserved originalmagmatic textures, such as zoned, subhedral crys-tals of mafic minerals, and magmatic mafic miner-als included in alkali-feldspars. The amphibolewas identified as riebeckite (Conceicao, 1990; PlaCid, 1994), constituting dark-blue to brownish-blue poikilitic phenoblasts, and interstitial subhe-dral grains. Leite et al. (1991) argues for twodifferent generations of riebeckite. The interstitialcrystals, recrystallized during the tectonic defor-mation and surrounded by aegirine, are parallel tothe rock foliation, while the phenoblasts are notoriented, generated by recrystallization during lateor post-tectonic stages. Poikilitic riebeckite en-closes remnants of aegirine, as well as alkali-feldspar, quartz and albite. It is associatedintimately with reddish, fibrous aenigmatite andeuhedral titanite crystals on the borders. Riebeck-ite–aenigmatite–titanite association is assumed tobe the latest mafic minerals to recrystallize.
Subhedral, zoned pyroxene phenocrysts are alsopresent, which are crosscut by the foliation. Theyare interpreted as preserved magmatic pyroxene,
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7554
Tab
le1
Tex
tura
lan
dm
iner
alog
ical
feat
ures
ofth
eC
ouro
deO
nca
intr
usiv
esu
itea
Blu
eam
phib
ole
Qua
rtz
Bio
tite
Alk
ali-
feld
sper
Olig
ocla
seB
row
aam
phib
ole
Subh
edra
lto
Aci
cula
ran
dSu
bhed
ral
toan
hedr
alSu
bhed
ral
Subh
edra
lto
anhe
dral
Gen
eral
lyan
hedr
al;
Euh
edra
lism
pris
mat
ican
hedr
alsu
bhed
ral
whe
nin
clud
edin
feld
spar
Ann
ite
Ede
nite
AnB
13%
Has
ting
site
––
Com
posi
tion
Clo
tsan
ddi
sper
sein
Occ
urre
nce
Clo
tsC
lots
––
–th
egr
anit
esW
avy
exti
ncti
onP
oiqu
iliti
ccr
ysta
ls,
Alb
ite
and
Bro
wn
tobr
own
Per
thit
ic,
wit
hIn
clot
s,ac
icul
arC
hara
cter
isti
csco
mm
onin
the
folia
ted
wit
hzo
ning
tobl
ueal
bite
-Car
lsba
dgr
eeni
shco
lor,
exso
lved
albi
tein
and
euhe
dral
inte
rnal
bord
ers,
and
twin
ned;
bord
ers
are
amph
ibol
ein
clot
sco
mm
only
gran
ites
pris
mat
icgr
ains
loca
llyco
rrod
ed,
wit
hsu
bsti
tuti
ngal
bite
–per
iclin
etw
in,
amph
ibol
esep
idot
egr
owth
inth
esl
ight
wav
yex
tinc
tion
core
Ori
enta
tion
Not
nece
ssar
ilyF
requ
entl
y,P
oiqu
iliti
cgr
ains
are
Fre
quen
tly
elon
gate
dN
otor
ient
edN
otor
ient
edfla
me-
type
pert
hite
s,fr
eque
ntly
orie
nted
orie
nted
(anh
edra
lgr
ains
),w
ith
para
llel
toth
esu
bgra
ins
inth
ebo
rder
sfo
liati
onT
hecl
osts
are
Som
etim
espa
rtof
the
Inte
rnal
rim
sof
Irre
gula
rm
ilim
etri
cP
arti
alsu
bsti
tuti
onIn
clot
spr
esen
tsT
extu
ral
rela
tion
ssu
bgra
ins
and
the
clot
s,w
ith
tita
nite
,ex
solv
edal
bite
insu
rrou
nded
orby
biot
ite
atth
ene
edle
s,so
met
imes
oxid
ean
deu
hedr
alin
ters
titi
alto
alka
life
ldsp
ar,
inas
soci
ated
wit
hco
ntac
tsar
est
raig
ht,
bord
ers
and
alon
gco
ntac
tw
ith
pre
the
clea
vage
s;m
icro
gran
ular
alka
li-fe
ldsp
aran
dep
idot
e,w
ith
brow
nw
ith
poly
gona
lte
xtur
esu
bsol
vus
quar
tzco
mm
onas
relic
tsin
quar
tz;
spon
gyce
llula
rco
ncen
trat
ions
ofam
phib
ole
relic
ts;
blue
amph
ibol
e,as
soci
ated
tozi
rcon
,te
xtur
esu
gges
tpa
rtia
lbi
otit
ean
dbl
ueam
phib
ole
wit
hin
quar
tzan
dbi
otit
ere
sorp
tion
(Hib
bard
,al
lani
tean
dap
atit
e19
95)
clot
sC
onta
ctre
lati
ons
Wit
hfe
ldsp
ars
are
Wit
hol
igoc
lase
and
Fre
quen
tly
stra
ight
Wit
hal
kali-
feld
spar
Fre
quen
tly
stra
ight
Fre
quen
tly
stra
ight
quar
tzar
eir
regu
lar
and
quar
tzar
efr
eque
ntly
sinu
ous
and
sinu
ous
irre
gula
ran
dsi
nuou
s;m
utua
lem
baym
ent
Aci
cula
rbl
ueO
ligoc
lase
relic
tsan
dIn
the
elon
gate
dgr
ains
,A
mph
ibol
ere
licts
Not
obse
rved
Qua
rtz,
Incl
usio
nsqu
artz
roun
ded,
alka
li-fe
ldsp
ar,
alka
li-fe
ldsp
aran
dam
phib
ole
and
olig
ocla
se;
apat
ite
olig
ocla
sean
dbr
own
zirc
onan
dbr
own
euhe
dral
biot
ite
clos
eam
phib
ole
grai
ns;
amph
ibol
e;ap
atit
ein
toth
ecl
ots
and
zirc
ongr
ains
are
com
mon
incl
usio
nsfo
liate
dte
rms
and
apat
ite-
orie
nted
zirc
onin
mas
sive
ones
term
s;al
lani
te-i
sotr
opic
ones
aT
hem
iner
alco
mpo
siti
ons
wer
eob
tain
edby
elec
tron
mic
ropr
obe.
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 55
Tab
le2
Sum
mar
yof
the
mai
nig
neou
san
dm
etam
orph
icfe
atur
esob
serv
edin
the
Serr
ado
Mei
osu
itea
Hig
hte
mpe
ratu
rem
etam
orph
icfe
atur
esIg
neou
ste
xtur
eL
owte
mpe
ratu
rem
etam
orph
icfe
atur
es
Pat
ch-p
erth
ite;
som
etim
esw
ith
sub-
grai
nsP
erth
itic
(in
vein
s)ph
enoc
ryst
core
sA
lkal
i-fe
ldsp
arcl
ear
bord
ers
Alk
ali-
feld
spar
pres
erve
dsu
rrou
ndin
gph
enoc
ryst
sQ
uart
zP
olyg
onal
text
ure
(ann
ealin
g),
Qua
rtz
‘rib
bons
’so
met
imes
wit
hal
ibit
ean
dm
icro
clin
egr
anob
last
icte
xtur
eN
ewly
-for
med
grai
ns,
wit
hse
cond
ary
twin
Gra
nobl
asti
cte
xtur
eM
icro
clin
e+al
bite
Seco
ndge
nera
tion
ofri
ebec
kite
,M
afic
min
eral
sF
irst
gene
rati
onof
rieb
ecki
te,
inte
rsti
tial
Ti-
bear
ing
aegi
rine
zone
dph
enoc
ryst
s;an
dor
ient
edal
ong
folia
tion
;th
isty
peis
cons
titu
ted
bypo
rphy
robl
asti
cgr
ains
mic
ro-i
nclu
sion
sof
aegi
rine
and
ofra
ndom
orie
ntat
ion;
aegi
rine
–aug
ite
inal
kali-
feld
spar
;su
rrou
nded
byae
giri
ne,
whi
chca
noc
cur
tran
sfor
mat
ion
ofth
ese
crys
tals
toas
indi
vidu
aliz
edgr
ains
alon
gfo
liati
on;
subh
edra
lzo
ned
mafi
cgr
ains
,en
clos
edin
aeni
gmat
ite+
tita
nite
biot
ite
orie
ntat
ion
alka
li-fe
ldsp
ar,
wit
har
fved
soni
tic
core
sIn
terp
reta
tion
san
dte
mpe
ratu
reQ
uart
zri
bbon
s,se
cond
ary
twin
inA
nnea
ling
ofqu
artz
grai
nsis
anH
iper
solv
usco
ndit
ions
duri
nges
tim
atio
nfe
ldsp
ars
poin
tto
adu
ctile
char
acte
rof
alka
li-fe
ldsp
arcr
ysta
lliza
tion
;ev
iden
ceof
recr
ysta
lliza
tion
ina
this
defo
rmat
iona
lev
ent
(Hib
bard
,19
95);
stat
icen
viro
nmen
t(H
ibba
rd,
1995
);ae
giri
ne–a
ugit
ean
dar
fved
soni
tem
iner
als
indi
cati
ngcr
ysta
lliza
tion
tem
pera
ture
sof
esti
mat
ete
mpe
ratu
reto
this
stag
ear
ound
tem
pera
ture
esti
mat
ions
betw
een
400
and
450°
C55
0°C
atle
ast
700°
C
aM
ore
deta
ilsar
ede
scri
bed
inth
ete
xt.
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7556
Fig. 3. Pressure (kb) vs. temperature (°C) diagram, showingthe limit between greenschist and amphibolite metamorphicfacies, and solidus curve of wet granite (Winkler, 1979). Sub-solidus reaction of riebeckite+quartzlaegirine+mag-netite+Quartz (Ernst, 1962) is suggested as the upper limit ofmetamorphism in the SMS.
of metamorphism for the alkaline granites. Asshown in Table 2, riebeckite grains surrounded byaegirine are common syntectonic features. Thereaction observed in Fig. 3 indicates that theparagenesis present in the granites is compatiblewith a progressive metamorphic origin, in temper-ature slightly below 600°C. Temperatures below600°C are assumed to be the strongest register ofmetamorphism in the SMS. At such temperatures,Brasiliano-age metamorphism has affected thesubsolidus minerals, and igneous features are par-tially preserved.
4.2. Couro-de-Onca intrusi6e suite
These rocks comprise alkali-feldspar granites tosyenogranites, with subordinate monzogranites.The associated dykes have similar compositions.Quartz-monzodioritic rocks are only found asrelicts of partially digested dyke material, as seenin outcrops 01 and 12 (Fig. 3). Isotropic rocksgenerally correspond to more evolved petro-graphic types, as syenogranites or alkali-feldspargranites. The foliated rocks are richer in plagio-clase, ranging from syenogranites to monzogran-ites. The main petrographic features of thesegranites are summarized in Table 1.
The Couro-de-Onca granites are medium tocoarse, locally fine-grained, inequigranular rocks.Alkali-feldspar and quartz grains may reach up to3.0 mm, and occasionally, brown amphibole andoligoclase occur up to 2.0 mm. The groundmass isconstituted by medium-grained alkali-feldspar,oligoclase, quartz, albite, brown and blue amphi-bole, biotite, opaque minerals and titanite. Zir-con, apatite, allanite, and epidote are accessoryphases. Alotriomorphic to hypidiomorphic tex-ture is characteristic, and sinuous contacts arecommon, mainly between quartz and feldspars,with embayment of quartz grains. The orientationis marked by poikilitic, brown amphibole andmesoperthitic alkali-feldspar, with exsolved lamel-lae being parallel to the foliation. Occasionally,amphibole is concentrated in millimeter-thick, ir-regular levels. The dykes are coarse grained topegmatitic, and alkali feldspar–quartz–oligoclaseexhibit glomeroporphyritic texture.
and microprobe data indicate Ti-aegirine and Ti-aegirine–augite composition for the dark-bluecore zones, with titanium decreasing towards thedark-green borders. Thus, original, magmatic Ti-pyroxene, which has reacted during the Brasil-iano-age deformation would result in alate-tectonic secondary mineral assemblage, con-stituted by riebeckite+ titanite+aenigmatite.
Annitic biotite is subhedral, reddish-brown incolour, and occurs usually at the borders of alka-line minerals (Conceicao, 1990; Pla Cid, 1994).Occasionally, dark amphibole included in alkali-feldspar and surrounded by aegirine–augite, pre-sents riebeckite–arfvedsonite mixed composition,which suggests a magmatic origin. Euhedral mag-netite blasts can be observed along the foliation,suggesting high fO2 conditions during the defor-mational event, due to Fe-rich fluids, as previ-ously described. Subhedral and anhedral,interstitial fluorite is a common accessory phase,associated with the sodic mafic minerals.
Fig. 3 shows the boundary between green-schist–amphibolite metamorphic facies condi-tions, the wet granite solidus, and the stabilityfield for riebeckite, which reacts with quartz+fluid and is transformed into aegirine+mag-netite+quartz+fluid. This reaction, occurringunder amphibolite conditions, is the upper limit
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 57
Mafic clots, with ca. 2.0-mm diameter, arewidespread. They are composed of quartz in thecentral region, surrounded by acicular, prismatic,blue amphibole, and euhedral biotite. Remnantsof brown amphibole are found within blue-amphi-bole grains. Blue amphibole, biotite, and felsicminerals frequently show microgranular texture inthe outer portions of these clots. Euhedral allan-ite, subhedral zircon and apatite, as well as euhe-dral oxide minerals, are accessory phases. Biotiteis sometimes intergrown with mesoperthitic alkali-feldspar.
Textural relations, such as plagioclase partialresorption, blue-amphibole acicular inclusions inoligoclase, microgranular agglomerations, and re-verse zoning in amphibole are evidence that asimple magmatic history is not observed for thesegranites. In order to understand these features, itis necessary to remember that the magmatic sys-tem has been disturbed by synchronic intrusion ofdykes having similar geochemical features. Thus,even if these dykes have not caused modificationof geochemical signature, it is probable that thedecompression undergone by the crystallizinggranitic magma at the time the dykes were em-placed, as well as the fluid circulation promotedby coarse to pegmatitic dykes intrusion, haveinduced some changes in Pfluids and total P of thegranite host. Microgranular textures, in associa-tion with large grainsize of the clots, would attestto a heterogeneous distribution of Pfluids. Somecompositional change in amphibole may also berelated to a new F−, H2O or Cl− contents foundsin magma.
5. Geochemistry
Whole-rock chemical analyses were obtained atthe laboratories of Centro de Estudos em Petrolo-gia e Geoquımica, Universidade Federal do RioGrande do Sul, Brazil, Activation LaboratoriesLtd., Ont., Canada, and Centre de RecherchesPetrographiques et Geochimiques (CRPG),Nancy, France. For the Couro-de-Onca granites,the analytical data of Figueroa and Silva Filho(1990) were used. Analytical results are presentedin Tables 3 and 4.
Both granitic suites are characterized by highalkali contents, confirming their alkaline signa-ture. The COIS exhibits composition rangingfrom quartz-monzonite to granite, whereas domi-nant granitic composition is present in the SMS,with subordinate syenite and quartz-monzonite(Fig. 4). The COIS has dominantly metaluminouscharacter, with slightly peralkaline and peralumi-nous compositions less common (Fig. 5). TheSerra do Meio granites are essentially metalumi-nous and peralkaline, with riebeckite-bearing
Fig. 4. Total alkalis vs. silica (TAS) diagram, after Le Maitreet al. (1989), with chemical classification and nomenclature ofplutonic rocks according to Middlemost (1994). Magmaticseries after Le Maitre et al. (1989).
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7558
Tab
le3
Who
le-r
ock
anal
yes
ofth
eSe
rra
doM
eio
suit
e
CL
-090
PP
B-6
4G
A-6
8C
L-8
8P
PB
-28a
JP-4
9C
L-8
7G
A-1
3JP
-50
CL
-74
Sam
ple
PP
B-1
4G
A-1
4P
PB
-63
PP
B-1
4aG
A-1
4bC
L-7
7m
etsl
ight
met
slig
htm
etsl
ight
met
slig
htsl
ight
met
met
met
met
met
met
slig
htF
acie
sgr
anit
epe
ralk
pera
lkgr
anit
egr
anit
egr
anit
epe
ralk
gran
ite
gran
ite
gran
ite
gran
ite
pera
lkgr
anit
egr
anit
epe
ralk
pera
lk
SiO
270
.772
69.3
72.9
75.1
74.1
70.2
69.5
73.9
61.5
73.2
71.7
73.2
74.5
67.3
74.5
0.43
0.22
0.53
0.1
0.38
0.58
0.51
0.16
0.22
0.4
TiO
20.
230.
230.
440.
150.
370.
2112
Al 2
O3
13.5
12.4
1212
.110
.812
.213
.212
.414
.413
.413
.211
12.1
13.4
11.8
4.2
3.1
3.2
0.58
3.8
1.9
2.2
0.92
1.4
7.5
0.64
1.5
0.37
4.3
Fe 2
O3
2.3
3.7
2.2
1.8
1.5
1.68
1.7
1.4
1.8
3.5
3.2
1.7
1.7
1.5
3.2
1.5
1.8
1.8
FeO
0.13
0.1
0.17
0.1
0.2
0.32
0.12
0.1
MgO
0.32
0.1
0.19
0.1
0.12
0.1
0.14
0.21
0.77
0.92
0.07
0.67
0.38
1.2
1.1
0.72
0.57
1.6
0.65
0.42
0.76
CaO
1.3
0.79
0.45
3.4
4.5
4.4
4.8
1.8
3.7
2.6
44.
73.
66
3.7
4.2
3.9
3.8
5.1
Na 2
O5
56.
74.
94.
94.
44.
15.
64.
25
K2O
55.
14.
65
5.1
4.9
0.18
0.09
0.1
0.05
0.05
0.22
0.19
0.04
0.07
0.43
0.16
0.08
0.13
MnO
0.14
0.21
0.06
0.05
0.05
0.05
0.05
0.05
0.06
0.05
0.05
P2O
50.
050.
050.
050.
050.
050.
050.
050.
050.
110.
390.
50.
420.
540.
250.
280.
350.
330.
39H
2Op
0.22
0.17
0.11
0.35
0.23
0.28
0.19
0.28
0.24
0.63
0.11
0.41
0.08
0.77
0.73
0.18
1.1
0.19
0.5
0.43
0.48
0.64
CO
2
99.7
199
.78
99.8
399
.58
99.6
899
.699
.88
99.7
299
.84
99.5
999
.81
99.6
899
.61
Tot
al99
.84
99.8
299
.76
2300
2100
620
300
860
3900
2300
1800
640
3100
4100
2400
1800
930
1800
2300
F20
2023
2020
2020
2020
2020
2044
2020
20C
l0.
921.
050.
990.
850.
940.
890.
930.
921.
040.
971.
041
0.86
0.93
0.96
Agp
aiti
c1.
04in
dex
3.87
3.46
5.28
4.47
4.58
1.92
5.22
5.21
5.18
2.53
8.45
2.85
3.53
4.83
2.38
FeO
t5.
6726
019
061
022
030
099
013
10B
a14
077
190
220
330
240
160
710
180
130
210
110
140
170
120
7512
013
023
015
016
012
0N
b19
012
020
05
55
85
55
85
55
55
55
5C
s12
076
200
280
160
120
8121
012
016
013
0R
b15
013
019
015
016
023
3516
822
826
1212
25H
f15
1521
1034
2830
7433
4225
120
120
43Sr
2617
3540
3635
140
3992
180
9820
018
079
9517
098
110
Y13
017
011
017
016
017
066
011
6091
013
0069
026
010
5040
012
0046
096
079
060
014
4043
081
0Z
r38
4129
4436
3336
4133
3853
3840
Ga
4146
338
88
811
88
148
88
88
85
8V
2019
818
877
195
1226
2228
1227
2722
Th
1010
1010
1010
1010
1010
1010
1010
U10
109
67
127
1214
75
59
105
1110
13T
a15
1111
117
1115
15C
u15
1115
1519
2215
1123
3445
3428
4534
3945
5123
4534
Co
5128
5112
5924
1824
2924
4118
2418
1829
1859
29N
i39
5252
3952
7726
2652
26C
r77
116
5211
652
3911
1.6
50.1
316
6.7
70.8
714
5.5
100.
122
4.6
155.
111
4.4
132.
5L
a20
8.2
213.
811
4.8
373.
616
1.7
303.
2C
e47
4.7
244.
430
5.3
436.
633
5.2
215.
110
5.9
5617
0.2
76.7
813
0.2
93.8
218
9.2
Nd
99.3
912
9.6
144.
417
9.1
17.1
418
.84
13.2
630
.68
14.6
324
.39
30.7
918
.05
23.6
529
.26
24.3
2Sm
2.57
90.
563.
962.
340.
951.
313.
271.
341.
47E
u2.
613.
2213
.82
13.8
213
.04
25.9
810
.46
19.7
220
.16
14.3
318
.48
20.3
918
.26
Gd
Dy
10.7
216
.222
.34
7.86
18.2
115
.74
13.1
517
.316
.89
16.2
312
.95
1.98
53.
364.
131.
853.
512.
512.
972.
493.
343.
18H
o3.
186.
344.
551
9.68
9.45
3.36
8.79
7.11
6.03
8.36
7.6
8.2
Er
3.21
78.
126.
422.
347.
125.
224.
42Y
b6.
415.
157.
044.
780.
404
0.92
0.82
0.32
0.87
0.62
0.65
0.87
0.53
0.77
Lu
0.62
�Rar
e71
5.41
468.
4948
7.42
286.
0781
4.28
352.
5166
2.46
974.
4151
8.53
647.
1891
0.21
eart
hel
emen
ts
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 59
Tab
le3
(Con
tinu
ed)
CL
-54
CL
-55
PP
B-7
8aG
A-1
3G
A-4
6C
L-1
6T
R-7
2G
A-4
6aG
A-1
4cC
L-1
05C
L-0
5bSa
mpl
eG
A-3
4G
A-3
9N
Em
etN
Em
etst
rong
NE
met
stro
ngN
Em
etst
rong
stro
ngst
rong
stro
ngN
Em
etst
rong
faci
esst
rong
pera
lkpe
ralk
pera
lkpe
ralk
pera
lkpe
ralk
pera
lkpe
ralk
75.1
74.1
68.3
73.3
74.2
66.9
71.4
72.3
72.4
73.3
7475
.7Si
O2
73.3
0.59
0.54
0.39
0.62
0.43
0.44
0.39
0.36
0.38
0.28
TiO
20.
420.
340.
2710
.811
.611
.510
.411
.314
.412
11.3
11.7
11.8
12.3
9.4
11.6
Al 2
O3
13.
84
4.2
2.7
2.9
3.7
34.
23.
20.
63
Fe 2
O3
1.3
3.6
1.4
1.4
2.5
1.8
1.9
1.8
1.1
1.8
2.2
2.5
FeO
2.1
2.5
0.2
0.1
0.22
0.1
0.1
0.12
0.1
0.1
0.2
0.1
0.1
0.16
0.23
MgO
1.2
0.34
0.4
0.23
0.64
0.58
0.36
0.31
CaO
0.89
0.7
0.46
0.89
0.38
5.2
4.6
4.3
5.8
4.8
4.6
5.3
4.1
2.6
3.7
Na 2
O3.
53.
14
4.9
4.1
4.3
4.3
4.5
5.7
4.3
4.8
2.7
54.
24
4.6
K2O
0.3
0.18
0.21
0.13
0.18
0.22
0.15
MnO
0.13
0.13
0.1
0.17
0.11
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
P2O
50.
050.
050.
050.
250.
290.
380.
420.
260.
190.
180.
55H
2Op
0.3
0.28
0.54
0.3
0.54
0.27
0.19
0.05
0.08
0.19
0.45
0.36
0.05
0.08
CO
20.
930.
370.
660.
5399
.68
99.8
899
.78
99.8
999
.98
99.8
599
.85
99.9
399
.79
100.
0599
.76
99.4
999
.73
Tot
al93
083
082
025
013
0081
054
025
023
0017
00F
1200
1100
1300
2020
2020
2020
2020
2020
2020
20C
l0.
930.
910.
891.
151.
171.
061.
091.
051.
130.
991.
030.
91
Agp
aiti
cin
dex
7.2
5.18
3.83
5.11
5.13
4.6
5.58
3.67
3.98
3.04
4.9
5.22
FeO
t3
840
300
740
600
440
670
510
340
690
300
1080
200
Ba
390
5052
5388
7750
110
Nb
5681
9535
013
017
05
76
185
55
55
6C
s5
55
160
110
7575
6877
8872
6484
110
160
82R
bH
f10
158
913
128
178
2168
2722
5518
3314
4019
5913
25Sr
7329
8768
6391
4393
8044
120
37Y
120
210
270
9518
039
037
036
059
060
034
083
040
010
50Z
r10
6026
9093
072
036
3241
3742
5237
4043
4134
2935
Ga
88
88
88
88
88
V8
88
1916
55
58
65
95
1445
18T
h10
1010
1010
1010
1010
1010
1010
U6
65
55
55
514
265
Ta
75
711
1111
711
117
1115
1115
11C
u23
3423
3428
2839
28C
o28
2828
2828
2418
129
4147
1818
2424
29N
i41
1818
5252
5239
258
9065
5239
3926
7710
3C
rL
a79
.88
76.2
995
.46
119.
316
6.7
182.
917
0.9
208.
737
3.6
Ce
226.
917
0.2
87.1
470
.188
.14
116
Nd
15.5
814
.54
16.8
2Sm
25.1
130
.68
2.43
2.25
3.01
3.96
Eu
3.16
25.9
810
.86
10.8
113
.05
22.6
7G
d7.
49.
1611
.14
Dy
21.0
522
.34
1.33
1.72
2.07
4.13
Ho
3.96
9.45
2.83
4.08
4.77
9.41
Er
1.85
3.3
3.09
Yb
7.02
6.42
0.23
0.44
0.39
0.82
Lu
0.89
�Rar
e81
4.28
392.
4336
3.59
446.
6455
5.47
eart
hel
emen
ts
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7560
Tab
le4
Who
le-r
ock
anal
yses
ofth
eC
ouro
deO
nca
intr
usiv
esu
ite
01D
5701
F01
C1
dyke
01B
04A
01C
dyke
Sam
ple
04B
dyke
1205
/II
704
A/I
I13
04/I
I56
02/I
Iis
otro
pic
orie
nted
isot
ropi
cis
otro
pic
isot
ropi
cis
otro
pic
isot
ropi
cor
ient
edor
ient
edis
otro
pic
isot
ropi
cfa
cies
isot
ropi
cdi
orit
icdy
ke
SiO
276
.479
.63
71.8
970
.43
74.7
774
.55
77.8
374
.22
71.8
476
.25
76.7
367
.877
.04
73.5
670
.84
72.4
711
.83
14.5
213
.82
11.6
812
.43
11.3
512
.46
12.2
615
.211
.51
11.2
811
.37
14.8
9A
l 2O
313
15.7
711
.89
0.14
0.53
0.84
0.5
1.15
22.
592.
191.
313.
992.
092.
254.
30.
30.
150.
83F
e 2O
3
1.27
1.93
3.57
00
01.
320
0F
eO0
0.91
01.
351.
143.
260.
840.
351.
261.
580.
680.
880.
480.
591.
192.
10.
65C
aO0.
431.
231.
331.
120.
20.
20.
040.
140.
280.
060.
130.
020.
670.
16M
gO0.
140.
040.
090.
170.
090.
090.
450.
250.
030.
040.
070.
020.
030.
030.
010.
060.
020
MnO
00.
170.
060.
010.
020.
020.
070.
180.
230.
180.
590.
210.
290.
210.
380.
390.
210.
200.
460.
180.
130.
46T
iO2
2.93
2.75
2.51
2.56
3.38
3.05
3.16
3.36
7.2
3.24
1.52
2.98
4.3
Na 2
O5.
074.
312.
771.
374.
774.
965.
334.
456.
185.
45.
084.
965.
064.
985.
395.
664.
82.
884.
4K
2O0.
02P
2O5
0.1
0.01
0.11
0.03
0.06
0.03
0.08
0.13
0.05
0.09
0.11
0.01
0.04
0.1
0.07
00.
150.
130
00
0.15
00
00.
040
0H
2Om
00.
020.
20.
150.
470.
410.
450.
460.
440.
70.
42L
OI
0.56
0.09
0.3
0.5
0.48
0.43
0.37
0.48
99.0
599
.26
99.1
98.6
410
0.2
100.
7110
0.01
98.8
610
0.21
99.6
8T
otal
99.8
399
.47
100.
2199
.83
99.1
599
.18
0.97
1.39
2.03
2.38
4.6
1.8
2.33
1.97
2.5
3.59
1.88
2.02
3.87
1.62
1.27
4.01
FeO
T0.
880.
860.
710.
650.
930.
920.
930.
851.
010.
90.
680.
930.
940.
89A
gpai
tic
0.65
0.82
inde
x89
*87
**
1315
14*
3.94
3.86
2.88
2.63
**
Cr
*17
1920
1718
160
018
1.39
2.09
2.72
1.84
1719
Ni
16*
**
**
**
3.59
*0.
96*
0.57
4.85
*V
**
7829
936
316
249
734
336
443
339
044
246
746
656
428
653
365
Zr
9511
384
127
136
138
7479
.7C
o15
493
136
80.9
8418
256
94*
5750
048
040
6.7
907
*3.
38.
3C
u44
5657
2019
2423
2420
2425
2226
.226
.225
.832
.726
2625
Ga
4525
338.
519
1916
35.3
530
.83
24.7
730
.31
Nb
1338
1240
617
1922
2144
3318
25.9
1727
Pb
20.1
24.8
2819
1820
8778
107
4983
117
2310
9Z
n10
245
103
110
9554
112
3312
111
913
012
917
613
312
813
7.9
2615
1.3
Rb
152.
915
7.6
142
7318
515
017
056
2167
115
52.4
48.8
14.7
8976
.929
.415
.985
.712
8989
Sr24
848
676
044
253
721
465
162
426
727
741
515
068
5B
a72
126
915
5*
430
190
**
1600
640
1000
F16
.77
**
20.4
18.7
20.4
25.3
2T
h21
.39
22.6
917
.79
15.0
417
.39
7.77
13.5
2*
0.92
**
1.57
2.22
1.05
1.65
*1.
82U
0.97
1.24
1.71
0.36
1.77
1.13
56*
*27
5341
71.3
Y53
3551
5562
2474
**
**
9.9
1112
12.1
*13
.8H
f13
14.7
**
**
**
**
*1.
052.
571.
762.
672.
441.
871.
99*
**
Ta
115.
9*
*95
.973
.760
.290
.5*
124.
811
6.1
117.
593
.14
La
56.7
35.4
72.8
*26
6.3
266.
3*
*18
613
823
918
0.8
242.
323
0.3
187.
117
2.6
81.1
137.
6C
e25
.4P
r*
24.7
7*
19.8
915
.83
13.7
20.9
227
.49
26.3
421
.216
.75
7.62
13.6
5*
111.
7*
*75
6252
.776
.84
*98
.56
63.9
95.9
575
.73
Nd
105.
176
33.8
18.6
**
12.6
12.1
11.2
14.4
9Sm
17.0
317
.517
.75
13.7
614
.86.
412
.6*
1.02
**
1.39
11.
397
0.87
31.
9*
1.17
Eu
0.89
2.25
21.
070.
671
*15
.316
.7*
*8.
9810
.49.
4112
.74
13.2
514
.311
.72
13.7
5.8
12.8
Gd
2.2
**
1.27
1.75
1.65
2.06
*2
1.9
2.06
1.89
Tb
20.
92.
1*
7.8
11.2
**
6.31
9.95
8.81
11.6
910
.610
.88
10.0
511
4.5
11.8
Dy
2H
o*
1.3
*1.
041.
891.
622.
62.
222.
082.
112.
10.
82.
4*
5.7
**
2.83
5.56
4.61
6.62
*5.
082.
34.
825.
527.
5E
r3.
26
*0.
350.
69*
*0.
359
0.80
80.
613
10.
730.
620.
840.
730.
290.
97T
m4.
2*
*1.
694.
83.
665.
894.
673.
77Y
b4.
641.
94.
21.
77
*0.
64*
*0.
243
0.66
10.
509
0.92
*0.
681.
170.
620.
69L
u0.
310.
630.
2618
1.94
�Rar
e33
2.09
562.
83*
582.
25*
*39
0.94
318.
5739
0.97
402.
3951
8.14
496.
7840
4.6
394.
08ea
rth
elem
ent
s
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 61
Fig. 5. Al(Na+K) vs. Al/(Ca+Na+K) diagram (Maniarand Piccoli, 1989) to discriminate peralkaline, metaluminous,and peraluminous rocks.
alkaline suites is due to their high FeOt con-tents, associated with extremely low MgO con-centrations in these magmas (Sorensen, 1974;Whalen et al., 1987), as evidenced by the ab-sence of Mg-rich minerals, since biotites are nor-mally Fe-rich.
5.1. Major elements
The extremely high SiO2 contents represent anobstacle to determinate the geochemical affinityof granitic rocks, as observed by Mahood andHildreth (1983). On the other hand, major ele-ment contents are diagnostic for alkaline gran-ites (Whalen et al., 1987; Nardi, 1991).
Fig. 6. Agpaitic index vs. FeOt/(FeOt+MgO) ratio diagram,according to Nardi (1991), showing the usual field for alkalinegranites.
rocks belonging to the strongly peralkalinegroup, as shown by agpaitic index values inTable 3. Alkaline saturated to oversaturatedcomplexes display associated metaluminous andperalkaline types (Bonin, 1988). In this context,the SMS represents a typical example of thisassociation.
The agpaitic index and FeO/(FeO+MgO) ra-tio are very useful indicators of alkaline signa-ture in silica-oversaturated rocks (Fig. 6). In thiscontext, both suites plot in the field of alkalinegranites. Two samples from the COIS show sub-alkaline character, 01F (quartz monzodioriticdyke) and 57 (biotite syenogranite; Fig. 6).Macroscopic and petrographic observations indi-cate that sample 01F represents contaminatedliquid, whilst the biotite syenogranitic sample isstrongly transformed by hydrothermal processes,evident as plagioclase and biotite transformationto epidote, and occasionally into white mica.Then, the chemical composition of these samplesprobably does not represent the liquid composi-tion during crystallization. Other six samplesplot below the field of alkaline granites, al-though they were considered to have alkalineaffinity. As evidenced by Nardi (1991), this fieldincludes more than 80% of the alkaline granites,although it is usual to find some granites whichplot below. The high FeO/(FeO+MgO) ratio in
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7562
Fig. 7. Harker type diagrams (in wt.%) for SMS, showing in some cases the parallel evolutionary lines between peralkaline andmetaluminous trends.
Saturated to oversaturated rocks of this magmaticseries have very low contents of MgO and CaO,medium concentrations of Al2O3, and high con-tents of Na2O, K2O, FeOt, and FeOt/(FeOt+MgO). The COIS present generally higher Al2O3
(11.28–15.77 wt.%) and CaO (0.20–1.58 wt.%)contents than the SMS (9.4–13.5 and 0.3–1.2wt.%, respectively), for samples with similar SiO2
contents. Bonin (1987) showed that calcium, ingranites, progressively decreases from late to
anorogenic phases of a tectonic event. In fact, aspreviously discussed, COIS granites have a late-Transamazonian age, while the SMS, is probablya typical alkaline anorogenic suite.
Major elements show similar behaviour to thatdescribed by Mahood and Hildreth (1983), Millerand Mittlefehldt (1984) for high-SiO2 granitic liq-uids, with most decreasing during magmatic evo-lution (Figs. 7 and 8). In Harker-type diagrams(Fig. 7), the SMS evolves according to two parallel
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 63
paths in terms of Al2O3, FeOt, CaO, and Na2O,for similar SiO2 contents, (i) metaluminous andslightly peralkaline granites, forming the met-aluminous trend, and (ii) strongly peralkalinegranites, forming the peralkaline trend. Their evo-lution, considering similar range of SiO2 contents,is indicative of the presence of two different alka-line parental liquids, and/or probably reflectingdifferences in source composition. The peralkalinetrend has higher contents of FeOt, Na2O, andTiO2, with lower CaO and Al2O3, for similar SiO2
values, relative to the metaluminous trend (Fig. 8).The parallelism between the two trends is indica-tive of similar magmatic evolution during crystal-lization of metaluminous and peralkaline rocks,
although the Al2O3 versus SiO2 plot suggests moreintense feldspar fractionation in the metaluminousliquids to produce the more differentiated granites.
In the COIS, Al2O3 behaviour reflects a con-stant decrease of plagioclase modal proportionswith increasing differentiation (Fig. 8). The dykeshave higher Al2O3 contents and lower concentra-tions of FeOt, MgO and TiO2 than the granites forsimilar SiO2 values, which is in agreement withtheir more leucocratic character. Most elementsshow linear or exponential decrease during evolu-tion from foliated to isotropic granites (Fig. 8),which indicates fractionation of plagioclase (Ca,Na, and Al decrease) and amphibole (Ca, Fe, andMg).
Fig. 8. Harker type diagrams of the COIS.
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7564
5.2. Trace elements
Alkaline granites present high HFSE concentra-tions, such as Zr, Nb, Y, Ga, and REE (Collins etal., 1982; Whalen et al., 1987; Bonin, 1988), asso-ciated with depletion of Sr, Ba, Cu, V, Cr and Ni,due to their strongly evolved character. As studiedby Watson (1979), in metaluminous liquids, zir-con saturation is attained at less than 100 ppm ofZr, whereas Zr solubility increases in peralkalineliquids (Watson, 1979; Watson and Harrison,1983). These results have pointed out that Zrsolubility is a function of agpaitic index. Accord-ing to the same authors, high temperature of themelt can increase Zr solubility.
5.2.1. Serra do Meio suiteTrace elements at these granites were previously
studied by Leite (1993), Pla Cid (1994), Pla Cid etal. (1997a). These rocks are characterized by highcontents of Zr (340–2690 ppm), Ga (29–53 ppm),Y (37–270 ppm), Nb (50–350 ppm), Hf (8–68ppm), F (250–4100 ppm), and Ce (114–474 ppm);medium concentrations of Ba (77–1310 ppm) andRb (64–280 ppm), and depletion of Cr (26–258ppm), Ni (12–129 ppm), and Sr (13–140 ppm).Incompatible elements display contents generallyabove the average concentrations of alkalinegranites, when compared with typical suites stud-ied by Bonin (1980), Bowden and Kinnaird(1984), Whalen et al. (1987).
Zirconium exhibits unusual behaviour, withhigher contents in the metaluminous trend gran-ites (260–2690 ppm) than in the peralkaline one(340–830 ppm; Pla Cid et al., 1997a,b). Zr hasbeen used by Bowden and Turner (1974), Bonin(1988) as a peralkalinity index indicator, a proce-dure also followed in this paper (Fig. 9). Thisunusual behaviour is followed by strong enrich-ment in HFSE and rare earth elements (REE) inthe metaluminous trend relative to the peralkalineone (Fig. 9). This behaviour is completely oppo-site to that expected for typical alkaline graniticsuites, where incompatible elements are enrichedin the peralkaline terms. This unusual behaviourmay be explained by two different possibilities, (i)metaluminous and peralkaline granites are pro-duced from different sources, or (ii) both granite
types were produced from liquids generated bysuccessive extraction from the same source.
The behaviour of Ga is related, according toWhalen et al. (1987), to the agpaitic index ofgranites, with higher contents present in peralka-line types. In the Serra do Meio granites, Ga isabove the average values for peralkaline granites,such as the peralkaline granites from Corsica(Bonin, 1980, 1982, 1988), alkaline granites fromNigeria (Bowden and Kinnaird, 1984), and from aSiO2-saturated peralkaline volcanic–plutonic as-sociation from New Mexico (Czamanske and Dil-let, 1988; Fig. 9). In Fig. 9, the SMS hascomparable composition to these typical alkalinesuites, although it presents higher amounts ofincompatible elements. This confirms that theSMS granites have a chemical composition typicalof alkaline magmatic suites, in spite of deforma-tion/recrystallization processes to which they havebeen submitted during the Brasiliano cycle.
Rb/Sr versus Sr, Rb and Ba plots (Fig. 10)permit separation of the metaluminous and peral-kaline trends. The parallelism between thesetrends suggests that the magmatic evolution isvery similar. Since these granites do not containplagioclase crystals, it is probable that Sr decreasereflect alkali-feldspar fractionation. These plotsmay be interpreted as suggestive of a metalumi-nous trend generated from a parental liquid richin Rb and Sr. The strong Ba decrease in themetaluminous trend, less intense than in the peral-kaline one, indicates a stronger alkali-feldsparfractionation in this evolutionary line.
5.2.2. Couro-de-Onca intrusi6e suiteThese granites, excluding samples 01F and 57,
show trace-element concentrations which are con-sistent with the alkaline series. They are character-ized by medium to high values of Rb/Sr (1–15),Rb (73–218 ppm), Ga (19–32 ppm), Nb (12–45ppm), Zr (53–688 ppm), Y (24–84 ppm), F (190–1600 ppm), and Ce (81–266 ppm), and low con-tents of Sr (12–89 ppm), Ba (155–760 ppm), Cr(B89 ppm), and Ni (B20 ppm).
LIL-element variation confirms the alkali-feldspar and plagioclase importance during theCOIS magmatic evolution (Fig. 8). This evolutionis controlled by plagioclase fractionation, princi-
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 65
Fig. 9. Zr (in ppm) vs. Ga, Y, Nb, Th, and REE (in ppm) plots showing the different contents between peralkaline andmetaluminous trends in the SMS. Corsican magmatic province analysis are from Bonin (1980), in the Liruei complex from Bowdenand Kinnaird (1984) and to the New Mexico volcanic–plutonic association from Czamanske and Dillet (1988).
pally around 73 wt.% of SiO2, as demonstratedby Rb/Sr and Sr evolution with differentiation(Figs. 8 and 11). Sr shows strong compatiblebehaviour, which is in agreement with plagio-clase fractionation during the magmatic evolu-tion of this suite. Rb/Sr versus MgO, FeOt, andTiO2 diagrams (Fig. 11) suggest amphibole frac-tionation, as well as Fe–Ti oxide, during mag-matic evolution.
5.3. Rare earth elements
Rare earth elements (REE) were normalisedto the chondritic values of Evensen et al. (1978),and patterns displayed in Fig. 12. Serra doMeio and Couro-de-Onca intrusive suites havevery similar patterns, with heavy REE fractiona-tion relative to the light ones, and pronounced,negative Eu anomalies.
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7566
5.3.1. Serra do Meio suiteThe REE patterns are parallel or roughly paral-
lel, with pronounced, negative Eu anomaly (Fig.12). CeN/YbN fractionation ratios are lower in theNS-oriented metaluminous plutons (CeN/YbN,8.97–16.16) than in the strongly peralkaline gran-ites (14.38–27.46). The slightly peralkaline andmetaluminous, magnetite-bearing facies have frac-tionation ratios similar to those of riebeckite-bearing facies (Pla Cid, 1994). Nevertheless, themetaluminous rocks have higher total REE con-centrations than the strongly peralkaline granites(Fig. 12). The negative Eu anomalies increasefrom the strongly peralkaline granites (Eu/Eu*=0.52–0.60) to the NS-oriented metaluminous plu-tons (Eu/Eu*=0.40–0.42). These anomalies aremore pronounced in the slightly peralkaline facies(0.25–0.38), the metaluminous, magnetite-bearingplutons presenting larger variation, from 0.13 to0.57. The metasomatically altered granites showlight REE-enrichment and high (CeN/YbN) ratios,which according to Pla Cid (1994), is due to REEtransport by fluids associated with the tectonicevents related to the Brasiliano cycle.
The Serra do Meio granites are REE enrichedwhen compared with the average contents pre-sented by Bowden and Whitley (1974) for alkalinemetaluminous and peralkaline granites. The nega-tive Eu anomalies increasing from peralkaline tometaluminous trend, associated with trace ele-ments evolution (Fig. 9), suggest differences eitherin the magmatic evolution, or in the sources, ofperalkaline and metaluminous magmas of thissuite.
5.3.2. Couro-de-Onca intrusi6e suiteRare earth element patterns in the COIS (Fig.
12) show an evolution from the foliated to theisotropic types, characterized by higher (CeN/YbN) ratio in the more evolved rocks (11.41–18.13), whereas in the granites with SiO2 around70 wt.%, it ranges from 5.46 to 11.20. This varia-tion is due to heavy REE-enrichment, possiblyattributed to higher modal concentrations of ap-atite, that may readily accommodate these ele-ments (Wilson, 1989), and as suggested bypositive correlation between P2O5 and Yb (Fig.13). The total REE-contents are extremely vari-
Fig. 10. Rb/Sr vs. Sr, Ba, and Rb (in ppm) plots for the SMS.
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 67
Fig. 11. Rb/Sr vs. FeOt, MgO, and TiO2 plots of the COIS, suggesting fractionation of amphibole, Fe–Ti oxide, and plagioclaseduring the magmatic evolution.
able, and are not related to SiO2 increase. Bothfoliated and isotropic facies have negative Euanomalies, which are lower in the foliated rocks,and increase with differentiation (Fig. 12). This isprobably related to plagioclase modal decreasefrom foliated types to less or non-foliated ones,suggesting plagioclase fractionation. These REEpatterns are similar to those found in the alkalineand Rapakivi granites from the Amazonianprovince (Fig. 12; Dall’Agnol et al., 1994).
6. Discussion
6.1. Emplacement and tectonic setting
Alkaline granites are commonly emplaced dur-ing the late stages of an orogenic event, or imme-diately after it, and in anorogenic settings, asdescribed by Bonin (1987), Nardi and Bonin(1991), Don Hermes and Zartman (1992). In the
COIS, the foliation of the granites parallel to theregional deformation structures is suggestive oftheir late-tectonic emplacement. According toFigueroa and Silva Filho (1990), these granitesare emplaced during the late stages of a transcur-rent event attributed to the Transamazonian cy-cle. On the other hand, the Rb–Sr isochron forthe SMS (850–500 Ma), associated with fielddata, is compatible with a pre-Brasiliano emplace-ment of this suite (Pla Cid, 1994). Its associationwith Transamazonian carbonatitic rocks andtholeiitic to transitional basalts with preservedcumulate textures (Couto, 1989) points to an ex-tensional, intracontinental tectonic setting duringemplacement of this magmatic province. The en-richment of incompatible elements, such as Zr,Nb, Y, Ga, and light REE, relative to the averageof anorogenic granites (Whalen et al., 1987), andto anorogenic suites, such as the Corsican (Bonin,1980, 1988) and the Amazonian provinces (Dal-l’Agnol et al., 1994) put in evidence the anoro-
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7568
Fig. 12. REE patterns of the SMS (A, B) and COIS (C), normalised to chondritic values (Evensen et al., 1978). Amazonian granitespatterns are from Dall’Agnol et al. (1994).
genic character of these alkaline granites and theiranomalous enriched source. Pearce et al. (1984)used trace-element diagrams to discriminate dif-ferent tectonic settings. For both suites, the trace-element contents plot in the field of within-plategranites (Fig. 14). In the COIS, the dykes havevery low trace-element concentrations, due to in-tense fractionation of accessory phases.
6.2. Crystallization and recrystallization
The SMS rocks are affected strongly by solid-state deformation, with intense recrystallization.However, some magmatic textures are preserved,such as alkali-feldspar phenocrysts and somemafic minerals, such as arfvedsonitic amphiboles,aegirine–augite, and Ti-aegirine, which have sub-hedral igneous forms and primary mineral com-positions. According to Silva (1987), Jardim de Sa(1994), the metamorphic temperatures estimatedfor the Riacho do Pontal fold belt, and particu-larly in the Campo Alegre de Lourdes region,range from greenschist to low amphibolite facies.
This is confirmed by the textural relations ob-served in these granitic rocks. The annealing ofquartz, observed in all facies, reflects static, post-deformational, thermal effects (Hibbard, 1995),and represents the lower limit of recrystallizationtemperatures. On the other hand, the newly-formed feldspars, and recrystallization of sodic
Fig. 13. P2O5 (in wt.%) vs. Yb (in ppm) plot of the COIS.
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 69
Fig. 14. Nb (in ppm) vs. Y (in ppm) plot, with tectonic settingfields for granitic rocks based on Pearce et al. (1984).
enhanced ductility during deformation.Irregular-shaped remnants of plagioclase in al-
kali-feldspar grains are taken as evidence of par-tial resorption texture in the COIS. Furthermore,the subhedral character of alkali-feldspar and pla-gioclase grains, and the relative absence of inclu-sions, suggest they have crystallized at the sametime in the less differentiated rocks. Based onexperimental studies by Luth and Tuttle (1966),the occurrence of perthitic feldspar with micro-cline borders is compatible with a minimumsolidus temperature of 650°C. According toNeksvasil (1992), the simultaneous crystallizationof alkali-feldspar and plagioclase, with intensiveplagioclase partial resorption, in SiO2-saturatedliquids, occurs at about 950°C, whereas for pla-gioclase initial crystallization the inferred temper-atures may reach up to 1100°C. Theseexperiments have been conducted at 5 kb, and atlower pressures would result in less than 10%variation. The mineral orientation, with irregular,curved and interpenetrating contacts, point tocrystallization controlled by a regional stress field,which is corroborated by Pb–Pb in zircon iso-topic data of 215795 Ma, confirming the late-orogenic character of this suite.
In the COIS, disequilibrium textures might re-sult from changes in conditions of Pfluids and totalpressure, caused by intrusion of the dyke swarm,with intense fluid percolation in the host granites.Since reversal zoning in amphibole, as well asmicrogranular concentrations are more commonlyobserved in more differentiated granites, it is evi-dent that the granitic liquid was partially crystal-lized. Then, during differentiation of this graniticmagma, fracturing was probably controlled byregional deformation. This supposition is sup-ported by parallelism between coarse grain dykesand banding in granites, as well as by a similarorientation of amphiboles in dykes and granites.
6.3. Possible nature of the source
Trace elements have been employed by Pearceand Norry (1979), Pearce et al. (1984), Wilson(1989), among others, to establish constraints onthe possible mantle or crustal derivation ofgranitic magmatism, as well as on its tectonic
mafic minerals represent the highest temperaturesattained during the Brasiliano cycle. This limitcan be estimated from the stability conditions ofalkali amphiboles (Ernst, 1962). This authorshows that, for high fO2 conditions, riebeckite isthe only sodic amphibole stable around 520°C, asalso observed by Bonin (1982), which providesevidence that riebeckite is a typical alkali-amphi-bole formed at subsolidus temperatures. Thus, theupper metamorphic temperature limit is probablybelow 550°C. The occurrence of calcite crystalssuggests that recrystallization condition was inpresence of CO2-rich fluids, which might have
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7570
setting. Spiderdiagrams normalised to the oceanridge granite (ORG) values (Pearce et al., 1984)for the SMS granites are shown in Fig. 15. Theirhigh Nb and Ta contents, and the similar concen-trations of Rb and Th with respect to the ORGvalues, are typical OIB-type mantle features(Pearce et al., 1984; Wilson, 1989). Pla Cid et al.(1997a,b) concluded that they derived from anincompatible-element enriched mantle source. Theperalkaline trend of the SMS is less enriched inHFSE, whereas the metaluminous one is stronglyenriched in Rb, Th, Nb, and Ta, with large,negative Ba anomaly (Fig. 15). The negative Baanomaly explains the more pronounced Eu nega-tive anomaly in the metaluminous trend granites,which is probably due to intense feldspar fraction-ation during magmatic evolution, less evident inperalkaline ones. The high Rb and Th contentsare suggestive of crustal contamination (Thirwall
and Jones, 1983), although, as shown in tectonicsettings diagrams of Pearce et al. (1984), within-plate granites exhibit enrichment in Rb with in-creasing Nb contents. Dooley and Patino Douce(1996), Patino Douce and Beard (1996) showed byexperimental studies the improbability of crustalsources to generate peralkaline graniticcompositions.
Trace element contents of COIS, normalized tothe ORG (Pearce et al., 1984; Fig. 15), show atypical within-plate pattern. There are some ex-amples of granites in Brazil with similar geochem-ical affinities, generally emplaced in extensionalcontinental settings, such as the Proterozoicanorogenic granitoids from Central Amazonianprovince (Dall’Agnol et al., 1994), and the post-orogenic monzo-syenogranites of Jaguari pluton,from the Saibro Intrusive Suite, southern Brazil(Gastal and Nardi, 1992).
Fig. 15. Spidergrams of the Serra do Meio (a, b) and Couro de Onca (c) suites normalised to Ocean Ridge granites (Pearce et al.,1984).
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 71
Jacobson et al. (1958) proposed a geneticmodel, modified by Bowden and Kinnaird (1978),to explain the evolution of alkaline suites, mainlyfor their acid components. This model proposesan evolution from olivine-gabbro to syenite andfayalite-granite through plagioclase dominatedfractionation, generating anorthosite cumulates.From fayalite-granite, two different paths begindue to fractional crystallization processes, a peral-kaline trend, producing arfvedsonite and riebeck-ite granite, and an aluminous line, responsible forhastingsite and biotite granites. Bonin (1982) pre-sents a similar scheme, where syenitic/monzoniticrocks evolve to peralkaline and peraluminouslines. In this scheme, hastingsite granites are theless evolved rocks in the peralkaline trend.
Two different hypotheses can explain the traceelement concentrations of metaluminous andstrongly peralkaline granites from the SMS.Firstly, both granitic types were produced bypartial melting from different sources. This modelwould hardly explain the linear trend observed forsome pairs of trace elements in Fig. 9. Further-more, different sources neither explain the alka-line signature of both granite types, nor similarratios of elements such as Zr/Nb, Y/Zr, La/Sm,and Dy/Yb. Secondly, two different parental basicmelts were produced by extraction, through asmall degree of partial melting, of the mantle atdifferent stages. If this is accepted, an originalmelt responsible for the evolution towards meta-luminous trend was the first magma type ex-tracted, with the solubility of Zr enhanced by hightemperature and F contents. This initial magmaextraction could have caused a relative depletionin these elements at the source, directly affectingthe composition of peralkaline liquids generatedduring a second partial-melting event.
In the specific case of granites from the SMS,the unusual concentrations of incompatible ele-ments in the metaluminous trend can be explainedin terms of F contents. In fact, the granites ofmetaluminous trend are strongly enriched in in-compatible elements when compared with typicalanorogenic metaluminous granites from Nigeria(Bowden and Kinnaird, 1984), but they are alsoenriched in F. According to Harris (1980), highalkali contents in magmas promotes F enrichment
and, consequently, the presence of halogens isresponsible for stabilization of Zr, Nb, Y, and Thcomplexes, which are the enriched elements in thestudied metaluminous granites. According toWatson (1979), high alkali contents in liquids withlow F promote the formation of alkali-zirconosili-cate complexes (Na4Zr(SiO4)2) and, as assumed byBonin (1988), the presence of ns (Na-silicate) inthe CIPW norm is a good indicator of alkali-zir-conosilicate complexes. In the SMS, ns absence inthe CIPW norm (Table 5) suggests the probableformation of F–Zr complexes. As can be seen inTable 3, the metaluminous trend is F-richer thanthe strongly peralkaline ones, where F content inthe former ranges from 300–4100 ppm (values areusually \1100 ppm), and from 250 to 1300 ppmin the latter. This may be suggestive of metalumi-nous trend being related to the first partial melt-ing stage, which promoted the stabilization ofF-rich complexes due to high alkali contents ofthe original melt, consequently transporting ele-ments such as REE, Nb, Zr, Y, and Th.
Similar control on HFSE and REE behaviourby F-rich complexes was demonstrated by Charoyand Raimbault (1994) in biotite granites of alka-line affinity from eastern China, extremely en-riched in these elements. According to theseauthors, the magmatic evolution of these graniteswas controlled by F-rich, water-poor, high-tem-perature magmas, which become volatile-satu-rated late in the crystallization sequence. Thepositive correlation between HFSE and F indi-cates that these fluids have complexed and trans-ported HFSE as soluble components.
It is concluded that SMS metaluminous andperalkaline trends reflect the evolution of twoparental magmas, successively extracted from thesame enriched mantle source. Both alkaline linesevolved from alkali-basalt liquids, mainly throughcrystal fractionation processes. Metaluminous-trend rocks evolved from a Ca-richer magma,whereas the strongly peralkaline granites wereproduced from a Na, Ti, and Fe enriched liquid.
Finally, it is concluded that the two differentalkaline granites suites, which occur at the north-ern border of the Sao Francisco Craton, havebeen emplaced in a within-plate tectonic setting,during extensional tectonic regimes related to the
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7572
Tab
le5
CIP
Wno
rmof
the
Serr
ado
Mei
osu
ite
Sam
ple
CL
-16
CL
-05b
GA
14c
GA
-34
GA
-39
GA
-46
GA
-46a
TR
-72
PP
B-7
8aC
L-1
05C
L-5
4C
L-5
5G
A-1
3
39.6
235
.28
19.8
931
.57
31.9
511
.226
.827
.21
31.1
130
.76
31.7
841
.71
32.7
2Q
uart
z–
––
––
––
–0.
65C
orun
dum
––
0.02
–29
.23
24.4
625
.625
.56
26.7
133
.925
.56
28.5
716
.08
29.7
125
.11
23.9
827
.6O
rtho
clas
e35
.45
29.7
133
.19
42.5
837
.99
31.6
45.1
833
.01
22.2
130
.06
Alb
ite
31.6
134
.25
26.4
4–
––
––
–0.
15–
Ano
rthi
te2.
953.
18–
2.38
1.57
7.82
8.32
2.95
62.
536.
69–
1.65
–A
cmit
e–
0.15
––
––
––
––
––
––
––
–ns
4.97
1.18
1.43
0.72
2.44
2.26
––
–D
iops
ide
1.54
1.7
–1.
043.
080.
550.
443.
06–
1.67
0.5
0.25
Hyp
erst
hene
2.54
4.13
0.9
2.85
0.5
1.92
1.96
2.45
1.22
4.13
1.03
5.19
2.94
4.91
Mag
neti
te1.
914.
340.
881.
46–
––
––
–0.
650.
62H
emat
ite
––
––
0.45
1.13
1.03
0.74
1.19
0.82
0.84
0.75
0.69
0.73
0.81
Ilm
enit
e0.
540.
520.
660.
110.
110.
110.
110.
110.
110.
110.
11A
pati
te0.
110.
110.
110.
110.
1111
.11
4.71
5.07
6.18
6.99
5.79
7.09
4.5
6.59
Col
.In
dex
7.11
7.6
5.53
5.58
80.9
486
.84
91.8
587
.68
90.3
587
.38
92.3
793
.48
Dif
f.In
dex
91.3
491
.14
92.1
290
.38
91.0
6P
PB
-14
PP
B-1
4aC
L-7
7C
L-8
8G
A-1
3G
A-1
4G
A-1
4bP
PB
-28a
JP-5
0P
PB
-63
GA
-68
PP
B-6
4C
L-0
9C
L-8
7JP
-49
Sam
ple
CL
-74
25.4
333
.61
26.7
221
.18
36.8
133
.11
39.6
226
.32
23.2
129
.68
10.2
731
.14
24.5
133
.531
.19
Qua
rtz
18.6
8–
––
–1.
79–
0.65
––
––
1.26
––
–C
orun
dum
–29
.74
29.9
239
.929
.32
29.2
326
.37
24.5
33.3
629
.15
25.3
Ort
hocl
ase
29.7
629
.86
30.4
129
.85
27.4
330
.36
35.6
328
.96
34.0
841
.12
15.3
531
.722
.21
34.3
340
.22
30.7
151
.64
31.5
435
.92
31.2
332
.48
Alb
ite
40.8
7–
0.52
0.02
1.95
1.57
2.36
2.83
1.13
2.47
––
1.77
2.41
Ano
rthi
te–
–1.
192.
99–
––
––
––
Acm
ite
0.1
2.41
––
1.82
–2.
36–
––
––
––
––
––
ns–
––
––
–0.
042.
56–
3.15
–0.
95–
2.85
2.08
1.88
––
0.95
–2.
14–
Dio
psid
e0.
32–
0.42
1.8
0.5
3.69
2.82
1.46
3.17
0.81
Hyp
erst
hene
1.77
5.24
0.35
1.79
0.3
0.93
4.19
4.55
4.3
0.85
4.91
2.79
3.23
1.34
Mag
nrti
te5.
822.
152.
190.
544.
480.
945.
092.
040.
3–
0.26
–0.
45–
––
–3.
6H
emat
ite
––
0.02
–0.
01–
0.82
0.42
1.01
0.19
0.73
1.12
0.98
0.31
0.77
Ilm
enit
e0.
440.
40.
440.
710.
290.
840.
420.
110.
110.
110.
110.
110.
130.
110.
110.
110.
110.
110.
110.
11A
pati
te0.
110.
110.
115.
647.
136
3.79
6.59
10.4
59.
14.
9811
Col
.In
dex
4.4
6.04
7.17
5.5
5.16
6.3
5.68
90.5
392
.22
92.0
694
.13
91.0
687
.03
87.9
493
.75
87.2
91.7
392
.44
Dif
f.In
dex
90.2
991
.42
92.1
689
.97
93.5
2
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 73
Transamazonian event. The COIS is a typicalexample of a late-Transamazonian suite, Ca- andAl-richer than the SMS, and according to thegenetic scheme suggested by Bonin (1982), thesegranites are earlier than the peralkaline types. TheSMS has composition very similar to anorogenicsuites. The strong deformation caused by conti-nental collision during the Brasiliano event, aswell as the Transamazonian age of carbonatiticrocks in this region, suggest that the SMS isrelated to continental rift setting after the Transa-mazonian event.
Acknowledgements
This work was supported by CNPq andCAPES. J.P.C. thanks to Professor Jean-MichelLafon for isotopic determinations on the Couro-de-Onca and Serra do Meio Suits, accomplishedat the Geochronological Laboratory from FederalUniversity of Para, Brazil; the Fundacao Coorde-nacao de Aperfeicoamento de Pessoal de NıvelSuperior (CAPES-no. 1772/95-14), the Centro deEstudos em Petrologia e Geoquımica-UFRGS,Programa de Pesquisa e Pos-Graduacao em Ge-ofısica, PPPG-UFBA, Departement des Sciencesde la Terre, Centre d’Orsay, Universite de Paris-Sud, and his wife, for her patience.
References
Almeida, F.F.M., Hasui, Y., Brito Neves, B.B., 1977. Provın-cias Estruturais Brasileiras. Simp. Geol. Nord., CampinaGrande-PB, Anais, pp. 363–391.
Ashley, P.M., Cook, N.D.J., Fanning, C.M., 1996. Geochem-istry and age of metamorphosed felsic igneous rocks withA-type affinities in the Willyama supergroup, Olary Block,South Australia, and implications for mineral exploration.Lithos 38, 167–184.
Barbosa, J.S.F., Dominguez, J.M.L., 1996. Mapa geologico doEstado da Bahia. Texto Explicativo. Secretaria da Indus-tria, Comercio e Mineracao do Estado da Bahia. PPPG-UFBA, Salvador, Brasil.
Bonin, B., 1980. Le complexes acides alcalins anorogeniquescontinentaux: l’exemple de la Corse. Doct. Etat es-Sci.Thesis, Universite Pierre et Marie Curie, Paris, France.
Bonin, B., 1982. Les granites des complexes annulaires.B.R.G.M. (Ed.) Manuels et Methodes no. 4, Orleans, p.147.
Bonin, B., 1987. From orogenic to anorogenic magmatism: apetrological model for the transition calk-alkaline–alkalinecomplexes. Internatinal Symposium on Granites and Asso-ciation of Mineral. Salvador, Brazil, Extended Abstracts,pp. 27–31.
Bonin, B., 1988. Peralkaline granites in Corsica: some petro-logical and geochemical constraints. Rend. Della Soc. Ital.Miner. Petrol. 43 (2), 281–306.
Bowden, P., Turner, D.C., 1974. Peralkaline and associatedring-complexes in the Nigeria — Niger province, westAfrica. In: Sorensen, H. (Ed.), The Alkaline Rocks. Wiley,London, pp. 330–351.
Bowden, P., Whitley, J.E., 1974. Rare-earth patterns in peral-kaline and associated granites. Lithos 7, 15–21.
Bowden, P., Kinnaird, J.A., 1978. Younger granites of Nigeria— a zinc-rich tin province. Trans. Sec. B Inst. Miner. Met.87, 66–69.
Bowden, P., Kinnaird, J.A., 1984. The petrology and geo-chemistry of alkaline granites from Nigeria. Phys. EarthPlanet Int. 35, 199–211.
Brito Neves, B.B., 1975. Regionalizacao geotectonica do Pre-Cambriano nordestino. Tese de Doutorado, Instituto deGeociencias, Universidade de Sao Paulo, Brasil.
Charoy, B., Raimbault, L., 1994. Zr-, Th-, and REE-richbiotite differentiates in the A-type granite pluton ofSuzhou (Eastern China): the key role of fluorine. J. Petrol.35 (4), 919–962.
Collins, W.J., Beams, S.D., White, A.J.R., Chappell, B.W.,1982. Nature and origin of A-type granites with specialreference to southeastern Australia. Contrib. Miner. Petrol.80, 189–200.
Conceicao, H., 1990. Petrologie du massif syenitique d’Itiuba:contribution a l’etude mineralogique des roches alcalinesdans l’Etat de Bahia (Bresil). Tese de Doutorado, Univer-site Paris-Sud, Centre d’Orsay, France.
Conceicao, R.V., 1994. Petrologia dos sienitos potassicos domacico de Santanapolis e alguns aspectos do seu embasa-mento granulıtico. Dissertacao de Mestrado, CPGG-UFBA, Salvador, Brasil.
Couto, L.F., 1989. Estudo petrologico do complexo mafico-ul-tramafico de Campo Alegre de Lourdes e dos oxidos deFe, Ti, (V) associados. Dissertacao Mestrado, Universi-dade de Brasılia-UNB, Brasılia, Brasil.
Czamanske, G.K., Dillet, B., 1988. Alkali amphibole, tetrasili-cic mica, and sodic pyroxene in peralkaline siliceous rocks,Quest caldera, New Mexico. Am. J. Sci. 288-A, 358–392.
Dall’Agnol, R., Lafon, J.-M., Macambira, M.J.B., 1994.Proterozoic anorogenic magmatism in the central Amazo-nian province, Amazonian craton: geochronological, petro-logical and geochemical aspects. Miner. Petrol. 50,113–138.
Dalton de Souza, J.A., Fernandes, F.J., Guimaraes, J.T.,Lopes, J.N., 1979. Projeto colomi: geologia da regiao doMedio Sao Francisco. Relatorio final, DNPM/CPRM, Sal-vador, Brasil.
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–7574
Don Hermes, O., Zartman, R.E., 1992. Late proterozoic andsilurian alkaline plutons within the southeastern New Eng-land Avalon zone. J. Geol. 100 (4), 477–486.
Dooley, D.F., Patino Douce, A.E., 1996. Fluid-absent meltingof F-rich phlogopite+rutile+quartz. Am. Miner. 81,202–212.
Ernst, W.G., 1962. Synthesis, stability relations, and occur-rence of riebeckite and riebeckite-arfvedsonite solid solu-tions. J. Geol. 70, 689–736.
Evensen, N.M., Hamilton, P.J., O’Nions, R.K., 1978. Rareearth abundances in chondritic meteorites. Geochim. Cos-mochim. Acta 42, 1199–1212.
Ferreira, V.P., Sial, A.N., Whitney, J.A., 1994. Large scalesilicate immiscibility: a possible example from northeastBrazil. Lithos 33, 285–302.
Figueroa, I., Silva Filho, I., 1990. Geologia da FolhaPetrolina. In: Prog. Levant. Geol. Bas. do Brasil (PLGBB)Folha SC.24-V-C-III Petrolina. DNPM/CPRM, Rio deJaneiro, pp. 11–56.
Gastal, M.C.P., Nardi, L.V.S., 1992. Petrogenese e Evolucaodo Granito Jaguari: um tıpico representante metaluminosoda Suıte Intrusiva Saibro, RS. Geochem. Brasil 6 (2),169–189.
Harris, N.B.W., 1980. The role of fluorine and chlorine in thepetrogenesis of a peralkaline complex from Saudi Arabia.Chem. Geol. 31, 303–310.
Hibbard, M.S., 1995. Petrography to Petrogenesis. Prentice-Hall, Englewood Cliffs, NJ.
Jacobson, R.R.E., Mac Leod, W.N., Black, R., 1958. Ring-complexes in the Younger granite province of northernNigeria. Geol. Soc. London Mem. 1, 1–72.
Jardim de Sa, E.F., 1994. A faixa Serido (Provıncia Bor-borema, NE do Brasil) e o seu significado geodinamico nacadeia Brasiliana/Pan-Africana. Tese de Doutorado, Uni-versidade de Brasılia, Brasil.
Leite, C.M., 1987. Projeto Remanso II. Relatorio Final.CBPM-SME, Salvador, Brasil.
Leite, C.M., 1993. Caracterizacao dos macicos alcalinos comogranitos tipo-A na Provıncia Toleıtica Alcalina de CampoAlegre de Lourdes (Bahia). IV Cong. Bras. Geoq., SaoPaulo, SP, Bol. Res. Exp., pp. 38–40.
Leite, C.M., 1997. Programa de Levantamentos GeologicosBasicos do Brasil, Folha-SC.23-X-D-IV (Campo Alegre deLourdes), escala 1E. :100000. CBPM/CPRM/SICM, Sal-vador, Brasil.
Leite, C.M., Froes, R.J.B., 1989. Caracterısticas petroquımicasdo granito alcalino Serra do Meio (Campo Alegre deLourdes-Estado da Bahia). II Cong. Bras. Geoq., Rio deJaneiro-RJ, Anais, 1, pp. 157–167.
Leite, C.M., Conceicao, H., Cruz, M.J., 1991. Plutonismohiperalcalino supersaturado da Provıncia de Campo Alegrede Lourdes: evolucao mineraloquımica e suas implicacoes.III Cong. Bras. Geoq., Sao Paulo, Anais, 2, pp. 717–721.
Leite, C.M.M., Santos, R.A., Conceicao, H., 1993. A provın-cia toleiıtica-alcalina de Campo Alegre de Lourdes: geolo-gia e evolucao tectonica. II Simposio sobre o Craton do
Sao Francisco. SBG/SGM, Salvador-Brasil, Anais, 1, pp.56–59.
Le Maitre, R.W., Bateman, P., Dubek, A., Keller, J.,Lameyre, J., Le Bas, M.J., Sabine, P.A., Schimid, R.,Sorensen, H., Streckeisen, A., Wooley, A.R., Zanettin, B.(Eds.), 1989. A Classification of Igneous Rocks and Glos-sary of Terms: Recommendations of the InternationalUnion of Geological Sciences Subcommission on the Sys-tematics of Igneous Rocks. Blackwell (Basil), Oxford.
Luth, W.C., Tuttle, O.F., 1966. The alkali-feldspar solvus inthe system Na2O–K2O–Al2O3–SiO–H2O. Am. Miner. 51(9/10), 1359–1373.
Mahood, G., Hildreth, W., 1983. Large partition coefficientsfor trace elements in high-silica rhyolites. Geochim. Cos-mochim. Acta 47 (1), 11–30.
Maniar, P.D., Piccoli, P.M., 1989. Tectonic discrimination ofgranitoids. Geol. Soc. Am. Bull. 101, 635–643.
Middlemost, E.A.K., 1994. Naming materials in the magma/igneous rock system. Earth Sci. Rev. 37, 215–224.
Miller, C.F., Mittlefehldt, D.W., 1984. Extreme fractionationin felsic magma chambers: a product of liquid-state diffu-sion or fractional crystallization? Earth Plan. Sci. Lett. 68(1), 151–158.
Nardi, L.V.S., 1991. Caracterizacao petrografica e geoquımicados granitos metaluminosos da associacao alcalina: re-visao. Pesquisas 18 (1), 44–57.
Nardi, L.V.S., Bonin, B., 1991. Post-orogenic and non-oro-genic alkaline granite associations: the Saibro intrusivesuite, southern Brazil — a case study. Chem. Geol. 92,197–212.
Neksvasil, H., 1992. Ternary feldspar crystallization in high-temperature felsic magmas. Am. Miner. 75 (5/6), 592–604.
Patino Douce, A.E., Beard, J.S., 1996. Effects of P, f(O2)andMg/Fe ratio on dehydration melting of model meta-greywackes. J. Petrol. 37 (5), 999–1024.
Pearce, J.A., Norry, M.J., 1979. Petrogenetic implications ofTi, Zr, Y e Nb variation in volcanic rocks. Contrib. Miner.Petrol. 69, 33–47.
Pearce, J.A., Harris, N.B.W., Tindle, A.G., 1984. Trace ele-ment discrimination diagrams for the tectonic interpreta-tion of granitic rocks. J. Petrol. 25, 956–983.
Pla Cid, J., 1994. Granitogenese alcalina de Campo Alegre deLourdes (Norte da Bahia): Petrografia, Mineraloquımica eGeoquımica. Dissertacao de Mestrado, CPGG-UFBA,Salvador, Brasil.
Pla Cid, J., Conceicao, H., Nardi, L.V.S., 1995. As micastri-octaedricas da suıte granıtica de Campo Alegre deLourdes (N da Bahia). V Cong. Bras. Geoq. and III Cong.Geoq. dos Paıses de Lıngua Port., CD-Rom, Niteroi-Brasil.
Pla Cid, J., Nardi, L.V.S., Conceicao, H., Bonin, B., 1997a. Omagmatismo alcalino da faixa de dobramentos Riacho doPontal e da borda noroeste do Craton do Sao Francisco,norte do Estado da Bahia, Brasil: uma sıntese. X Semanade Geoq. e IV Cong. Geoq. dos Paıses de Lıngua Por-tuguesa, Braga, Portugal, Actas., pp. 127–130.
Pla Cid, J., Nardi, L.V.S., Conceicao, H., Bonin, B., 1997b.Alkalic plutonic activity within in the Riacho do Pontal
J. Pla Cid et al. / Precambrian Research 104 (2000) 47–75 75
fold belt, NE Brazil. Second ISGAM, Ext. Abst. and Prog.,Salvador-Brasil, pp. 143–144.
Rosa, M.L.S., 1994. Magmatismo shoshonıtico e ultrapotas-sico no sul do cinturao movel Salvador-Curaca, macico deSao Felix: Geologia, mineralogia a geoquımica. Dissertacaode mestrado, CPGG-UFBA, Salvador, Brasil.
Scogings, A.J., 1986. Trans. Geol. Soc. South Afr. 89, 361–365.
Silva, M.E., 1987. O sistema de dobramentos Rio Preto e suasrelacoes com o Craton do Sao Francisco. Dissertacao deMestrado, Instituto de Geociencias, Universidade deBrasılia, Brasil.
Silva, A.B., Liberal, G.S., Grossi Sad, J.M., Issa Filho, A., Ro-drigues, C.S., Riffel, D.F., 1988. Geologia e petrologia docomplexo Angico dos Dias (Bahia, Brasil), uma associacaocarbonatıtica precambriana. Geochem. Bras. 2 (1), 81–108.
Sorensen, H., 1974. The Alkaline Rocks. Wiley, London.Thirwall, M.F., Jones, N.W., 1983. Isotope geochemistry and
contamination mechanics of tertiary lavas from Skye,Northwest Scotland. In: Hawkesworth, C.J., Norry, M.J.(Eds.), Continental Basalts and Mantle Xenoliths. Shiva,
Nantwich, pp. 186–208.Van Schmus, W.R., Brito Neves, B.B.D., Hackpacher, P.,
Babinski, M., 1995. U–Pb and Sm–Nd geochrnologicstudies on eastern Borborema province, northeasternBrazil: initial conclusions. J. South Am. Earth Sci. 8 (3/4),267–288.
Watson, E.B., 1979. Zircon saturation in felsic liquids: experi-mental results and applications to trace element geochem-istry. Contrib. Miner. Petrol. 70, 407–419.
Watson, E.B., Harrison, T.M., 1983. Zircon saturation revis-ited: temperature and composition effects in a veriety ofcrustal magma types. Earth Plan. Sci. Lett. 64, 295–304.
Wernick, E., 1981. The Archean of Brazil. Earth Sci. Rev. 17,31–48.
Whalen, J.B., Currie, K.L., Chappell, B.W., 1987. A-type gran-ites: geochemical characteristics, discrimination and petro-genesis. Contrib. Miner. Petrol. 95, 407–419.
Wilson, M., 1989. Igneous Petrogenesis: A Global TectonicApproach. Unwin Hyman, London, UK.
Winkler, H.G.F., 1979. In: Winckler, H.G.F. (Ed.), Petrogene-sis of Metamorphic Rocks. Springer, New York.
.