Antamina Love

0361-0128/01/3439/887-30 $6.00 887

The Lithologic, Stratigraphic, and Structural Setting of the Giant Antamina Copper-Zinc Skarn Deposit, Ancash, Peru

DAVID A. LOVE, ALAN H. CLARK,Department of Geological Sciences and Geological Engineering, Queens University, Kingston, Ontario, Canada K7L 3N6

AND J. KEITH GLOVER*Glover Consulting Ltd., 146 Simcoe St., Victoria, B.C., Canada V8V 1K4

AbstractAntamina, located at latitude 9 32' S and longitude 77 03' W in the Ancash Department of north-central

Peru, is the largest known Cu-Zn skarn ore deposit. It incorporates a mineral reserve of 561 Mt, which has anaverage grade of 1.24 percent Cu, 1.03 percent Zn, 13.71 g/t Ag and 0.029 percent Mo, calculated at a 0.7 per-cent Cu equiv cutoff grade. The grandite-dominated calcic skarn formed in and around an upper Miocene por-phyritic monzogranite stock emplaced into Upper Cretaceous carbonate strata that had experienced thin-skinned, northeast-verging thrusting and folding in the late Eocene Incaic orogeny. The exoskarn Cu-Zn ore isdiscordant to the strata of the Jumasha and overlying Celendn Formations, which comprise, respectively, mas-sive to thick-bedded, relatively pure limestones and thin-bedded, predominantly marly limestones. The Ju-masha Formation, the upper contact of which is locally defined as the top of the uppermost thick-bedded lime-stone or marble unit, hosts approximately three-quarters of the known exoskarn. Approximately the samefraction of the contiguous endoskarn Cu ore occurs adjacent to this formation. The overlying Celendn For-mation is less extensively mineralized but, because it is widely metamorphosed to hornfels and locally con-verted to diopsidic skarnoid, may have inhibited the upward and outward migration of hydrothermal fluids,thereby promoting the development of the unusually large endoskarn ore zone. Ore also occurs in late hy-drothermal breccias emplaced during the formation of mineralized endoskarn.

The preskarn thermal metamorphic aureole around the ore deposit is expressed differently in the two hostformations. Jumasha Formation limestone is coarsened and bleached to banded gray marble and locally towhite marble peripheral to the intrusion and skarn. Minor scapolite occurs in dark gray bands in marble, con-centrated in a discontinuous halo tens of meters wide and commonly separated from the skarn by tens of me-ters. Three facies of calc-hornfels are recognized in the marl beds of the Celendn Formation adjacent to theintrusion extending hundreds of meters beyond sulfide-bearing skarn: a peripheral, very fine grained, lightbrown phlogopitic facies; an intermediate, fine-grained, gray tremolitic facies; and a proximal, medium-grained,light green diopsidic facies. At an XCO2 of 0.1 to 0.9 and P = 100 MPa, these zones reflect temperatures in-creasing to circa 495C adjacent to the intrusion. In addition, in nodular beds of the Celendn Formation thathave been metamorphosed to hornfels, diagenetic calcite nodules are selectively replaced by diopside for dis-tances of tens of meters beyond the skarn front. Such calc-silicate formation through both metamorphism andmetasomatism, together with a 9 km2 cluster of Pb-Zn-Ag vein deposits, provides district-scale vectors to ore.

The Antamina deposit lies on a newly recognized cross-strike structural discontinuity in the segmented In-caic Maran thrust and fold belt, the northeast-trending Querococha arch. Southeast of the arch, Incaic foldsand thrust faults strike north-northwest, but northwest of the arch they strike northerly. The plunge of fold axesconcomitantly changes from south-southeast to north. Stratigraphic relationships indicate that the arch was apaleohigh, at least in the Jurassic and possibly throughout the late Paleozoic-early Mesozoic interval. The mid-dle Miocene Carhuish pluton is exposed on the arch 30 km southwest of Antamina, whereas coeval Calipuy Su-pergroup volcanic units lie at similar altitudes to the north and south. Only scattered hydrothermal centers oflate Miocene age are known in the Cordillera Negra, but an apparent swarm of intrusions, including the Anta-mina stock, occurs along the Querococha arch.

Antamina is situated where the locus of changes in the strike of folds and faults and the plunge of folds stepsleft along the arch. At Antamina, a pair of fault-bend folds above frontal thrust ramps show approximately 500m of dextral apparent offset across the deposit and are inferred to have been separated by a northeast-strik-ing transfer fault or lateral ramp, itself localized by a left-stepping jog in the Valley fault, an underlying, sim-ilarly oriented transverse structure. The jog in the Valley fault is inferred to have also controlled intrusion andskarn development. This local-scale jog in the Valley fault mimics the regional step along the arch. The archmay reflect a transform segment of the originally jagged, rifted continental margin, which persisted as a trans-verse basement weakness. Northeast-striking, originally sinistral, basement structures affected regional-scalesedimentation and structural patterns, including articulation of the thrust and fold belt. At a local scale, they

2004 by Economic GeologyVol. 99, pp. 887916

Corresponding author: e-mail, [email protected]*In Memoriam. On July 17, 2001, Keith Glover died unexpectedly but peacefully in his sleep, immediately after returning from fieldwork. He had an in-

fectious passion for rocks and great patience in teaching about them. He was a skilled structural geologist and a professional ore deposit specialist with acuteperception and a love for walking the rocks. Keith had been consulting internationally for 14 years and was respected for his intellect and breadth of geolog-ical understanding. He also managed admirably to balance his love for geology with that for his family. Keith was well liked for his generous spirit and hu-manity, and he is greatly missed.

IntroductionDESPITE their potential ore genetic and metallogenic impor-tance, the lithologic, stratigraphic, and structural settings ofskarn mineralization have rarely been comprehensively docu-mented. It is therefore difficult to assess their influence onthe localization of skarn-generating hydrothermal systemsand, in particular, to envisage the specific environments inwhich exceptional deposits have developed. In this paper, wedescribe the host rocks and structural relationships of the An-tamina Cu-Zn(-Ag-Mo) deposit, north-central Peru, and pro-pose a model for the stratigraphic and tectonic environmentin which this largest known Cu-Zn skarn orebody formed.These aspects are controversial, in part because of the poorlydefined local stratigraphic succession and because of the de-formation, metamorphism, and metasomatism imposed onthe ore-hosting strata. Following a brief summary of the geol-ogy of the deposit, we document the regional-scale (ca. 5,000km2) geologic and geodynamic setting of the mineralizationbefore focusing on the district scale (ca. 120 km2).

Copper mineralization was known at Antamina (anta: cop-per in Quechua) in pre-Colonial times, but only modestamounts of Pb and Ag are known to have been produced inthe district prior to 2001 (Redwood, 1999). The skarn con-tains proven and probable reserves of 561 Mt with an averagegrade of 1.24 percent Cu, 1.03 percent Zn, 13.71 g/t Ag, and0.029 percent Mo (calculated at a 0.7% Cu equiv cutoffgrade). Compaa Minera Antamina S.A., which operates theAntamina open-pit mine, is owned by BHP Billiton (33.75%),Noranda (33.75%), Teck Cominco (22.5%), and Mitsubishi(10%). Production of copper-silver and zinc concentrates, aswell as lead, molybdenum, and bismuth byproducts, began inJuly 2001 (Zuzunaga, 2003).

The Antamina mine is located at approximately 9 32' S and77 03' W, 270 km north of Lima and 130 km from the Pacificcoast, in Ancash Department in north-central Peru (Fig. 1). Itlies in the eastern part of the Cordillera Occidental, east ofthe Cordillera Blanca and west of the Ro Maran valley.The skarn is exposed between approximately 4,200 and 4,800m a.s.l., at the head of a southwest-draining glacial valley, butprior to mining much of the orebody was covered by LagoAntamina, a glacial tarn (Fig. 2). The history of exploration atAntamina and the general geology of the deposit are summa-rized by Redwood (1998, 1999). OConnor (2000) reviewedthe geologic and geophysical approaches that delineated theore and outlined the development of the mine, metallurgicaltesting, and resource calculations. Both authors generallysupported previous geologic descriptions and interpretations(e.g., Petersen, 1965).

The Antamina deposit formed at 9.86 to 10.18 Ma(40Ar/39Ar step-heating data of Love et al., 2003) around asmall monzogranitic porphyry intrusion. It is hosted by UpperCretaceous carbonate strata within the Maran thrust andfold belt, formed by the late Eocene Incaic orogeny (Fig. 1;Noble et al., 1979; Mgard, 1984). The western Andes ofPeru were the site of episodic arc magmatism from the LateTriassic to the late Miocene (Cobbing et al., 1981). However,

from latitudes 2 S to 15 S they are now underlain by a flatsubduction zone widely ascribed to underthrusting of theNazca Ridge (Barazangi and Isacks, 1976; Pilger, 1981; Ham-pel, 2002) and, in the northern part, the postulated IncaPlateau (Gutscher et al., 1999). This major flat-slab domainseparates the Northern and Central Volcanic zones of theAndes and has been apparently amagmatic since emplace-ment of the last phase of the Cordillera Blanca batholith at6.3 to 8.2 Ma (Mukasa, 1984; McNulty et al., 1998). Intrusionand mineralization at Antamina took place shortly before thisterminal magmatism. Although the carbonate rocks that hostthe deposit have long been recognized as Upper Cretaceous,they have been assigned to various formations. The strati-graphic relationships in the mine area are herein clarifiedthrough examination of the carbonate rocks around the skarnand their comparison with well-described measured sectionselsewhere in north-central Peru. This analysis permits bothelucidation of the structure of the area and characterization ofthe types of rocks replaced by the skarn.

The Antamina deposit shares numerous common featureswith other large porphyry-related Cu skarns (Einaudi, 1982a,b), but it differs from most in the exceptional development ofmineralized endoskarn and the association of ore-grade Cu,Zn, and Mo in contiguous zones. An additional unusual fea-ture is the widespread development of chalcopyrite-rich hy-drothermal breccias, the extent and above-average Cu con-tent of which were unrecognized prior to our research. Manyof the observations on the deposit- and district-scale geologyrecorded herein, and their interpretation, were introduced inunpublished reports prepared by the authors (D.A. Love andA.H. Clark, 1998a, b, 2000, unpublished reports to CompanaMinera Antamina S.A., Lima; J.K. Glover, 1998a, 1998b,1998c, unpublished reports to Compana Minera AntaminaS.A., Lima).

The Antamina deposit

The Antamina deposit comprises endoskarn and exoskarn,with subordinate breccia bodies that cut both skarn and in-trusion within the perimeter of skarn. The mineralized skarnis dominated by grandite garnet, which grades from brown togreen nearer the host limestone (Petersen, 1965). Thus thedeposit conforms to the oxidized calcic clan of Einaudi et al.(1981). The geology of the deposit and its immediate sur-roundings is summarized in Figure 3a on three northwest-southeast drill-hole cross sections, spaced 50 m apart,through the middle of the orebody. The simplified surface ge-ology map in Figure 3b was constructed by projecting thelithologic boundaries defined on these and thirty other crosssections. The area delimited by the mineralized skarn front isapproximately 1.18 km2 overall. The skarn zone straddles theoriginal intrusive contact and surrounds a circa 0.24 km2 coreof porphyry that forms a crudely parallelogram-shaped prismwith a vertical axis (Fig. 3b). The outer boundary of the min-eralized skarn is more elongated northeast-southwest thanthis core because it expands around faults and dikes extendingto the east, northeast, and southwest. It is therefore roughly

888 LOVE ET AL.

0361-0128/98/000/000-00 $6.00 888

influenced lateral ramp formation and related fracture development in the overlying thrust sheets. In the pro-posed model, they also localized later uplift and the rapid transit of small volumes of productive melt into ashallow crustal setting, conditions favorable for formation of a giant magmatic-hydrothermal ore deposit.

elliptical in plan and has a northwest-southeast width of up to1,000 m, and a northeast-southwest length of more than 2,500m. The long axis parallels the Antamina valley and is perpen-dicular to the regional structural grain of the deformed car-bonate host rocks. As recognized by Petersen (1965) theouter limit of skarn is generally subvertical. The skarn nar-rows with depth as the core of porphyry widens (Fig. 3a), butthere is no significant change in the Cu and Zn grades todepths of at least 400 m below the original valley floor. Exceptat high elevations in the eastern part of the deposit, the ex-oskarn is almost everywhere mineralized.

In detail, the individual skarn facies and breccia bodies arecomplexly shaped and discontinuous, but the deposit can besimplified as comprising an inner shell of endoskarn, stock-work, breccia, and brown garnet exoskarn that contains thecopper molybdenum ore and an outer shell of green garnet

exoskarn comprising the copper-zinc ore. Molybdenite is dis-seminated in irregular zones within and at the margin of theintrusion. Chalcopyrite is the dominant copper mineral ex-cept at shallow depths in the southwestern part of the deposit,where bornite predominates in wollastonitic exoskarn, whichforms an enclave in green garnet Cu-Zn exoskarn.

Hydrothermal breccia is common at or near the endoskarn-exoskarn contacts along the northwest and southeast sides ofthe deposit. Breccia also cuts the porphyry core as anasto-mosing sheets and pipes that are commonly enveloped byfine-grained maroon garnet endoskarn (Fig. 3). The brecciazones contain minor chlorite and were originally described byPetersen (1965) as chlorite skarn. The breccias are generallypoorly sorted and comprise angular to rounded fragments ofthe skarn and metallic minerals supported in a sand-sizedmatrix of similar composition. As in skarn, sphalerite and

LITHOSTRATIGRAPHY AND STRUCTURE, ANTAMINA COPPER-ZINC DEPOSIT, PERU 889

0361-0128/98/000/000-00 $6.00 889

FIG. 1. Location map of the Antamina deposit and general geology of part of Ancash and La Libertad Departments, Peru.Compiled and modified after Egeler and De Booy (1956), Cosso (1964), Wilson and Reyes (1964), Cosso and Jan (1967),Wilson et al. (1967, 1995), Myers (1976, 1980), Reyes (1980), Snchez (1995), Allende (1996), Cobbing et al. (1996), Jacay(1996), Snchez et al. (1998), INGEMMET (1999), and Strusievicz et al. (2000). The Tapacocha axis delineates the westernedge of the Cretaceous shelf (Myers, 1974, 1975). Other deposits (Pierina and Pasto Bueno) and prospects (Magistral) men-tioned in the text are shown, as are the areas illustrated in Figs. 5 and 6. MTFB = Maran thrust and fold belt.

molybdenite do not commonly occur together in the breccias;sphalerite is found in breccias in or near exoskarn, but molyb-denite occurs in breccias in endoskarn. The common metallicminerals in the breccias are pyrite, chalcopyrite, and mag-netite, which are mostly comminuted but also occur as veins,massive bodies, and large fragments. This mineral assemblageis rare in exoskarn but forms widespread stockworks andsheeted vein swarms in fine-grained maroon garnet en-doskarn, in many places grading into crackle, mosaic, and ma-trix-dominated breccia. We estimate that approximately one-third of the ore at Antamina may have formed during this latebrecciation, veining, and endoskarn-forming stage.

Widely spaced, late calcite-tetrahedrite sphalerite galena veinlets are common and cut all skarn and brecciatypes, although they are also locally dismembered in breccia,probably because of settling. Scarce realgar veinlets occur incalc-hornfels above and peripheral to the skarn.

Regional Geologic SettingThe Upper Cretaceous strata enclosing the Antamina de-

posit are part of a metallogenically important Albian andUpper Cretaceous package of carbonate rocks, the MachayGroup, that hosts many ore deposits in the polymetallic skarnand carbonate-replacement belt of central Peru (Soler et al.,1986). These rocks formed during the later of two Permian toPaleocene episodes of basin development that deposited asuccession of alternating siliciclastic and carbonate facies in

western South America (Sempere et al., 2002). This LateJurassic to Paleogene subsidence formed the West Peruviantrough, which separated the magmatic arcs to the west fromthe eastern geanticline now represented by the Maranmetamorphic complex (Benavides, 1956; Wilson, 1963;Atherton et al., 1983; Mgard, 1987). The Cretaceous sedi-mentary rocks that crop out in the Antamina area accumu-lated in the shallow-water portion of this trough, the Yaulishelf (Szekely, 1967). The Tapacocha axis, now a north-north-westtrending high strain zone, separates the western,deeper-water portion of the trough from the shelf sedimen-tary rocks (Fig. 1; Myers, 1974, 1975).

Stratigraphic relationships

The host Machay Group (Figs. 1 and 4) includes all Albianto mid-Campanian carbonate rocks south of 9 S and east ofthe Tapacocha axis in north-central Peru. Szekely (1967) in-troduced the Machay Group in central Peru and Samam-Boggio (1980) applied it throughout Peru. These rocks alsohave been referred to informally as the middle Cretaceouslimestone series (Harrison, 1940), the upper Cretaceousand Albian carbonate series (Mgard, 1984), and the uppercarbonate sequence (Manrique, 1998). The group is under-lain by a predominantly siliciclastic sequence comprisingUpper Jurassic marine black shales of the Chicama Groupand Lower Cretaceous (Berriasian to Aptian) continental toshelf sandstones, shales, and minor limestones of the Goyllar-isquisga Group (Fig. 4). It is widely overlain, conformably orslightly unconformably, by red beds (Wilson, 1963), which aremainly Campanian to Paleocene but as old as Santonian incentral Peru (Jaillard, 1987). Not preserved in the immediateAntamina area, these nonmarine, coarse clastic rocks havebeen variously described as the Pocabamba Formation, 25km southeast of Antamina at La Unin (Wilson, 1963, afterMcLaughlin, 1924), the Chota Formation, 20 km north of An-tamina (Benavides, 1956: after Broggi, 1942), and the Cas-apalca Group, 25 km southwest of Antamina in the CordilleraHuayhuash (Coney, 1971, after McLaughlin, 1924).

The Machay Group contains two transgressive sequencesseparated by a disconformity ascribed to late-middle Albianuplift and erosion related to the Mochica orogeny (Mgard,1984). In the lower part of the group, the successive Pari-ahuanca, Chulec, and Pariatambo Formations (Fig. 4) recorda transition from near-shore, calcareous sandstone and mas-sive, shelly limestone, through thin-bedded limestone andmarl, to deep-water, thin-bedded, bituminous, dark gray marland limestone (Benavides, 1956, 1999; Wilson, 1963; Jaillard,1987). Following the late-middle Albian hiatus, carbonatesedimentation on the platform resumed with the depositionof the shallow-water, upper Albian to upper Turonian Ju-masha Formation (Jaillard, 1987), originally defined byMcLaughlin (1924) in central Peru. This formation is overlainby the muddier, deeper-water Celendn Formation (Bena-vides, 1956), largely Coniacian to Santonian in age (Jaillard,1987) but attaining the mid-Campanian in northern Peru(Mourier et al., 1988). The upper part of the group, compris-ing the Jumasha and Celendn Formations (Fig. 4), thus rep-resents a second major transgressive sequence. The lower tomiddle Albian carbonate strata (the first transgressive sequence)are similar in lithology and thickness in both northern and

890 LOVE ET AL.

0361-0128/98/000/000-00 $6.00 890

FIG. 2. Premine physiography of the Antamina area, with place names re-ferred to in the text, illustrating the clustering of Ag-bearing Pb-Zn vein de-posits around Lago Antamina documented by Bodenlos and Ericksen (1955).Also indicated are the Contonga Pb-Zn-Cu-Ag mine 5 km to the north-north-west of Lago Antamina and veins about 500 m northeast of Contonga. B =Barrn, C = Casualidad, Cc = Condorcoccha, F = Fortuna, JE = Julia Eloisa,P = Poderosa, Pp = Putapuquio, R = Recompensa, RdO = Rosita de Oro, SF= San Francisco, SR = Santa Rosa, UP = Usu Pallares. The viewpoints forphotographs in Figures 7, 8, 12, and 15 are shown as eyes. (Contour inter-val is 200 m.)


0361-0128/98/000/000-00 $6.00 891

FIG. 3. Simplified cross sections (a) and projected surface geology map (b) of the Antamina deposit, illustrating thecrudely elliptical, vertical zones of endoskarn and exoskarn developed between a core of largely skarn-free porphyry and thelimestone host rocks. Simplified surface geology map is based on drill-hole geology projected to surface and on surface map-ping by D.A.L. and J.K.G., combined with that by L. Hathaway (Inmet) and M. Wunder (Noranda). Prominent northwest-striking Incaic folds and the left-stepping, transverse Valley fault (VF) are indicated. The location of the proposed Valley lat-eral ramp (VLR), inferred to be responsible for the apparent dextral offset of the Antamina anticline (AA), is also shown.

central Peru (Benavides, 1956; Jaillard, 1987). However,north of approximately 9 S, the overlying upper Albian tomid-Campanian carbonate rocks are much thicker, more fos-siliferous, and lithologically more variable than in centralPeru, and the Jumasha Formation interval is divided into fiveformations (Benavides, 1956). Jaillard (1987) provides corre-lations between these stratigraphic sections in northern andcentral Peru.

Tectonic relationships

Published descriptions of the structural setting of the Anta-mina deposit (Bodenlos and Ericksen, 1955; Terrones, 1958;Petersen, 1965; Redwood, 1999) have focused on theMaran thrust and fold belt, which developed circa 30 m.y.before mineralization, with scant consideration of the re-gional tectonic environment that existed in the mid-Miocene.We argue, however, that Antamina lies athwart a large-scale,cross-strike (northeast-southwest) structural discontinuity(Wheeler, 1978) that was tectonically active at the time ofskarn formation and hence has metallogenic significance.

Love et al. (2001) termed the structure the Querococha archbecause its southwestern limit at the margin of the Callejonde Huaylas lies close to Laguna Querococha (Figs. 5 and 6).Its influence on the abundance of Neogene intrusions, the re-gional strikes in the Maran thrust and fold belt, and thestratigraphic relationships of underlying Mississippian toLower Jurassic strata (Fig. 4) is described below.

The overall strike of the Maran thrust and fold beltchanges, and the common plunge directions reverse, acrossthe proposed northeast-trending cross-strike structural dis-continuity (Fig. 5). These thin-skinned Eocene structures, at-tributed to the Incaic orogeny, are the dominant tectonic ele-ments in the region, although two regional unconformities inthe Cretaceous succession in central Peru, one at the base ofthe Jumasha Formation and the other below the CasapalcaGroup and its equivalents, represent earlier emergence dur-ing, respectively, the Mochica and Peruvian tectonic phases(Noble et al., 1979; Mgard, 1984). The Maran thrust andfold belt extends from 5 S to 12 30' S, generally strikingnorth-northwestsouth-southeast, parallel to the present

892 LOVE ET AL.

0361-0128/98/000/000-00 $6.00 892

Uchupata Type-Sections

Celendn Formation (163 m, from Section 19)Marl: light gray, nodular, soft, white weathering Shale: calcareous, slightly silty, yellowish, with a few interbeds of dark-brown limestoneMarl: light gray, nodular, white weathering, with sparse interbeds of limestoneMarl: light gray, nodular, soft, white weathering Marl: light gray to tan, nodular, with few interbeds of massive light-gray limestone Jumasha Formation (820 m, from Section 20)

Limestone: medium gray, thick bedded, weathering dark dove gray, Foraminifera-bearing

Limestone: argillaceousLimestone: medium gray, massive, thick bedded, weathering dark-brownish gray Dolostone: thin bedded, brown

Dolostone: light gray to orange-brown, massive, thick-bedded, karstic, weathering dark orange-brown

Dolostone: silty, medium gray, somewhat nodular

Antamina Mine Area

Upper SequenceThin-bedded shaly limestone: dark gray to black, weathers medium to light gray, thin- to thick-bedded, generally very fine grained, dominated by biomicrite to microsparite, ranges in carbonate content from muddy calcisiltite with 50 - 75 % calcite to calcareous siltstone with < 50 % calcite, variably nodular, 10-90% nodules 1-15 cm diameter, nodules are rounded to multilobate and have sharp to indistinct contacts with the matrix Lower SequenceImpure limestone interbedded near the top: rare, medium to thick beds (20-50 cm), very dark gray to black, silty limestone, weathers medium grayPredominantly thick-bedded, relatively pure limestone:dark gray, medium- to very thick-bedded, ranges in grain size from mudstone to wackestone, weathers light to pale grayLocally fossiliferous bioclastic wackestone: with broken pelecypods and gastropods, but no apparent diagnostic fauna

SKARN

Campanian- ? - - - ? - - - ? -Santonian

- ? - - - ? - - - ? -Coniacian

Turonian

- ? - - - ? - - - ? -

Cenomanian

- ? - - - ? - - - ? -

Upper Albian

0

100

200 m

Outline of Regional Stratigraphy M Eocene - M Miocene Calipuy Supergroup Campanian - Paleocene Casapalca Group Albian - U Cretaceous Machay Group Coniacian - Santonian Celendn Formation U Albian - Turonian Jumasha Formation M Albian Pariatambo Formation L - M Albian Chulec Formation L Albian Pariahuanca Formation L Cretaceous Goyllarisquisga Group (Berriasian - Aptian)U Jurassic Chicama Group U Triassic - L Jurassic Pucar Group

L Permian - L Triassic Mitu Group Mississippian Ambo Group

FIG. 4. Inferred stratigraphic column in the mine area compared with that for the Jumasha and Celendn Formations,compiled from observations (Benavides, 1956) on the measured sections in the Ro Puchca valley, approximately 20 km northof Antamina. The contact between the Jumasha and Celendn Formations coincides with the boundary between the Turon-ian and Coniacian stages and is indicated with a solid line; unknown stage boundary locations have been approximated andindicated with dashed lines and question marks. The stratigraphic interval interpreted to host the skarn is shown. Inset showsan outline of the regional stratigraphic section.


0361-0128/98/000/000-00 $6.00 893

FIG. 5. Structural geology of the Maran thrust and fold belt in the vicinity of Antamina, illustrating the marked changein the orientations of thrust faults and fold axes across a northeast-trending zone through Antamina (after Egeler and deBooy, 1956; Wilson et al., 1967, 1995; Cobbing et al., 1996; Jacay, 1996; and Strusievicz et al., 2000). The northeast-south-westtrending loci of these changes in structural attitude is indicated by the heavy dashed line, which is offset in a left-step-ping sense in the vicinity of Antamina. This area is the location of the proposed cross-strike structural discontinuity discussedin the text. Inset shows the location of the map area relative to the Huari (19-i), Singa (19-j), Requay (20-i) and La Unin(20-j) quadrangles. No attempt has been made to establish continuity between the map units of Wilson et al. (1967, 1995)north of 9 30' S and those of Cobbing et al. (1996) to the south. The undated volcanic rocks of Pampa Junn have been as-signed to the Calipuy Supergroup (see text for discussion).

894 LOVE ET AL.

0361-0128/98/000/000-00 $6.00 894

FIG. 6. Simplified geology of approximately 8,000 km2 of the Maran thrust and fold belt in the region around Antam-ina; the dominantly upper Miocene Cordillera Blanca batholith is removed (cf. Figs. 1 and 5), reflecting the geology at thetime of intrusion and mineralization in the late Miocene. The northeast-southwest trending cross-strike structural disconti-nuity that passes through Antamina is delimited by heavy dashed lines. Mississippian to Lower Jurassic sedimentary rocksare absent beneath the Cretaceous Goyllarisquisga Group northeast of Antamina along the cross-strike structural disconti-nuity but are present north and southeast of the cross-strike structural discontinuity, except where cut out by faulting. An un-usual abundance of igneous bodies intrudes the Maran Belt along the cross-strike structural discontinuity, compared totransects to the north and south. After Egeler and de Booy (1956), Wilson et al. (1967, 1995), Cobbing et al. (1996), Jacay(1996), and Strusievicz et al. (2000).

plate boundary, and comprises structures that predomi-nantly verge northeast (Mgard, 1984; Fig. 5). However,southeast of the cross-strike structural discontinuity, foldsand thrust faults strike north-northwest, whereas to thenorthwest of it, they strike northerly (Figs. 1, 5, and 6).Moreover, fold plunges are reversed across this zone: to thesoutheast, most major anticlines and synclines plunge to thesouth-southeast, whereas to the northwest, they plungenorth (Fig. 5). The locus of changes in strike and plunge ex-tends northeast from Laguna Querococha, but about 5 kmsouthwest of Antamina it steps 8 km to the north before con-tinuing northeastward (Fig. 5). Faults with the same overallnortheast strike as the cross-strike structural discontinuitycontrol some present-day drainages, such as the northeast-trending Ro Puchca valley, 20 km north of Antamina, whichis discordant to the overall north to north-northwest grain ofthe terrain (Fig. 5).

The deflection in the Antamina area is one of several thatarticulate the Maran thrust and fold belt. Northerly strikescontinue to Llamalln, 50 km to the north of Antamina in theeastern part of the belt (Fig. 1). Still farther north, the over-all north-northwest regional strike of the Maran thrust andfold belt resumes. A comparable sharp deflection to north-south strikes occurs at the northern end of the CordilleraBlanca (Fig. 1), 175 km to the north-northwest of Antamina,where the Casma-Pasto Bueno fault zone intersects the re-gional north-northwest strikes (Rivera, 1996). Benavides(1999) identified many segments in the fabric of the Maranthrust and fold belt, including the two described above, al-though he proffered a different mechanism for their forma-tion, as discussed below.

Stratigraphic variations across the cross-strike structuraldiscontinuity: The contact relationships of the upper Paleo-zoic and Mesozoic strata (Fig. 4) to the lower PaleozoicMaran metamorphic complex vary in accordance with thesegmentation of the thrust and fold belt. Figure 6 shows therelationships along the western margin of the Maran com-plex throughout an area more extensive than that shown inFigure 5. Mississippian to Lower Jurassic strata that normallyseparate the pre-Ordovician metamorphic rocks from theCretaceous sedimentary rocks are absent near the proposedcross-strike structural discontinuity. In the north-striking seg-ment of the fold belt immediately north of the Antamina area,the eastern limit of the Mississippian to Cretaceous succes-sion is a subhorizontal unconformity that strikes north overallbut has an irregular surface trace owing to the incised topog-raphy (Figs. 5 and 6). In the north-northweststriking seg-ments of the belt, however, the eastern extent of the Meso-zoic rocks is generally delimited by north-northweststrikingfaults (e.g., southeast of Antamina, and northwest of Llamal-ln; Fig. 6). Southeast of Antamina, the southwest margin ofthe Maran metamorphic complex is largely defined by a se-ries of major northeast-verging reverse faults that involvedbasement, but the Mississippian to Lower Jurassic strata arepreserved between the pre-Ordovician metamorphic rocksand the Cretaceous sedimentary rocks (Fig. 6). However, inthe north-northweststriking segment northwest of Llamal-ln, the Maran complex is backthrust over the Cretaceousrocks and the thickness of the Mississippian to Lower Juras-sic strata is only locally apparent (Fig. 6).

Northeast of Antamina, Mississippian to Lower Jurassicstrata (Fig. 4) are absent along the proposed cross-strikestructural discontinuity, and the clastic rocks of the LowerCretaceous Goyllarisquisga Group lie unconformably on thepre-Ordovician Maran metamorphic complex (Figs. 5 and6). In contrast, north and southeast of the cross-strike struc-tural discontinuity, a relatively thick sequence of Mississip-pian to Lower Jurassic strata separates these units (Fig. 6).Sandstones and shales of the Mississippian Ambo Group lo-cally unconformably overlie the Maran complex near itswestern limit. Similarly, continental sedimentary rocks and al-kaline to subalkaline volcanic rocks of the Lower PermianMitu Group commonly overlie the Ambo Group and also lo-cally lie unconformably on the Maran complex along itswestern edge but are absent northeast of Antamina. Bothnorth and south-southeast of the cross-strike structural dis-continuity, these siliciclastic sedimentary rocks are overlain byUpper Triassic to Lower Jurassic carbonate rocks and shales(Pucar Group). The absence of these three groups directlynortheast of Antamina records either a sub-Cretaceous ero-sional unconformity or nondeposition from Mississippian toLate Jurassic times. The cross-strike structural discontinuitytherefore also coincided with a topographic high or arch, atleast in the Middle to Late Jurassic, but possibly persistingthroughout the Mississippian to Jurassic interval.

The cross-strike structural discontinuities probably also in-fluenced the distribution of the uppermost Cretaceous to Pa-leocene red beds that unconformably overlie the Cretaceousstrata (Fig. 4). These are absent near the proposed cross-strike structural discontinuity through the Antamina area, andthey also thin significantly near the more northerly Casma-Pasto Bueno deflection, but they attain a considerable thick-ness between these two transverse zones as well as to thesouth of the arch (Fig. 1).

Igneous activity along the cross-strike structural disconti-nuity: Between the Cordillera Blanca and the Maran meta-morphic complex, the 1:100,000 quadrangle maps of Cobbinget al. (1996) and Wilson et al. (1967, 1995) record a greaterabundance of Tertiary hypabyssal and extrusive rocks alongthe proposed cross-strike structural discontinuity than to thenorthwest or southeast (Fig. 6). In addition, west of the thrustand fold belt, volcanic rocks of the Calipuy Supergroup(Strusievicz et al., 2000) are abundant northwest and south-east of this cross-strike structural discontinuity (e.g., in theNevado Huantsn area and the Cordillera Huayhuash; Fig.6). These rocks are interpreted to be lower-middle Miocenebecause hypabyssal intrusions associated with similar volcanicrocks elsewhere in the region (Huaraz Group of the CalipuySupergroup; Fig. 5) have been shown by 40Ar/39Ar step-heat-ing geochronology to have persisted to 14.2 Ma (Strusievicz etal., 2000; Love et al., 2001). However, the 115 km2 middleMiocene granodioritic Carhuish pluton crops out on the axisof the proposed cross-strike structural discontinuity (Fig. 5),representing deeper-seated rocks of broadly equivalent age(13.7 Ma U/Pb zircon date on the main phase, Mukasa, 1984;16.5 Ma K/Ar date on the marginal phase, Cobbing et al.,1981). The volcanic rocks of Pampa Junn (Egeler and deBooy, 1956) northeast of the Carhuish pluton (Fig. 5) havenot been dated, so their significance with regard to the trans-verse structure is uncertain. However, we propose that the


0361-0128/98/000/000-00 $6.00 895

cross-strike structural discontinuity focused the distributionof igneous activity as it diminished through the mid-Miocene,prior to intrusion and mineralization at Antamina. In Figure6 the arch is depicted with a width of approximately 20 km toincorporate the left-stepping locus of changes in Incaic strikesand plunges (Fig. 5), the absence of Mississippian to LowerJurassic rocks beneath the sub-Cretaceous unconformity atthe northeast end of the arch, the diameter of the Carhuishpluton at its southwestern margin, and the array of Mioceneintrusions.

Local Geologic Setting

Sedimentary host rocks

The contact of the Jumasha and Celendn Formationswithin the Machay Group has not previously been defined inthe Antamina mine area, owing to the intense skarn and horn-fels development and to the paucity of biostratigraphic mark-ers. The cliff-forming strata surrounding the Antamina depositwere mapped initially by Bodenlos and Ericksen (1955) as ma-rine limestones of the Jumasha Formation. J.J. Wilson (un-published report to Cerro de Pasco Corp., Lima, 1959, in Pe-tersen, 1965) observed that the beds on the northwest slopes

of Quebrada Antamina were stratigraphically higher thanthose at similar elevations on the southeast side, but he didnot locate the interformational contact. Petersen (1965) de-scribed the carbonate units as the Machay Formation, andthey were subsequently reassigned to the Jumasha Formationby Cobbing et al. (1996).

The skarn-hosting limestone strata at Antamina are hereinsubdivided into two sequences, the upper markedly moreshaly than the lower, and assigned to the Celendn and Ju-masha Formations, respectively. These sequences are de-scribed and compared with a section of the relevant strati-graphic interval constructed from two sections measured byBenavides (1956) at Uchupata in the Ro Puchca valley, 20 kmnorth of Antamina (Fig. 4). A measured section has not beenestablished for the Antamina minesite because of the struc-tural complexity of the area (described below) and the ab-sence of marker beds in the exposed host rocks. The inferredstratigraphic interval occupied by the skarn is also indicatedin Figure 4. The upper sequence of thin-bedded rocks thatprior to mining underlay the ridge crests flanking QuebradaAntamina (Fig. 2) is widely altered to hornfels adjacent to theAntamina intrusive center and grades into shaly limestonewith increasing distance from it (Fig. 7). The lower sequence

896 LOVE ET AL.

0361-0128/98/000/000-00 $6.00 896

FIG. 7. The two distinct host rock types of the Antamina skarn system: an upper sequence of thin-bedded, silty limestones,here largely converted to calc-hornfelses, assigned to the Celendn Formation (Ce), and a lower sequence of thick-bedded,relatively pure limestones that form marbles, interpreted as the Jumasha Formation (J). Locations from which the pho-tographs were taken are indicated in Fig. 2. Photographs taken in 1997 through 1999; this ridge has now been largely re-moved by mine development. (a) The southeast side of Quebrada Antamina, looking southeast. The contact (long blackdashes) between the Jumasha and Celendn Formations is placed at the top of the uppermost massive thick-bedded lime-stone (pale gray band) deformed by the Antamina anticline. Note the irregular upper, southeastern contact of the skarn (longwhite dashes), here largely confined to the Jumasha Formation. (b) Looking northeast at the head of Quebrada Antaminashowing the distinct control of bedding in the Celendn Formation on skarn development.

of calcitic marble contains only subordinate intercalated diop-side-rich units and grades outward into predominantly thick-bedded, relatively pure limestone (Fig. 7).

The lower sequence (Figs. 4, 7a, and 8a) generally containsmore than 75 percent calcite and comprises calcitic, variably

bioclastic limestones that range in grain size from mudstoneto wackestone. This sequence is generally thick-bedded andmassive and displays karstic weathering (Fig. 8b). Toward thetop of this sequence, medium to thick beds (2050 cm) of im-pure, silty limestone are interbedded with the pure limestone


0361-0128/98/000/000-00 $6.00 897

FIG. 8. The Jumasha and Celendn Formations. (a) Cliff-forming thick beds of massive limestone on the southeast flankof Quebrada Callapo (see Fig. 2 for location). (b) Massive limestone of the Jumasha Formation, showing fluted weathering(south of Yanacancha area, Fig. 2). (c) Upper: massive (i.e., unlaminated and with no preferred orientation of fossils) pele-cypod carbonate wackestone of the Jumasha Formation (DDH CMA-039, 155 m). Lower: sheared fossiliferous carbonatewackestone, equivalent to upper piece, not marmorized, but with a strong preferred orientation (DDH CMA-039, 160 m).(d) More recessive weathering, vegetated, thin-bedded limestones and marls of the Celendn Formation cropping out on thesoutheast slopes of Quebrada Ayash near its intersection with Quebrada Tucush (see Fig. 2 for location). Photograph takenin 1997; the lower half of this area is now obscured by construction of the tailings dam. (e) Nodular limestone of the Ce-lendn Formation, consisting of light gray, coalescing, carbonate-rich nodules separated by wisps of dark gray calcareous silt-stone (Fortuna Mine area, see Fig. 2; hammer for scale). (f) Upper: Celendn Formation nodular limestone (DDH CMA-C3, 221 m). Lower: lenticular bedding interpreted as sheared nodular limestone of the Celendn Formation (DDHCMA-C5, 87.5 m).

(Figs. 4 and 7a). This sequence hosts ore at the surface and atdepth in the southwest part of the deposit, but only in thesubsurface in the northeast. Few whole fossils are preserved(Fig. 8c) and no ammonites were observed in the mine area,precluding biostratigraphic correlations. However, we assignthis sequence to the Jumasha Formation on lithologicgrounds. Sheared fossiliferous carbonate wackestone occurslocally in this sequence (Fig. 8c). In north-central Peru, Be-navides (1956) described the Jumasha Formation as domi-nated by massive, thick-bedded, light orange-brown to yel-lowish-brown and gray, fossil-poor dolostones and limestonesthat weather dark yellowish-brown to brownish-gray. The for-mation has been further described as comprising topographi-cally prominent, cliff-forming, light-gray limestones and yel-lowish dolostones that are characteristically bioclastic(Wilson, 1963). In the Uchupata section (Fig. 4), the upper437 m is limestone, and the lower 353 m is dolostone. Unlikethe overlying Celendn Formation, this formation only rarelycontains calcareous siltstones, although it incorporates marlylimestone beds near its top (Jaillard, 1987). The upper limit ofthe Jumasha Formation was defined lithostratigraphicallywhere medium- or thick-bedded limestones pass upward intothin-bedded marls and limestones (Wilson, 1963).

The upper sequence in the mine area (Figs. 4, 7, and 8d)comprises thin- to thick-bedded impure limestone-marl thatvaries in carbonate/silicate ratio from relatively calcite-richmuddy limestone (generally 5075% carbonate) to calcareoussiltstone (less than 50% carbonate). Many units contain light-gray, calcitic nodules, composing 10 to 90 percent of the rock,enclosed by dark, silty calcareous mudstone (Fig. 8e, f). Nolimestone beds with siliceous nodules have been observed.The nodular texture is interpreted to be diagenetic. Shearedlimestone with lenticular bedding exposed at the northeasthead of Quebrada Antamina represents deformed nodularlimestone in which the nodules have been flattened intolenses during folding and faulting (Fig. 8f). The upper se-quence lacks identifiable fossils but is assigned to the Ce-lendn Formation on lithologic grounds. In the Uchupata sec-tion, 20 km north of the mine, the Celendn Formationcomprises very soft, friable, fossil-poor, light greenish-gray,nodular, moderately silty marls and calcareous shales (Fig. 4;Benavides, 1956). This formation is generally medium-bed-ded (0.30.8 m) and variably dolomitic, and it ranges fromfine-grained to pelletal (Wilson, 1963).

The contact between the Jumasha and Celendn Forma-tions in the mine area is conformable and generally grada-tional throughout several meters and is marked by upward-in-creasing siltiness and decreasing bed thickness. This contactis interpreted to be the top of the uppermost thick-beddedlimestone or marble unit on the basis of lithology and beddingcharacteristics. This stratigraphic position is illustrated in Fig-ure 7a for outcrops on the southeast wall of the Antamina val-ley. The contact is folded by the anticlines exposed on the val-ley walls adjacent to the southwestern part of the deposit, andit dips northeast in the subsurface in the northeast part of thedeposit, as do the overlying strata exposed in that area. Thethin-bedded calc-hornfelses that predominate on the ridgecrests flanking Quebrada Antamina grade laterally away fromskarn into the marly, variably nodular sequence assigned tothe Celendn Formation, but they do not contain the shales

characteristic of its upper part. The relatively pure calciticmarble of the Jumasha Formation that hosts ore at surface inthe southwest part of the deposit and at depth in the north-east sector does not contain the thick dolostones characteris-tic of the lower part of that formation, and no magnesianskarn has been found. On this basis, we conclude that theskarn developed in the upper part of the Jumasha Formationand the lower part of the Celendn Formation (Fig. 4). How-ever, we estimate that approximately three-quarters of themineralized exoskarn, and all of the wollastonite exoskarn, de-veloped in the Jumasha Formation, and that a similar propor-tion of the contiguous and genetically related endoskarn wasformed in the stock adjacent to that formation. It is notknown whether magnesian skarn developed in dolostonebelow the limit of exploration and development drilling.

Alteration of sedimentary rocks adjacent to the skarn

The outer contact of coarse-grained mineralized exoskarn isabrupt, convoluted, and uninfluenced by bedding in the mar-bles formed from the limestones of the Jumasha Formation(Fig. 7a), but it is more gradational and stratigraphically con-trolled in the fine-grained calc-silicate rocks developed in theinterbedded limestones and marls of the overlying CelendnFormation (Fig. 7b). In the Jumasha Formation, the mineral-ized skarn is juxtaposed with marbles containing local hori-zons rich in calc-silicate minerals that may be ascribed tothermal metamorphism. In contrast, the calc-silicate-bearingrocks developed in the marly Celendn Formation includerock types that are interpreted as the products of either meta-morphism or metasomatism. The latter do not host sulfideminerals and differ radically in texture and mineralogy fromthe exoskarn ore. Their wide distribution around the upperpart of the orebody constitutes a significant exploration tar-get.

Jumasha Formation: On approaching skarn, dark gray Ju-masha limestone is converted to coarsely crystalline, gray towhite calcitic marble (Table 1, Fig. 9), forming an aureoleranging from tens to hundreds of meters wide. Within the au-reole, local development of sparse diopside, wollastonite, orscapolite porphyroblasts (Fig. 9a) or slight differential erosionand color variation (Fig. 9b) in calcite marble record minorcompositional variations in the limestone. Rare, fine-grained,light green, diopsidic layers in calcite marble (Fig. 9c), theporcellanite of Terrones (1958), are interpreted as calc-horn-fels, representing thermally metamorphosed, medium tothick beds of dolomitic, muddy, fine-grained limestone nearthe top of the formation.

Within the marble aureole, medium-grained, mottled, orbanded gray marble predominates, locally grading inward tomedium- to coarse-grained, pure white marble, reflecting ei-ther degraphitization or metamorphism of organic matter toclear vitrinite. Rare, medium- to coarse-grained, thin, buffgarnet layers in both white and mottled facies of marblemimic the forms of folded silicate-rich laminae (Fig. 9d) andappear to have replaced them. This type of garnet develop-ment is interpreted as a bimetasomatic reaction skarn. In graymarble, scapolite is locally developed in some darker-graybands (Fig. 9e) but does not persist into white marble orskarn (Table 1). Scapolite therefore forms a discontinuoushalo up to tens of meters wide and commonly separated from

898 LOVE ET AL.

0361-0128/98/000/000-00 $6.00 898

the skarn front by tens of meters. White marble develops inpatches (Fig. 9d and f) and is concentrated along stylolitesand fractures (Fig. 9f). The latter probably represent fluidpathways, suggesting that development of white marble in-volved the local channeling of fluids. Such pathways were notconsistently used by later skarn-related fluids and are locallycut by fractures with wollastonite-rich selvages (Fig. 9f).

Immediately southwest of the skarn, the Jumasha Forma-tion has a distinct planar banded fabric, locally a spaced cleav-age, that dips northeast at a high angle to bedding (Fig. 9a).To the northeast, this fabric steepens and becomes verticaladjacent to the skarn and along the strike of a major fold axis.The consistent strike, systematically changing dip, and rela-tionship to folds visible in the adjacent limestones indicatethat this is an upward-fanning cleavage associated with localfolding. In the adjacent skarn, alternating andradite-rich andsphalerite-rich bands, generally 5 to 50 cm wide, are subpar-allel to this fabric, suggesting that at least some preexistingstructures influenced mineralization. This banding may cor-respond with the coarse-grained garnet-defined fabric ob-served by Terrones (1958).

Celendn Formation: In the Celendn Formation, on theridges around Lago Antamina, the skarn has a halo of miner-alogically diverse fine-grained calc-silicate rocks with subcon-choidal to conchoidal fractures. With increasing distancefrom the skarn, these rocks grade into interbedded limestoneand marl that reacted differentially to thermal metamorphismaround the porphyry intrusion and orebody, thus highlightingthe bedding (Fig. 7b).

Dark gray massive units change in color, mineralogy, andgrain size proximal to monzogranitic dikes. Although we havenot mapped the various facies of these rocks in detail, we rec-ognize the development of three distinct zones: a distal very

fine grained, light-brown phlogopitic facies; an intermediatefine-grained, light-gray, tremolitic facies; and a proximalmedium-grained, light-green diopsidic facies (Table 1, Fig.10a, b). The boundaries between these facies are commonlysharp and smooth but are locally irregular where controlledby stockwork fractures (Fig. 10a, b). The outer limit of thedistal brown facies is typically hundreds of meters from theboundary of sulfide-bearing skarn, and the progression frombrown through gray to green facies occurs over distancesranging from tens of meters adjacent to the main intrusion totens of centimeters (Fig 10b) adjacent to dikes extending be-yond the main intrusion.

Although local fluid channeling and probably metasoma-tism occurred at the boundaries of the facies, the systematicmineralogic zonation (Table 1) most likely reflects increasingtemperature. The first appearance of light brown calc-horn-fels corresponds with the phlogopite-in reaction, Dol + Kfs =Phl + Cal, which occurs in the range 350 to 465C under ge-ologically reasonable conditions (i.e., P = 1,000 bars and XCO2= 0.1 to 0.9: Tracy and Frost, 1991). The gray calc-hornfelscontains tremolite with or without phlogopite, whereas browncalc-hornfels is tremolite free (Table 1). The transition frombrown to gray calc-hornfels therefore is interpreted to repre-sent a phlogopite-out reaction that coincides with the forma-tion of tremolite, probably through the reaction Phl + Cal +Qtz = Tr + Kfs (Tracy and Frost, 1991; Table 1). The light-green calc-hornfels records the first appearance of massivediopside (Table 1), although veinlets and small blebs of thismineral locally occur farther from the intrusive contacts. Thiszone probably formed through the reaction Tr + Qtz + Cal =Di, which under similar pressure and XCO2 occurs at temper-atures of between 405 and 495C (Tracy and Frost, 1991).The local narrowness of the gray calc-hornfels facies and the


0361-0128/98/000/000-00 $6.00 899

TABLE 1. Mineral Assemblages in Calc-Hornfels in the Celendn Formation, and in Marbles of the Jumasha Formation, Adjacent to Ore-Bearing Garnet Skarn, Antamina, Peru1

Formation, rock type, Faciesand mineralogy Distal Intermediate Proximal

Celendin Formation2Calc-hornfels Brown, very fine grained Gray, fine grained Green, fine to medium grained

Common minerals Calcite, anorthite, calcite, Calcite, quartz, tremolite Diopside, calcite, quartz, microclinequartz, phlogopite

Minor minerals Diopside Diopside, Vesuvianite, wollastonite, grossular, phlogopite, hedenbergite orthoclase, ferrosilite

Jumasha FormationMarble Gray; fine to medium grained, Gray; medium grained, banded Gray; medium to coarse grained, massive

banded

Common minerals Calcite Calcite Calcite

Minor minerals Vitrinite or other organic matter Scapolite, vitrinite, graphite, Colorless vitrinite, or other organic matter wollastonite,

diopside

1Determined by X-ray powder diffraction using a Philips PANalytical XPert PRO diffraction system with XPert Plus and XPert HighScore software forpeak matching and phase identification

2Distal, intermediate, and proximal facies zones are broader in the Celendn Formation than in the Jumasha Formation

900 LOVE ET AL.

0361-0128/98/000/000-00 $6.00 900

FIG. 9. Marmorized Jumasha Formation. (a) Subtle bedding, defined by porphyroblast abundance (outlined) in calciticmarble, dips moderately southwest and is overprinted by a northeast-dipping planar fabric (looking northwest, portal of bulksample adit, 450 m southwest of west shore of Lago Antamina; hammer for scale). (b) Subtle bedding, defined by slight colordifference and differential erosion (outlined), reflecting cryptic difference in porphyroblast abundance in calcitic marble(looking northwest, 350 m southwest of west shore of Lago Antamina; 20 cm notebook in foreground for scale). (c) Medium-bedded layer of light green, fine-grained, diopsidic calc-hornfels defines bedding within coarse-grained calcitic marble nearthe top of the formation (looking northeast, 300 m southwest of the western shore of Lago Antamina; 15 cm ruler for scale).(d) Upper: mottled gray calcitic Jumasha Formation marble with tightly folded, boudinaged, dark-gray silty laminae. Lower:mottled white and gray marble with tightly folded buff garnet-rich layer, similar in shape to the silty layer in the upper core(DDH CMA-086, 218 m). (e) Banded gray Jumasha Formation marble showing development of approximately 10 percentblack scapolite crystals within darker layers (DDH CMA-136, 232 m). (f) Upper: white marble in gray Jumasha Formationmarble both as patches throughout and locally around veinlets and stylolites (DDH CMA-136, 228 m). Lower: veinlet-con-trolled white marble in gray Jumasha Formation marble cut by irregular veinlets with chalky-white wollastonitic selvages(DDH CMA-236, 441 m). Scale bars in centimeters.


0361-0128/98/000/000-00 $6.00 901

FIG. 10. Hornfels, skarnoid, and wrigglite developed in the Celendn Formation. (a) Irregular stockwork of sealed frac-tures controlling development of light-gray tremolitic calc-hornfels in light-brown phlogopitic calc-hornfels (on the southeastridge crest, center of Fig. 12c; 15 cm pencil for scale). (b) Calc-hornfels facies with distal, light-brown, very fine grained phl-ogopitic calc-hornfels at upper left, light-gray tremolitic in the center, and proximal, light-green medium-grained diopsidiccalc-hornfels at lower right. These facies are unusually closely spaced because they are on the margin of a narrow dike, ap-proximately 350 m southeast of the main intrusion (same location as [a]; 55 mm lens cap for scale on the far left). (c) Sparse,large, medium-green diopsidic nodules in pale gray calc-hornfels (base of cliffs, northeast of Lago Antamina; 10 cm knife forscale). (d) Very abundant coalescing diopsidic nodules with calcite-bearing calc-hornfels matrix (base of cliffs, east of LagoAntamina; 15 cm pencil for scale). (e) Irregular concentric banding, interpreted by Bodenlos and Ericksen (1955) to be ofalgal origin, but reinterpreted as wrigglite texture of metasomatic replacement origin (base of cliffs, east of Lago Antamina;15 cm pencil for scale). (f) Banding developed in calc-hornfels and apparently controlled by northeast striking fractures (onthe ridge crest southeast of Antamina valley, just east of the east end of Fig. 12c; 55 mm lens cap for scale).

absence of tremolite in the phlogopite facies suggest that themetamorphism took place at high XCO2. Under such condi-tions the diopside-in and phlogopite-in reactions are sepa-rated by only circa 10C. Also, the development of tremolitethrough the elimination of phlogopite is promoted underthese conditions, but its formation within the field of phlogo-pite stability through the reaction Dol + Qtz = Tr + Cal is in-hibited.

In contrast to these calc-hornfelses that formed under es-sentially thermal metamorphic conditions, calc-silicate devel-opment through metasomatism is revealed by development ofskarnoid in some beds of the Celendn Formation, extendingtens of meters beyond the mineralized skarn front. Skarnoidis a descriptive term for calc-silicate rocks that are relativelyfine grained, calcium rich, and iron poor, and that reflect, atleast in part, an aluminosilicate component in the protolith(Zharikov, 1970). It is genetically intermediate between apurely metamorphic hornfels and a purely metasomatic,fluid-controlled skarn, which is typically coarser grained anddoes not as closely reflect the composition or texture of theimmediately surrounding rocks (Einaudi, 2000). In somebeds in the Celendn Formation, diopsidic nodules stand outin relief against the calcite-bearing calc-silicate matrix (Fig.10c, d). The nodules, locally concentrically banded, rangefrom >10 cm to


0361-0128/98/000/000-00 $6.00 903

FIG. 11. Local geology of the Antamina area (modified after Cobbing et al.,1996; Glover, 1997, unpublished report toCompana Minera Antamina S.A., Lima; Palomino, 1997, unpublished report to Compana Minera Antamina S.A., Lima; andLove and Clark, 1998, unpublished report to Compana Minera Antamina S.A., Lima). The physiography of this area isshown in Fig. 2. The areas depicted in Fig. 13 are indicated by the three diagonal rectangles. AT = Antamina thrust, FT =Fortuna thrust, VF = left-stepping Valley fault, YBPT = Yaquirsh-Buque Punta thrust.

904 LOVE ET AL.

0361-0128/98/000/000-00 $6.00 904

FIG. 12. Photomosaics illustrating the structure and stratigraphy of the Antamina mine area. Locations from which thephotographs were taken are indicated in Fig. 2. The views in (a) and (b) look northwest, whereas those in (c) and (d) looksouth. Photographs taken in 1997 through 1999; much of (a), (b), and (c) has now been removed by mine development, andthe skyline of (d) has been modified. (a) The northwest side of the Antamina valley; the irregular upper or northwestern con-tact of the skarn (long white dashes) cuts off the stratigraphic contact (long black dashes) of the Jumasha (J) and Celendn(Ce) Formations. (b) The southeast flank of the ridge extending northeast of Cerro Buque Punta (i.e., the opposite side ofthe ridge illustrated in [c]), viewed from the Yanacancha area. (c) The southeast side of the Antamina valley; the irregularupper, southeastern contact of the skarn (long white dashes) cuts off the stratigraphic contact (long black dashes) of the Ju-masha (J) and Celendn (Ce) Formations. (d) The northwest flank of the ridge extending northeast of Cerro Yaquirsh (i.e.,the opposite side of the ridge illustrated in [a]). Faults and fold axes are shown by solid black lines, the stratigraphic contactbetween the Jumasha (J) and Celendn (Ce) Formations by long black dashes, the contact of the skarn by long white dashes,intrusive contacts by solid white lines, intrusions by crosses. AT = Antamina thrust, Ce = Celendn Formation, FT = Fortunathrust, J = Jumasha Formation, OT = Oscarina thrust, VF = surface expression of the northeastern segment of the Valleyfault, YBPT = Yaquirsh-Buque Punta thrust.

broader and more open upward but the intervening synclinesare tight. Such concentric folds are common elsewhere in theMaran thrust and fold belt (e.g., Coney, 1971). The largestof the parallel folds visible on the flanks of the Antamina val-ley has been widely referred to as the Antamina anticline, theaxial plane of which has an apparent dextral offset of approx-imately 500 m in a northeast-southwest direction across thevalley. This offset has been variously ascribed to local bendsin the strike of the folds (Bodenlos and Ericksen, 1955), dex-tral offset on a postulated northeast-striking Valley fault (Ter-rones, 1958), transverse normal faulting (Petersen, 1965,) andcross-folding (McKee et al., 1979). However, in this study weshow that the offset is best explained by a lateral ramp model.In this model (Fig. 14), the anticlines represent fault-bendfolds related to hanging-wall cutoffs folded over the top ofnorthwest-striking frontal ramps that are offset because anortheast-striking, 500-m-long transfer fault or lateral rampseparates the frontal ramps in an apparent dextral sense. Thispostulated structure, the Valley lateral ramp, would accountfor the absence of corresponding offsets of other structures tothe southwest and northeast (Fig. 11).

Minor diking and skarn alteration extend beyond the mainintrusion and orebody to the southwest down Quebrada An-tamina and to the northeast and east at the head of the valley.The vertical dikes that extend east from the main porphyrymass to the Rosita de Oro area (Figs. 11 and 12c) offset twominor forethrusts and backthrusts, indicating that there wasminor, apparently vertical, faulting between the time of thrust

faulting and folding and that of intrusion and mineralization.The orientation of these dikes changes considerably alongstrike; near the main stock they are nearly vertical, strike east,and cut strata (Fig 12c), whereas farther east beyond theridge crest they become sills and follow either strata or bed-ding-parallel thrust flats, strike south, and dip moderately tothe west (Fig. 12b). The dikes become progressively more al-tered to endoskarn toward the main intrusion, and althoughthey transect zoned exoskarn they do not appear to have in-truded it. A similarly variable orientation is shown by a dikethat extends northeast from the northern shore of Lago Anta-mina; it locally crosses strata at a steep angle but also followsflat and moderately dipping strata (Fig. 12a). The parallel fea-tures at opposite ends of the main intrusion and orebody sug-gest an offset, left-stepping, northeast-southwest fracturezone that is longer than the 500 m lateral ramp, the postu-lated Valley fault (Figs. 3b and 11).

A well developed northeast-striking, widely spaced, nearlyvertical fracture set occurs in the sedimentary host rocksthroughout the area (e.g., on the southwest flank of CerroRacpe; Fig. 15a), and we interpret it as regional a-c jointingrelated to folding. A similarly oriented fracture set is in-tensely developed near the deposit, especially along thesoutheast side of the Antamina valley (Fig. 15b, c). In manyplaces, calc-hornfels is more intensely developed and wide-spread where these fractures are closely spaced. All of thenortheast-striking fractures may have been related to Eocenefolding and thrust faulting, but they were better developed in


0361-0128/98/000/000-00 $6.00 905

Ce

J

J

Ce

Ce

J

Ce

CeOT

AT

YBPT

OTNE

SW

CeJ

Ce

Ce

Ce

J

0

500

1000 m

Ce

J

J

J

J

Ce

Ce

J

J

Ce

AT

AA

AA

Ce

J

J

VF

VF

VLR

Ce

Ce

OT

J

J

OT

Ce

Ce

YBPT

AT

FT

OT

AA

Ce

J

J

Ce

Ce

Ce

Ce

Ce

Ce

Ce

NE

FT

AT

OT

YBPT

Ce

SW

J

J

Ce

Ce

J

CeCe

FIG. 13. Schematic, isometric block diagram of the pre-Miocene geology of the Antamina area, looking north, summa-rizing the major Eocene Incaic folding and thrust faulting in the host Cretaceous strata. In the late Miocene, intrusion andformation of the Antamina deposit occurred in the central block on this diagram, localized by the Valley lateral ramp (VLR)and the step in the Valley fault (VF). The locations of the three blocks are indicated by rectangles in Fig. 11. The Jumasha(J) and Celendn (Ce) Formations are indicated. Faults and fold axes are shown by solid black lines, bedding by black dashes.Other features abbreviated as in Figures 11 and 12.

the mine area because of flexure and tearing of strata abovea transfer fault or lateral ramp (discussed below). Followingintrusion, they provided access for fluids to the hornfels-bearing strata.

In the Antamina district, therefore, the Jumasha and Ce-lendn Formations have been thrust-faulted, folded, and jux-taposed into a thick, complex thrust stack. The total thicknessof rock that overlay the site of mineralization in the lateMiocene cannot be estimated because the local thickness ofoverlying uppermost Cretaceous and Paleocene red beds isunknown and because it is unclear if the subaerial volcanicrocks and associated sedimentary rocks of the Eocene to mid-dle Miocene Calipuy Supergroup (Strusievicz et al., 2000) ex-tended into this area.

Structural control of mineralization: As in most intrusion-related skarn deposits, the most obvious feature controllingmineralization at Antamina is the contact between the mainstock and the host limestones (Terrones, 1958; Petersen,1965; Redwood, 1999). However, the southeast and northwestintrusive contacts are themselves parallel to, and along strikefrom, the two segments of the proposed northeast-striking

Valley fault. The hydrothermal breccia sheets that are com-mon at or near the endoskarn-exoskarn contact also have thisorientation. The irregular zones of breccia and endoskarnwithin the intrusion are interpreted to strike predominantlynorth-south (Fig. 3b) and may have been controlled by cross-cutting structures. The western end of the main body of theintrusion also generally strikes north-south, as does the skarnfront in that area (Fig. 3b). The parallelogram shape, in plan,of the main body of the intrusion and the surrounding skarnis complicated by the network of anastomosing dikes with en-velopes of fine-grained garnet skarn extending to higher ele-vations to the east of Lago Antamina (Fig. 12c). At least oneof these dikes intrudes a normal fault on which movement oc-curred between the time of thrust faulting and folding andthat of intrusion and mineralization (Fig. 12c). Several otherdikes with skarn envelopes also extend beyond the main massof porphyry and most are controlled by Incaic structures.Along the southern edge of the porphyry and skarn, minor pe-ripheral rhyodacitic dikes and sills and their skarn envelopeslocally follow bedding and thrust faults. Several minor Pb-Zn-Ag veins (Fig. 2), some of which have been mined on a small

906 LOVE ET AL.

0361-0128/98/000/000-00 $6.00 906

Foot-wall Block

Flat

b

Frontal Ramp

Lateral R

amp

transport direction

Anticline

c d

Lateral A

nticline

Fault bend fold Anticline

Hangingwall Cutoff Anticline

FutureHanging-wallBlock

FutureFoot-wallBlock

Trace of Future Fault

a

FIG. 14. Schematic diagram illustrating the lateral ramp model for the structural setting of the Antamina deposit. Look-ing south so that the face of the lateral ramp is exposed, showing offset anticlines produced in the hanging wall of a thrustfault by two frontal ramps separated by a lateral ramp. (a) Geometry of the fault prior to movement. (b) Geometry of thefootwall (hanging-wall block removed), showing the southwest-dipping frontal ramps linked by a northeast-striking lateralramp or transfer fault. (c) After minor thrust movement, offset ramp-cutoff anticlines produced in the hanging wall of thethrust fault, with more intense northeast-striking fracturing above the lateral ramp, represented by dashed lines. (d) Hang-ing-wall cutoff anticlines separated from fault-bend anticlines by additional thrust movement. Extensive fracturing developedin the thrust sheet where it flexed over the lateral ramp.

scale, occur at or near the contacts between these dikes andthe host limestone, describing a circa 3 3 km area of dis-persed hydrothermal activity centered on the Antamina skarn(Bodenlos and Ericksen, 1955).

Discussion

Regional stratigraphic and tectonic relationships

The postulated Valley lateral ramp at Antamina may havebeen controlled by underlying, northeast-striking basement

faults. Lateral ramps are thought to be largely the result ofthe interaction of thrust sheets and old basement fracture sys-tems (Pohn, 2000). Both Mgard (1987) and Benavides(1999) proposed that the segmentation and articulation of theMaran thrust and fold belt probably reflect control by base-ment structures, although they differ in their interpretationsof the mechanism. Benavides (1999) attributed the segmen-tation of the Maran thrust and fold belt to major northeast-striking, dextral basement faults, without specifying when thefaults were active. Further, Bussell and Pitcher (1985) sug-gested that a well developed set of northeast-striking, en ech-elon, dextral faults in the Cretaceous-Paleogene Coastalbatholith may have controlled some of the contacts of earlyPaleocene intrusive ring complexes (Fig. 16). However, nosignificant tear faulting parallel to the northeast-trending de-flections is apparent in the deformed Mesozoic cover strata,suggesting that the basement faults have not been active sincethe Jurassic, and that any dextral offsets are apparent, notreal. Mgard (1987), in contrast, proposed that en echelonsinistral growth faults in the basement were the commonboundary of the miogeosyncline and Yauli shelf in the easternpart of the West Peruvian trough, and controlled the attitudeof the present thrust and fold belt.

The effects of the cross-strike structural discontinuities onsedimentation have varied through time and with rock type.A compilation of documented stratigraphic columns (Fig.16b) shows that from Huancayo in the south to 50 km northof Antamina the Cretaceous strata vary in aggregate thicknessfrom approximately 500 to 1,500 m and thicken overall to thenorth. Generally, only the easternmost sections have beenused in this compilation, minimizing the effect of thicknessvariations perpendicular to the margin, and thus emphasizingalong-strike variations that may be related to segmentation ofthe margin. It is evident that the Cretaceous strata exhibit noclear relationship between thickness and proximity to a cross-strike structural discontinuity.

A model for the geologic history of the Antamina region

The geologic history of the Antamina region from the for-mation of the West Peruvian trough in the Jurassic to the


0361-0128/98/000/000-00 $6.00 907

FIG. 15. Northeast-striking, nearly vertical fracturing in the eastern andsoutheastern part of the Antamina mine area. Locations from which the pho-tographs were taken are indicated in Fig. 2. Photographs taken in 1998; muchof (b) and (c) has now been removed by mine development. (a) Looking east-northeast at the southwest flank of Cerro Racpe where all bedding dipssouthwest, toward the point of view. Widely spaced northeast-striking frac-tures are expressed as lineaments that converge in the distance and are per-pendicular to the strike of the Celendn (Ce) and upper Jumasha (J) Forma-tions on the lower slope, but are more obvious as vertical fractures in thelower Jumasha Formation on the steep upper slope. Long black dashes indi-cate the stratigraphic contact between the Celendn and Jumasha Forma-tions. (b) Looking north-northeast along the ridge crest on the southeast sideof the Antamina valley in the immediate vicinity of the mine. The more in-tensely developed northeast-striking fracture set in the Celendn Formationcan be seen as closely spaced vertical fractures in the cliff face. (c) Lookingnortheast along the ridge bounding the southeast side of the Antamina valley,near its head, in the immediate mine area. The closely spaced northeast-striking fractures in the Celendn Formation strike away from the point ofview across the steep slope in the lower-right foreground. The fractures areless strongly developed along strike to the northeast on Cerro Aparina (upperright).

Miocene intrusion and formation of the orebody is summa-rized schematically in Figure 17. In this model, the margin-parallel West Peruvian trough formed during the Middle orLate Jurassic by extension on en echelon, northwest-strikingnormal faults separated by northeast-striking transform faults(Fig. 17a, b; Mgard, 1987). The normal faults on the westernmargin of the basement are thus right-stepping, but the seg-ments experienced no relative displacement and each north-east-striking transform fault experienced only sinistral move-ment. This distribution of growth faults in the Jurassic wouldresult in promontories and reentrants in the margin of theWest Peruvian trough, similar to those of the larger-scaleearly Paleozoic eastern margin of North America (Thomas,1977).

The Querococha arch coincided with an intermittent topo-graphic high that developed at least in the Middle Jurassic, oreven throughout the Mississippian to Middle Jurassic inter-val, and which also influenced the distribution of Cretaceouscarbonate rocks, Paleocene red beds, and Miocene igneousrocks. It is apparent that the development of the northeast-striking basement structures predated Jurassic-Cretaceoussedimentation. The Querococha arch apparently influencedthe distribution of Mississippian to Lower Jurassic rocksnortheast of Antamina, resulting in either local nondepositionin the Mississippian to Early Jurassic or a Middle to LateJurassic erosional unconformity (Fig. 17b). Additional minorextension at any angle discordant to the northeast-trendingfaults could have resulted in reactivation of the originally

908 LOVE ET AL.

0361-0128/98/000/000-00 $6.00 908

0 100 200 kms

0 100 200 miles

YANACOCHA

Lima

Huaraz

Huarmey

TrujilloChimbote

Hunuco

Huancayo

La Oroya

Huancavelica

Cerro de Pasco

79W

78W 7

7W

76W

75W

Maran

Coastal Batholith

Cordillera

6 911 14

17

20 22 D

B-14,15

B-16 B-19

Cajamarca

Rio

ANTAMINA

Blanca

12S

10S

8S

Casma

PASTOBUENO

KEYSectionIntrusionAnticlineFaultSegment boundaries discussed in textOther segment boundaries of Benavides (1999)

7PIERINA

11S

9S

N

Pacific Ocean

QuerocochaArch

Casma - PastoBueno zone

a

b 14

JatunhuasiBasin,S of La Oroya& SW of Huancayo

Cenoma

nian

Albian (97

.0 2)

clasticsequence

Creta

ceou

s

Juras

sic (1

44.8 3

.1)

calcareoussequence

Campanian and younger (< 83 4 Ma)

Ks-Ce558m

Ks-Ca800m

B-16

KsP-Ch

(equivalent totop 1/5 of Ks-J )

>920m

Ks-R (equivalent tolowest 1/5 of Ks-J)

B-14, 15

Ki-Cr613m

Ki-G>516m

135m

Paleogene~~~~~~~~~~

CretaceousUpper

~~~~~~~~~~Lower

163m

750m

B-19

190m

>413m

KsP-Ch

Ki-Cr

>180m

Ks-J

22

44m

280m

851m

337m

KsP-C>250m

20

230m

167m

59m

135m

>200m

17

115m

232m

138m

185m

>200m

52m

11

44m

266m

>250m

617m

>500m

9

122m

600m

70m

174m25m>150m

6

KsP-C

Ki-Pt

Ki-Cl 138m

>485m

25m

433m

131m

>200m

D

Pucar Gp.~700m

Carcapuquio Fm.Chunumayo Fm.

Ki-G~850m

Ki-Cl ~280m

Ki-Pt ~50m

Ks-J ~340m

KsP-P ~250m

>1000m

443m

39m

>280m

107m

25m

ANTAMINA

NORTH (north of approx. 930' S)

Chota Fm. (KsP-Ch),

Celendn Fm.(Ks-Ce)

Cajamarca Fm. (Ks-Ca)Coor Fm.Romirn Fm.Mujarrn Fm.Rosa Fm. (Ks-R)

Crisnejas Fm. (Ki-Cr)

Inca Fm.

Goyllarisquisga Gp. (Ki-G)

NORTH-CENTRAL(south of approx. 930' S)

Casalpalca Gp. (KsP-C)

Celendn Fm.(Ks-Ce)

Jumasha Fm. (Ks-J)

Pariatambo Fm. (Ki-Pt)

Chulec Fm. (Ki-Cl)

Pariahuanca Fm.

Goyllarisquisga Gp. (Ki-G)

Otuzco Gp.

Quilquian Gp.

Pulluicana Gp.

CENTRAL

Pocobamba Fm. (KsP-P)

Machay Group(north-central and central)

FIG. 16. Segmentation and articulation of the Maran thrust and fold belt and its effect on Cretaceous sedimentation.(a) Simplified geology of the western Andes of central and north-central Peru, showing the major anticlinal axes, faults andsegment boundaries in the Maran thrust and fold belt (Benavides, 1999), the major plutonic centers of the Coastalbatholith and Cordillera Blanca batholith (Pitcher et al., 1985), and the numbered locations of the sections used in (b). Themajor producing mines, Yanacocha, Pierina, Antamina and Cerro de Pasco are also shown. (b) Longitudinal fence diagramof the Cretaceous stratigraphic section of central and north-central Peru along the eastern margin of the Maran belt. Sec-tions B-14, B-15, B-16, and B-19 from Benavides (1956), sections 6, 9, 11, 14, 17, 20, and 22 from Wilson (1963), and sec-tion D from Manrique (1998). Antamina is located between sections 6 and B-19.

sinistral faults as north-side-down normal faults and henceerosion of preexisting sedimentary rocks from the transversehighs (Fig. 17b). The effect of the Querococha arch in theLate Jurassic cannot be deduced because the Upper JurassicChicama Group is not exposed in the eastern part of theMaran thrust and fold belt.

In the Cretaceous, the Querococha arch and other trans-verse structures that segment the Maran thrust and foldbelt had little apparent effect on the thickness of clastic(Goyllarisquisga Group) and carbonate (Machay Group) sed-imentary rocks on the Yauli shelf (Figs. 16b and 17c), but itappears to have affected the distribution of the carbonate


0361-0128/98/000/000-00 $6.00 909

a b

c

basement

Middle Ju

rassic

middle-La

te Cretace

ous

N

QA

CPB

f Miocene

e Eocene

dPaleo

cene

Miocene intrusionsPaleocene Casapalca Gp.Albian - Upper Cretaceous Machay Gp.Upper Jurassic - Lower Cretaceous Chicama & Goyllarisquizga Gps.Upper Triassic - Lower Jurassic Pucar Gp.Mississippian - Lower Triassic Ambo & Mitu Gps.pre-Ordovician Maran metamorphic complex

Faults transform normal thrust

Fold axis, with regional plunge Regional tectonic forces

Q

CPB

FIG. 17. Regional-scale schematic diagrams, looking east, illustrating the proposed structural evolution of the Antaminaarea. (a) In the Middle or early Late Jurassic, the West Peruvian trough started to develop through formation of an en ech-elon pattern of left-lateral growth faults on the western edge of the Maran metamorphic complex. The Mississippian toLower Jurassic sedimentary rocks that overlay the metamorphic complex are removed from the diagram for clarity (after M-gard, 1987). (b) Also in the Middle or Late Jurassic, minor extension at an angle to the transform faults segmented the con-tinental basement, producing on the promontories structural highs parallel to the original offsets, such as the Querocochaarch, along which the Mississippian to Lower Jurassic sedimentary rocks were eroded. In the Late Jurassic the ChicamaGroup was deposited in the western, deeper-water part of the basin, not illustrated. (c) Cretaceous clastic and carbonate(Goyllarisquisga and Machay Groups) sedimentation on the Yauli shelf was not strongly controlled by the segmentation al-though carbonate facies may have extended farther seaward adjacent to promontories. (d) Latest Cretaceous to Paleocenedeposition of red beds (checkerboard patterned) was controlled by the segmentation, with depocenters in the reentrants. (e)In the Eocene Incaic orogeny, major strike deflections and en echelon fold patterns that conform to the segmentation of thebasement developed in the Maran thrust and fold belt. Reentrants became structural salients, and promontories becamestructural recesses. (f) In the Miocene, additional uplift of the Querococha arch reactivated basement structures, which al-lowed intrusions to extend farther inland along the arch and to reach shallow levels. A = Antamina mine, CPB = Casma-Pasto Bueno zone, Q = Querococha arch.

rocks. The Machay Group does not extend as far west to thenorth of the arch as it does on the arch and to the south of it.From the arch north the westernmost limit of outcrop of theMachay Group therefore crosses the belt in a northeast-southwest direction (Fig. 1). Thomas (1977) noted a similarrelationship between the extent of carbonate facies and thelocations of structural salients and recesses in the Appalachi-ans. The distribution of latest Cretaceous and Paleocene redbeds has also been influenced by the transverse structures(Fig. 17d). The red beds pinch out toward both the Quero-cocha arch and the Casma-Pasto Bueno zone, and they at-tain their greatest thickness between these arches in a struc-tural salient, which would have been a reentrant anddepocenter in the margin of the West Peruvian trough at thetime of sedimentation.

Shortening and variations in the regional attitudes of foldsand thrust faults generated in the Eocene Incaic orogeny arerepresented in Figure 17e. The folds and thrusts in the An-tamina region constitute an articulated structural recess inthe margin of the craton, which probably formed through de-formation around a basement promontory, the northwesternedge of which is now delineated by the Querococha arch.Because the orientation of folds and thrust faults in thin-skinned tectonic belts generally reflects the underlyingramps rather than the translation direction of deformation(Pohn, 2000), the strike and articulation of the Maranthrust and fold belt mimic the geometry of the basement, de-spite variations in the Cenozoic plate convergence directionand rate (Pardo-Casas and Molnar, 1987; Somoza, 1998;Norabuena et al., 1999). Old transform faults in the marginof the West Peruvian trough did not experience extensivelater strike-slip movement because the maximum shorteningdirection was not parallel to the orientation of movementduring rifting. We propose that the sinuous configuration ofthe mountain belt generally reproduces the original zig-zagmargin, although the articulation is pronounced on the east-ern side of the belt, and shortening in the interior of the belthad a smoothing effect on the segmentation. Thus, the re-gional strike at the western extent of the Cretaceous carbon-ate strata gradually changes throughout a distance of approx-imately 175 km along strike, from northerly near Antaminato north-northwesterly at the northwest end of the CordilleraBlanca (Fig. 1). As Thomas (1977) concluded in the contextof eastern North America, the Maran thrust and fold belthas formed a best-fit curve around old promontories andreentrants.

The persistence of Miocene igneous rocks farther from themain axis of the magmatic arc into the foreland thrust andfold belt along a proposed basement transverse structure(Fig. 17f) is consistent with observations in other belts. Ig-neous intrusions have been documented directly over, andelongated parallel to, three of the four lateral ramps in theAppalachians analyzed in detail by Pohn (2000). Further-more, the difference in exposure level of the Miocene ig-neous rocks on and adjacent to the southwest end of the archimplies that post-Eocene uplift enhanced the plunge rever-sals of the Eocene folds across the arch (Love et al., 2001).Thus, faults parallel to the arch may have accommodated itsuplift and furnished the structural anisotropies that providedconduits for magma ascent in the late Miocene.

Regional effects on local structure and mineralization

At Antamina, the northeast-trending fracture set, the Valleyfault and the Valley lateral ramp are parallel to the regionalcross-strike structural discontinuity, the Querococha arch,and may have been controlled by similarly oriented underly-ing basement structures. At the local scale, the Valley lateralramp and the Antamina intrusion have been localized by theleft-stepping jog in the Valley fault. Regionally, about 5 kmsouthwest of Antamina, the northeast-trending locus ofchanges in strike and plunge in the Maran thrust and foldbelt steps left by approximately 8 km (Fig. 5). Thus the stepin the Valley fault developed within, and mimics, a similar,larger-scale, left-stepping jog within the Querococha arch.

The local-scale structural evolution of the Antamina area,including intrusion and formation of the orebody in the lateMiocene, is summarized schematically in Figure 18. Whereasother models, such as en echelon folding, could explain theapparent dextral offset of the Antamina anticline, the lateralramp model is preferred here because it provides a locus forlater, northeast-elongated intrusion and hydrothermal activ-ity. The northeast-striking fracture set in the host rocks pe-ripheral to the ore at Antamina (Figs. 15 and 18a) was, wecontend, formed by deformation associated with thrust trans-lation along the underlying and similarly oriented transferfault or lateral ramp (Fig. 14). The overall form of a thrustsheet does not record its passage over a lateral ramp beyondthe limit of that ramp but, in overlying thrust sheets, longitu-dinal fractures would form in the lateral anticline over theramp. These fractures would have sheared vertically in anorth-sidedown sense or opened owing to flexure above thelateral ramp or transfer fault and would persist beyond it (Fig.15d), providing evidence that a thrust sheet had traversedsuch a ramp. At Antamina, the inferred Valley lateral ramp ex-tends only 500 m from one offset anticlinal axis to the other,but strong, northeast-striking, nearly vertical fracturing, in-terpreted herein as evidence of tearing in the overridingthrust sheet, extends farther northeast (Figs. 15 and 18a) andis interpreted as trace evidence of the Valley lateral ramp.The upward-fanning axial planar cleavage related to the An-tamina anticline also formed at this time (Fig. 18a).

In this model, the left-stepping jog in the transcurrent Val-ley fault localized the Eocene development of the lateralramp that resulted in formation of the offset anticlines andalso focused the

Antamina Love

Documents

Transcript of Antamina Love