THE COMPOSITION OF ZIRCON IN THE PERALUMINOUS HERCYNIAN GRANITES OF THE SPANISH CENTRAL SYSTEM BATHOLITH

509

The Canadian MineralogistVol.45,pp.509-527(2007)DOI:10.2113/gscanmin.45.3.509

THE COMPOSITION OF ZIRCON IN THE PERALUMINOUS HERCYNIAN GRANITES OF THE SPANISH CENTRAL SYSTEM BATHOLITH

CeciliaPÉREZ-SOBA§,CarlosVILLASECAandJoséGONZÁLEZDELTÁNAGO

Departamento de Petrología y Geoquímica, Faculdad de Ciencias Geológicas, Universidad Complutense, E–28040 Madrid, Spain

LutzNASDALA

Institut für Mineralogie und Kristallographie, Universität Wien, A–1090 Wien, Austria

Abstract

WehaveinvestigatedthezirconfromgranitesoftheSpanishCentralSystem(SCS)batholith.ThisbatholithiscomposedofplutonsofbothI-andS-typegranite,withdifferentdegreesoffractionation.Theassociationprovidesanaturallaboratoryforthestudyofcompositionalevolutionofzircon.Werecognizetwotypesofmagmaticzircon.Type-1zirconconsistsofeuhedralandelongatecrystalsshowinganoscillatoryzoning,usuallyhostedintheearlymajorminerals,beingmoreabundantingranodioriteandmonzograniteplutons.Type-2zirconischaracterizedbysmallerandmoreequantcrystals,unzoned,orirregularlyzoned,andusuallyinterstitialtotherock-formingminerals,andformingclusterswithotheraccessoryminerals.Thissecondtypeismoreabundantinleucogranites,aplitesandpegmatites.Type-2zirconshowsthehighestconcentrationsofHf(16.58wt.%HfO2),U(4.83%UO2),Th(4.87%ThO2),Y(8.51%Y2O3),andHREE(3.78%HREE2O3),andthelowestanalyticaltotals(downto91.2%).ThereplacementofZrbytetravalentcations(Th,UandHf)isfavoredinthemostevolvedgranites.Micro-areasofzirconwithlowanalyticaltotals(91–98wt.%)haveahighlevelofcationsubstitution.Moreover,theirhighTh–Ucontentsinduceelevatedlevelsofradiationdamage.Disturbancesintheoriginalstructure,combinedwithseveremetamictizationofU–Th-richmicro-areas,mayhavecausedanenhancedsusceptibilitytosecondaryalteration.Thisprocess,inturn,isbelievedtohaveformedthe observed strongly damaged and cation-deficient zircon. The range of Zr/Hf values in zircon crystals of the granites studied andthoseofrelatedmetamorphicrocks(aspossiblecrustalsources)aresimilar.ThecompositionofzirconfromSCSgranulitexenoliths overlaps with the compositional field of zircon from granite, supporting previous models of crustal recycling for the originoftheSCSbatholith.

Keywords:zircon,composition,traceelements,I-typegranite,S-typegranite,peraluminousgranulites,IberianHercynianBelt,SpanishCentralSystem.

Sommaire

NousavonsétudiélezirconprovenantdesgranitesdubatholiteduSystèmeCentralEspagnol.Cebatholitecomprenddesplutons de type I aussi bien que de type S, avec des degrés de fractionnement variables. L’association fournit un excellentlaboratoirenaturelpourl’étudedel’évolutioncompositionnelduzircon,quipeutavoirundedeuxaspects.Lezircondetype1seprésenteencristauxidiomorphesetallongésmontrantunezonationoscillatoire,généralementenglobésdanslesminérauxmajeursprécoces;ilestdoncplutôtabondantdanslesplutonsgranodioritiquesetmonzonitiques.Lezircondetype2seprésenteencristauxpluspetitsetpluséquidimensionnels,nonzonésoubienzonésdefaçonirrégulière,etgénéralementtardifs,enposi-tioninterstitielleparmilesminérauxmajeursdelaroche,etregroupésavecd’autresminérauxaccessoires.Lezircondetype2estplusabondantdanslesleucogranites,aplitesetpegmatites.IlmontrelesteneurslesplusélevéesenHf(16.58wt.%HfO2),U(4.83%UO2),Th(4.87%ThO2),Y(8.51%Y2O3),etterresrareslourdes(3.78%desoxydes),etlestotauxd’analyseslesplusfaibles(jusqu’à91.2%,poids).LeremplacementduZrparlescationstétravalentsTh,UetHfseraitfavorisédanslesgraniteslesplusévolués.Lesmicro-régionsduzirconayantunfaibletotalanalytique(91–98%)possèdentunniveauélevéderemplace-mentparcescations.Deplus,leursteneursélevéesenThetUontproduitunniveauélevédedommagedûàlaradiation.Lesdistorsionscauséesàlastructureduzircon,enplusdelamétamictisationavancéedesmicro-régionslesplusrichesenUetTh,pourraientavoircausélasusceptibilitéaccrueduzircondetype2àl’altérationsecondaire,quirendraitcomptedescompositionsfortement déficitaires du zircon endommagé. Les valeurs du rapport Zr/Hf des cristaux de zircon dans les granites étudiés et dans lesrochesmétamorphiquesapparentées(commesourcespossiblesdanslacroûte)sontsemblables.Lacompositionduzircon

§ E-mail address:[email protected]

510 thecanadianmineralogist

danslesenclavesgranulitiquesdecebatholitechevauchecelleduzircondesgranites,étayantlesmodèlesd’unrecyclagedesrochesdelacroûtepourexpliquercellesdubatholite.

(TraduitparlaRédaction)

Most-clés: zircon, composition, éléments traces, granite de type I, granite de typeS, granulites hyperalumineuses,CeintureHercynienneIbérique,batholiteduSystèmeCentralEspagnol.

granitewiththoseofzirconfromlower-crustgranuliticxenoliths, the presumed residual counterpart of thegraniticSCSbatholith(Villasecaet al.1999).

DescriptionoftheGraniticPlutonsoftheSpanishCentralSystem

TheSCSbatholith,locatedincentralSpain,extendsover an area of about 10,000 km2. It was emplacedduringthelatestagesoftheHercynianorogeny,mainlyfrom325to285Ma(Villaseca&Herreros2000).TheSCSbatholith is almost entirelycomposedofperalu-minousgraniticrocks.Threetypesofgranitehavebeendistinguishedonthebasisoftheirdegreeofperalumi-nosity: i) moderately peraluminous cordierite-bearinggranites of S-type affinity, ii) weakly peraluminous tometaluminousgraniticrockswithlocallyaccessoryamphibole and affinity with I-type granites, and iii) transitional types of biotite-bearing granitic rocks ofintermediate peraluminous character (Villaseca et al.1998).FiveplutonsrepresentativeofthisI-andS-typevariability has been selected for this study focusedon zircon.The mineralogical data gathered on thesegraniticplutonsaresummarizedinTable1.

Of the five plutons, three are of I-type composition (Fig. 1).TheAtalaya Real pluton is a small circularmassif (18 km2) showing a wide lithological range.It ismainlycomposedofequigranularcoarse-grainedamphibole and biotite-bearing granitic rocks rangingfrom granodiorite to monzogranite. In the outermostmargin, a leucogranitic facies with associated apliticbodies is locallydeveloped.ARb–Sr isochronageof284±13Mawasreportedforthispluton(Villasecaet al.1995).

The La Cabrera pluton, covering 160 km2, isanother I-typemassif thatexhibitscomplexcomposi-tionalvariations. Itshowsanevolutionary trendfromamphibole-bearing granodiorite at the margin towardbiotite monzogranite as the dominant facies at thecenter(Bellido1979).Someleucograniticfaciesoccurinthecore.ARb–Srisochronageof310±14Mawasobtainedforthemainmonzogranite.

The La Pedriza pluton is one of the most felsicI-type plutons in the SCS. It is composed of coarse-grained biotite leucogranite outcropping in a 75 km2massif.Itshowsacrypticgeochemicalzoning,withastrongincreaseinrareelementstowardthenortheasternmargin,whereitreachesthehighestbulk-rockcontentsof Na–F–Rb–Hf–Th–U–Nb–Ta–Y–HREE. In some

Introduction

Zircon is one of the most widespread accessorymineral in granites, and particularly in peraluminousgranites and granulitic metamorphic rocks (Wark &Miller1993,Bea1996,Villasecaet al.2003).Itischar-acterizedbyitshighdurabilitytoweathering,alteration,metamorphism and resistance to high-temperaturediffusivere-equilibration(e.g.,Watson1996).Favoredbytheseuniqueproperties,zirconhasreceivedconsid-erable attention in the past decades as a thermochro-nologicalindicator(e.g.,Hanchar&Hoskin2003,andreferencestherein)andforitspotentialapplicationasahostmaterialforthesafestorageofnuclearwaste(e.g.,Mathieuet al.2001,Ewinget al.2003).Zirconisusedas a crude thermometer for crystallization conditionsandaspetrogeneticindicator(e.g.,Pupin1980,Harrison&Watson1983,Vavra1994,Hanchar&Watson2003,Valley2003,Kempeet al.2004,Watsonet al.2006).Moreover, zircon is widely used in geochronologicalstudies (e.g., Mezger & Krogstad 1997, Solar et al. 1998,Parrish&Noble2003,Hanchar&Hoskin2003,Jacksonet al.2004).

The composition of zircon varies widely owingto solid solution with xenotime (Y, HREE), and thesubstitution Zr–1(Hf,Th,U)1. Many studies have beendevoted to zircon composition in a wide variety ofrocks(e.g.,Pupin2000,Caironiet al.2000,Rubattoet al.2001,Hoskin&Schaltegger2003,Fedoet al.2003,Belousova et al. 2002, 2006). Indeed, trace-elementcontentsandelementalratioshavebeensuggestedbothasindicatorofmagmafractionation(e.g., Černy et al.1985,Irberet al. 1997, Uher & Černy 1998, Wang et al. 2000, Koreneva & Zaraisky 2004, Breiter et al.2006),andasacrudeestimateofmagmasource(e.g.,Uheret al.1998,Hoskin&Ireland2000,Pupin1992,2000,Belousovaet al.2006,Loweryet al.2006).Sometraceelements(Hf,Th,U,Y,HREE)attainaveryhighconcentration in zircon from some felsic products offractionation (Wanget al. 1996,Heamanet al. 1990,Hoskin & Schaltegger 2003) and especially in somepegmatites(e.g.,Wanget al. 1996, Uher & Černy 1998, Seidleret al.2005).

In this paper, we study the composition of zirconfrom a group of Hercynian plutons of the SpanishCentral System (SCS) batholith, comprising S- andI-type granites with variable degrees of fractionation(Villasecaet al.1998,Villaseca&Herreros2000).Wealso compare the geochemical features of zircon in

zirconinthespanishcentralsystembatholith 511

marginal areas, aplitic topegmatiticbodies appear.ARb–Srisochronageof307±3Mawasreportedforthepluton(Pérez-Soba1992).

TwoS-typeplutonshavebeenselectedforthisstudy.The cordierite-bearingAlpedrete pluton, a typical S-typegranite,isanirregularmassifcoveringanareaofaround350km2(Fig.1).Itiscomposedofequigranularcordierite-bearinggranodioritetomonzogranite.Someattempts at Rb–Sr dating have failed, suggesting aheterogeneousisotopiccharacter(Villasecaet al.1995).Itisintrudedbyyoungercircularplutons,suchastheAtalayaRealI-typepluton.TheCabezaMedianaplutonisa two-mica leucogranitealso intrusive in theAlpe-dretepluton(Fig.1);itformsasmallcircularplutonof7km2andisdatedat291±6MabyRb–Srisochron.

AnalyticalMethods

Wavelength-dispersionelectron-microprobe(EPMA)analysesofzirconwerecarriedoutusingaJEOLSuper-probeJXA8900–Mequippedwithfourspectrometers

attheMicroscopíaElectrónicaCAI,oftheUniversidadComplutensedeMadrid.TheEPMAwasoperatedwithanaccelerationvoltageof20kVandaprobecurrentof 50 nA. Counting times were varied from elementto element, in the range 10–30 s (peak) and 5–15 s(background).Zircongrainswithtoolowananalyticaltotal(discussedindetailbelow)wereanalyzedagain,withabeamcurrentof150nAandcountingtimesforZr,Hf,REE,UandThextended to100s (peak)and50s(background).Underthelatterconditions,detec-tionlimits(calculatedfromthethree-foldbackgroundnoise)wereontheorderof100ppmformostoftheseelements.Thebeamdiameterwasbetween2and5mmtominimizedamage to thegrains.Natural standards,includingalbite,kaersutite,almandine,zircon,vanadi-niteandREEphosphates,wereusedformostelements(Jarosewichet al.1980, Jarosewich&Boatner1991)exceptUandTh, forwhichcommercialglasseswereused.An on-line ZAF correction program was used.Toexaminethecompositionalheterogeneityofzircon,back-scattered-electron (BSE) imaging was used.

Fig.1. GeologicalmapoftheSierradeGuadarrama,intheSpanishCentralSystem,showingthedistributionandtypeofgraniteplutons considered in this study (modified after Villaseca et al.1998).


Cathodoluminescence(CL)imagingalsoisapowerfultechnique to reveal internal textures of zircon (e.g.,Hanchar&Miller1993,Corfuet al.2003).Wefound,however, that the majority of our zircon samples arevirtually non-luminescent, which is why CL imagingwasnotappliedinthisstudy.

The degree of radiation damage of zircon wasestimated from the FWHM (full width at half band-maximum)ofthen3(SiO4)Ramanband(B1g-typevibra-tion)ofthismineral,whichisobservedatabout1000cm–1(Nasdalaet al.1995,2001a).RamanspectrawereobtainedbymeansofaRenishawRM1000research-

grade Raman spectrometer. This notch-filter-basedsystemisequippedwithaLeicaDMLMseriesopticalmicroscope,agratingwith1200groovespermillimeter,andaSi-basedcharge-coupleddevicedetector.Becauselaser-inducedphotoluminescencemay,asananalyticalartefact,affectRamanspectra,especiallyinthegreen-yellow range of the electromagnetic spectrum (e.g.,Nasdala&Hanchar2005), spectrawere excitedwiththe6328ÅemissionofaHe–Nelaser.TheLeica1003objective(numericalaperture0.95)wasused.Confocalmeasurements were done with a lateral resolution of


~2–3mm.Thewavenumberaccuracywas0.5cm–1,andthespectralresolutionwas~2cm–1.

PetrographicFeaturesandPatternofZoningintheZircon

WerecognizetwomaintypesofzirconintheSCSplutons studied, comparing both I- and S-type gran-ites, according to zoning and textural characteristics.Type 1: Euhedral to subhedral crystals exhibit finely concentric oscillatory zoning in BSE imaging.Thesecrystals commonly show an unzoned inner part andouter oscillatorybands (Figs. 2A toD).Wehavenotfoundaclearinheritedcore(residualzircon?),andonlyrarelydoesthistypecontaininclusionsofotherminerals(Fig.2C).About75%ofthezirconcrystalsareisolated,whereastheothersformclusterswithxenotime,thorite,monaziteorapatite.Type-1crystalsaremainlyincludedin biotite, occasionally in quartz or K-feldspar, andexceptionally they are hosted in plagioclase (4%).Thistypeofzirconappearsmainlyingranodioriteandmonzograniteplutons,butalsocanbefoundinminoramountsinleucogranites(Table2).

Type-2zirconformssubhedraltoanhedralunzonedorirregularlyzonedcrystals(Figs.2EtoI).About60%of these crystals form clusters with other accessoryminerals,suchasxenotime,thorite,apatiteormonazite.Thesecrystalsoccurmostlyasinclusionsattherimofthe biotite (less frequently than in type 1) and sodicplagioclase (20%),oras intergranularcrystalsamongquartz, the feldsparsorbiotite.Type-2zircon is typi-cally found in leucogranites, aplites and pegmatites,andislessabundantingranodioritesandmonzogranites(Table2).Threesubtypesaredistinguishedaccordingtotheirpatternofzoning.Type2a,thealveolarsubtype,is

themainone,characterizedbythepresenceofdomainsrichinmicro-inclusionswithhomogeneousorirregularzoning.Thesemicro-inclusionsmayextendalloverthegrain(inabout80%ofthecrystals)(Figs.2H,E),andmorerarelyarerestrictedtoaninnerdomain(Fig.2G)ortoanouterrim.Type2aistexturallysimilartozircondescribedbyRomanset al.(1975),Rubinet al.(1989)and Uher & Černy (1998). The micro-inclusions are irregularcrystals(usually<2mminlength)of thoriteorxenotime,orlargercrystalsofrock-formingminerals(biotiteandplagioclase),rangingupto20mminlength.Type-2azircon is includedmainly inbiotiteand, lesscommonly, in plagioclase. In plagioclase, the zirconcrystals are isolated, whereas in biotite they usuallyappear to be associated with xenotime, thorite andmonazite.Thistypeofzirconmainlyappearsinleuco-graniteandapliticgranite,beingespeciallywidespreadintheLaPedrizaleucogranite(Table2).

Type2bcrystalsarehomogeneous,unzonedorwithvery weak (diffuse) sector-zoning.The sections aresubhedral to euhedral. Crystals usually occur insidebiotiteandplagioclase,andcanbefoundbothisolatedcrystalsorgroupedwithxenotimeandthorite,orlesscommonlywithapatiteandmonazite.Thissubtypeismostcommoninmonzogranitesandgranodiorites.

Type 2c crystals constitute the complex subtype;the crystals exhibit areas with oscillatory, simple orpatchyzoning.Moreover,someoftheinteriorregionsmay be alveolar, suggesting transitions between allthesesubtypesofzircon(Fig.2I).IntheAtalayaRealaplites,someofthesecrystalsshoweccentricandstrik-ingly corroded cores (Fig. 2F), although commonlythedistinctsector-zoninghasaconcentricdistribution.Mostofthesetype-2ccrystalsarehostedinbiotite.

Zircon of types 1 and 2 has not been found inthe same host mineral, but could appear in the samethin section. These types of zircon also contrast interms of two morphological parameters, the longestdimension of the grain and the elongation (width /length). Geometrical features based on thin-sectionstudies underestimate the real greatest dimension ofthezirconcrystalsandalso theirelongationratio,butfor comparativepurposes, thesemaybe considered agoodestimate.Figure3showsgeometricalparametersfor the two types of zircon and for the three mainsubgroups of lithologies.Although in some cases thenumberofdatapoints is lowowingtothescarcityofzirconinthesamples,sometendenciescanbededucedusingsuchagraphicalanalysis,asdiscussedinHiggins(1994).Thefrequencydistributionofbothparametersin type-1 zircon from monzogranite and granodioriteplutons (Fig. 3) is indicative of elongate and largecrystals.The type-2 zircon from leucogranites showsmoreequantandshortersectionsthanthoseoftype-1zircon,althoughavalueof250mmforasingle-crystallengthwasestablished.Moreover,theshortestcrystalsof the zircon population (≤20 mm) are present in theLa Pedriza leucogranite.Type-2 zircon in aplites is


Fig.

2.

Bac

k-sc

atte

red-

elec

tron

imag

eso

fdiff

eren

ttyp

esa

ndsu

btyp

eso

fzirc

ond

istin

guis

hed

inth

isw

ork.

Typ

e-1

zirc

on:A

)Elo

ngat

ecr

ysta

lsho

win

gos

cilla

tory

zon

ing

(Ata

laya

R

ealg

rano

dior

ite,s

.959

22).

B)C

omm

onc

ryst

alo

ftyp

e1

(La

Pedr

iza

leuc

ogra

nite

,s.6

7065

).C

)Euh

edra

ltra

nsve

rsal

sec

tion

ofty

pe-1

zirc

on,w

ithc

hara

cter

istic

osc

illat

ory

zoni

ng.D

)Sub

hedr

als

ectio

nof

type

-1z

ircon

,sho

win

gos

cilla

tory

zon

ing,

wea

kly

cont

rast

ed.T

ype-

2zi

rcon

:E)S

ubty

pe2

b(a

lveo

lar)

:com

mon

cry

stal

ofa

lveo

lars

ubty

pe,

inth

isc

ase

with

irre

gula

rinc

lusi

ons

ofp

lagi

ocla

se.T

hea

rrow

poi

nts

toth

esp

oto

fthe

com

posi

tion

with

SiO

2=3

1.65

wt.%

(Tab

le3

);(L

aPe

driz

ale

ucog

rani

te,s

.870

59)F

)Su

btyp

e2c

(com

plex

):cr

ysta

lwith

alv

eola

rcen

ters

how

ing

ane

ccen

tric

nucl

eus

(arr

ow)w

itha

com

posi

tion

SiO

2=

24.8

3w

t.%in

Tab

le3

,in

aw

ide

unzo

ned

cent

rals

ecto

r(A

tala

yaR

eala

plite

,s.9

6922

).G

)Sub

type

2b

(alv

eola

r):z

oned

cry

stal

with

an

alve

olar

cen

ter(

rich

inb

right

and

opa

que

incl

usio

ns,c

ompo

sitio

nw

ithS

iO2=

28.

84w

t.%in

Ta

ble

3)a

nda

hom

ogen

eous

rim

(SiO

2=3

1.35

wt.%

inT

able

3)(

LaP

edriz

ale

ucog

rani

te,s

.872

25).

H)S

ubty

pe2

b(a

lveo

lar)

:euh

edra

ltra

nsve

rsal

sec

tion

ofz

ircon

rich

in

irreg

ular

incl

usio

ns; a

rea

(mar

ked

by th

e ar

row

) with

a d

efici

ent a

naly

tical

tota

l (co

mpo

sitio

n w

ith S

iO2=

28.

76w

t.%in

Tab

le3

)(La

Ped

riza

leuc

ogra

nite

,s.8

7166

).I)

Sub

type

2c

(com

plex

):zo

ned

crys

talw

itha

smal

lalv

eola

rsec

tori

na

mai

nly

hom

ogen

eous

inne

rzon

ean

da

rimw

ithsu

btle

osc

illat

ory

grow

th-z

onin

g;th

ebr

ight

cry

stal

isx

enot

ime

(La

Pedr

iza

leuc

ogra

nite

).


mostly equidimensional, but isnot shorter than thosebelonging to leucogranites. This tendency clearlychangesfortype-2zirconfromLaCabrerapegmatites(not includedinthesediagrams),whichreachlengthsup to800mm,withelongationratiobetween0.1and0.2(acicularvarieties).

CompositionandStructuralStateofZirconintheGraniticRocks

Nearly 125 EPMA analyses of type-1 zircon andabout240analysesoftype-2zirconweremade.Repre-sentativeEPMAresultsareshowninTable3.InternalzoningismainlyrelatedtovariationsofHf,Y,HREE,UandThconcentrations;theseelementsarethereforeconsidered to be the main indicators of degree ofevolutionanddifferencesamongtypesofzircon.ThecontrastinthecompositionalrangeoftypeofzirconisexpressedinFigure4accordingtoplutonsanddegreeofdifferentiation.

In granodiorite and monzogranite, both types ofzircon define low values in the oxides considered,whereasthemostfelsicvarieties(leucogranites,aplitesand pegmatites) have higher medians and reach thehighest contents of Hf,Y, HREE,Th and U.This isdisplayed by I-type granites and, to a lesser degree,by S-type granites, perhaps related to the lack ofaplite–pegmatitedykesintheseplutons.

Type-1zirconusuallyhasthelowestcontentsinU,Th,Y,HREEandHf,anddisplaysanarrowercompo-sitional range than type-2 zircon.The compositionalrange of type-1 zircon is independent of their host(leucogranitetogranodiorites,Fig.4),asfoundinotherstudies (Wang et al. 2000). Compositional ranges intype-1zirconare(inwt.%):0.51–5.75%HfO2,0–4.03%UO2,0–1.65%ThO2,0–4.12%Y2O3and0.21–2.35%HREE2O3. The zoning defined in type-1 single crystals

isdifferentinthelessandthemoreevolvedgranites.Inzirconfromgranodioriteandmonzogranitesamples,apositivecorrelationofHf,U,Th,YandHREEconcen-tration isusuallyobserved,whichgenerally increasesfrom core to rim, whereas in the evolved granites,levelsofU,Th,YandHREEdecreaseandHfincreasetowardtherim.

Thetype-2zirconfromleucograniteandaplitehasamorevariablecomposition(Fig.4),withhighercontentsof HfO2 (13.60 wt.%), UO2 (4.83%),ThO2 (4.87%),Y2O3 (8.51%)andHREE2O3 (3.78%)and the lowestanalyticaltotals(91.17wt.%)(Table3).ZirconfromtheLaCabrerapegmatitesshowsthehighestHfO2contents(16.58wt.%).Intype-2zircon,thecompositionalrangesamongthethreesubtypesingeneralshowagreatoverlap(Table3),althoughtype-2bzirconfromtheLaCabrerapegmatitesshowsthehighestHfcontent,andtype-2czirconfromtheAtalayaRealaplites(Fig.4)showsthehighestTh,YandHREEcontents.Alveolar subtype-2aisricherinThandpoorerinHfinthoseareaswiththoritemicro-inclusions.Abouthalfofthepopulationofzoned type-2 zircon displays significant covariation in U,Th,YandHREEcontents(similartothatdescribedby Pointer et al. 1988,Wang et al. 2000, Mathieu et al.2001,Fowleret al.2002),whereas inmostcases,Hfdisplaysanoppositebehavior,whichcontrastswithfindings in other studies (e.g., Uher & Černy 1998, Wanget al.2000,Belousovaet al.2002).

The compositional zoning in subtype-2 zirconusually shows the same trend, defined by a decrease in levelsofU,Th,Y,HREE toward rim,whereasHfincreases.Nevertheless,alveolarcrystalsareirregularlyzoned;onlywherethealveolardomainisrimmedbyahomogeneousgrowth,theinnerzoneisricherorsimilarin content to the rim (Fig. 2G). Subtype 2b usuallyshows the general compositional path defined for type 2, showingthelargestrangeofUcontents.Insubtype-2c

Fig.3. Histogramsoflength(Lmax)andelongation(width/length)inzirconoftype1and2.Foreachcategory,thepercentageofthemainrockstothetotalofthepopulationiscompared.


zircon,thecorrodedcoresfromtheAtalayaRealaplitesshowthehighestTh(4.7to4.9wt.%ThO2),Y(7.6to8.5wt.%Y2O3)andHREE(3.7to3.8wt.%HREE2O3)contents,whereasHfincreasestowardtherim.ThisHfincreasetowardtherimofasinglecrystalofzirconiscommon inourcollectionof typesand is alsoexten-sively recognized in other studies (e.g., Hanchar &Miller1993,Uheret al.1998,Pupin2000).

Transgressiverimsintype-2zircondonotmodifythe common zoning defined for this type of zircon; only Ca andAl are appreciably lost.This contrastswith the loss of trace elements in transgressive rimsdescribedbyPidgeonet al.(1998),andalsobyGeisleret al. (2003a)inexperimentsofzirconalterationatlowtemperatures.


awell-crystallizedsyntheticcrystalofzircon.Wefoundthatourzirconsamplesdogenerallyshowhighlevelsof radiation damage, which is indicated by FWHMvalues[n3(SiO4)Ramanband]exceeding30cm–1,andlow band-intensities (Fig. 6). Such extensive band-broadeningisconsideredtocorrespondtoanamorphousvolume-fractionexceeding60–70%(Zhanget al.2000,Zhang&Salje2001),whichcharacterizesourareasoflowanalyticaltotalasstronglymetamict.

Discussion

Cation substitutions

Zircon conforms to the general formula ABO4,where the position A represents the relatively largezirconium ion in eight-fold coordination with O, andpositionBrepresentsthesiliconionintetrahedralcoor-dinationwithO.InpositionA,Zrcanbereplacedbytetravalent(M4+=Hf,Th,U),trivalent(M3+=REE,Y,Fe),anddivalentcations(M2+=Ca,Fe,Mg,Mn).AssummarizedbyHoskin&Schaltegger(2003),therearemultiplemechanismsofsubstitution,includingsimpleisovalentreplacementofZrbyothertetravalentcations,andcoupledsubstitutionswheredivalentandtrivalentcationsmaybeincorporated.

Fig.4. Compositionalcomparisonbetweenzirconoftypes1and2fromrepresentativeSCSgraniticplutons.Barshowsthecomplete range of zircon composition (vertical line), including the median (short horizontal line). Oxides expressed inwt.%.

Figure 5 shows the different Th/U paths in ourzircon.Mostzircongrainsfrommonzograniteorgrano-diorite(irrespectivelyoftheirzircontype)showaTh:Uratioofapproximately0.1.Type-1zirconingrano-dioriteandmonzograniteshowsamorerestrictedrangeofvaluescomparedwiththezircon(types1and2)inmoreevolvedgranites.Zirconof these felsicgranitesusuallyshowahigherTh:Uratio,whichseems tobepluton-dependent.Thus,intheLaPedrizapluton,mostofthezirconcrystals(irrespectivelyoftype)alsoshowaTh:Uratioclose to0.1,whereas type-1zirconfromtheLaCabreraaplite showsaTh:Urationear1, andtype-2czirconfromtheAtalayaRealapliteshowsaTh:Urationear2(Fig.5).

Shortfalls in analytical totals (considered as lessthan98wt.%)arefoundinabout30%oftype-1zirconpopulation and in 50% of the type-2 population; thelatter population shows the lowest analytical totalsfound.The mean of the analytical totals of corrodedcoresofsubtype2cshowslowervalues(96.11wt.%)than types 2a (~97.7 wt.%). In these crystals, thestructuralstateofzircon,i.e.,thedegreeofmetamicti-zation,wasstudiedbymeansofRamanspectroscopy(Nasdalaet al.1995).Figure6showsatypicalRamanspectrumasobtainedinthisstudy,incomparisonwithtworeferencespectraofthepoorlyradiation-damagedCurtinUniversity(Perth)ion-probestandardCZ3and


Solidsolutiontowardxenotimeplaysanimportantrole in zircon composition. The proportion of M3+cationsattheAsitemustbesimilartotheproportionofP5+attheBsite,accordingtothecoupled-substitutionmechanism(Y+HREE)PZr–1Si–1(e.g.,Speer1980,Hanchar et al. 2001).The close correlation betweenZr + Hf andTh + U allows us to consider total M4+cations(Zr+Hf+U+Th)insteadofZr(ashasbeenproposedbyFowleret al.2002).Inthatway,mostofthezirconsamplesplotalongthevector1:1representativeofA-sitesubstitutionforM3+cations(Fig.7A).Ontheotherhand,thecoupled(P+Al)forSisubstitutionplotsclosertothe1:1linethanPalone,whichissuggestiveofpreferentialAlincorporationatthissite(Fig.7B),asstated by Uher & Černy (1998) and Akhtar & Wassem (2001).This substitution is shown by all the groupsof zircon, although some type-2 zircon samples aresituatedbelowthisline,indicatinginthesecasessomedeficiency in P.

Divalent cations like Ca, Fe and Mn also may bepresent in zircon, occupying interstitial sites.TheseM2+ cations have been considered as involved in thexenotime–zircon series, providing a mechanism ofcharge balance for the HREE andY in excess of P:(Mg, Fe)2+

(int) + 3(Y, HREE)3+ + P5+ = 3Zr4+ + Si4+(Romans et al. 1975, Hoskin et al. 2000).This fact

reflects the inability of the Si site to incorporate enough Ptocharge-balancetheREE3+ions(Finchet al.2001).Ontheotherhand,studiesofalterationinzirconshowthat these elements (speciallyCa andFe, but alsoAland Ti) may be secondarily enriched in the alteredzones of zircon crystals (Geisler & Schleicher 2000,Mathieuet al.2001,Zhanget al.2003,Geisleret al.2003a,b,Johan&Johan2004,Kempeet al. 2004).Forexample,Geisleret al.(2003a)studiedexperimentallythe interaction of metamict zircon with hydrothermalfluids, and they observed a gain of solvent cations(e.g.,Ca,Ba)fromthehydrothermalsolution,andthelossofvariableamountsofM4+cations(Zr,Si,Hf,Th,U)aswell asM3+ cations (REE).On theotherhand,theREEincorporated inzirconhavebeenconsideredessentiallyimmobilebyCherniaket al.(1997).Inthissense, significant correlations of REE versus (Ca +Fe) contents may indicate a primary (i.e., magmatic)originofthedivalentcationsCa2+andFe2+.Figure8Ashowssuchcorrelationformostoftheplutonsstudied,except for someanalysesofmonzogranitesor aplites

Fig.5. VariationinamountofUversusThinzirconcrystals.DiagonallinesrepresentingTh:Uratiosof0.1,0.6and2are shown.The compositional rangeof zirconof type1andtype2frommonzogranitesandgranodiorites(MzGr),andzirconoftype-1fromleucogranitesandaplites(LAP)are enclosed in areas. Arrows define the compositional zoninginasinglecrystal.Symbolsfor type-2zirconareasfollows:✩LaCabrerapegmatite,+LaCabreraaplite, 3 AtalayaRealaplite,AtalayaRealleucogranite,LaPedrizaleucogranite,CabezaMedianaleucogranite.

Fig.6. Ramanspectrumofamicro-areainzirconinaplitefromtheAtalayaRealpluton,whoseelectron-microprobeanalysis gave a deficient total (96027c in Table 3). It is shownincomparisonwithspectraofthemoderatelyradia-tion-damagedCZ3zircon(Nasdalaet al.2001b)andwell-crystallized,syntheticzircon(Nasdalaet al.2002).Spectraare stacked for clarity. Broadening and extremely lowintensitiesofRamanbandscharacterizesample96027casstronglyradiation-damaged.


andagroupofanalysesof type-1and-2zirconfromLaPedriza.Inthesecases,partoftheCa+Fecontentscouldbegainedinthealterationprocesses.Exceptforthesegroupsofcompositions(around40),weconsiderthat Ca and Fe are mainly magmatic in origin.Thisargumentissupportedwherethexenotimesubstitution

isrepresented,includingallM4+andM2+cationsattheAsite(Fig.8B).Inthisplot,mostofthezircondatafallalongthe1:1line,suggestingthattheincorporationofM2+cationsisrelatedmainlytoamagmaticmechanismofsubstitution.

Fig.7. DiagramsshowingtheextentofsolidsolutiontowardxenotimeinzirconcrystalsinSCSgranites(inatomsperformulaunit:apfu).A)A-sitesubstitution:(Zr,Hf,Th,U)4+

versus(Y,REE)3+diagram.B)B-sitesubstitution:Si4+versus(P5++Al)

plot.Areaencloseszirconcrystalsoftypes-1and2frommonzogranitesandgranodiorites,andthediagonallineindicatesthe1:1substitutionline.SamesymbolsasinFigure5.

Fig.8. Zirconplots.A)(Ca2++Fe2+)versusREEplot.B)3(Zr,Hf,Th,U)4+versus(Ca,Fe)2++3(Y,REE)3+substitutiondiagraminzirconoftypes1and2.Thediagonallineindicates1:1substitutionline.Thecompositionsareexpressedinatomsperformulaunit(apfu).SamesymbolsasinFigure5.


Some differences in the degree of replacement ofZrbyM2+,M3+andM4+cationsseemtodependupontheirhost-granitetype.Thus,zirconfromS-typegran-itesdoesnotshowappreciablesubstitutionof(Th+U)forZr,andthedegreeofanycationsubstitutionislessimportant.Adecrease inThcontent isa typical trendinS-typegranites,contrary to that inI-typevarieties,which combined with their higher P content, favorsxenotimecrystallizationinsteadofThincorporationinzircon(Villasecaet al.1998).

Zircon growth: implications from their textural and chemical characteristics

Textural and chemical contrasts between the twotypes of zircon and their different modal proportionswith granite evolution suggest differences in growthconditions. Oscillatory zoning in zircon has beenintensively studied (e.g.,Vavra 1994, Hoskin 2000),and a magmatic origin is suggested. Type-1 zirconseemstohavecrystallizedduringchangingmagmaticconditions,andisprobablynotanearlyliquidusphase(itrarelyappearsincludedinthecoreofplagioclase).

Ontheotherhand,mostcrystalsoftype-2zirconlikelygrewatalate-magmaticstagebecausetheyshowsomefaces(i.e.,aresubhedral)andtheygrewsimultaneouslywith themajorphases (e.g.,biotite, the feldsparsandquartz).

The Th-, U-, Y- and HREE-rich zircon is confined mainly to type-2 crystals from leucogranites, aplitesandpegmatites.Itscrystallizationinthemostevolvedgranitic magmas, which show enrichment in theseelements,explainswhyitscompositiondepartsprogres-sively from the ideal one.At the same time, theTh,U,Y, HREE and Hf contents are broadly correlatedwith the whole-rock composition (Fig. 9).Thus, themost-enrichedzirconappearsinthemostfractionatedgranites.

Asthelevelofincompatibletrace-elementsincreasesin felsic granitic melts, zircon incorporates otherelements in addition to Hf.The decrease ofTh–U–Y–HREE contents usually shown by type-2 zirconwouldsuggestcocrystallizationwithaccessoryphasesrich inYandHREE (garnet, xenotime),or rich inUandTh(uraninite, thorite,monazite,allanite).Indeed,the presence of micro-inclusions of thorium-bearingmineralsinthecoreareaofthesezirconcrystals(Fig.2G) suggests an environment saturated inTh-bearingphases (Černy & Siivola 1980). The cocrystallization ofthoritecontributestoamarkeddecreaseofthetotalTh content in the outer part of the zircon crystals,althoughtoalowerdegreethanstatedinotherstudies(Lumpkin&Chakoumakos1988,Wanget al. 1996),asspotanalysesin“alveolar”sectorsgivestillhigherTh contents (up to 1.82 wt.%).The various patternsof trace-element zoning in zircon may also reflect the delicatebalanceinthemoreimmediateenvironmentofcrystallization(e.g.,Rubinet al.1989,Heamanet al. 1990,Wanget al.1996),inviewofcompetitionwithotheraccessoryphases.

Finally,postmagmaticevents(hydrothermalaltera-tionofmetamictzircon)mayintroduceCa,Fe(Geisler& Schleicher 2000), together with U (Kempe et al.2004)andwithHfandY(Pointeret al.1988,Uheret al.1998).Aswehaveseenabove,thepositivecorrelationbetweensomeofthesecations(inourstudy,Ca,FeandMg)andthoseconsideredasbeingmagmaticformostofplutons(Fig.8A)suggeststhatthesecationsmaybemainlyincorporatedatthemagmaticstage.

The question of low analytical totals

Intheplutonsstudied,zirconcrystalsormicro-areaswithincrystalswithlowanalyticaltotals,intherangeof 91–98 wt.%, have been found.This was observedinbothtype-1andtype-2zirconsamples,withtype-2zirconshowingthephenomenonmorecommonly.Suchdomains (especially in the core and inner regions ofzonedtype-2crystals)revealthehighestcontentsofY,HREE,UandTh(Table3).

Fig.9. PlotofThinwholerockversusamountofThO2inzircon fromgranitesof theAtalayaRealandLaPedrizaplutons.DatapointsrepresentthecompositionalrangeofThO2 contents in zircon crystals from a single thin-sec-tion.


Zircon samples yielding such unusually lowanalytical totalsseemtooccurquiterarely.Exampleshavebeendescribedfromseverallocalitiesworldwide(Petermanet al.1986,Pointeret al.1988,Smithet al.1991,Irberet al.1997,Kempeet al.2000,Geisleret al.2003b,Breiteret al.2006).Lowtotalshavemostlybeenreported from altered and significantly metamict zircon. Ashasalreadybeendiscussed, the“low-total”zirconiscommonlycharacterizedbyaparticularlylowBSEintensity(Kempeet al. 2000), which was confirmed in ourmaterial(Fig.2).

CausesoftheanalyticalshortfallandtherelatedlowBSEintensitiesarestillhotlydebated. Inmostof theabovepapers,authorshaveattemptedtoexplainthelowtotalsby the incorporationofhydrousspecies (whicharenot analyzed in the electronmicroprobeandmayinadditioncausephasedegradationundertheelectronbeam).Alternatively,enhancedquantitiesofvacancies(Kempeet al.2000)andsubmicrometricvoids(Pointeret al. 1988) have also been discussed as potentialcauses.We consider that the phenomenon cannot beexplainedbythehighcontentofhydrouscomponentsalone.Detailed studiesonhydrous species incrystal-line and metamict zircon indicate that zircon rarelyincorporatesmorethan1wt.%ofhydroxylgroupsorH2Omolecules(Woodheadet al.1991,Nasdalaet al.2001a), which is clearly insufficient to explain our low analytical totals. We found deficiencies up to 9 wt.% (Table3),whichwouldcorrespondtoanelevatedH2Ocontentexceeding30mol.%;thisappearsunlikely.Ontheotherhand,samplesdescribedintheabovepapersaddressing the question of hydrogen incorporation inzirconhavenotbeenaffectedbyextensivesecondaryalterationprocesses,andmaythereforenotbedirectlycomparable to our samples; this limits any conclu-sionsbyanalogy.Consequently,micro-analysesofthesamples for hydrogen in zircon showing deficient totals needtobedonebeforethisproblemcanberesolved.

Using the equation of Murakami et al. (1991),time-integrateda-doseswerecalculatedforthezirconsamples studied from present U and Th concentra-tions and assuming a self-irradiation period of ~300million years.This value is based on the Hercynianage of emplacement of the batholith (Villaseca &Herreros2000)andtheassumptionthatsampleshavenot experienced significant annealing since. Results suggest that most areas of the crystals yielding lowanalytical totals, which are rich in actinides (UO2 +ThO2>1wt.%;Table3),haveexperiencedself-irradia-tiondosesexceeding8 3 1018a-eventspergram.Suchhigha-fluences are definitely sufficient to cause severe radiation-induceddamage(Holland&Gottfried1955,Murakamiet al.1991,Nasdalaet al. 2001a).Thiswasconfirmed by Raman microprobe analyses: Raman band FWHM >30 cm–1 (Fig. 6) characterize these micro-areasasstronglyradiation-damaged.

However,itisalsonotpossibletosimplyrelatethelow totals to enhanced degrees of radiation damage.

There are many examples for strongly radiation-damaged, up to virtually amorphous, zircon samplesthatgave“normal”electron-microprobetotalsof~100wt.%(Zhanget al. 2000,Nasdalaet al.2002).Never-theless, these non-deficient samples of metamict zircon haveextremelylowY,REE,CaandFecontents.Onthecontrary,mostofthecasesofmetamictzirconfromtheSCSfelsicgranitesarerichinUandTh,butalsorichintrivalentcations(Y,HREE)anddivalentcations(Ca,Fe,Mn).Theincreaseinextentofcationsubstitutionsinzirconisclearlyrelatedwiththemarkeddecreaseofanalytical totals: the greater the sum ofThO2 + UO2(Fig.10A),andoxidesoftrivalentanddivalentcations,thelowertheanalyticaltotal(Fig.10B).Thelowanaly-ticaltotalsthusareobtainedinareasofzirconcrystalsnotonly rich inactinides,which triggermetamictiza-tion, but also with a high abundance of heterovalentcationswhich,insomeway,mayenhancethedisorderedstructureattainedbythezircon,assuggestedbyKöppel&Sommerauer(1974).Inthisway,Rioset al.(2000)observed that high concentrations of point defects inradiation-damagedsamplesstronglyaffectthestructuralpropertiesofcrystallinezircon.Indeed,theabundanceofthesenon-tetravalentcationsthatarenotcorrelatedwithPsubstitutioninthezirconstructurewouldrequirealso, to some degree, oxygen defects, which maycontributetothelowanalyticaltotals.

Zr/Hf values of zircon as a fractionation index of the magma

Heaman et al. (1990) considered that changes inzircon composition must be a function of changes inmelt composition and degree of fractional crystal-lization.Asa typicalsubstitutionisHfZr–1, theZr:Hfratio has been used either in matters of terminology(e.g., Correia Neves et al. 1974) or as an index ofmagmatic differentiation (Černy et al. 1985). Manyauthors(Wark&Miller1993,Irberet al.1997,Wanget al. 2000) have pointed out that the concentrationofHfinzirconisvariable,butdependsmainlyonthedegreeofevolutionandthecompositionoftheoriginalgraniticmagma.ThegreatvariationinZr/Hfvaluesinzircon from subaluminous to peraluminous membersof plutonic suites reveals a decoupling of Zr and HfinthedirectionofenrichmentinHfwiththedegreeofdifferentiation of the host granite (Linnen &Keppler2002,Loweryet al.2006).

In the diagram of Černy et al. (1985), the zircondisplays an asymptotic curve (Fig. 11) inwhichbothtypesofzirconareconcentratedatintermediatevalues(70 – 25) in terms of Zr:Hf ratio.Type-1 zircon ofgranodiorites and monzogranites have Zr/Hf valuesfrom110to25;inthatofleucogranite,itvariesfrom80tolessthan10,andincrystalsfromaplite,therangeis 40 – 30.The range of Zr:Hf in granodiorites andmonzograniteoftype-2zirconisshorterandlower(63–31)thanintype1,andtheoppositeisobservedforthe


moreevolvedgranites(53<Zr/Hf<3inleucogranites,and 71 < Zr/Hf < 14 in aplites).Type-2 zircon fromtheLaPedrizaleucograniteandLaCabrerapegmatiteshasthelowestZr/Hfvalues(10–2.4)(Fig.11).Thereare no significant differences in Zr/Hf between zircon of I- and S-type granites, as is also the case whencomparingI-andA-typegraniteswithasimilardegreeofdifferentiation(Wanget al.2000).

We believe that the ratio Zr:Hf in zircon may beonlyacrudeindexofmagmaticdifferentiation.AsZrin zircon is also replaced by other elements (e.g.,Y,HREE,ThandU)andasHfmayincreaseordecreasein a zoned single crystal of the same rock, theZr:Hfratiowilldisplayalargevariation,evenatthescaleofasinglecrystal.Forinstance,insomegrainsofzirconfromtheLaPedrizaleucogranite,Zr/Hfvaluesdecreasefrom 14 (core) to 2.4 (rim) (Table 3, no. 67263). IntheAtalayaRealpluton,asingletype-2zirconcrystalshowsasimilarrangeofZr/Hfvalues(62–17)(Table3,no.96027)asthewholepopulationofzirconcrystalsintheplutonicbody.Furthermore,inasamethinsectionfromtheLaPedrizaleucogranite(Table3,no.87165),type-2 zircon shows a Zr/Hf range of 49.4 to 4.6, aslargeasthevariationfoundattheplutonscale.

Zircon composition and magma sources in the crust

InthecompilationofzirconcompositionsmadebyPupin(2000),thesignatureofzirconinthecontinentalcrust ismarkedbyaZr/Hfvalue in the range36–45,which differs significantly from zircon in mantle-

derivedrocks,withZr/Hfaround60–68,orevenhigherifbasicandundersaturatedalkalinerocksareincluded(Pavlenko et al. 1957, Erlank et al. 1978). In Figure12, we compare the previous Zr/Hf values collectedinFigure3ofPupin(2000)withthosefromdifferentperaluminous metamorphic rocks (migmatites, ortho-gneisses,granulitexenoliths)ofcentralSpain(Villasecaet al.2003)andthosefromthemostprimitivemembersof the S- and I-type granites studied. Following thecriteria of Pupin (2000), we have excluded composi-tions from the rare-element-rich rim in fractionatedgranites.We obtain averaged Zr/Hf values of 46 forlower-crustgranulitexenoliths,49formigmatitesfromthe metasedimentary anatectic complex of Toledo,39–45formigmatitesfromSCS,and56forSCSaugenorthogneisses. Consequently, purely crustal sourcesseemtohaveaslightlyhigherZr/Hfratio,intherange36–56, rather than 36–45 as previously suggested(Pupin2000).

TheparentalmagmasoftheplutonicsuitesstudiedhaveZr/Hfvaluesinthesamerangeasthat inzirconfrom crustal sources, and closely match those valuesofzirconcrystalsfromthepreviouslysupposedrestitickeel of the SCS batholith, the lower-crust granulitexenoliths(Villasecaet al.1999).Thechemicaldataonzircon are in accordance with previous geochemicaland isotopic arguments for theminorparticipationofmantle-derivedcomponentsinthegenesisoftheperalu-minousSCSgraniticrocks(Villasecaet al.1998,1999,Beaet al.1999,Villaseca&Herreros2000).

Fig.10. A)Plotof(Th+U)versus(Y+HREE)forzirconoftypes-1and2fromalltheplutonsstudied.B)Analyticaltotalversus(Y2O3+HREE2O3)+(CaO+FeO)inwt.%inzirconcrystals.Seetextforfurtherexplanations.SamesymbolsasinFigure5.


Conclusions

1. Two types of magmatic zircon crystals havebeendistinguished.The type-1zirconappearsmainlyas isolated crystals in the least-evolved magmas. Itexhibits oscillatory zoning in large idiomorphic crys-tals,anditscompositionisclosetothatofpurezircon,withincreasingHf,Th,U,YandHREEcontentsfromcore to rim.The second type of zircon predominatesin the most evolved magmas, as equant and smallercrystals,usuallyformingclusterswithotheraccessoryphases.Texturally,thecrystalsaremainlyalveolarwithmicro-inclusions of other accessory minerals. Suchtype-2zirconshowshighercontentsofHf,Th,U,YandHREE,withadecreasingtrendofTh,U,Y,HREEandM2+fromcoretorim,associatedwithanoppositeHftrend.Type-2zirconisusuallymetamictandcharacter-izedbyalowanalyticaltotal.

2.Mechanismsofsubstitutionmaybebroadlycorre-latedwiththedegreeofmagmadifferentiation.Inthemostfelsicgranites,theZrsiteinzirconispreferentiallyoccupiedby(Th+U)ratherthan(Y+HREE).Divalentcationsarealsoprogressivelyintroduced,andthetotaldegree of substitution is high. Most zircon crystalsshowthedominant“xenotime”substitution,especially

zirconfromS-typegranites.ZirconcrystalsfromsomeofthemostevolvedI-typegranitesshowahighdegreeof(Th+U)incorporation.

3.MetamictzirconfromtheSCSgranitesshowsalowanalytical total,mostly in the range91–98wt.%.The deficiency of analytical totals correlates with the contentof“non-formula”elements(especiallythesumofM2+andM3+cations).TheincorporationofH2Oinzircon is considered insufficient to explain such low analyticaltotals.

4.Thecompositionofzirconisbroadlycorrelatedwith thatof theirhostgranitecomposition.Neverthe-less,theuseofZr:Hfratioasanindexofmagmafrac-tionationshouldbeusedwithcaution.

5.ThemostprimitivegranitesstudiedhavezirconZr:Hfvalueswiththesamerangeasthatofzirconfromcrustalsources(e.g.,residualgranulites),supportingthehypothesis that the peraluminous SCS batholith wasderivedmainlyfromrecycledcrustalrock-types.

Acknowledgements

The authors thankAlfredo Fernández Larios forhis assistance with the electron-microprobe analyses,and J.M. Caballero for providing microprobe data

Fig.11. ThecompositionofzirconintermsofHf(wt.%)versusZr:Hf.Inbothdiagrams,theenclosedareaisthecompositionalfield of zircon crystals in granodiorites and monzogranites. Compositional growth-zoning (core to rim) is also marked for somezirconcrystalsoftype2.SamesymbolsasinFigure5.


forsomesamplesfromtheLaPedrizapluton.WeareindebtedtoJohnCobbingforhisdetailedcommentsona preliminary version of the manuscript, and to JohnM.Hancharandananonymousexpertformosthelpfulreviews.ThanksareduetoRobertMartinforeditorialhandling and his constructive suggestions.This workhas been financially supported by the projects DGYCIT BTE2000–0575andCGL2004–0575oftheMinisteriodeEducaciónyCiencia,SpanishGovernment.

References

Akhtar, M.J., &Wassem, S. (2001):Atomistic simulationstudiesofzircon.Chem. Phys.274,109-120.

Bea,F.(1996):Controlsonthetraceelementcompositionofcrustalmelts.Trans. R. Soc. Edinb.: Earth Sci.87,33-41.

Bea,F.,Montero,P.&Molina,J.F. (1999): Mafic precur-sors, peraluminous granitoids, and late lamprophyres intheAvilaBatholith:amodelforthegenerationofVariscanBatholithsinIberia.J. Geol.107,399-419.

Bellido, F. (1979): Estudio petrológico y geoquímico del plutón de La Cabrera. Ph.D. thesis, Universidad Com-plutensedeMadrid,Madrid,Spain.

Belousova, E.A., Griffin,W.L. & O’Reilly, S.Y. (2006):Zircon crystalmorphology, trace element signatures andHf isotope composition as a tool for petrogenetic mod-elling: examples from EasternAustralian granitoids. J. Petrol.47,329-353.

Belousova, E.A.,Walters, S., Griffin,W.L., O’Reilly,S.Y.&Fisher,N.I.(2002):Igneouszircon:trace-elementscomposition as indicator of source rock type. Contrib. Mineral. Petrol.143,602-622.

Breiter, K., Förster, H.-J. & Škoda, R. (2006): ExtremeP-,Bi-,Nb-,Sc-,U-andF-richzirconfromfractionatedperphosphorousgranites:theperaluminousPodlesígranitesystem,CzechRepublic.Lithos88,15-34.

Caironi,V.,Colombo,A.,Tunesi,A.&Gritti,C.(2000):Chemicalvariationsofzirconcomparedwithmorphologi-calevolutionduringmagmaticcrystallization:anexample

Fig.12. ComparisonoftheZr:Hfvaluesinzirconofdifferentcrustalsourcesandtypesofgranite.Boxplotsshowthe25thand75thpercentiles(limitsofthebox),themedian(thinline),themean(coarseline)andthemaximumandminimumvaluesexcludingoutliers.ValuesforSCSrocksaretakenfromthisstudyandfromVillasecaet al.(2003);othervaluesaretakenfromPupin(2000).


from theValle del Cervo pluton (WesternAlps). Eur. J. Mineral.12,779-794.

Černy, P., Meintzer, R.E. &Anderson,A.J. (1985):Extremefractionationinrare-elementgraniticpegmatites:selectedexamplesofdataandmechanisms.Can. Mineral.23,381-421.

Černy,P.&Siivola,J.(1980):TheTancopegmatiteatBernicLake, Manitoba. XII. Hafnian zircon. Can. Mineral. 18,313-321.

Cherniak, D.J., Hanchar, J.M. &Watson, E.B. (1997):Rare-earth diffusion in zircon. Chem. Geol. 134, 289-301.

Corfu, F., Hanchar, J.M., Hoskin, P.W.O. & Kinny, P.(2003):Atlasofzircontextures.InZircon(J.M.Hanchar& P.W.O. Hoskin, eds.). Rev. Mineral. Geochem. 53,469-500.

CorreiaNeves, J.M.,LopesNunes, J.E.&Sahama,T.G.(1974): High hafnium members of the zircon–hafnonseriesfromthegranitepegmatitesofZambezia,Mozam-bique.Contrib. Mineral. Petrol.48,73-80.

Erlank,A.J., Smith, H.S., Marchant, J.W., Cardoso,M.P. &Ahrens, L.H. (1978): Hafnium, In HandbookofGeochemistry(K.H.Wedepohl,ed.).Springer,Berlin,Germany(72D-E).

Ewing,R.C.,Meldrum,A.,Wang,Lumin,Weber,W.J.&Corrales,L.R.(2003):Radiationeffectsinzircon.InZir-con(J.M.Hanchar&P.W.O.Hoskin,eds.).Rev. Mineral. Geochem.53,387-425.

Fedo, C.M., Sircombe, K.N. & Rainbird, R.H. (2003):Detritalzirconanalysisofthesedimentaryrecord.InZir-con(J.M.Hanchar&P.W.O.Hoskin,eds.).Rev. Mineral. Geochem.53,277-303.

Finch, R.J., Hanchar, J.M., Hoskin, P.W.O. & Peters,C.B.(2001):Rare-earthelementsinsyntheticzircon.2.Asingle-crystal X-ray study of xenotime substitution. Am. Mineral.86,681-689.

Fowler,A.,Prokoph,A.,Stern,R.&Dupuis,C.(2002):Organization of oscillatory zoning in zircon: analysis,scaling,geochemistryandmodelofazirconfromKipawa,Quebec, Canada. Geochim. Cosmochim. Acta 66, 311-328.

Geisler,T.,Pidgeon,R.T.,Kurtz,R.,vanBronswijk,W.& Schleicher, H. (2003a): Experimental hydrothermalalterationofpartiallymetamictzircon.Am. Mineral.88,1496-1513.

Geisler,T., Rashwan,A.A., Rahn, M.K.W., Poller, U.,Zwingmann, H., Pidgeon, R.T., Schleicher, H. &Tomaschek, F. (2003b): Low-temperature hydrothermalalteration of natural metamict zircon from the EasternDesert,Egypt.Mineral. Mag.67,485-508.

Geisler,T. & Schleicher, H. (2000): Improved U –Th–totalPbdatingofzirconsbyelectronmicroprobeusinganewbackgroundmodellingprocedureandCaasachemi-cal criterion of fluid-induced U–Th–Pb discordance in zircon.Chem. Geol.163,269-285.

Hanchar,J.M.,Finch,R.J.,Hoskin,P.W.O.,Watson,E.B.,Cherniak,D.J.&Mariano,A.N.(2001):Rareearthele-mentsinsyntheticzircon.1.Synthesis,andrareearthele-mentandphosphorusdoping.Am. Mineral.86,667-680.

Hanchar,J.M.&Hoskin,P.W.O.(2003):Zircon.Rev. Min-eral. Geochem.53.

Hanchar,J.M.&Miller,C.F.(1993):Zirconzonationpat-terns as revealed by cathodoluminescence and backscat-tered electron images: implications for interpretation ofcomplexcrustalhistories.Chem. Geol.110,1-13.

Hanchar, J.M. &Watson, E.B. (2003): Zircon saturationthermometry.InZircon(J.M.Hanchar&P.W.O.Hoskin,eds.).Rev. Mineral. Geochem.53,89-112.

Harrison,T.M.&Watson,E.B.(1983):Kineticsofdissolu-tionandzirconiumdiffusioningraniticmeltsofvariablewatercontent.Contrib. Mineral. Petrol.84,66-72.

Heaman,L.M.,Bowins,R.&Crocket,J.(1990):Thechem-icalcompositionofigneouszirconsuites:implicationsforgeochemical tracer studies. Geochim. Cosmochim. Acta54,1597-1607.

Higgins,M.D.(1994):Numericalmodellingofcrystalshapesinthinsections:estimationofcrystalhabitandtruesize.Am. Mineral.79,113-119.

Holland, H.D. & Gottfried, D. (1955): The effect ofnuclearradiationonthestructureofzircon.Acta Crystal-logr.8,291-300.

Hoskin,P.W.O.(2000):Patternsofchaos:fractalstatisticsandtheoscillatorychemistryofzircon.Geochim. Cosmochim. Acta64,1905-1923.

Hoskin,P.W.O.&Ireland,T.R.(2000):Rareearthelementchemistryofzirconanditsuseasaprovenanceindicator.Geology28,627-630.

Hoskin, P.W.O., Kinny, P.D.,Wyborn, D. & Chappell,B.W. (2000): Identifying accessory mineral saturationduringdifferentiationingranitoidmagmas:anintegratedapproach.J. Petrol.41,1365-1396.

Hoskin,P.W.O.&Schaltegger,U.(2003):Thecompositionof zirconand igneousandmetamorphicpetrogenesis. InZircon(J.M.Hanchar&P.W.O.Hoskin,eds.).Rev. Min-eral. Geochem.53,27-62.

Irber,W.,Förster,H.-J.,Hecht,L.,Möller,P.&Morte-ani,G.(1997):Experimental,geochemical,mineralogicalandO-isotopeconstraintsonthelate-magmatichistoryoftheFichtelgebirgegranites(Germany).Geol. Rundsch.86,Suppl.,S110-S124.


Jackson,S.E.,Pearson,N.J.,Griffin,W.L.&Belousova,E.A.(2004):Theapplicationoflaserablation–inductivelycoupledplasma–massspectrometryto in situU–Pbzircongeochronology.Chem. Geol.211,47-69.

Jarosewich,E.&Boatner,L.A.(1991):Rare-earthelementreferencesamplesforelectronmicroprobeanalysis.Geo-stand. Newslett.15,397-399.

Jarosewich,E.,Nelen,J.A.&Norberg,J.A.(1980):Refer-encesamplesforelectronmicroprobeanalysis. Geostand. Newslett.4,43-47.

Kempe, U., Gruner,T., Nasdala, L. &Wolf, D. (2000):Relevanceof cathodoluminescence for the interpretationofU–Pbzirconages,withanexampleof anapplicationtoastudyofzircons fromtheSaxonianGranuliteCom-plex, Germany. In Cathodoluminescence in Geosciences(M.Pagel,V.Barbin,P.Blanc&D.Ohnenstetter, eds.).Springer-Verlag,Berlin,Germany(415-455).

Kempe,U.,Gruner,T.,Renno,A.D.,Wolf,D.&René,M.(2004):DiscussiononWanget al. (2000):ChemistryofHf-richzirconsfromtheLaoshanI-andA-typegranites,easternChina.Mineral. Mag.68,669-675.

Köppel,V. & Sommerauer, J. (1974):Trace elements andthebehaviouroftheU–Pbsystemininheritedandnewlyformedcrystals.Contrib. Mineral. Petrol.43,71-82.

Kretz, R. (1983):Symbols for rock-forming minerals. Am. Mineral.68,277-279.

Linnen, R.L. & Keppler, H. (2002): Melt compositioncontrol of Zr/Hf fractionation in magmatic processes.Geochim. Cosmochim. Acta66,3293-3301.

Lowery,L.E.,Miller,C.F.,Wooden,J.L.,Mazdab,F.K.&Bea,F.(2006):HfandTizoninginzircon:detailedrecordsofmagmaticprocesses.Geophys. Res. Abstr.8,01258.

Lumpkin,G.R.&Chhakoumakos,B.C.(1988):Chemistryand radiation effects of thorite-group minerals from theHardingpegmatite,TaosCounty,NewMexico.Am. Min-eral.73,1405-1419.

Mathieu, R., Zetterström, L., Cuney, M., Gauthier-Lafaye,F.&Hidaka,H.(2001):Alterationofmonaziteandzirconand leadmigrationasgeochemical tracersoffluid paleocirculation around the Oklo–Okélobondo and Bangombé natural nuclear reaction zones (FrancevilleBasin,Gabon).Chem. Geol.171,147-171.

Mezger, K. & Krogstad, E.J. (1997): Interpretation ofdiscordantU–Pbages:anevaluation.J. Metam. Geol.15,127-140.

Murakami,T.,Chakoumakos,B.C.,Ewing,R.C.,Lumpkin,G.R.&Weber,W.J.(1991):Alpha-decayeventdamageinzircon.Am. Mineral.76,1510-1532.

Nasdala,L.,Beran,A.Libowitzky,E.&Wolf,D.(2001a):Theincorporationofhydroxylgroupsandmolecularwaterinnaturalzircon(ZrSiO4).Am. J. Sci.301,831-857.

Nasdala,L.&Hanchar,J.M.(2005):Commenton:Appli-cationofRamanspectroscopytodistinguishmetamorphicandigneouszircon(Xianet al.,Anal. Lett.2004,v.37,p.119).Anal. Lett.38,727-734.

Nasdala,L., Irmer,G.&Wolf,D. (1995):Thedegreeofmetamictizationinzircons:aRamanspectroscopicstudy.Eur. J. Mineral.7,471-478.

Nasdala, L., Lengauer, C.L., Hanchar, J.M., Kronz,A.,Wirth,R.,Blanc,P.,Kennedy,A.K.&Seydoux-Guillaume,A.-M. (2002):Annealing radiation damageand the recovery of cathodoluminescence. Chem. Geol.191,121-140.

Nasdala,L.,Wenzel,M.,Vavra,G.,Irmer,G.,Wenzel,T.&Kober,B.(2001b):Metamictisationofnaturalzircon:accumulation versus thermal annealing of radioactivity-induceddamage.Contrib. Mineral. Petrol.141,125-144.

Parrish, R.R & Noble, S.R. (2003): Zircon U–Th–Pbgeochronology by isotope dilution – thermal ionisationmass spectrometry (ID–TIMS). InZircon (J.M.Hanchar& P.W.O. Hoskin, eds.). Rev. Mineral. Geochem. 53,183-213.

Pavlenko,A.S.,Vainshtein, E.E. & Shevaleevskii, I.D.(1957): On the hafnium–zirconium ratio in igneous andmetasomaticrocks.Geochemistry,411-431.

Pérez-Soba,C.(1992):Petrología y geoquímica del macizo granítico de La Pedriza, Sistema Central Español.Ph.D.thesis, Universidad Complutense de Madrid, Madrid,Spain.

Peterman,Z.E.,Zartman,R.E.&Sims,P.K.(1986):Apro-tractedArchean history in theWatersmeet gneiss dome,northernMichigan.U.S. Geol. Surv., Bull.1622,51-64.

Pidgeon, R.T., Nemchin,A.A. & Hitchen, G.J. (1998):InternalstructuresofzirconfromArchaeangranitesfromtheDarlingRangebatholith:implicationsforzirconstabil-ity and the interpretation of zircon U–Pb ages. Contrib. Mineral. Petrol.132,288-299.

Pointer,C.M.,Ashworth, J.R.& Ixer,R.A. (1988):Thezircon–thorite mineral group in metasomatized granite,Ririwai, Nigeria. 2. Zoning, alteration and exsolution inzircon.Mineral. Petrol.39,21-37.

Pupin, J.P. (1980): Zircon and granite petrology. Contrib. Mineral. Petrol.73,207-220.

Pupin, J.P. (2000): Granite genesis related to geodynamicsfrom Hf–Y in zircon. Trans. R. Soc. Edinb.: Earth Sci.91,245-256.

Ríos,S.,Malcherek,T.,Salje,E.K.H.&Domeneghetti,C.(2000):Localizeddefectsinradiation-damagedzircon.Acta Crystallogr.B56,947-952.

Romans, P.A., Brown, L.L. & White, J.C. (1975):Anelectron microprobe study of yttrium, rare earth, and


phosphorusdistributioninzonedandordinaryzircon.Am. Mineral.60,475-480.

Rubatto, D.,Williams, I.S. & Buick, I.S. (2001): Zirconand monazite response to prograde metamorphism intheReynoldsRange,centralAustralia.Contrib. Mineral. Petrol.140,458-468.

Rubin,J.N.,Henry,C.D.&Price,J.G. (1989):Hydrother-malzirconsandzirconovergrowths,SierraBlancaPeaks,Texas.Am. Mineral.74,865-869.

Seidler,K.,Lentz,D.R.,Hall,D.C.&Susak,N.(2005):Zircon-richTa–Nb–REEmineralizationintheradioactiveMcKeelLakepegmatite–aplitesystem,WelsfordIgneousComplex, southwesternNewBrunswick.Explor. Mining Geol.14,79-94.

Smith,D.G.W.,deSt.Jorre,L.,Reed,S.J.B.&Long,J.V.P.(1991):ZonallymetamictizedandotherzirconfromThorLake,NorthwestTerritories.Can. Mineral.29,301-309.

Solar, G.S., Pressley,R.A., Brown, M. &Tucker, R.D.(1998):Graniteascentinconvergentorogenicbelts:testingamodel. Geology26,711-714.

Speer,J.A.(1980):Zircon.InOrthosilicates(P.H.Ribbe,ed.).Rev. Mineral.5,67-112.

Uher,P.,Breiter,K.,Kleka,M.&Pivec,E.(1998):ZirconinhighlyevolvedHercynianHomolkagranite,Moldanu-bianzone,CzechRepublic:indicatorofmagmasourceandpetrogenesis.Geol. Carpathica49,151-160.

Uher,P.&Černy,P. (1998):Zircon inHercyniangraniticpegmatites of the western Carpathians, Slovakia. Geol. Carpathica49,261-270.

Valley, J.W. (2003): Oxygen isotopes in zircon. In Zircon(J.M. Hanchar & P.W.O. Hoskin, eds.). Rev. Mineral. Geochem.53,343-385.

Vavra, G. (1994):Systematics of internal zircon morphol-ogyinmajorVariscangranitoid types.Contrib. Mineral. Petrol.117,331-344.

Villaseca, C., Barbero, L. & Rogers, G. (1998): Crustalorigin of Hercynian peraluminous granitic batholiths ofcentralSpain:petrological,geochemicalandisotopic(Sr,Nd)constraints.Lithos43,55-79.

Villaseca,C.,Downes,H.,Pin,C.&Barbero,L.(1999):NatureandcompositionofthelowercontinentalcrustincentralSpainandthegranulite–granitelinkage:inferencesfromgranuliticxenoliths.J. Petrol.40,1465-1496.

Villaseca,C.,Eugercios,L.,Snelling,N.,Huertas,M.J.&Castellón,T.(1995):Nuevosdatosgeocronológicos

(Rb–Sr, K–Ar) de granitoides hercínicos de la Sierra deGuadarrama.Rev. Soc. Geol. España8,137-148.

Villaseca, C. & Herreros,V. (2000):A sustained felsicmagmaticsystem:theHercyniangraniticbatholithoftheSpanishCentralSystem.Trans. R. Soc. Edinb.: Earth Sci.91,207-219.

Villaseca,C.,MartínRomera,C.,delaRosa,J.&Bar-bero, L. (2003): Residence and redistribution of REE,Y, Zr,Th and U during granulite-facies metamorphism:behaviourofaccessoryandmajorphasesinperaluminousgranulitesofcentralSpain.Chem. Geol.200,293-323.

Wang, Ru-Cheng, Fontan, F., Xu, Shi Jin, Chen, XiaoMing&Monchoux,P.(1996):Hafnianzirconfromtheapical part of the Suzhou granite, China. Can. Mineral.34,1001-1010.

Wang, Ru-Cheng, Zhao, G.T., Lu, J.J., Chen, X.M., Xu,S.J.&Wang,D.Z. (2000):ChemistryofHf-rich zirconfrom theLaoshan I- andA-typegranites, easternChina.Mineral. Mag.64, 867-877.

Wark, D.A. & Miller, C.F. (1993):Accessory mineralbehaviourduringdifferentiationofagranitesuite:mona-zite,xenotimeandzirconintheSweetwaterWashpluton,southeasternCalifornia,U.S.A.Chem. Geol.110,49-67.

Watson, E.B. (1996): Dissolution, growth and survival ofzirconsduringcrustalfusion:kineticprinciples,geologicmodelsandimplicationsforisotopicinheritance.Trans. R. Soc. Edinb.: Earth Sci.87,43-56.

Watson,E.B.,Wark,D.A.&Thomas,J.B.(2006):Crystal-lizationthermometersforzirconandrutile.Contrib. Miner. Petrol.151,413-433.

Woodhead, J.A., Rossman, G.R. &Thomas,A.P. (1991):Hydrousspeciesinzircon.Am. Mineral.76,1533-1546.

Zhang,M.&Salje,E.K.H. (2001): Infraredspectroscopicanalysis of zircon: radiation damage and the metamictstate.J. Phys: Condens. Matter13,3057-3071.

Zhang, M., Salje, E.K.H. & Ewing, R.C. (2003): Oxida-tion state of uranium in metamict and annealed zircon:near-infraredspectroscopicquantitativeanalysis.J. Phys. Condens. Matter15,3445-3470.

Zhang,M.,Salje,E.K.H.,Farnan,I.,Graeme-Barber,A.,Daniel, P., Ewing, R.C., Clark,A.M. & Leroux, H.(2000):Metamictizationof zircon:Raman spectroscopicstudy.J. Phys: Condens. Matter12,1915-1925.

Received March 18, 2005, revised manuscript accepted Sep-tember 27, 2006.

Documents

THE COMPOSITION OF ZIRCON IN THE PERALUMINOUS HERCYNIAN GRANITES OF THE SPANISH CENTRAL SYSTEM BATHOLITH