RENDJCONTI DEllJ\ SOCIETA ITAllANA DJ MINERALOGlA E PETRQLOGlA, i9S8, \101. 4}.2, pp. 27'·280

Andean batholiths and marginal basins

WALLACE S. PITCHERUniversity of Liverpool and 14 Church Road, Upton, Wirnl, Merseyside, L49, 6]Z, U.K., Td. No. OS 1·677 6896

Ae.sTv.CT. - 'The evolution of the Mcsozoic-urlyCenozoic buholiths of rhc Andes models the origin ofone type of granitoid at an active, continental platemargin. Cruna! extension is associated both wilh thegeneration of new, thick cruSt by the deep·seatedextraction of a basic progenitor from the mande, andwith the formation of ..back-are- voleanodanic basinswithin the edge of the continental plate. In such a zoneof high heat flux episodic movements or:J decp-~ing

ttanscut'f'Cnt faults trigger mnclti'll of the basicundcrpl.te, whilst in relaxation these faalts provide hotconduits for the fl"lCtionating magmas to introde highinto the voleanidutic infill of the marginal basins.

'The chain of great granitoid batholiths tha[cores the Western Cordilleras of the centraland southern Andes provides a prime exampleof silicic magmatism at an active plate edgeof a continent (ZEIL, 1979; PITCHER et aI.,1985). During the Mesozoic and Palaeogenethe primary source of the magmas lay withinthe upper mantle, and despite emplacementinto the continental margin their compositionhad little to do with the old crust; rather thenew magmas made major contributions to thecontinental plate.

The generative process is associated withthe subduction of the oceanic Nazca (Farallon)plate with the establishement of a marginalcontinental volcanic arc (e.g. lAMES, 1977,1978). The easterly migration of this arc has

been modelled in terms of changinginclination of the dip of the subducting slaband also frontal erosion, whilst the episodiccharacter and varying activity of magmatismcan be related to changes in the plateconvergence rate (FRUTOS, 1981; $oLER,1987). However, complications are introd.~by the obvious segmentation of the Andes,and the new importance given to strike-slipfaults with the likelihood of major terranedisplacement.

The early part of the Mesozoic Andeancycle was everywhere a time of crustalextension of the continental edge, albeit witha varied time scale along its length (DALZIEL,1981; LEVl and AGUIRRE, 1981; AGtJllUlE,1987; ABERG et al., 1984). In the centralAndes, from the Triassic to the LowerCretaceous, such extension led to blockfaulting, subsidence and the growth of fault­controlled, marginal sedimentary basins(MYERS, 1975a; Cobbing, in PITcHER et al.,1985). The latter show a polarity of theirsedimentary infill, with clastics derived fromthe continental interior contrasting with thesubmarine lavas and detritus of a volcanic arccomplete with submarine exhalative mineraldeposits (CAROOZO & WAUSCHUHN, 1984;VIDAL, in PrraiER et al., 1985). The greatthickness of such volcanic deposits, especiallythose of Lower Cretaceous age, attests to

276 'ilI.s. PI1t:HER

rapid subsidence within a «back-arc» basinalenvironment, whilst a primitive compositionindicates a mantle source and, furthermore,features of the non-deformative burialmetamorphism confirm a contemporary highheat flux (COBBlNG, ATI-lERTON et al., AGUIRREand OFFLER, in PITCHER et al., 1985),

The continental crust underpinning thest:marginal basins is represented in southernEcuador, Peru and northern Chile by anancient Precambrian massif, whilst in centraland southern Chile the basement consists ofslate belts representing aecreted foreaeeterranes of Palaeozoic age. It was this old crustthat was extended, differentially thinned,even ruptured and puUed apart during theMesozoic, with the result that basic magmasreleased from the mantle wedge ascendedrapidly, making little contact with old crust.In Peru there is the additional geophysicalevidence for a deep-crustal arch of high­density material rising into the continentaledge underlying the volcanogenic basin(WILSON, ATHERTON et al., in PITcHER et al.,1985). This may rep~nt a new basaltic crustextracted ftom the mantle wedge during theextensional phase. In north-eentral Chilespreading combined with subsidence, againassociated with the eruption of mantle.derivedbasalts, is more evident than in Peru (LEVIand AGUIRRE, 1981; AGUIRRE, 1985; p. 333).However in neither case is it thought thatoceanic crust was ever generued whichcontrasts with an analogous situation inPatagonia where the marginal basin ismodelled as having bttn Ooorc:::l. by new maficcrust (DALZIEL, 1981).

Within this particular marginal continentalenvironment the multiple granitoid batholithslie parallel to the tectonic trend and along ornear to the axes of precursor voIcanogenicbasins (PrrcHER et al., 1985; AGU1RRE, 1985).The earliest intrusions were often gabbroic,quickly succeded by the granitoids andsynplutonic, basic dyke swarms. The wholeprocess conforms to a magmatoctonic cycle ofvulcanicity, burial metamorphism, mildcompression with open folding andtranscurrent faulting, followed by plutonicemplacement, then relaxation and dykeintrusion, and finally uplift (Ct--wwER, 1973;

AGUIRRE, 1985 p. 332). This cycle ofoverlapping events can be repeated withwaning intensity from the mid-Cretaceous tothe mid-Cenozoic. Possibly compressioninhibited extrusion, extending the residencetime sufficiently for the ponded magmas tofractionate.

The Coastal Batholith of Peru illustrateswell the nature of this type of silicicmagmatism (PITcHER et al., 1985). A 1600km-long linear array of hundreds of plutonswas stoped out of the axial zone of one ofthese precursor «back-arc» basins of LowerCretaceous age. Characteristically the firstintrusions were largely of gabbro followed,over the next fJ) Ma, by the episodi.c inrrusionof short-lived p.dses of granitoid magma ofdecreasing volume. Calc-alkaline, magnetite­bearing, I-type tonalites and granodioritespredominate, though the compositionalspectrum is locally widened to include bothK-rich diorites and evolved granites. Therocks naturally group into well-defined, timeseparated, consanguineous rock suites, eachwith its own idenriry as defined in terms ofchronology, modal and chemical composition,textural characteristics, enclave populationsand dyke-swarm association (PITcHER, 1974).The identity of the suites and the number ofthe successive rhythms varies along thesegmented batholith but individual suitesmaintain their character over hundreds ofkilometers despite being distributed withinseparated plutons (COBBING et al., 1977;PITCHER et al., 1985). Whilst some suites weredifferentiated in situ within a single plutonmany others represent the multi-pulseintrusion from depth of magmas alreadylargely differentiared, possibly in deep-seatedand laterally extensive melt cells (PITCHER,1978; TAYLOR, 1976; ATHERTON, 1981). Eachsuite conforms to a simple pattern of calc­alkaline variation more characteristic ofmagmas undergoing crystal fractionation,albeit with variations in the proportions ofthe separating crystal crops, than of magmasdeveloped by fractional remelting (McCouRT,1981; ATIlERTON and SANDERSON, in PITCHERer al., 1985).

A characteristic feature of this andanalogous bathol.iths is the p~nce of coeval,

ANDEAN IlATIiOUTHS AND MARGlNAL IlASINS 277

basic dyke swarms (PITcHER & BUSSELL, inPITCHER et aL, 1985, cl. PiCHOWIAK &BREITKREUZ, 1984). In this connexion theubiquitous presence of dioritic endaves,often demonstrably derived from thedismemberment of such synplutonic basicdykes, shows that mixing between new basicmagma and various stages of the rest magmawas always possible, especially so within thering-complexes where mixing took place in afluidized media (BuSSELL, in PITCHER et al.,1985). .

Petrographically these granitoids showsimple textures and their mineralogy conformsto relatively high-temperature, water­undersaturated magmas lacking in latepegmatitic differentiates and vein systems.Such findings are consonant with the evidentupwelling of the magmas to a high, sub­volcanic level in the crust where they wereemplaced by stoping aided by brittlefracturing. Appropriately the high­temperature thermal aureoles are relativelynarrow and magma cooling is likely to havebeen rapid (ATIIERTON and BIIDolCHLEY, 1972;MVERS, 1975b; PITcHER, 1978; MUKASA andTILTON, in PrrcHER et al., 1985). Goodexamples of the plutono-volcanic interface areprovided by the centred complexes andporphyry·Cu breccia pipes, yet it seems thatfew plutons represent substantial feedermagma chambers - rather it was thesynplutonic dyke swarms that provided theconduits for any coeval extrusives (BuSSELL,PITcHER and WILSON, 1976; BUSSEU andPITCHER, in PrrcHER et al., 1985).

The compositional and isotopic data arewholly in accord with a primary mantle sourcefor the magmas. This is especially so in theLima sc=gment where the batholith is so clearlyaxial to a marginal basin, but where, as insouthern Peru, the granitoids enter into directcontact with the ancient massif, a mild degreeof crustal contamination is recorded by boththe Pb and Sr isotopes (BECKINSALE inPITCHER et aL, 1985; MUKASA, 1986).However an increase in Sri' varying in degre1::during each of the separate magmaticepisodes, has been interpreted by SOLER(1987) as due less to direct crustalcontamination than to sub-crustal

heterogenity - recording a competitionbetween subduction-related alteration of themantle wedge, a process producingheterogeneity, and mantle-wedge convection,which redresses that process.

Central to the understanding of the originof the granitoid magmas is the close acid-basicrelationship as revealed by the earlyproduction of a new dense crust, the precursorbasaltic volcanicity, the early appearance ofgabbro and the continued synplutonicintrusion of basic dykes. It is envisaged that,during the rapid plate convergence of earlyCretaceous times, basalt was melted out froma diapiric upwdling of the mantle wedge toform new crust. This underplate wassubsequently locally remelted during episodesof resurgent movement, che resultant batchesof magma fractionating to produce thetonalites (COBBING and PITcHER, 1983, Fig.7). Throughout the process the continuingintrusion of basic magma maintained a highheat flux.

The Coastal Batholith of Peru representsthe generation of magma along a single fault­margined, mega-lineament over the long timespan of 102·37Ma, and with but limitedeastward migration (PITcHER & BUSSEll,1977). Thus it is suggested that the batchesof melt were tapped and channelled by de1::pfaults of long-standing; indeed. it may be thatit was the episodic movement on such majorfaults that triggered remelting: a thesis thatplaces less emphasis on sulxiuction mechanics.

Such a lineamental control is also obviousin the even longer-lasting PatagonianBatholith, 155-10Ma. with its likedimensions, near identical compositions, asimilar history, and again with a closetemporal and spatial relationship with thedevelopment of a transcurrent-fault­controlled, «back-arc» basin (SUAREZ, 1977;AGUIRRE, 1985; RAPEu., 1978; NFJ..SON et al.,1988). As in Peru the volume of magma wasgreatest between 100-75Ma which correlateswith the pericx:l of high spreading rates. Here,however, it is the earliest intrusions that showevidence of crustal contamination of themantle-derived magmas, whilst the laterintrusions seem to have been insulated fromcontact with the old crust (NELSON et al.,

278 W.s. PITCHER

1978).Essentially similar batholiths occur in

between, in north and central Chile, but incontrast to other segments of the Andes boththe plutonic and volcanic belts show a verymarked easterly migration with time, and oneaccompanied by a compositional changewhich, as in the more subdued example inPeru, is thought to be related more to a mantlesource changing in composition with time ordistance than [0 an inettasing influence of theold crust (LoPEZ·Esco8AR et al., 1979;ATHERTON and SANDERSON, 1987). Innorthern Chile the most westerly batholith isactually emplaced within the basementflanking the compound volcanic belt soproviding, in cross-cutting relationships, adirect contrast with the very different, post­tectonic granites belonging to the UpperPalaeozoic magma-tectonic cycle (BERG andBRElTKREuz, 1983; DAMM and PICHOWlAX.1983).

In the Central Cordillera of Colombia, theold continental margin hosts Mesozoicbatholiths similar to those of the centralAndes. However this margin is now so slicedup by the sutures associated with the dockingof a Pacific island arc that the relationshipsare obscured between «back-arc. basinaldeposits and possible plutonic correlatives(ALVAREZ, 1983; MCCOURT et aI., 1984).Nevertheless the granitoids within the island­arc domain of the Western Cordillera showa fundamental contrast to those: of the truecontinental margin, viz. a much lesser volumeand a greater compositional restriction toquartz diorites and tonalites which arecandidates for the M-type characteristic ofisland arcs (PrrCHER, 1983).

Overall, the evolution of the Mesozoic-earlyCenozoic batholiths of the Andes provides amodel for the origin of one important typeof granitic rocks at an active, continental platemargin (PITcHER, 1987, Fig. 1). In thisscenario crustal extension is connected withthe generation of new, thick crust byextraction of a basic progenitor from themantle, and with the siting of a volcanic arcwithin the edge of the COntinental plate. Insuch a zone of high heat flux episodicmovements on deep-seated, transeW'I'etlt faults

trigger remelting at sub-crustallevels, but theassociated compression sufficiently seals thecontinuously thickened crustal carapace toprovide the increased travel distances andtimes necessary to advance fractionationprocesses; that is until relaxation allows thedeep-faults to tap the new magmas (cf. DAMMand PrCHOWIAK, 1983), permitting them tointrude high into the volcaniclastic infill ofthe marginal basins.

It is doubtful whether such a Pacific-typemarginal environment with its voluminousproduction of K-poor, Ca-rich granitoidswithin huge, Iineamental batholiths, everformed an element in the geological evolutionof the European crust.

REFERENCES

ABERG G., AGUlRRE L., LEVl B., NySlltOMj.O. (1984)- Spr(!ading - subsidmu and gmt!TQtion 0/ tnsialicmarginal basins: an ~ampk from th~ ~arly Cretaceous0/ctntral Chik. In: Marginal Basin Geology: Volomicand usociated sedimentary and ta::lonic processes inmodern and ancient marginal basins. Kokdaar B.P.,& Howdls M.F. (ech.). Geo1. Soc. London Spa::.Publ., 16, 18'·193.

AGUUUl£ L. (198:5) - T~5<Jllthmr Andn. In: The~Basins and Margins. Vol. 7A, The Pacific <Xean.Nairn A.E.M., Stehll F.G., & Uycda S. (cds.). PlenumPress, 26:j·J76.

AGUIII.R£ L. (1987) • Andean moJd/ing. In: Geodigest,Geolotly loday, 3, 47-48.

ALVAStEZ A.]. (1983) - G~/otU1 tk Iiz ConiU/t'1# Ontral, tl Octi6m1lt Coilu"bi;Jno , pdrOqllimial tk /osintrisiuos u-niteUks mn«moloioos. Bot Gcol.Ingeominas Colombia, 26, 2,1-7S.

ATIiERTON M.P., BRENCHLEY P.). (1972) - A prrlimilUll"jstudy 0/ t« struetUrt, stratigraph, a"J m~tamorphism

0/ some contact rodes 0/ th~ UJntern Andes, n~ar t«Qu~brad. VmadoMllmo, PmJ. GeaI.)., 8,161·178.

AYHEJlTON M.P. (1981) . Horizonlill and vmiul zoningill W Pnivua1I CotuJaJ Batho/iJh.J. Gcol. Soc. London,U8, 343-349.

AntEJl"I'()H M.P., SANDEaSa'l L.M. (1987)· T~CordiiJmlBlam:" Batho/ith: a Shldy 0/ uanik intrwsio" and tJNrtlation 0/ crusul thkkming to t-alllminosity. GeoJ.Runcb,h., 76, 231·232.

BERG K., BREITKR£uZ C. (1983) - Mesoloiu:he Plutonein JeT nonkhi/nliKMn Kus~kordilkrt: Petrogtnest,G~ochonologie, G~ochemi~ und Geodynamik1IUInklbnontcM""".tik. Geolekt. Fonch., 66, 1·11,1-107.

BlISSUL M.A., Pncllu W.S., Wn.sa;o P.A. (1976)·Ring comp/a.es 0/tht Pmmian CotuJaJ Batho/ith: a long

ANDEAN BATHOUTHS AND MARGINAL BASINS 279

J14ndinl JIIb-voialnk "&imt. Can. J. Earth. 50, n,1020-IO}(l.

CAlOOZOM., WAUSOiKlIHN A. (1974)- ThtCoputJ_ndP,Up mtul/ot«t 0" t«~ Jilk of Ct71tm/ Pnu.In; Syngenesis and Epigenesis in the Formation ofMineral Deposits. W.uschkuhn A. el al., (cds.).Springer Verbg, 616-646.

CHAaJ.lEa R. (1973) -lntm'utpiom of J1"Nding ,nd t«comprrnivt t«tonic p"'~ of tIN mmdionaf An,u,.Earlh Planet. Sci. Lett., 20, 242·249.

C0881NCl E.)., PrrcJIER W.S. (1983) - Andtan plutonismin Ptru and in rtlrJtionJhip 10 liokaniJm and JtgmtnttdplrJu tdgt. Mon. Geol. Soc. Am., 159, 277.292.

COBBING E.}., PlTcHEI W.S., TAYLOR W.P. (1977) .$qmt71ts and Jupn·units in t« Co4stoI &tho/ith ofPtnI.}I. GeoL, 85, 625-631.

DALZIEL LW.D. (1981) - .&d:41C1t txmf)io" i" t«JOllthml Andn, il rtVi~ ilnd critiuf rupfJ11liJil/. Phll.Tram. R_ Soc. London, A300, 319-33'.

DAM.'\1 K.W., PJCHOWIAI: S. (1981) . Gmdynsmilt llMMil".m~ in dn KIU1nIItordiI~ Nol'lkhiWJz.wischm T./Iill und CMfIll1JIl. Geotcclr.l. Forsch., 61,J·U, 1·166.

FJ.tn'OS j. (1981) . AndNn t«toniQ IIJ il con~~tof_./kJor J1"Nding. Tectonophysics, 72·T21-TJ2.

JAMES D.E. (1971) • PlIlu t«tonic makljar t«tvOlutionof tht Cn.tTtJ/ Andn. Bull. Geol. Soc. Am., 82,3325·3346.

j,v,.,,:.5 D.E. (1978)· Subduction oftht Nazes PlIlu btntathut/trill Ptru. Geology 6, 174·178.

LEvl B., AGUIRRE L. (1981)· Emis/ic rprtading'Jubs~t

in IIN Mnoxoic and Pakogmt Andts ofcmfTQf Chik.J. Geol. Soc. LoMon, 138, 75-81.

LoPEZ·EscoBAlI L., FREY F.A., OYARZUN J. (1979) .G«JClNmica/CN1fICtnisticJ ofCtntml Chik (J) 0·)4 OS)",,,itow. Contrib. Mineral.. Petrol., 70, 439·4'0.

McCouaT W.). (1981)· 'J'« If!OCbmrisIrJ and~oftM CoIIJl4/ &lho/ith ofPnu, Lima ~mt.}. GcoLSoc. London, 138, 407·420.

McCoun W.}., ASPDEN}.A., BIIOOK M. (I984)· N~gto/ogiul and~l d.lJI from tM C%mbUm~: continct14/vowth by mllltipk~. j. Geol.Soc. London, 141,831·845.

MUICASA S.B. (I986) . Zircon U·Ph IJgts ofJllptT,"nil$ inlht CosJtal batho/ith, Ptru; lmplialtiom for magmaticand t«tonic procmn. BuI. Geol. Soc. Am., 97,241_2'4.

MYERS J.S. (1975a) . Vroical crusta! movemtnn in tMAndes in Ptru. Nature, 254, 672·674.

MyERS J.S. (1975\)) . UJuldton w~eand /luidit.ation:mtr:haniJl1lJ ofintru.JWns oftM Co4stoI &tho/ith ofPnuinto in own IJoktlnic tj«14. Bull. GeoI.. Soc. Am., 86,

1209·1220.NELSON E., BaUCE B., E.tTHOH D., KMu.1EJI, D.,

WUV£I. S. (1988) . Rq,i.mull /itholofjul IJtlriatio"J intk PlItagonian kholiJh. }.S.A. Earth 50., 1, 239·247.

NELSON E., BRUCE B., WEAvt:a S.G., Lux D.,BRUECKNEI H.K. (1987) . Vtrtic.l cruJtll!cnnJamiflllticn in tM PatqpnUm &tho/ith. In. Circum·Pacific Phaneorozoic Granites. RJPd. C.W. (eel.).Decimo Congrno GcoIOgico Atgentino, Tucuman,Argentina, 4, 24·26.

PICHOWlAK S., BRElTOREUZ C. (1984) . Vokanic dyltnin tht North Chilean CQ/l$t Ranu. Gee!. Rundsch., 73,8'3·868.

PITCHER W.S. (1974) . TIN Mt~oic and CenoloicBatholithJ of Pnu. Pacific GeoIOBY, 8, 51·62.

PrrcHa W.S. (1978) . TIN afllllo"" of. batho/ith. j.Gee!. Soc. Lend., 1J5, 1'7·182.

Prro-ta W.S. (1984) . Phantr'Ozoic Plutonism in tMpennJimr Attdn. In: Andean Magmatism Chemical andIsotopic u-trainu, Harmon, R.S. &: BlrTiero, B.A.(cds.). Shiva Publishing Lld., 152·167.

PnoiEJr. W.S. (1987) . Gnlllites and1# _ f1IVIilafort1~n on. Geol. Rundsch., 76, '1-79.

PITcHER W.S., ATHEaTON M.P., CoBBING E.].,BECKINSALE R.D. (£US.). (l98')· AUfItUlIWn 111. P/meEdgt: IIN PnuliUzn Andes. Blaclcie Halstead Press, p.328.

PITcHER W.S. (1983) . Grani~ fypt and t«tonicmvironmeflt. In. Mountain Buildi~ Processes. K. Hsii(ed.l. Academic. Press, 19·40.

PlTcUER W.S., BUSSELL M.A. (1971) . Structuralcontrolofbalholilhic nnplllument in Ptru: a rtlIitw. j. Geo!.Soc. London, IB, 249-256.

R.APt:ur. C.W. (1987) • El BawJito PaJilg6nicom~ 40°)0'y 4 J°Jj '5. EmJio uoquimico prtlimi1l4r. In; Circum·Pacific Phanerowic Granites. Rapda C.W. (eel.).Decimo Conirem GeoJ6gico Argentina, Tucuman,ArJentina, 4, 21·23.

Sou:J. P. (1987)· Chronology.nd sptIwl distriblltion of11Ulgnalk activity dluing Upper cmaaous IlJId Ccro.toictimes in u"tml PnuIJUm Attdts. A plllu dynamicintepm4ti~ scbntr. In: Circum·Pacific PhaneroroicGranites. Rapda C.W. (ed.). Decimo CongresoGeol6gico Argemioo, Tueuman, Arsentina, 4, '4·'6.

SUAltEZ M. (1977) . NotaJ gtO<JulmicllJ prtfimi1l4rtS utBato/iro Pa14g6nico al Sur de TittTa uf Fu~o, Chik.Rev. Geo!. Chile, 4, 15·33.

TAYLQR W.P. (1976) - IntruJion and dilfermtiation ofgranitic magmas at a high ~l in the cruSI; the Pu!Caopluton, Lima Prrwinu, Ptru. j. Petrol., 17, 194·218.

ZEa W. (1979) . TIN Andes. A Geological Review.Gebrider Bomtreger, Berlin, p. 23 I.