Geological Society of America Bulletin 2004 Loewy 171 87

Geological Society of America Bulletin

doi: 10.1130/B25226.1 2004;116, no. 1-2;171-187Geological Society of America Bulletin

Staci L. Loewy, James N. Connelly and Ian W.D. Dalziel Andean margin of South AmericaAn orphaned basement block: The Arequipa-Antofalla Basement of the central

Email alerting servicescite this article

to receive free e-mail alerts when new articleswww.gsapubs.org/cgi/alertsclick

SubscribeAmerica Bulletin

to subscribe to Geological Society ofwww.gsapubs.org/subscriptions/click

Permission request to contact GSAhttp://www.geosociety.org/pubs/copyrt.htm#gsaclick

viewpoint. Opinions presented in this publication do not reflect official positions of the Society.positions by scientists worldwide, regardless of their race, citizenship, gender, religion, or politicalarticle's full citation. GSA provides this and other forums for the presentation of diverse opinions and articles on their own or their organization's Web site providing the posting includes a reference to thescience. This file may not be posted to any Web site, but authors may post the abstracts only of their unlimited copies of items in GSA's journals for noncommercial use in classrooms to further education andto use a single figure, a single table, and/or a brief paragraph of text in subsequent works and to make

GSA,employment. Individual scientists are hereby granted permission, without fees or further requests to Copyright not claimed on content prepared wholly by U.S. government employees within scope of their

Notes

Geological Society of America

on November 30, 2011gsabulletin.gsapubs.orgDownloaded from

http://gsabulletin.gsapubs.org/cgi/alerts

http://gsabulletin.gsapubs.org/subscriptions/index.ac.dtl

http://www.geosociety.org/pubs/copyrt.htm#gsa

http://gsabulletin.gsapubs.org/

For permission to copy, contact [email protected] 2004 Geological Society of America 171

GSA Bulletin; January/February 2004; v. 116; no. 1/2; p. 171–187; DOI 10.1130/B25226.1; 10 figures; Data Repository item 2004028.

An orphaned basement block: The Arequipa-Antofalla Basement ofthe central Andean margin of South America

Staci L. Loewy†

James N. ConnellyDepartment of Geological Sciences, The University of Texas at Austin, Austin, Texas 78712, USA

Ian W.D. DalzielDepartment of Geological Sciences, The University of Texas at Austin, Austin, Texas 78712, USA and Institute for Geophysics,The University of Texas at Austin, 4412 Spicewood Springs Road, Building 600, Austin, Texas 78759, USA

ABSTRACT

The Arequipa-Antofalla Basement, aProterozoic crustal block exposed along thecentral Andean margin, provides a key tointerpreting the pre-Andean history ofSouth America. New U/Pb geochronologyand whole-rock Pb and Nd isotope geo-chemistry from the Arequipa-AntofallaBasement refine the tectonic history and de-lineate three distinct crustal domains thatdecrease in age from north to south. Thenorthern domain of southern Peru andwestern Bolivia contains juvenile Paleopro-terozoic 2.02–1.79 Ga intrusions that weremetamorphosed at 1.82–1.79 Ga. TheMesoproterozoic central domain in north-ernmost Chile contains a significant Meso-proterozoic juvenile component that incor-porates Paleoproterozoic crust from thenorthern domain. Rock units from both thenorthern and central domains were meta-morphosed between 1.20 and 0.94 Ga, withcoeval magmatism occurring only in thecentral domain. The southern domain innorthern Chile and northwestern Argenti-na comprises Ordovician rocks, derivedfrom a mix of juvenile material and oldercrust. Similar Ordovician magmatism(476–440 Ma) also occurred in the northernand central domains followed by metamor-phism at ca. 440 Ma.

Based on this refined geologic and tec-tonic characterization of the Arequipa-Antofalla Basement and comparison withthat of Amazonia, we conclude that: (1) theisolated exposures of the Arequipa-AntofallaBasement comprise a single basement block

†E-mail: [email protected].

with multiple domains, (2) the Arequipa-Antofalla Basement was not derived fromAmazonia, and (3) the Arequipa-AntofallaBasement accreted onto Amazonia duringthe 1.0 Ga Sunsas Orogeny.

Keywords: U/Pb, whole-rock lead, Sm/Nd,Arequipa-Antofalla Basement, Amazonia,isotopes, absolute age.

INTRODUCTION

The Paleoproterozoic to early PaleozoicArequipa-Antofalla Basement is situatedalong the western margin of South Americabetween the Andean Cordillera and present-day Peru–Chile trench (Fig. 1). The currentposition and southward crustal growth of theArequipa-Antofalla Basement (Dalmayrac etal., 1977; Lehmann, 1978; Shackleton et al.,1979; Damm et al., 1990; Wasteneys et al.,1995; Worner et al., 2000) are at odds with itsformation in a simple model of westwardgrowth (Tassinari et al., 2000) of the Amazoncraton. These basic observations have inspiredat least two models to explain its existence:(1) the Arequipa-Antofalla Basement is al-lochthonous with respect to Amazonia andwas accreted to the pre–Andean margin (Coiraet al., 1982; Nur and Ben-Avraham, 1982; Ra-mos, 1988; Litherland et al., 1989; Dalziel,1992, 1993; Ramos et al., 1993; Dalziel, 1994;Bahlburg and Herve, 1997), and (2) the Are-quipa-Antofalla Basement is parautochthon-ous with respect to Amazonia, emplaced alongtranscurrent faults (Sadowski and Bettencourt,1996; Tosdal, 1996). An improved under-standing of the crustal growth, isotopic char-acter, and tectonic history of the Arequipa-Antofalla Basement will facilitate evaluation

of the relationship between the Arequipa-Antofalla Basement, Amazonia, and othercontinents and, thus, provide constraints forpaleogeographic reconstructions.

This paper integrates new U/Pb geochro-nologic and isotopic data from southern Peruand northern Chile with the existing data torefine the tectonic history and isotopic char-acter of the Arequipa-Antofalla Basement.With these data, we address the followingquestions: (1) Is the Arequipa-Antofalla Base-ment a single crustal block?, (2) Is Amazoniathe parent craton?, and (3) If allochthonous,when was the Arequipa-Antofalla Basementaccreted to Amazonia?

REGIONAL GEOLOGY

Proterozoic and Early Paleozoic rocks ofthe Arequipa-Antofalla Basement are exposedalong the present-day Andean margin in atleast 10 known inliers in southern Peru, west-ern Bolivia, northern Chile and northwesternArgentina (Fig. 2). Additional pre–Andean in-liers, apparently unrelated to the Arequipa-Antofalla Basement, occur in Colombia (theGarzon, Santander, and Santa Marta massifs)and Argentina (the Precordillera Terrane) (Fig.1). The Colombian inliers are interpreted to bethe northwestern continuation of the Mesopro-terozoic Sunsas Orogen of Amazonia (Alva-rez, 1981; Kroonenberg, 1982; Alvarez, 1984;Litherland et al., 1989; Priem et al., 1989;Restrepo-Pace et al., 1997). The ArgentinePrecordillera Terrane is interpreted to haveoriginated in or near the Ouachita Embaymentof Laurentia and accreted to South Americaduring the Paleozoic (Abbruzzi et al., 1993;Astini et al., 1995; Dalziel et al., 1996; Kayet al., 1996). The Arequipa-Antofalla Base-



172 Geological Society of America Bulletin, January/February 2004

LOEWY et al.

Figure 1. Map of South America with generalized age provinces. Age of province reflectsmost recent metamorphic event to affect region. Poorly constrained extent of Rio de laPlata Craton is shown as dashed line. Three Proterozoic basement blocks along Andeanmargin are: (1) Arequipa-Antofalla Basement, (2) Precordillera Terrane, and (3) Colom-bian inliers (Garzon (a), Santander (b), Santa Marta (c) massifs). Adapted from Cordaniet al. (2000) and Tassinari et al. (2000).

ment is similar to the Colombian and Argen-tine inliers in areal extent, minimal exposure,and proximity to the Andean Margin, but thereis less agreement regarding its origin andevolution.

The Arequipa-Antofalla Basement

The Arequipa-Antofalla Basement is ex-posed through Andean volcanic and sedimen-tary rocks along the Arica Embayment (Fig.1). Although the inliers contain dissimilarrock types, previous workers cited a coherentwhole-rock Pb isotopic signature throughoutto define a single crustal block (Tilton andBarreiro, 1980; Barreiro and Clark, 1984;Worner et al., 1992; Aitcheson et al., 1995;

Tosdal, 1996; Worner et al., 2000). The rela-tionship of the individual inliers to each otherand to Amazonia, the cratonic core of SouthAmerica, is obscured by the intrusion of youn-ger rocks, deposition of overlying sequences,and deformation of the Andean Orogeny.

Based on ages of the oldest rock units ex-posed, the inliers of the Arequipa-AntofallaBasement are divided into three domains (Fig.2) that young from north to south. The north-ernmost domain exists in southern Peru fromSan Juan to Mollendo and east into westernBolivia. The central domain extends fromthe Peru-Chile border to Quebrada Choja innorthern-most Chile. The southern domainis exposed as far south as northwesternArgentina.

According to previous work, the oldestrocks in the northern domain formed at ca.2.0–1.9 Ga (Wasteneys et al., 1995; Worneret al., 2000) but there has been a debate overthe timing of metamorphism of these rocks.Rb/Sr and early U/Pb studies implied granuliteto amphibolite facies metamorphism between1.9 and 1.8 Ga (Cobbing et al., 1977; Shack-leton et al., 1979), but more recent U/Pb datafrom gneisses at Quilca, Mollendo, and CerroUyarani (Fig. 2) indicated high-grade meta-morphism at ca. 1.2–1.0 Ga (Wasteneys et al.,1995; Worner et al., 2000). Alternatively, Dal-mayrac et al. (1977) proposed that these rocksexperienced metamorphism during both thePaleoproterozoic (1.9 Ga) and the Neoproter-ozoic (0.7 Ga).

Ages of the oldest rocks in the central do-main have been poorly constrained but appearto be Mesoproterozoic. Metavolcanic rocksexposed at Belen (Fig. 2) yielded an imprecisewhole-rock Sm/Nd isochron age of 1460 6448 Ma (Damm et al., 1990). At QuebradaChoja (Fig. 2), U/Pb zircon analyses from a‘‘migmatite’’ and an ‘‘orthogneiss’’ suggestedcrystallization ages of 1254 197/–94 Ma and1213 128/–25 Ma, respectively (Damm et al.,1990).

In the southern domain, maximum protolithages range from ca. 476 Ma to 434 Ma, coevalwith ca. 440 Ma metamorphism (Mpodozis etal., 1983; Omarini et al., 1984; Damm et al.,1990; Lork and Bahlburg, 1993). Two granitesat Cordon de Lila (Fig. 2) yielded U/Pb crys-tallization ages of 450 112/–11 Ma and 4346 2 Ma (Damm et al., 1990). Rb/Sr whole-rock analyses and K/Ar analyses of horn-blende and biotite from granodiorite at Cordonde Lila yielded metamorphic ages of 441 614 Ma and 429 6 11 Ma, respectively (Mpo-dozis et al., 1983). Intrusions in the Puna in-lier (Fig. 2) were dated by U/Pb analyses ofmonazite fractions at 476 6 1 Ma, 473 6 1Ma, 472 6 1 Ma, and 467 6 1 Ma (Lork andBahlburg, 1993) and by Rb/Sr whole-rockanalyses at 471 6 12 Ma (Omarini et al.,1984).

Paleozoic magmatism and metamorphismhave also been identified in the northern andcentral domains. Intrusions in Peru yielded anRb/Sr whole-rock isochron age of 44067 Ma(Shackleton et al., 1979) and a 444614 MaK-Ar age from Belen (Pacci et al., 1980) anda lower intercept age of 415 136/–38 Mafrom Quebrada Choja (Damm et al., 1990)were interpreted to represent the timing ofmetamorphism. Younger magmatism has alsobeen identified in Peru, with data including,U/Pb crystallization ages of 42564 Ma and388 113/–18 Ma (Mukasa and Henry, 1990),



Geological Society of America Bulletin, January/February 2004 173

AREQUIPA-ANTOFALLA BASEMENT

Figure 2. Map of Arequipa-Antofalla Basement and adjacent crustal provinces. Arequipa-Antofalla Basement is divided into three domains (northern, central, and southern). Solidblack regions are exposed inliers in Arequipa-Antofalla Basement. Overlapping patternsin Arequipa-Antofalla Basement indicate multiple tectonic events. RNJ—Rio Negro–Juruena Province. Adapted from Damm et al. (1990), Tosdal (1996), Pankhurst and Ra-pela (1998), and Keppie and Bahlburg (1999).

an Rb/Sr whole-rock isochron age of 3926 22Ma (Shackleton et al., 1979) and K-Ar mineralages of 37466 Ma, 36566 Ma, and 33965Ma from one intrusion (Cobbing et al., 1977).These younger ages (425–339 Ma) may rep-resent post-tectonic magmatism, but have alsobeen interpreted to reflect a separate event as-sociated with the convergence of the ChileniaTerrane (Fig. 1; Ramos et al., 1986; Dalziel,1997).

In summary, there is a general youngingfrom north to the south in the Arequipa-Antofalla Basement, where protoliths range inage from Paleoproterozoic in southern Peruand western Bolivia through Mesoproterozoicin northernmost Chile to Paleozoic in Chileand northwestern Argentina. Geochronologicdata from the northern domain suggest that itwas probably metamorphosed during the Me-soproterozoic, but it may have been metamor-phosed earlier, during the Paleoproterozoic.All domains experienced Ordovician/Silurianmetamorphism and/or magmatism.

Relationship to Amazonia

Previously published data and consequentmagmatic and metamorphic history of theArequipa-Antofalla Basement allow a prelim-inary evaluation of the relationship betweenthe Arequipa-Antofalla Basement and Ama-zonia. Southwest Amazonia comprises a seriesof progressive age domains that young west-wards away from an Archean core (Fig. 1;Tassinari et al., 2000; Geraldes et al., 2001).In contrast, the Arequipa-Antofalla Basementyoungs southward, away from the Paleopro-terozoic (ca. 2.0–1.9 Ga) basement of thenorthern domain (Fig. 2). In its current posi-tion, the age and southward growth of theArequipa-Antofalla Basement are incongruouswith westward growth of Amazonia, implyingthat the Arequipa-Antofalla Basement is eitherallochthonous or parautochthonous with re-spect to Amazonia.

Allochthonous ModelsAllochthonous models were based on 2.0–

1.9 Ga crust in the Arequipa-Antofalla Base-ment, west of the ca. 1.3–1.0 Ga Sunsas Prov-ince and active Andean tectonism (Fig. 1; Nurand Ben-Avraham, 1982; Litherland et al.,1989). According to Dalziel (1994), pervasive1.2–1.0 Ga metamorphism in the northern andcentral domains of the Arequipa-AntofallaBasement supports accretion of the Arequipa-Antofalla Basement to Amazonia during theSunsas Orogeny. Based on the presence ofNeoproterozoic to Cambrian (0.7–0.5 Ga) oro-genic belts east of the Arequipa-Antofalla

Basement (the Pampean orogen and Tucavacabelt; Litherland et al., 1989), several authorshave proposed early Cambrian accretion of theArequipa-Antofalla Basement to Amazonia(Ramos, 1988; Ramos et al., 1993). Alterna-tively, Coira et al. (1982) cited pervasive Or-dovician intrusions and deformation through-out the Arequipa-Antofalla Basement tosupport Ordovician accretion. Based on paleo-magnetic data, an alternative model has thenorthern domain and northern half of the cen-tral domain of the Arequipa-Antofalla Base-ment accreting to Amazonia in the Protero-zoic, with the southern half of the centraldomain and southern domain colliding as aseparate block during or after the OrdovicianFamatinian Orogeny (Rapalini et al., 1999).

Parautochthonous ModelsTosdal’s (1996) parautochthonous model is

based primarily on a similarity of the whole-rock Pb isotopic signatures of the Arequipa-Antofalla Basement and the Rondonia–San Ig-nacio and Rio Negro–Juruena provinces of

southwestern Amazonia and the GuyanaShield (northern Archean core of Amazonia).If correct, this model implies that the ages ofrocks and tectonic events in the Arequipa-Antofalla Basement and Amazonia shouldalso be similar, an assertion correct to a firstapproximation. According to this model, em-placement as a parautochthonous block oc-curred at an unspecified time prior to thePaleozoic.

RESULTS—U/Pb GEOCHRONOLOGY

A better understanding of the pre-Sunsasand Sunsas magmatic and tectonic history ofthe Arequipa-Antofalla Basement would helpidentify the parent craton and distinguish be-tween the different hypotheses. We focused onthree inliers of the northern and central do-mains of the Arequipa-Antofalla Basement,where Paleo- and Mesoproterozoic rock unitshad previously been identified: (1) coastalsouthern Peru from San Juan to Mollendo; (2)an area around the town of Belen, Chile; and




LOEWY et al.

Figure 3. Concordia diagrams for samples from northern domain. Ages are defined by linear regression through data except whereindicated. See text and Table DR1 for details about individual data points and interpretations. POF—probability of fit.





Figure 3. (Continued.)

(3) Quebrada Choja, Chile (Fig. 2). The fol-lowing section outlines field observations andgeochronological results for each inlier. Ana-lytical results with 2-sigma errors are in TableDR1. Table DR1, geologic maps of each inlierwith sample locations (Figs. DR1 and DR2Aand B), and a description of analytical meth-ods are in the GSA Bulletin Data Repository.1

Northern Domain, Southern Peru

Field ObservationsThe Peruvian Arequipa Massif extends

along the coast from San Juan to Mollendo(Caldas, 1978). We focused our work in theSan Juan region, where the record is mostcomplete, but samples were also collected inOcona and Mollendo (Fig. DR1).

San Juan Area. In San Juan, the oldestunits are granitic, banded gneisses with am-phibolite layers, some of which may have sed-imentary protoliths. The entire sequence wasrecrystallized during at least amphibolite fa-cies metamorphism (M1) indicated by the for-mation of the gneissic layers comprising elon-gate, deformed quartz and feldspar ribbonswith abundant rotational recrystallization ofboth minerals and aligned mafic minerals(now mostly retrograde chlorite). Banding inthe gneisses is cut by potassium-feldspar me-gacrystic granite that contains a weak am-phibolite-facies fabric (M2/S2) defined byalignment of elongate quartz grains and maficminerals (now mostly retrograde chlorite) andminor rotational recrystallization of feldspar.The S2 fabric in the granites is typically par-allel to that of the host gneisses. The differ-ence in the degree of fabric development andthe lack of migmatization in the crosscuttingmegacrystic granite indicates that the mainbanding in the gneisses developed prior to theintrusion of the megacrystic granite and thatboth rock types were deformed after the gran-ite intruded.

These gneisses and granites form the base-ment below sediments that were deposited af-ter M2. The lowest member in the sequenceis the Chiquerıó Formation, a well-beddedsiltstone tillite, the lower portions of whichcontain abundant pink granitic and gneissicclasts ranging in size from meters to milli-meters. The larger clasts depress well-preserved bedding in the matrix, implying ice-rafted dropstones of glacial origin (Harland et

1GSA Data Repository item 2004028, analyticalmethods, geologic maps with sample locations(Figs. DR1 and DR2A and B) and data tables (Ta-bles DR1 and DR2), is available on the Web at http://www.geosociety.org/pubs/ft2004.htm. Requestsmay also be sent to [email protected].




LOEWY et al.

al., 1966). In the upper portion, across anabrupt transition, clasts change from granite tocarbonate.

The Chiquerıó Formation is immediatelyoverlain by 1–2 m of finely-laminated, alter-nating pink and gray carbonate layers that are,in turn, overlain by a thick, intensely alteredsequence of carbonates. The finely laminatedpink and gray carbonate layers strongly re-semble globally observed Neoproterozoic capcarbonates (Fairchild and Hambrey, 1984), in-cluding their characteristic negative d13C val-ues (Frank Corsetti, 2001, personal commun.).Published geologic maps identify two distinctcarbonate units with similar character, theNeoproterozoic San Juan Formation and thePaleozoic Marcona Formation (Caldas, 1978).More recent mapping by the Marcona MiningCorporation (Juan Aranibar Loaysa, 2000,personal commun.) reinterprets them to be asingle formation that they name the San JuanFormation. Unable to identify significant dif-ferences between rocks mapped as San Juanand Marcona formations, we agree with theirinterpretation.

Deformation in both the Chiquerıó and SanJuan formations includes folding, elongationof carbonate clasts, and formation of agreenschist facies cleavage (M3) defined bythe alignment of micas and chlorite drapedaround larger grains with quartz-rich pressureshadows. Gneissic and granitic basementrocks were pervasively retrogressed togreenschist facies during this event.

The entire sequence is cut by undeformedintrusions, which, in the San Juan region, arebimodal, fine-grained granite and diabasedikes.

Ocona and Mollendo. Portions of the geo-logical record at San Juan are exposed to thesouth at Ocona and Mollendo (Fig. DR1). InMollendo, the inlier contains pink and gray-green, banded gneiss metamorphosed to atleast amphibolite facies and undeformedpotassium-feldspar granite porphyry. Thegneiss is correlated with the banded gneiss atSan Juan and similarly contains alternatingbands of aligned chlorite and elongate quartzand feldspar ribbons. At Ocona, foliated me-gacrystic granite is intruded by amphibolitedikes. Although the megacrystic granite lookssimilar to the foliated megacrystic granite ex-posed at San Juan (our observations and thoseof Shackleton et al., 1979), U/Pb ages pres-ented below demonstrate that the granites atOcona are significantly younger.

Geochronology

Basement. Two samples of the bandedgneisses were analyzed, one from San Juan

(U/Pb-1) and one from Mollendo (U/Pb-2). U/Pb-1 is a pink, granitic gneiss with well-developed gneissic layering. Most of the feld-spar was sericitized and mafic minerals wereretrogressed to chlorite and epidote duringM3. The sample represents a felsic layer in anexposure of heterogeneous banded gneiss withbands that ranged in composition from gab-broic to granitic. Five zircon fractions from U/Pb-1 define a line between 1819 117/–16 Maand 1033 6 31 Ma (14% probability of fit(POF); Davis, 1982; Fig. 3A). Z3 was a singleprism with no visible inclusions and containedconcentric zonation (visible in plane polarizedlight) typical of magmatic growth. The otherzircons analyzed were similar in morphologyto Z3. Thus, the upper intercept is interpretedto be the crystallization age of the graniticprotolith. The lower intercept is interpreted torepresent the time of Pb loss or overgrowth ofzircon during metamorphism.

U/Pb-2 is a sample of pink and gray-greenlayered gneiss from Mollendo, in whichquartz and sericitized feldspar-rich bands areinterlayered with epidote and chlorite-richbands. Eight zircon fractions from U/Pb-2 de-fine a line between 1851 6 5 Ma and 935 614 Ma (72% POF; Fig. 3B). Four other frac-tions (Z4, Z7, Z9, Z11) that lie close to, butnot on, this line are interpreted either to con-tain older inheritance or to have been affectedby a younger thermal event. Z1 compriseselongate euhedral prisms consistent with mag-matic growth and plots closest to the upperintercept. Z6, a single grain imaged by cath-odoluminescence (CL) before analysis, exhib-ited concentric zonation typical of magmaticgrowth (Fig. 3B, inset). Z2, Z3, Z4, and Z5were similar in shape and size to Z6, sug-gesting similar growth histories. Because themorphologies and internal zonation of thesefractions that lie on the line near the upperintercept are consistent with magmaticgrowth, we interpret the upper intercept torepresent the crystallization age of the gneissprotolith. Z10, Z11, and Z12 lie closest to thelower intercept. Z11 and Z12 comprised verysmall anhedral equant grains, typical of zircongrown during metamorphism. Z10, a singlegrain imaged by CL before analysis, showeda large, unzoned rim around a small, zonedcore (Fig. 3B, inset). The lack of zonation inthe rim suggests metamorphic growth of therim around a magmatic core. The position ofthese three fractions nearest the lower inter-cept suggests that the lower intercept repre-sents the timing of zircon growth duringmetamorphism.

U/Pb-3 is a sample of the foliated mega-crystic granite at San Juan. It is a pink,

potassium-feldspar-rich granite porphyry witha weakly developed foliation defined by theelongation of quartz-feldspar lenses and align-ment of chlorite and muscovite. Feldspars aremostly sericitized with muscovite rims andmafic minerals have retrogressed to chlorite,presumably during M3. Analyses of eight zir-con fractions from U/Pb-3 yield a discordialine between 1793 6 6 Ma and 1052 6 12Ma (23% POF; Fig. 3C). That eight fractionsof zircons with morphologies consistent witha magmatic growth fall on a single line im-plies that the upper intercept represents a crys-tallization age. The high degree of discordancetoward the lower intercept is consistent withmetamorphism and Pb loss or zircon over-growth at ca. 1 Ga.

Chiquerıó Tillite. Three samples of graniticclasts from the Chiquerıó Tillite were ana-lyzed (U/Pb-4, -5, -6). U/Pb-4 and U/Pb-5 areboth round clasts of pink, potassium-feldspar-rich, megacrystic granite that is sim-ilar in appearance to the megacrystic granitein the basement but with poorly-developed fo-liations. U/Pb-6 is similar in composition tothe other two but has a gneissosity that wasfolded prior to its incorporation into the tillite.Three zircon fractions (Z1, Z2, Z3) from U/Pb-4 define a line with intercepts of 1168 19/–6 Ma and 147 1238/–260 Ma (90% POF;Fig. 3D). Z4 is interpreted to contain a minoramount of inherited zircon. The zircon popu-lation from U/Pb-5 is more complex than thatfrom U/Pb-4. Z1, Z4, and Z5 define a discor-dia line from 1162 6 6 Ma to 367 6 76 Ma(16% POF; Fig. 3E). Z2, Z3, Z6, and Z7 ap-parently contain inherited Paleoproterozoiczircon and were not included in the regression.

All five fractions of U/Pb-6 have middleMesoproterozoic 207Pb/206Pb ages but do notdefine a discordia or mixing line. Two zirconfractions (Z4 and Z5) from U/Pb-6 suggest anupper intercept age of ca. 1165 Ma projectedfrom a lower intercept of 440 Ma (Fig. 3F),the metamorphic age of the tillite (see below).Z1, Z2, and Z3 lie close to this upper interceptbut are interpreted to contain minor inheri-tance from an older source.

Late Granites. Three samples of late gran-ite were collected: an undeformed, pinkpotassium-feldspar granite porphyry at Mol-lendo (U/Pb-7); a weakly-foliated, megacrys-tic potassium-feldspar granite at Ocona (U/Pb-8); and an undeformed, fine-grained granitewith abundant partially resorbed feldspar andquartz xenocrysts at San Juan (U/Pb-9). Thelate granite at Mollendo cuts the basementgneiss. The late granite at San Juan cuts boththe gneisses and the Chiquerıó Tillite. AtOcona, there are no other rocks exposed.





Four zircon fractions from U/Pb-7 clusterca. 460 Ma (Fig. 3G). By pinning the lowerintercept at 0 Ma, a line regressed through Z1,Z2, and Z3 yields an upper intercept of 4686 4 Ma (55% POF).

Three zircon fractions from U/Pb-8 fall nearor at the bottom of a line with a lower inter-cept of ca. 464 Ma (Fig. 3H). One analysis(Z1) is concordant with an age and analyticalerror of 464 6 4 Ma (average of 206Pb*/238Uand 207Pb*/235U errors). Because all three frac-tions from this unmigmatized granite repre-sent the dominant population of large, beige,subhedral grains with minimal inclusions, weinterpret the lower intercept to be the crystal-lization age. Z2 and Z3 apparently contain aminor amount of inherited zircon of slightlydifferent age or Pb loss histories such that weuse the concordant fraction to define an ageof 464 6 4 Ma for this granite.

Z8 and Z2 from U/Pb-9 are both nearlyconcordant at ca. 500 Ma and ;1050 Ma, re-spectively (Fig. 3I). Given that the host sedi-ments were deposited and deformed after ca.1.03 Ga (the age of basement metamorphism),the 1050 Ma zircon must be inherited. CL im-ages of grains with similar morphologies toZ8 show euhedral, magmatic rims around in-herited cores (Fig. 3I, inset). Thus, Z8 is in-terpreted to approximate the crystallizationage with a minor amount of inheritance. Dif-fering degrees and ages of inherited compo-nents preclude derivation of a mixing line and,consequently, determination of a precise age.However, projection of mixing lines from eachfraction though Z8 constrains the range ofpossible lower intercepts between 468 and 440Ma (Fig. 3I). Thus, the San Juan granite be-longs to the same suite as the Ordoviciangranites at Mollendo (U/Pb-7) and Ocona (U/Pb-8).

ImplicationsProtoliths of the basement gneisses at San

Juan and Mollendo crystallized between ca.1851 and 1819 Ma. These ages are similar topreviously published ca. 1.9 Ga protolith agesfor the basement rocks of San Juan, Quilca,and Mollendo (Cobbing et al., 1977; Dal-mayrac et al., 1977; Shackleton et al., 1979;Wasteneys et al., 1995). The gneissic fabric(M1) formed prior to the ca. 1793 Ma intru-sion of the megacrystic granite. Mesoproter-ozoic metamorphism (M2) occurred betweenca. 1052 and 935 Ma, as recorded by zirconin both the gneisses and the megacrystic gran-ite at San Juan. Wasteneys et al. (1995) reportsa metamorphic age of 970 6 23 Ma at Mol-lendo and a slightly older age of 1198 16/–4Ma at Quilca. This range in metamorphic ages

may represent real temporal variation alongthe margin, variation in cooling histories be-tween sites, and/or timing of different zircon-forming reactions (Connelly, 2001). Most pre-vious work suggested that only oneProterozoic metamorphic event occurred inPeru, either at ca. 1.9 Ga or ca. 1.0 Ga (Cob-bing et al., 1977; Shackleton et al., 1979; Was-teneys et al., 1995). Only Dalmayrac et al.(1977) proposed two episodes of metamor-phism at 1.9 and 0.7 Ga. Our new data requirethat metamorphism and deformation occurredat both ca. 1.8 Ga and ca. 1.0 Ga with noevidence for a ca. 0.7 Ga event.

Despite the complex zircon systematics ofthese samples, there is a clear indication thatall three clasts fall in the range 1.17–1.16 Ga,requiring an extrabasinal source. No Mesopro-terozoic rocks have been identified in southernPeru, although evidence of Mesoproterozoiccrystalline basement at depth can be inferredfrom inherited zircons in the Ordovician gran-ites at San Juan and Ocona. In the central do-main, ca. 1.1 Ga rocks are discussed belowand similar-aged rocks may exist to the westin Bolivia (Lehmann, 1978). The Sunsas Prov-ince of Amazonia also contains ca. 1.1 Garocks (Tassinari et al., 2000).

No precise age has been determined for thedeposition of the tillite. It must be youngerthan the ca. 1.03 Ga M2 recorded in its base-ment and older than the Ordovician granite atSan Juan. Less definitive constraints include:(1) ‘‘stromatolite-like’’ structures in the over-lying San Juan Formation (Injoque and Rom-ero, 1986) that the authors correlate with lateNeoproterozoic–Early Cambrian stromatolites,and (2) a possible Neoproterozoic cap carbon-ate at the top of the tillite.

M3 metamorphism occurred after intrusionof the 464 6 4 Ma megacrystic granite atOcona but before emplacement of undeformedgranites at Mollendo (468 6 4 Ma). AlthoughM3 and Ordovician magmatism may havebeen diachronous across this belt, M3 is con-strained as Ordovician.

Central Domain, Belen, Northern Chile

Field ObservationsThe Arequipa-Antofalla Basement at Belen

is exposed in cores of breached anticlines oflayered Andean volcanic and sub-volcanicrocks. The largest single basement exposure is3 km (east-west) by 8 km (north-south) andextends north from Belen (Fig. DR2A). Threesmaller exposures lie south-southeast ofBelen.

The oldest rocks in these inliers are con-cordant layers of muscovite schist and

quartzo-feldspathic and mafic gneisses. Theconcordant, alternating, compositional layersand remnant sedimentary features, such ascross beds, preserved in less deformed layerssuggest that this unit comprises amphibolitefacies metasedimentary and mafic metavol-canic rocks. In the northwestern portion of theexposure, near Chapiquina (Fig. DR2A), theserocks are associated with partially serpentini-zed ultra-mafic units that crop out in a north-northeast linear trend. The majority of out-crops display an S1 foliation parallel toprimary S0 layering. In a few locations, S0/S1 is tightly folded (F2) and the muscoviteschist exhibits a second foliation, S2, that isaxial planar to folds. The only age constraintfor these sediments is a 1460 6 448 Ma agefrom a whole-rock Sm/Nd isochron from themetavolcanic layers (Damm et al., 1990).

Massive granodiorites, diorites, and gabbrosare the most abundant rocks in the Belen in-liers. They cut the S1 fabric in the metasedi-mentary and metavolcanic rocks and containxenoliths of these rock units. Fabric develop-ment in the intrusions is variable and rangesfrom rocks with no preferred orientation tothose with a well-developed foliation (includ-ing protomylonites) defined by alignment ofamphibole and/or biotite that is correlatedwith S2 in the metasedimentary rocks. Thisfabric (S2) in the younger intrusive rocksranges from straight to tightly folded (F3), butthere is no fabric developed axial planar tothese folds (no S3).

GeochronologyAges of clastic sedimentary rocks are dif-

ficult to determine directly by radiometricmethods, but ages of cross-cutting igneousunits and metamorphism constrain the timingof deposition. Two samples of the massivegranodiorite were collected: one from thelarge northern exposure, near Belen (U/Pb-10), and one from a smaller southern exposure(U/Pb-11) (Fig. DR2A).

Five zircon fractions from U/Pb-10 define aline from 1559 6 21 Ma to 473 6 2 Ma (43%POF; Fig. 4A). CL images indicate that zir-cons from this sample mainly have magmaticzonation, suggesting that the analyzed zirconsare igneous (Fig. 4A, inset). Four of the fivefractions analyzed are nearly concordant at thelower intercept, suggesting that this is thecrystallization age of the granodiorite. Z5 con-tained a significant inherited component.

Two fractions (Z1 and Z2) from U/Pb-11overlap concordia with an averaged 206Pb/238Uage of 47363 Ma, which is interpreted to bethe crystallization age. Z4 reflects a combi-




LOEWY et al.

Figure 4. Concordia diagrams for samples from Belen area. Ages are defined by linearregression through data except where indicated. See text and Table DR1 for details aboutindividual data points and interpretations. POF—probability of fit.

nation of inheritance and Pb loss (Fig. 4B),and Z1 may have lost Pb.

A felsic dike (U/Pb-12) intrudes a thick,folded unit of muscovite schist. The dike,composed mainly of quartz, feldspar, andmuscovite, cuts S1 in the schist and containsan F2-axial-planar foliation (S2). This samplewas collected to constrain the timing of D1and D2. Zircons from this sample yielded un-expected Paleoproterozoic 207Pb/206Pb ages andZ2, Z3, and Z5 define a line from 1866 6 2to 227 6 17 (Fig. 4C). CL images of zirconsfrom U/Pb-12 identify fractured, partially re-sorbed cores with subhedral to euhedral over-growths that commonly contain magmatic zo-nation (Fig. 4C, inset). This morphologysuggests that the overgrowths formed duringdike crystallization around inherited cores.Thus, the upper intercept age reflects inheri-tance from a ca. 1.9–1.8 Ga source, similar inage to the granitic gneisses in southern Peru.The lower intercept may reflect the intrusionage of the dike but more likely does not datea specific event and reflects a complex com-bination of overgrowth and Pb-loss. Cores inthese zircons are sizeable and U-rich, suchthat the rims are likely metamict and, as such,have lost Pb.

ImplicationsThe age of deposition and deformation (D1)

of the metasedimentary rocks at Belen havenot been well constrained with the addition ofthis new data. Definitive ages from the mas-sive granodiorite intrusions (473 6 2 and 4736 3 Ma) require the metasedimentary rocksand D1 to be older than ca. 473 Ma. The onlyother constraint on the age of the metasedi-mentary rocks is the poorly defined 1460 6480 Ma age from the amphibolite layers(Damm et al., 1990) that suggests depositionoccurred some time between 1.9 and 1.0 Ga,followed by D1 deformation.

Central Domain, Quebrada Choja,Northern Chile

Field ObservationsProterozoic rocks are exposed in the east-

west–trending Quebrada (Canyon) Choja (Fig.DR2), where the oldest units include a se-quence of migmatitic gneisses containingquartz-biotite paragneiss and granodioritic or-thogneiss. Both gneisses are included as xe-noliths in younger megacrystic granite, indi-cating that a phase of high-grademetamorphism (M1) occurred after formationof the gneiss protoliths but before intrusion ofthe megacrystic granite. The megacrysticgranite has a variably developed foliation (S2/





M2) defined by aligned biotite, muscovite,and elongate quartz. The paragneiss containstwo neosomes, an earlier wispy tonalitic type(N1) and a second, more coherent granitictype (N2). N1 neosomes occur in outcrops ofparagneiss and gneissic xenoliths in the me-gacrystic granite, indicating that they formedprior to crystallization of the granite and likelyduring M1/D1. N2 neosomes are folded andboudinaged in paragneiss outcrops and are notfound in paragneiss xenoliths in the mega-crystic granite. Thus, formation of N2 neo-somes postdates M1/D1 but predates M2/D2.Postdating both M1 and M2, a massive, un-deformed tonalite cuts both neosomes in theparagneiss.

West of the gneisses and granite, a ;500-m-thick metasedimentary sequence containsmetasiltstone with relict graded bedding,quartz-rich chlorite-muscovite paragneiss, in-terbedded muscovite schist and phyllite, lay-ered carbonates, and an irregularly bandedamphibolite unit. The protolith to the amphib-olite is uncertain, but banding suggests avolcanic-volcaniclastic protolith. This com-posite sedimentary package has a bedding par-allel foliation defined by aligned micas, chlo-rite, amphiboles, and elongate quartz.Similarities of this metamorphic fabric to thatin the megacrystic granite suggest that this isan S2 fabric. However, a sharp discordancebetween foliation orientations in this packageand those in the megacrystic granite suggestsa fault contact (Fig. DR2B).

The amphibolite unit is intruded by dacitedikes that cut across layering. The dacite dikescontain a moderate foliation that is parallel toS2 in the host amphibolites, indicating that asingle foliation (S2) in both rocks post-datesthe dacite dikes.

The rock units in the western portion of thecanyon are overlain by a variably deformedintrusive suite of diorites and tonalites that areintruded by late, undeformed hornblenditedikes. The boundary at the base of this intru-sive suite is a sharp, subhorizontal, planarcontact that we interpret to be a low-anglefault (Fig. DR2B). Because this boundary hasnot been deformed, faulting occurred afterM2. Based on similarities in mineralogy andtexture, tonalite in this intrusive suite is cor-related with the undeformed tonalite that cutsboth neosomes in the paragneiss. A suite ofpegmatites of varying composition cuts allunits in the area, except the amphibolites.

GeochronologySamples of the migmatized granodiorite or-

thogneiss (U/Pb-13), the foliated megacrysticgranite (U/Pb-14), a dacite dike that cuts the

layered amphibolite (U/Pb-15), a granite neo-some (N2) from the paragneiss (U/Pb-16), andthe undeformed tonalite from the fault-boundintrusive suite (U/Pb-17) were analyzed for U/Pb geochronology (Fig. DR2B).

Four of nine zircon fractions (Z1, Z5, Z6,Z9) from the orthogneiss (U/Pb-13) define aline between 1067 6 4 and 497 6 16 (78%POF; Fig. 5A). CL analyses of a representa-tive sample of the zircon population indicatethat ellipsoidal, pitted grains have the best-developed magmatic zonation compared to therest of the zircon population. Z1 was a palepink, ellipsoidal, pitted grain and is the mostconcordant point. Therefore, we interpret theupper intercept to represent the crystallizationage of the gneiss protolith. CL imaging ofmore euhedral grains commonly shows dis-continuous overgrowths on rounded coreswith magmatic growth zoning (Fig. 5A, inset).Fractions of these euhedral zircons plottedlower on the mixing line, suggesting varyingproportions of core and overgrowth. The over-growths are interpreted to be metamorphicsuch that the lower intercept represents thetiming of metamorphism. Fractions lying be-low the mixing line (Z2, Z3, Z7, and Z8) ap-parently contained an inherited component. Z4lies above the mixing line and may reflect Me-soproterozoic migmatization of the orthog-neiss. Migmatization occurred after crystalli-zation at 1067 Ma and prior to intrusion of thefoliated granite (1024 Ma, see below). A mix-ing line (not shown) from 497 Ma through Z4intersects concordia at ca. 1040 Ma, a reason-able age for migmatization.

Four zircon fractions (Z1, Z2, Z6, Z7) froma sample of the foliated megacrystic granite(U/Pb-14) define a line from 102465 Ma to44468 Ma (26% POF; Fig. 5B). Most frac-tions (Z1, Z2, Z3, Z4, Z5, and Z7) were pinkto light brown, euhedral to subhedral, multi-faceted, equant, clear single grains. Becausethis granite is not migmatized and the clear,euhedral-subhedral fractions lie near the upperintercept, we interpret the upper intercept torepresent the crystallization age. Z6 falls sig-nificantly lower on discordia and comprisedsmaller equant grains, typical of metamorphicgrowth. This analysis and the degree of dis-cordance of other fractions indicate that thelower intercept is the metamorphic age. Z8 ap-parently contained inherited zircon, and Z3,Z4, and Z5 apparently experienced minor re-cent Pb loss.

The heterogeneous zircon population of thedacite dike (U/Pb-15) includes equant to elon-gate grains that range from pale-beige to lightbrown, and CL images indicate a range of coreand rim textures. Zircon fractions (Z4, Z5, Z6,

Z7) representing a range of morphologies de-fine a line from 1697 6 48 Ma to 635 6 5Ma (42% POF; Fig. 5C). Concordant fractionZ7 was a single light-brown elongate grainwith no inclusions. Because this sample is un-migmatized, has a single metamorphic fabric,and most zircon fractions from this populationfall near the lower intercept, we interpret 6356 5 Ma to be the crystallization age. Z1, Z2,and Z3 contained a significant component ofinherited zircon of various ages.

Zircons from N2 (U/Pb-16) are mainly eu-hedral, brown prisms comprising dark, euhed-ral tips and pale, rounded cores. CL images ofthese grains show rims with concentric zona-tion around inherited cores (Fig. 5D, inset).Two analyses of tips broken off of euhedralgrains yielded nearly concordant points with207Pb/206Pb ages of 470 12/–1 Ma and 47013/–1 Ma (Fig. 5D). We interpret these to bethe formation age of the neosome.

Two zircon fractions (Z1 and Z2) from thetonalite (U/Pb-17) define a mixing line be-tween ca. 1070 Ma and ca. 450 Ma with athird fraction (Z3) lying only slightly belowthe line. CL analysis of representative zirconsindicates a heterogeneous population com-posed of zircons with zoned and unzonedcores and rims. The complex morphology andthe position of all three fractions equidistantfrom either intercept make interpretation dif-ficult. However, because the tonalite is unde-formed, it crystallized after Ordovician D2/M2 deformation. Also, the correlative tonalitein the paragneiss cuts the 470 Ma neosome.Thus, the lower intercept must approximatethe age of intrusion.

ImplicationsThe age of the paragneiss protolith is not

independently constrained but must have beenolder than 1067 Ma, the crystallization age ofthe orthogneiss protolith that intrudes it. Bothgneisses are included in the 1024 Ma unmig-matized megacrystic granite, requiring mig-matization (M1) to have occurred between1067 and 1024 Ma. Deposition of the meta-sedimentary rocks and extrusion/intrusion ofthe amphibolite protolith occurred betweenM1 and 635 Ma, the age of the crosscuttingdacite dikes. Metamorphism (M2) of all unitsoccurred during the Ordovician/Silurian, con-strained by zircon tips from N2 (ca. 470 Ma)and lower intercept ages from the samples oforthogneiss and megacrystic granite (ca. 497Ma and ca. 444 Ma, respectively). No preciseage was determined for the fault-bound intru-sive suite, but a lower intercept for the unde-formed tonalite suggests an Ordovician (ca.450 Ma) age. Because the main focus of this




LOEWY et al.

Figure 5. Concordia diagrams for samples from Quebrada Choja.Ages are defined by linear regression through data. See text andTable DR1 for details about individual data points and interpreta-tions. POF—probability of fit.





Figure 6. Summary of tectonic history of Arequipa-Antofalla Basement. Data are fromthis study and references cited in text.

work was the older tectonic history, unde-formed pegmatites were not dated, but theyare presumed to post-date Ordovician defor-mation and metamorphism. The fault-boundedintrusive suite was emplaced either during lateOrdovician deformation or, more likely giventhe undeformed nature of the fault, during lat-er Andean deformation.

Summary of the Sequence of Events in theNorthern and Central Domains

Figure 6 summarizes pre-existing and newU/Pb data. In the northern domain, Paleopro-terozoic intrusions with ages in the range of1851–1819 Ma and possibly as old as 2024Ma (Worner et al., 2000) were metamor-phosed during M1 to at least amphibolite fa-cies between 1819 and 1793 Ma. In the centraldomain, sediments and volcanics were depos-ited and plutons were emplaced during themid–late Mesoproterozoic. Both domains ex-perienced Grenvillian-age metamorphism be-tween 1200 and 935 Ma. During the Neopro-terozoic, sediments were deposited in bothdomains (San Juan and Quebrada Choja), butevidence of magmatism was only identified atone location in the central domain (QuebradaChoja). During the Ordovician, magmatism,metamorphism, and deformation occurred inboth the northern and central domains be-tween ca. 473 and 440 Ma with metamor-phism beginning as early as 497 Ma.

WHOLE-ROCK ISOTOPIC ANALYSIS

To provide a regional characterization of theArequipa-Antofalla Basement, whole-rock

samples were analyzed for Pb and Nd isotopiccompositions. Pb isotopic compositions ap-pear to define distinct signatures of crustalprovinces (for example, Kay et al., 1996; Tos-dal, 1996; Sinha and McLelland, 1997), andSm/Nd isotopic systematics («Nd(0) values andTDM ages) are commonly used to indicatesource reservoirs of intrusions and ages ofcrustal provinces (e.g., Farmer and DePaolo,1984; Bennett and DePaolo, 1987; Murphyand Nance, 2002; Payolla et al., 2002; Single-tary et al., 2003). This regional signature canbe used in conjunction with the ages and se-quences of events to compare inliers in theArequipa-Antofalla Basement and to comparethe Arequipa-Antofalla Basement withAmazonia.

Lead

Whole-rock powders from 12 samples fromthe northern domain, 39 samples from the cen-tral domain (22 from Belen and 17 from Que-brada Choja), and one sample from the south-ern domain were analyzed for Pb (Table DR2;see footnote 1). These samples were selectedand analyzed to complement existing whole-rock Pb isotopic analyses of basement rocks(Proterozoic boulders found in younger sedi-mentary rocks have not been included.) (Til-ton and Barreiro, 1980; Aitcheson et al., 1995;Tosdal, 1996; Bock et al., 2000; Worner et al.,2000).

Data from the three domains define distinct,but overlapping fields in uranogenic space(206Pb/204Pb versus 207Pb/204Pb; Fig. 7A). Pa-leoproterozoic samples from the northern do-

main have the highest 207Pb/204Pb ratios (rela-tive to 206Pb/204Pb), plotting above Stacey andKramers’ (1975) crustal evolution curve. Me-soproterozoic samples from the central do-main have lower 207Pb/204Pb ratios (relative to206Pb/204Pb) and overlap Stacey and Kramers’(1975) crustal evolution curve (Fig. 7A).Along the central domain field, Mesoprotero-zoic samples from Belen are less radiogenic(lower 206Pb/204Pb and 207Pb/204Pb) than thosefrom Quebrada Choja and overlap those of thenorthern domain. Ordovician samples fromthe southern domain (Bock et al., 2000) plotbetween Stacey and Kramers’ (1975) crustalevolution curve and overlap the central do-main samples (Fig. 7A). Neoproterozoic andOrdovician intrusions in both the northern andcentral domains plot along the trend of thecentral domain between the compositions ofthe Mesoproterozoic units (Fig. 7A). Mostrocks from all three domains have 208Pb/204Pbratios between 42 and 37. Thus, in thoro-uranogenic Pb space, Pb isotopic composi-tions define two distinct regions based on the206Pb/204Pb ratio (Fig. 7B). Samples from thenorthern domain and the northern central do-main (Belen) have low 206Pb/204Pb ratios,whereas samples from the southern central do-main (Quebrada Choja) and the southern do-main have high 206Pb/204Pb ratios.

Samarium-Neodymium

Whole-rock Sm/Nd analyses of 21 samplesfrom southern Peru, Belen, and QuebradaChoja (Table DR2; Fig. 8A, B, and C) yielda continuous range of TDM ages that, in gen-eral, young from north to south. Samples fromthe northern domain (with the exception of theOrdovician intrusions and one ca. 1.8 Ga am-phibolite) have TDM ages of 2.3–1.9 Ga and«Nd(0) values of233.18 to217.87 (Table DR2;Fig. 8A, B, and C). The similarity of TDM agesand crystallization ages (TU/Pb; Table DR2;Fig. 8C) suggests that the Paleoproterozoicunits were predominantly derived from juve-nile mantle. Samples from the central domain(including Mesoproterozoic and youngerunits) and samples of Ordovician intrusionsfrom the northern domain (all plotting alongthe central domain Pb trend) have youngerTDM ages (2.2–1.3 Ga with two as young as0.5 Ga and one anomalously old age of 3.6Ga) and more positive «Nd(0) values (-15.86to23.41 with one positive value of 2.15) (Ta-ble DR2; Fig. 8A, B, and C). For most ofthese samples, the TDM ages are older thantheir U/Pb crystallization ages (TU/Pb; TableDR2; Fig. 8C), suggesting contamination of ajuvenile mantle source by Paleoproterozoic




LOEWY et al.

Figure 7. Whole-rock Pb data from Arequipa-Antofalla Basement. Data arranged by lo-cation. New data (dark symbols) are distinguished from data previously published (lightsymbols). Previously published data are from Tilton and Barreiro (1980), Aitcheson et al.(1995), Tosdal (1996), Bock et al. (2000), and Worner et al. (2000). Stacey and Kramers’(1975) average crust evolution curve is given for reference with ages in Ga. (A) UranogenicPb. Fields distinguish domain signatures. Central domain field is divided to distinguishjuvenile Mesoproterozoic signature from rock compositions likely to contain a significantcomponent of Paleoproterozoic crust. Numbered data points are discussed in text andlisted in Table DR2. (B) Thoro-uranogenic Pb. Fields denote different signatures.

Figure 8. Sm/Nd data. Values are listed inTable DR2 and in Bock et al. (2000). TDM

ages for data from Bock et al. (2000) wererecalculated by method used in this study.(A) Histogram of TDM ages from Arequipa-Antofalla Basement. (Anomalously old TDM

ages of P3 and B8 are not included. TDM

ages could not be calculated for B3 andQ13.) (B) Histogram of «Nd(0) values fromArequipa-Antofalla Basement. (Anoma-lously low «Nd(0) of P3 is not included.) (C)Compositional fields defined by Sm/Nd iso-topic systematics of each domain. Symbolsare «Nd at TU/Pb (Table DR2). Thick dashedline separates compositions from Belen andQuebrada Choja. (Compositions of Q7 andQ8 are not included. They appear to bemantle-derived Neoproterozoic intrusionsand, thus, their compositions are not relat-ed to Mesoproterozoic crustal reservoir.)

and/or Mesoproterozoic crust or derivation en-tirely from older crust. Samples from Belenhave older TDM ages than those from QuebradaChoja, suggesting a greater Paleoproterozoiccrustal component. The distribution of TDM

ages from the Belen samples matches that ofthe northern domain (Fig. 8A). For some sam-ples from Quebrada Choja (for example, Q2

(U/Pb-13), Q7, and Q8), the TDM age approx-imates the U/Pb crystallization age, suggest-ing a juvenile source for these samples. Datafrom the southern domain (one sample fromthis study and samples from Bock et al.[2000], recalculated using the depleted mantlemodel curve of DePaolo [1981]), define theyoungest TDM age group (1.9 Ga–0.5 Ga) and

highest «Nd(0) values (–13.42 to 5.77) (TableDR2; Fig. 8A, B, and C). These TDM ages areolder than the known crystallization ages, alsoimplying contamination of a juvenile mantlesource by Paleoproterozoic or Mesoprotero-zoic crust or derivation entirely from oldercrust.





Implications of Whole-Rock Isotopic Data

The coherent, but different, Pb isotopic sig-natures of the northern and central domainsindicate that each domain had isotopically dis-tinct source reservoirs. Correlation of Paleo-proterozoic U/Pb and Sm/Nd TDM ages in thenorthern domain rocks suggests a Paleopro-terozoic mantle source. In the central domain,the oldest rocks are Mesoproterozoic and maybe as old as ca. 1.5 Ga. In addition to thepoorly-defined ca. 1.5 Ga Sm/Nd isochron ageof the metavolcanic rocks at Belen, an Ordo-vician granodiorite at Belen contains ca. 1.5Ga inherited zircon, suggesting that crust ofthis age exists at depth. Mesoproterozoicrocks (Q1, Q2, and Q5) from Quebrada Chojacommonly have Mesoproterozoic TDM ages.Thus, Mesoproterozoic mantle appears to bethe isotopically distinct Pb source at QuebradaChoja.

The lower 206Pb/204Pb ratios from Belen rel-ative to those from Quebrada Choja (Fig. 7A)suggest that the Mesoproterozoic Belen rocksmay have had a slightly different source. Pa-leoproterozoic Sm/Nd TDM ages of the Belenrocks and the overlap of the Belen and north-ern domain Pb isotopic composition supportincorporation of a significant component ofPaleoproterozoic crust from the northern do-main. As mentioned previously, U/Pb zirconanalyses indicate that a felsic dike at Belen(U/Pb-12; B9) contains inherited ca. 1.87 Gazircon (Fig. 4C), suggesting ca. 1.87 Ga crustat depth. Thus, the Mesoproterozoic Belenrocks may have been derived entirely from orwere strongly contaminated by 1.9–1.8 Ganorthern domain crust. The intermediate Pbisotopic compositions of the Neoproterozoicand younger rocks at both Belen and Quebra-da Choja and their Paleo- and Mesoprotero-zoic TDM ages suggest that these rocks com-prise a mixture of Belen and Quebrada Chojacrust.

The four Ordovician intrusions in the north-ern domain (P8, P9, P11, P12) have centraldomain Pb isotopic signatures (Fig 7A) and«Nd(0) values (Fig. 8B) and all four intrusions(three granites and one gabbro), three ofwhich are dated by U/Pb geochronology, yieldca. 1.5 Ga Sm/Nd TDM ages (Table DR2,Fig. 8A). U/Pb data from U/Pb-8 (P12) andU/Pb-9 (P8) provide evidence for Mesoproter-ozoic inherited components (Figs. 3H and I).«Nd(0) values, TDM ages, Pb isotopic composi-tions, and Mesoproterozoic ages of the north-ern domain Ordovician intrusions indicate thatthey were derived from a Mesoproterozoic(possibly ca. 1.5 Ga), central domain-like low-er crustal reservoir. Thus, the Mesoproterozoic

Pb isotopic reservoir that defines the centraldomain may extend beneath the northern do-main at depth.

The division of the Arequipa-AntofallaBasement into three domains was originallybased upon the ages of the oldest units ex-posed. Uranogenic Pb isotopic compositions(Fig. 7A), «Nd(0) values, and TDM ages (Fig. 8,A, B, and C) support the distinction of thesethree temporal domains. Previous work on Pbisotopic compositions of Andean volcanics inthe Arequipa-Antofalla Basement identifiedthe two signatures in thoro-uranogenic Pbcompositions (Fig. 7B) and used them to di-vide the Arequipa-Antofalla Basement intotwo terranes (Worner et al., 1992). However,as proposed above, incorporation of differentamounts of older crust in rocks at Belen andQuebrada Choja would account for the ob-served distinction in the 206Pb/204Pb ratios. Theexistence of three domains is supported bymapping, U/Pb geochronology, uranogenicwhole-rock Pb isotopes and Sm/Nd isotopicsystematics («Nd(0) values and TDM ages).

DISCUSSION

The main goals of this paper were to refinethe tectonic history and characterize the iso-topic signature of the Arequipa-AntofallaBasement to address the following questions:(1) Is the Arequipa-Antofalla Basement a sin-gle crustal block? (2) Is Amazonia the ‘‘parentcraton’’?, and (3) if allochthonous, when wasthe Arequipa-Antofalla Basement accreted toAmazonia? The next section summarizes theevolution of the Arequipa-Antofalla Basementas refined by the new data and then addresseseach of these questions.

Tectonic History

U/Pb ages of the oldest rocks exposed andPb and Sm/Nd whole-rock isotopic composi-tions delineate three distinct domains in theArequipa-Antofalla Basement that decrease inage from north to south. The northern domainof southern Peru and western Bolivia (Fig. 2)formed from juvenile material between 1.85and 1.79 Ga, may have components as old as2.02 Ga, and was first metamorphosed be-tween 1.82 and 1.79 Ga (Figs. 6 and 9).

A second stage of crustal growth occurredduring the Mesoproterozoic, possibly as earlyas ca. 1.5 Ga, along the southern margin ofthe northern domain (Fig. 9). Although no ex-posed rocks of this age have been preciselydated, Sm/Nd and Pb isotopic evidence sug-gest that 1.5–1.4 Ga crust forms the basementof the central domain and underplates the 1.9–

1.8 Ga basement of the northern domain (Or-dovician intrusive rocks in the northern do-main with ca. 1.5 Ga TDM ages and centraldomain Pb signature and «Nd(0)). Ca. 1.5 Gacontinental volcanism at Belen was part ofthis magmatic event, although isotopic com-positions indicate contamination from Paleo-proterozoic (northern domain) crust.

A third stage of growth began in the centraldomain by 1.07 Ga. It was associated withmetamorphism and deformation in both thenorthern and central domains between 1.20 Gaand 0.94 Ga (Fig. 9).

A fourth stage of growth formed the south-ern domain during the Ordovician (Fig. 9). Ig-neous rocks of the southern domain were de-rived from juvenile material variablycontaminated by Paleo- and/or Mesoprotero-zoic crust. Coeval intrusions in the northerndomain were derived from a Mesoproterozoicreservoir similar to that of the central domain.Coeval intrusions in the central domain werederived from a mix of the Mesoproterozoiccentral domain crust and low-m northerndomain-like Paleoproterozoic crust. All do-mains were metamorphosed at ca. 0.44 Ga.

Comparison Between Inliers

U/Pb geochronology indicates that juvenilematerial was added to the Arequipa-AntofallaBasement in discontinuous stages from northto south between at least 1.85 Ga to 0.44 Ga.Furthermore, deformation and metamorphismoccurred in the Arequipa-Antofalla Basementin three distinct pulses: (1) 1.82–180 Ga, (2)1.20–0.94 Ga, and (3) ca. 0.44 Ga (Fig. 9).Each successive thermotectonic event affectedall rock units in domains formed by that time.Whole-rock isotopic data indicate that youn-ger rock units incorporated components thatresemble older crust from other domains dur-ing this progressive growth (i.e., central do-main rocks with a northern domain signatureand Ordovician northern domain rocks with acentral domain signature). This observationsuggests that one domain was built upon theother. This history of systematic growth andprogressive deformation strongly suggests thatthe Arequipa-Antofalla Basement grew as asingle, coherent crustal block.

Relationship to Amazonia

Citing gross similarities in ages and simi-larities in Pb isotopic compositions, Tosdal(1996) interpreted the Arequipa-AntofallaBasement to be a parautochthonous block ofAmazonia. To a first approximation, ages ofrock units broadly correlate with known ages




LOEWY et al.

Figure 9. Schematic model of tectonic evolution of Arequipa-Antofalla Basement (AAB)shown in map view and north-south–striking cross section. ND—northern domain, CD—central domain, SD—southern domain.

of provinces in Amazonia. Paleoproterozoicprotolith and metamorphic ages of the north-ern domain of the Arequipa-Antofalla Base-ment are close to those of the Ventuari-Tapajos Province of Amazonia and rocks inboth are juvenile. Ca. 1.5–1.4 Ga and ca. 1.1–1.0 Ga rocks of the central domain of the Ar-equipa-Antofalla Basement are similar in ageto those of the Rondonia–San Ignacio andSunsas provinces in Amazonia (Tassinari etal., 2000; Geraldes et al., 2001). The Meso-proterozoic orogenic event recorded in thenorthern and central domains of the Arequipa-Antofalla Basement is coeval with events inthe Sunsas Province (Tassinari et al., 2000).However, existing protolith ages for units inthe Ventuari-Tapajos Province are between1943 and 1830 Ma (Gaudette et al., 1996; Tas-sinari et al., 2000), with metamorphism be-tween 1943 and 1883 Ma (Gaudette et al.,1996). Therefore, in detail, the Paleoprotero-zoic intrusive and metamorphic events in

Amazonia are generally older than those in theArequipa-Antofalla Basement (Fig. 6).

From a chronological perspective, fourproblems exist with correlating the Arequipa-Antofalla Basement with Amazonia: (1) Pa-leoproterozoic events occurred at distinctlydifferent times, (2) the Ventuari-Tapajos Prov-ince (2.0–1.8 Ga) did not experience Sunsas(1.3–1.0 Ga) metamorphism, (3) the Ventuari-Tapajos Province is not juxtaposed against ei-ther the Rondonia–San Ignacio or Sunsasprovinces in Amazonia, and (4) no units in theArequipa-Antofalla Basement have ages thatcorrespond to those of the Rio Negro–JuruenaProvince (1.8–1.5 Ga). The first point couldbe accounted for by diachroneity along thelength of a once-continuous Ventuari-Tapajosprovince or insufficient data from this prov-ince, but the second, third, and fourth wouldrequire a complex fault emplacement scenario.To produce the observed sequence in theArequipa-Antofalla Basement from Amazon-

ia, a piece of the Ventuari-Tapajos Provincewould have to have been emplaced next to theRondonia–San Ignacio margin during ca. 1.5–1.4 Ga magmatism, rotated and emplaced nextto the pre–Sunsas margin, deformed with out-board rocks during the Sunsas Orogeny, de-tached, and then rotated into its current posi-tion along the modern margin. Althoughphysically possible, it is a complex sequenceof events involving a significant amount oftranscurrent and rotational motion for whichthere is little evidence in the Arequipa-Antofalla Basement. Additionally, this wouldhave to be accomplished without incorpora-tion of the intervening Rio Negro–JuruenaProvince (1.8–1.5 Ga).

Based on previously published Pb data, sig-natures of the Arequipa-Antofalla Basementand Amazonia appear similar. However, thenewly constrained Pb isotopic signatures ofthe northern and central domains only partial-ly overlap data from Amazonia/Colombia andthe trends of the data sets are distinct (Fig.10). Thus, the Pb isotopic compositions of theArequipa-Antofalla Basement and Amazoniaare not sufficiently similar to provide evidencefor correlation. However, differences in theages and metamorphic histories of the rocksin the two data sets may account for differ-ences in the Pb isotopic compositions. Thus,the current Pb isotopic data sets are not suf-ficiently different to distinguish the Arequipa-Antofalla Basement from Amazonia.

Ambiguity in the interpretation of the ex-isting Pb isotopic data precludes the use of Pbisotopes to constrain the relationship of theArequipa-Antofalla Basement to Amazonia.Thus, we rely only on the chronological com-parison. Given the required complex historyand differences in the Paleoproterozoic ages,we prefer a model in which the Arequipa-Antofalla Basement is allochthonous toAmazonia.

Timing of Arrival of the Arequipa-Antofalla Basement

If the Arequipa-Antofalla Basement werenot derived from Amazonia, then it must havecollided with Amazonia and would likely havecaused deformation and metamorphism in theArequipa-Antofalla Basement (AAB). Assuch, there are three potential times for accre-tion: (1) Paleoproterozoic (D1/M1AAB; 1.82–1.80 Ga), (2) Mesoproterozoic (D2/M2AAB;1.20–0.94 Ga), and (3) Ordovician/Silurian(D3/M3AAB; 0.50–0.43 Ga) (Fig. 6). Lookingoutside the Arequipa-Antofalla Basement, Ra-mos (1988) and Ramos et al. (1993) notedNeoproterozoic-Cambrian deformation to the





Figure 10. Comparison of uranogenic Pb compositions from Amazonian and Arequipa-Antofalla Basement. Amazonia Pb data are from metamorphosed ca. 1.7–1.6 Ga basementrocks from Rio Negro–Juruena Province (RNJ) (Tassinari, 1984), ca. 1.6–1.0 Ga A-typegranites from RNJ and Rondonia–San Ignacio Province (RSI) (Tosdal and Bettencourt,1994), feldspars from a 1.45 Ga batholith in RSI with predicted modern whole-rock com-positions (Geraldes et al., 2001) and ca. 1.1 Ga rocks from Colombian Garzon and SantaMarta inliers (Ruiz et al., 1999). Independent evidence suggests that rocks in Proterozoicinliers of Colombia are along-strike equivalents of Sunsas Province of Amazonia (Alvarez,1981; Kroonenberg, 1982; Alvarez, 1984; Litherland et al., 1989; Priem et al., 1989;Restrepo-Pace et al., 1997). This hypothesis, although not directly testable by mapping, issupported by colinearity of Pb data from deformed Sunsas-aged rocks in Colombian in-liers with similar-aged A-type intrusive rocks in Amazonia. Individual data points fornorthern and southern domains of Arequipa-Antofalla Basement and Amazonian A-typegranites are not shown. Stacey and Kramers’ (1975) curve provided for reference withages in Ga.

east of the Arequipa-Antofalla Basement andproposed accretion of the Arequipa-AntofallaBasement during the Pampean Orogeny (0.6–0.5 Ga). Given arguments above for the in-dependent growth history of the Arequipa-Antofalla Basement between 1.9 and 1.0 Ga,accretion during the Paleoproterozoic is im-mediately discounted. Accepting this leavesonly the Mesoproterozoic Sunsas, the Neopro-terozoic Pampean, and the Ordovician Fama-tinian orogenies as possible times of accretion.These are discussed below, from youngest tooldest.

Famatinian AccretionThe consumption of ocean crust beneath the

South American plate related to the approachof the Precordillera Terrane fueled continental,subduction-related magmatism along the west-ern margin of South America (i.e., the Fa-matina province and the western Sierras Pam-peanas) between 515 and 450 Ma (Astini etal., 1995; Pankhurst et al., 1998; Saavedra etal., 1998; Quenardelle and Ramos, 1999). Theeventual collision of the Precordillera Terrane

caused widespread metamorphism between470 and 450 Ma (Pankhurst et al., 1998; Ra-mos et al., 1998; Quenardelle and Ramos,1999). In the Arequipa-Antofalla Basement,Famatinian arc rocks and related metamor-phism were previously recognized only in thesouthern domain. Our work identifies thismagmatism and metamorphism across the en-tire Arequipa-Antofalla Basement (Fig. 6), re-quiring the entire Arequipa-Antofalla Base-ment to have been part of the Famatiniancontinental arc founded on the western marginof South America by 515 Ma (Fig. 9). Con-sequently, accretion of the Arequipa-AntofallaBasement must have been pre-Famatinian.

Pampean AccretionWe believe that several lines of evidence

argue against accretion during the Neoproter-ozoic to early Cambrian Pampean Orogeny.First, no evidence of late Neoproterozoic toearly Cambrian metamorphism has been iden-tified anywhere in the Arequipa-AntofallaBasement (Fig. 6). Only the intrusion of dacitedikes in Quebrada Choja at 635 6 5 Ma and

a K-Ar age of 530 6 30 Ma from a meta-granite at the base of the San Andreas bore-hole in Bolivia (Lehmann, 1978) are roughlycoeval with the Pampean orogeny (0.7–0.5Ga). Second, paleomagnetic data suggest thatthe Arequipa-Antofalla Basement lay alongthe southwest margin of Amazonia throughoutthe Pampean orogeny (Forsythe et al., 1993).This interpretation is consistent with currentmodels explaining Pampean orogenesis thatmainly assume the Arequipa-Antofalla Base-ment to have been proximal to Amazonia atthe onset of this event (Rapela et al., 1998;Alkmim et al., 2001).

Sunsas AccretionElimination of all other possible collisions

implies that accretion of the Arequipa-Antofalla Basement to Amazonia occurredduring the Sunsas Orogeny (1.20–0.94 Ga),consistent with the current position of theArequipa-Antofalla Basement adjacent to thecollisional Sunsas Province (1.3–1.0 Ga) insouthwestern Amazonia. Two scenarios arepossible: (1) the Arequipa-Antofalla Basementcollided as part of a larger continent, or (2) itcollided as a microcontinent. The pattern ofpre-collisional growth from north to south,parallel to the length of the Arequipa-Antofalla Basement, suggests that the Arequipa-Antofalla Basement evolved within a largercontinent rather than as a microcontinent.Thus, we prefer the first scenario. Potentialcorrelations with other continents are dis-cussed in Loewy et al. (2003).

CONCLUSIONS AND IMPLICATIONS

In summary, we interpret that the Arequipa-Antofalla Basement formed as a single base-ment block. It comprises three age domainsthat young from north to south. An integrateddata set that includes precise U/Pb geochro-nology, growth polarity, and the sequence ofadjacent provinces suggests that the Arequipa-Antofalla Basement was not Amazonian andwas, therefore, accreted. We believe that dock-ing occurred during the Sunsas Orogeny at ca.1.05 Ga, when the combined juvenile ca. 1.9–1.8 and 1.5–1.4 Ga crust of the Arequipa-Antofalla Basement collided with Amazonia.The convergence and collision caused mag-matism and metamorphism in the Arequipa-Antofalla Basement and Sunsas Province. TheArequipa-Antofalla Basement likely collidedas part of a larger craton and, as such, it maybe a tectonic tracer (Dalziel, 1993) holdingclues to the identity of its parent craton.




LOEWY et al.

ACKNOWLEDGMENTS

This study was partially funded by National Sci-ence Foundation grant EAR 94 18236. Additionalsupport was provided by The University of TexasGeology Foundation, The University of Texas In-stitute of Latin American Studies, South CentralSection of the Geological Society of America, andthe Tectonics Special Research Centre at the Uni-versity of Western Australia. Field logistics greatlybenefited from the assistance of Rio Tinto Miningand Exploration Limited in northern Chile, thanksto Alasdaire Pope, George Steele, and OsvaldoPonce. We thank Victor Ramos of Rio Tinto andPatrick Mickler for their invaluable assistance in thefield, Kathryn Manser and Todd Housh for their as-sistance in the isotopic facilities, Lisa Gahagan forhelp with Rodinia reconstructions using PLATES,and Eric North and Johnathan Bumgarner for theirassistance with sample processing.

Sharon Mosher, Randy Van Schmus, and SamMukasa provided helpful and constructive reviewsof previous drafts. Comments from journal review-ers Kent Condie and Lang Farmer and associate ed-itor J. Doug Walker greatly enhanced the finalmanuscript.

REFERENCES CITED

Abbruzzi, J., Kay, S.M., and Bickford, M.E., 1993, Impli-cations for the nature of the Precordilleran basementrock Precambrian xenoliths in Miocene volcanicrocks, San Juan Province, Argentina: XII CongresoGeologico Argentino y II Congreso de Exploracion deHidrocarburos Actas, v. III, p. 331–339.

Aitcheson, S.J., Harmon, R.S., Moorbath, S., Schneider, A.,Soler, P., Soria-Escalante, E., Steele, G., Swainbank,I., and Worner, G., 1995, Pb isotopes define basementdomains of the Altiplano, central Andes: Geology,v. 23, p. 555–558.

Alkmim, F.F., Marshak, S., and Fonseca, M.A., 2001, As-sembling west Gondwana in the Neoproterozoic:Clues from the Sao Francisco craton region, Brazil:Geology, v. 29, p. 319–322.

Alvarez, J., 1981, Determinacıón de edad Rb/Sr en rocasdel Macizo de Garzon, Cordillera Oriental de Colom-bia: Geologıá Norandina, v. 4, p. 31–38.

Alvarez, J., 1984, Una edad K/Ar del Macizo de Garzon,Departamento del Huila (Colombia): Geologıá Nor-andina, v. 7, p. 35–38.

Astini, R.A., Benedetto, J.L., and Vaccari, N.E., 1995, Theearly Paleozoic evolution of the Argentine Precordil-lera as a Laurentian rifted, drifted, and collided ter-rane: A geodynamic model: Geological Society ofAmerica Bulletin, v. 107, p. 253–273.

Bahlburg, H., and Herve, F., 1997, Geodynamic evolutionand tectonostratigraphic terranes of northwestern Ar-gentina and northern Chile: Geological Society ofAmerica Bulletin, v. 109, p. 869–884.

Barreiro, B.A., and Clark, A.H., 1984, Lead isotopic evi-dence for evolutionary changes in magma-crust inter-action, Central Andes, southern Peru: Earth and Plan-etary Science Letters, v. 69, p. 30–42.

Bennett, V.C., and DePaolo, D.J., 1987, Proterozoic crustalhistory of the western United States as determined byneodymium isotopic mapping: Geological Society ofAmerica Bulletin, v. 99, p. 674–685.

Bock, B., Bahlburg, H., Worner, G., and Zimmermann, U.,2000, Tracing crustal evolution in the southern centralAndes from Late Precambrian to Permian with Geo-chemical and Nd and Pb isotope data: The Journal ofGeology, v. 108, p. 515–535.

Caldas, J., 1978, Geologıá de los Cuadrangulos de SanJuan, Acarı´ y Yauca, Hojas: 31-m, 31-n, 32-n: Lima,Peru, Instituto de Geologia y Mineria, 78 p.

Cobbing, E.J., Ozard, J.M., and Snelling, N.J., 1977, Re-connaissance geochronology of the crystalline base-ment rocks of the coastal Cordillera of southern Peru:

Geological Society of America Bulletin, v. 88,p. 241–246.

Coira, B., Davidson, J., Mpodozis, C., and Ramos, V.,1982, Tectonic and magmatic evolution of the Andesof northern Argentina and Chile, in Linares, E., Cor-dani, V.G., and Munizago, F., eds., Magmatic evolu-tion of the Andes: Earth Science Reviews, v. 18,p. 303–332.

Connelly, J.N., 2001, Constraining the timing of metamor-phism: U-Pb and Sm-Nd ages from a transect acrossthe northern Torngat orogen, Labrador, Canada: TheJournal of Geology, v. 109, p. 57–77.

Cordani, U.G., Sato, K., Teixeira, W., Tassinari, C.C.G., andBasei, A.S., 2000, Crustal evolution of the SouthAmerican platform, in Cordani, U.G., Milani, E.J., Fil-ho, A.T., and Campos, D.A., eds., Tectonic evolutionof South America: Rio de Janeiro, 31st InternationalGeological Congress, p. 19–40.

Dalmayrac, B., Lancelot, J.R., and Leyreloup, A., 1977,Two-billion-year granulites in the Late Precambrianmetamorphic basement along the southern Peruviancoast: Science, v. 198, p. 49–51.

Dalziel, I.W.D., 1992, On the organization of Americanplates in the Neoproterozoic and the breakout of Lau-rentia: GSA Today, v. 2, p. 237, 240–241.

Dalziel, I.W.D., 1993, Tectonic tracers and the origin of theProto-Andean margin: XII Congreso Geologico Ar-gentino y II Congreso de Exploracion de Hidrocar-buros Actas, v. III, p. 367–374.

Dalziel, I.W.D., 1994, Precambrian Scotland as a Laurentia-Gondwana link: Origin and significance of cratonicpromontories: Geology, v. 22, p. 589–592.

Dalziel, I.W.D., 1997, Neoproterozoic-Paleozoic geographyand tectonics: Review, hypothesis, environmentalspeculation: Geological Society of America Bulletin,v. 109, p. 16–42.

Dalziel, I.W.D., Dalla Salda, L.H., Cingalani, C., and Palm-er, A.R., 1996, The Argentine Precordillera: A Lau-rentian terrane?: GSA Today, v. 6, p. 16–18.

Damm, K.W., Pichowiak, S., Harmon, R.S., Todt, W., Kel-ley, S., Omarini, R., and Niemeyer, H., 1990, Pre-Mesozoic evolution of the central Andes; The base-ment revisited, in Kay, S.M., and Rapela, C.W., eds.,Plutonism from Antarctica to Alaska: Boulder, Colo-rado, Geological Society of America, p. 101–125.

Davis, D.W., 1982, Optimum linear regression and errorestimation applied to U-Pb data: Canadian Journal ofEarth Sciences, v. 19, p. 2141–2149.

DePaolo, D.J., 1981, A neodymium and strontium isotopicstudy of the Mesozoic calc-alkaline granitic batholithsof the Sierra Nevada and Peninsular Ranges, Califor-nia: Journal of Geophysical Research, v. 86,p. 10470–10488.

Fairchild, I.J., and Hambrey, M.J., 1984, The Vendian suc-cession of northeastern Spitsbergen: Petrogenesis of adolomite-tillite association: Precambrian Research,v. 26, p. 111–167.

Farmer, G.L., and DePaolo, D.J., 1984, Origin of Meoszoicand Tertiary granite in the western United States andimplications for pre-Mesozoic Crustal Structure 2.Nd and Sr isotopic studies of unmineralized and Cu-and Mo-mineralized granite in the Precambrian cra-ton: Journal of Geophysical Research, v. 89,p. 10,141–10,160.

Forsythe, R.D., Davidson, J., Mpodozis, C., and Jesinkey,C., 1993, Lower Paleozoic relative motion of the Ar-equipa Block and Gondwana; Paleomagnetic evidencefrom Sierra de Almeida of northern Chile: Tectonics,v. 12, p. 219–236.

Gaudette, H.E., Olszewski, W.J., and Santos, O.S., 1996,Geochronology of Precambrian rocks from the north-ern part of the Guiana shield, state of Roraima, Brazil:Journal of South American Earth Sciences, v. 9,p. 183–195.

Geraldes, M.C., Van Schmus, W.R., Condie, K.C., Bell, S.,Teixeira, W., and Babinski, M., 2001, Proterozoic geo-logic evolution of the SW part of Amazonian cratonin Mato Grosso state, Brazil: Precambrian Research,v. 111, p. 91–128.

Harland, W.B., Herod, K.N., and Krinsley, D.H., 1966, Thedefinition and identification of tills and tillites: EarthScience Reviews, v. 2, p. 225–256.

Injoque, J., and Romero, L., 1986, Estromatolitos (?) en laformacion San Juan, San Juan de Marcona: Evidenciade estructuras fosiles Precambricas en el Peru: De ReMetallica, v. Enero-Febrero, p. 4–5.

Kay, S.M., Orrell, S., and Abbruzzi, J.M., 1996, Zircon andwhole rock Nd-Pb isotopic evidence for a Grenvilleage and a Laurentian origin for the basement of thePrecordillera in Argentina: Journal of Geology, v. 104,p. 637–648.

Keppie, J.D., and Bahlburg, H., 1999, Puncoviscana For-mation of northwestern and central Argentina: Passivemargin or foreland basin deposit?, in Ramos, V.A.,and Keppie, J.D., eds., Laurentia-Gondwana Connec-tions before Pangea: Boulder, Colorado, GeologicalSociety of America, p. 139–144.

Kroonenberg, S., 1982, A Grenvillian granulite belt in theColombian Andes and its relations to the GuianaShield: Geologia Mijnbouw, v. 61, p. 325–333.

Lehmann, B., 1978, A Precambrian core sample from theAltiplano/Bolivia: Geologichse Rundschau, v. 67,p. 270–278.

Litherland, M., Annells, R.N., Darbyshire, D.P.F., Fletcher,C.J.N., Hawkins, M.P., Klinck, B.A., Mitchell, W.I.,O’Connor, E.A., Pitfield, P.E.J., Power, G., and Webb,B.C., 1989, The Proterozoic of eastern Bolivia and itsrelationship to the Andean mobile belt: PrecambrianResearch, v. 43, p. 157–174.

Loewy, S., Connelly, J.N., Dalziel, I.W.D., and Gower, C.F.,2003, Eastern Laurentia in Rodinia: Constraints fromwhole-rock Pb and U-Pb geochronology, in Sircombe,K.N. and McElhinny, M.W., eds, Orogenic belts, re-gional and global tectonics: A memorial volume toChris McAulay Powell: Tectonophysics, v. 375,p. 169–197.

Lork, A., and Bahlburg, H., 1993, Precise U-Pb ages ofmonazites from the Faja Eruptiva de la Puna Orientaland the Cordillera Oriental, NW Argentina: XII Con-greso Geologico Argentino y II Congreso de Explor-acion de Hidrocarburos Actas, v. IV, p. 1–6.

Mpodozis, C., Herve, F., Davidson, J., and Rivano, S.,1983, Los granitoides de Cerro de Lela, manifesta-ciones de un episodio intrusivo y termal del Paleo-zoico Inferior en los Andes del Norte de Chile: Re-vista Geologica de Chile, v. 18, p. 3–14.

Mukasa, S.B., and Henry, D.J., 1990, The San Nicolasbatholith of coastal Peru: Early Palaeozoic continentalarc or continental rift magmatism?: Journal of theGeological Society of London, v. 147, p. 27–39.

Murphy, J.B., and Nance, R.D., 2002, Sm-Nd isotopic sys-tematics at tectonic tracers: An example from westAvalonia in the Canadian Appalachians: Earth-Science Reviews, v. 59, p. 77–100.

Nur, A., and Ben-Avraham, Z., 1982, Oceanic plateaus,the fragmentation of continents, and mountain build-ing: Journal of Geophysical Research, v. 87,p. 3644–3661.

Omarini, R., Viramonte, J., Cordani, U., Salfity, J., and Ka-washita, K., 1984, Estudio geochronologico Rb-Sr dela Faja Eruptiva de la Puna en el sector de San An-tonio de los Cobres, Provincia de Salta: IX CongresoGeologico Argentino Actas, v. 3, p. 146–158.

Pacci, D., Herve, F., Munizaga, F., Kawashita, K., and Cor-dani, U., 1980, Acerca de la edad Rb-Sr Precambricade rocas de la Formacion Esquistos de Belen, Depar-tamento de Parinacota, Chile: Revista Geologica deChile, v. 11, p. 43–50.

Pankhurst, R.J., and Rapela, C.W., 1998, The proto-Andeanmargin of Gondwana: An introduction, in Pankhurst,R.J., and Rapela, C.W., eds., The Proto-Andean mar-gin of Gondwana, The Geological Society of London,p. 1–10.

Pankhurst, R.J., Rapela, C.W., Saavedra, J., Baldo, E.,Dahlquist, J., Pascua, I., and Fanning, C.M., 1998,The Famatinian magmatic arc in the central SierrasPampeanas: An Early to Mid-Ordovician continentalarc on the Gondwana margin, in Pankhurst, R.J., andRapela, C.W., eds., The Proto-Andean margin ofGondwana: London, Geological Society [London],p. 343–368.

Payolla, B.L., Bettencourt, J.S., Kozuch, M., Leite, W.B.,Jr., Fetter, A.H., and Van Schmus, W.R., 2002, Geo-logical evolution of the basement rocks in the east-





central part of Rondonia Tin Province, SW Amazoniacraton, Brazil: U-Pb and Sm-Nd isotopic constraints:Precambrian Research, v. 119, p. 141–169.

Priem, H.N.A., Kroonenberg, S.B., Boelrijk, N.A.I.M., andHebeda, E.H., 1989, Rb-Sr and K-Ar evidence for thepresence of a 1.6 Ga basement underlying the 1.2 GaGarzon-Santa Marta granulite belt in the ColombianAndes: Precambrian Research, v. 42, p. 315–324.

Quenardelle, S., and Ramos, V.A., 1999, Ordovician west-ern Sierras Pampeanas magmatic belt: Record of Pre-cordillera accretion in Argentina, in Ramos, V.A., andKeppie, J.D., eds., Laurentia-Gondwana connectionsbefore Pangea: Boulder, Colorado, Geological Societyof America, p. 63–86.

Ramos, V.A., 1988, Late Proterozoic–Early Paleozoic ofSouth America—A collisional history: Episodes,v. 11, p. 168–174.

Ramos, V.A., Dallmeyer, R.D., and Vujovich, G.I., 1998,Time constraints on the early Paleozoic docking of thePrecordillera, central Argentina, in Pankhurst, R.J.,and Rapela, C.W., eds., The Proto-Andean Margin ofGondwana: London, Geological Society [London],p. 143–158.

Ramos, V.A., Jordan, T.E., Allmendinger, R.W., Mpodozis,C., Kay, S.M., Cortes, J.M., and Palma, M., 1986,Paleozoic terranes of the central Argentine-ChileanAndes: Tectonics, v. 5, p. 855–880.

Ramos, V.A., Vujovich, G., Kay, S.M., and McDonough,M., 1993, La orogenesis de Grenville en las SierrasPampeanas Occidentales: La Sierra de Pie de Palo ysu integracion al supercontinente Proterozoico: XIICongreso Geologico Argentino y II Congreso de Ex-ploracion de Hidrocarburos Actas, v. T. III,p. 343–357.

Rapalini, A.E., Astini, R.A., and Conti, C.M., 1999, Paleo-magnetic constraints on the evolution of Paleozoicsuspect terranes from southern South America, inRamos, V.A., and Keppie, J.D., eds., Laurentia-Gondwana Connections before Pangea: Boulder, Col-orado, Geological Society of America, p. 171–182.

Rapela, C.W., Pankhurst, R.J., Casquet, C., Baldo, E., Saa-vedra, J., and Galindo, C., 1998, Early evolution of

the Proto-Andean margin of South America: Geology,v. 26, p. 707–710.

Restrepo-Pace, P.A., Ruiz, J., Gehrels, G., and Cosca, M.,1997, Geochronology and Nd isotopic data fromGrenville-age rocks in the Colombian Andes: Newconstraints for Late Proterozoic-Early Paleozoic pa-leocontinental reconstructions of the Americas: Earthand Planetary Science Letters, v. 150, p. 427–441.

Ruiz, J., Tosdal, R.M., Restrepo-Pace, P.A., and Murillo-Muneton, G., 1999, Pb isotope evidence for Colombia-southern Mexico connections in the Proterozoic, inRamos, V.A., and Keppie, J.D., eds., Laurentia-Gond-wana connections before Pangea: Boulder, Colorado,Geological Society of America, p. 183–199.

Saavedra, J., Toselli, A., Rossi, J., Pellitero, E., and Durand,F., 1998, The Early Paleozoic magmatic record of theFamatina System: A review, in Pankhurst, R.J., andRapela, C.W., eds., The Proto-Andean Margin ofGondwana: London, Geological Society [London],p. 283–296.

Sadowski, G.R., and Bettencourt, J.S., 1996, Mesoproter-ozoic tectonic correlations between eastern Laurentiaand the western border of the Amazon Craton: Pre-cambrian Research, v. 76, p. 213–227.

Shackleton, R.M., Ries, A.C., Coward, M.P., and Cobbold,P.R., 1979, Structure, metamorphism and geochronol-ogy of the Arequipa Massif of coastal Peru: Geolog-ical Society [London] Journal, v. 136, p. 195–214.

Singletary, S.J., Hanson, R.E., Martin, M.W., Crowley, J.L.,Bowring, S.A., Key, R.M., Ramokate, L.V., Direng,B.B., and Krol, M.A., 2003, Geochronology of base-ment rocks in the Kalahari Desert, Botswana, and im-plications for regional Proterozoic tectonics: Precam-brian Research, v. 121, p. 47–71.

Sinha, A.K., and McLelland, J.M., 1997, Lead isotope map-ping of crustal reservoirs within the Grenville Super-terrane: II. Adirondack Massif, New York, in Sinha,A.K., ed., Basement tectonics 13: Proceedings of theThirteenth International Conference on Basement Tec-tonics: Dordrecht, Kluwer Academic Publishers,p. 297–312.

Stacey, J.S., and Kramers, J.D., 1975, Approximation ofterrestrial lead isotope evolution by a two-stage mod-

el: Earth and Planetary Science Letters, v. 26,p. 207–221.

Tassinari, C.C.G., 1984, A porcao ocidental do CratonAmazonico: Evidencias isotopicas de Acrecao conti-nental no Proterozoico Medio: II Symposium Ama-zonico Anais, p. 439–446.

Tassinari, C.C.G., Bettencourt, J.S., Geraldes, M.C., Ma-cambira, M.J.B., and Lafon, J.M., 2000, The Amazoncraton, in Cordani, U.G., Milani, E.J., Filho, A.T., andCampos, D.A., eds., Tectonic evolution of SouthAmerica: Rio de Janeiro, 31st International GeologicalCongress, p. 41–95.

Tilton, G.R., and Barreiro, B.A., 1980, Origin of lead inAndean calc-alkaline lavas, southern Peru: Science,v. 210, p. 1245–1247.

Tosdal, R.M., 1996, The Amazon-Laurentia connection asviewed from the Middle Proterozoic rocks in the cen-tral Andes, western Bolivia and northern Chile: Tec-tonics, v. 15, p. 827–842.

Tosdal, R.M., and Bettencourt, J.S., 1994, U-Pb zircon agesand Pb isotopic compositions of Middle ProterozoicRondonian massifs, southwestern margin of the Am-azon craton, Brazil, in VII Congreso Geologico Chi-leno: Chile, Universidad de Concepcion in Concep-cion, p. 1538–1541.

Wasteneys, H.A., Clark, A.H., Farrar, E., and Langridge,R.J., 1995, Grenvillian granulite-facies metamorphismin the Arequipa Massif, Peru: A Laurentia-Gondwanalink: Earth and Planetary Science Letters, v. 132,p. 63–73.

Worner, G., Lezuan, J., Beck, A., Heber, V., Lucassen, R.,Zinngrebe, E., Rossling, R., and Wilke, H.G., 2000,Geochronology, petrology and geochemistry of base-ment rocks from Belen (N. Chile) and C. Uyarani (W.Bolivian Altiplano): Implications for the evolution ofthe Andean basement: Journal of South AmericanEarth Sciences, v. 13, p. 717–737.

Worner, G., Moorbath, S., and Harmon, R.S., 1992, AndeanCenozoic volcanic centers reflect basement isotopicdomains: Geology, v. 20, p. 1103–1106.

MANUSCRIPT RECEIVED BY THE SOCIETY 1 AUGUST 2002REVISED MANUSCRIPT RECEIVED 27 JUNE 2003MANUSCRIPT ACCEPTED 6 JULY 2003

Printed in the USA



Geological Society of America Bulletin 2004 Loewy 171 87

Documents

rio de janeiro

planetary

single grain

western united

rio negrojuruena

single crustal

bound intrusive

crustal evolution