ARTICLE

Geology of the Rosário–Neves Corvo antiform, Iberian PyriteBelt, Portugal: new insights from physical volcanology,palynostratigraphy and isotope geochronology studies

J. T. Oliveira & C. J. P. Rosa & Z. Pereira & D. R. N. Rosa &

J. X. Matos & C. M. C. Inverno & T. Andersen

Received: 18 May 2012 /Accepted: 28 December 2012# Springer-Verlag Berlin Heidelberg 2013

Abstract The lithostratigraphic sequence in the Rosário–Neves Corvo antiform comprises the Phyllite–QuartziteGroup, whose top is of Famennian age, the Volcanic Sedi-mentary Complex, of Strunian to upper Visean age, and theMértola Formation (the lower unit of the Baixo AlentejoFlysch Group) of upper Visean age. The volcanic sedimen-tary complex comprises a lower sequence of Strunian (LateFamennian) age and an upper sequence of lower to upperVisean age. Detailed mapping of the antiform towards NWof the Neves Corvo mine, supported by palynological dat-ing, identified two new lithostratigraphic units: the Barran-cão member (upper Famennian) ascribed to the Phyllite–Quartzite Group and made up of laminated dark shales withsiliceous lenses and nodules, and the Ribeira de CobresFormation of the Volcanic Sedimentary Complex, contain-ing shales, siltstones and fine volcaniclastic rocks. Based onzircon U–Pb isotope dating, five discrete felsic magmaticevents were identified at approximately 354, 359, 365, 373

and 384 Ma. This suggests that the volcanic activity in thearea has extended for about 30 Ma, in a context of highregional heat flow as indicated by the geochemical signa-tures of the felsic volcanic rocks. The characteristics ofmagmatism and the depositional environment indicated bythe sedimentary record should therefore have been highlyfavourable for massive sulphide formation. However, evi-dence of massive sulphide mineralization in the study area isstill to be found. Moreover, reconstruction of the volcanicfacies architecture demonstrated that the volcanic units inthe Rosário area are strongly dominated by coherent faciestypical of the inner part of thick lavas/domes. In fact, mostof their external part, the more favourable location forpossible massive sulphide mineralization, is missing. Paly-nological dating indicates a significant hiatus, recognisedbetween the lower and upper sequences of the volcanicsedimentary complex, which implies erosion of the top ofthe volcanic centre, where VHMS deposits could possibly

Editorial handling: F. Tornos

Electronic supplementary material The online version of this article(doi:10.1007/s00126-012-0453-0) contains supplementary material,which is available to authorized users.

J. T. Oliveira :C. J. P. Rosa : C. M. C. InvernoLNEG Estrada da Portela Bairro do Zambujal Apartado,7586 Alfragide, 2610-999 Amadora, Portugal

J. T. Oliveirae-mail: tomas.oliveira@lneg.pt

C. J. P. Rosae-mail: carlos.rosa@lneg.pt

C. M. C. Invernoe-mail: carlos.inverno@lneg.pt

Z. Pereira (*)LNEG Rua da Amieira, 4465-965 S. Mamede Infesta, Portugale-mail: zelia.pereira@lneg.pt

D. R. N. RosaDept. Petrology and Economic Geology, Nationale GeologiskeUndersøgelser for Danmark og Grønland (GEUS), Øster Voldgade10, 1350 København K, Denmarke-mail: dro@geus.dk

J. X. MatosLNEG Rua Frei Amador Arrais No 39 r/c, Apartado 104,7801-902 Beja, Portugale-mail: joao.matos@lneg.pt

T. AndersenDepartment of Geosciences, University of Oslo, P.O. Box 1047Blindern, 0316 Oslo, Norwaye-mail: tom.andersen@geo.uio.no

Miner DepositaDOI 10.1007/s00126-012-0453-0

have formed. However, lateral areas of this volcanic centre,eventually preserved at depth, have good potential to hostmassive sulphide mineralization.

Keywords Iberian Pyrite Belt . VHMS deposits . Rosário–Neves Corvo antiform . Physical volcanology .

Palynostratigraphy . U–Pb dating

Introduction

The Iberian Pyrite Belt (IPB) consists of a sequence ofvolcanic and sedimentary rocks of upper Devonian to lowerCarboniferous age, which hosts significant syngenetic mas-sive sulphide deposits. This belt extends from near Seville inSpain to the Atlantic coast in Portugal, in the southwest ofthe Iberian Peninsula (Fig. 1). The IPB belongs to the Iberiansegment of the Variscan fold belt and is part of the SouthPortuguese zone, which was accreted to the Ossa-Morenazone, presently located to the north, during the late PaleozoicVariscan orogeny (Quesada et al. 1994).

The IPB is known as one of the largest volcanogenicmassive sulphide provinces in the world and includes giantdeposits such as Aljustrel and Neves Corvo in Portugal andRio Tinto and Tharsis in Spain, which are related to long-lasting focused submarine hydrothermal systems (Carvalhoet al. 1999; Tornos 2006). Due to its high economic interest,the geology of the Neves Corvo deposit has been extensive-ly studied and several publications address the geology andmetallogeny of the deposit (Oliveira et al. 2004; Pereira etal. 2003; Relvas 2000; Relvas et al. 2001, 2002, 2006a, b;Rosa et al. 2008, 2010, among many others). An agreementbetween LNEG, acting as the Portuguese Geological Surveyand Lundin Mining, the parent to the Somincor companyoperating the Neves Corvo mine allowed a detailed study ofthe Rosário–Neves Corvo antiform which incorporates, inits SE termination, the Neves Corvo mine area (Fig. 1). Themain purpose of this study was focused on the recognitionof the Neves Corvo mine stratigraphic succession across theentire antiform, the characterization of the dominant bimod-al volcanism (Munhá 1983; Rosa et al. 2004), volcanicfacies, and their relationship to volcanic centres. It is ingeneral accepted that the massive sulphide mineralizationand associated hydrothermal alteration in the host rocks isclosely related to volcanic centres. Having this in mind, itwas expected that the identification of new volcanic centreswould allow comparisons with the geology of the mine areaand pinpointing of new targets for massive sulphide explo-ration. Insights were obtained through extensive mapping ofthe antiform and from physical volcanology, petrography,geochemistry, palynostratigraphy, and isotope geochronolo-gy studies, presented in this paper. The results obtainedrepresent a contribution towards a better geological

understanding of this important segment of the Portuguesepart of the Iberian Pyrite Belt and add new information forfuture exploration activities.

Geological setting

Previous geological studies in the Rosário–Neves Corvoantiform (Leca et al. 1983; Oliveira et al. 2004). haveidentified two major IPB stratigraphic units, from baseto top, the Phyllite–Quartzite Group (PQG) and thevolcanic sedimentary complex (VSC). The VSC was con-sidered to be overlain by the Mértola Formation (Mt), thelower unit of the Baixo Alentejo Flysch Group. As in mostparts of the Pyrite Belt, the regional metamorphic gradein the study area is within the prehnite–pumpellyite facies(Munhá 1983).

In the Neves Corvo mine region, at the SE termination ofthe Rosário–Neves Corvo antiform, the lithostratigraphicsuccession has been well established and dated (Oliveira etal. 2004). However, this detailed stratigraphic successionwas at the time restricted to the mine area and geologicalmapping by private companies to extend their identificationacross the antiform was unsuccessful. The detailed fieldmapping undertaken within the scope of the present workallowed the recognition of the stratigraphic succession of theRosário–Neves Corvo antiform which is as follows, frombase to top (Fig. 1).

The Phyllite-Quartzite Group as documented by fieldmapping and drill core investigation incorporates a thick(>100 m) succession of laminated dark shales with siliceouslenses and nodules (the newly defined Barrancão member)that grades upward to the Phyllite-Quartzite Formation com-posed of shales, siltstones and quartz sandstones, in generalshowing a thickening upward lithological trend. The latter isoverlain by a decametric thick shale band with interbeddedlimestone lenses and nodules, whose exposure is restrictedto small outcrops close to the Monte do Forno da Calfarm. The Barrancão member shales yielded moderatelypreserved upper Famennian palynomorphs, whereas theshales interbedded in the upper part of the Phyllite–QuartziteFormation provided Strunian (latest Famennian) palyno-morphs, and, conodont fossils, from the Monte Forno da Callimestones, indicate an upper Famennian age (Boogard andSchermerhorn 1971).

In the Neves Corvo mine area, Oliveira et al. (2004)identified local stratigraphic gaps within the VCS succes-sion, the most significant of these corresponding to theabsence of Tournaisian sedimentary rocks. This unconfor-mity has now been confirmed for the entire antiformexplaining the subdivision of the volcanic sedimentary com-plex into a lower sequence and an upper sequence. Thelower volcanic sedimentary sequence encompasses, from

Miner Deposita

base to top, mafic volcanic rocks (mostly diabases andspilites), felsic volcanic rocks (rhyolites and rhyodacites)and dark shales. The Neves–Corvo deposit (seven massivesulphide orebodies) only known at depth is associated to ablack shale unit, called Neves Formation by Leca et al.(1983), stratigraphically at the top of the lower sequence.Leca et al. (1983) assumed that this unit outcropped in thenorthern part of the mine region. Our detailed mapping andpalynostratigraphic ages have shown that what has been

mapped as Neves Formation is indeed part of the Barrancãomember and of the Phyllite–Quartzite Formation. The thick-ness of the Lower VSC Sequence varies laterally dependingon the volcanic facies distribution, reaching in places 300 m.Palynomorph fossils assigned derived from shales from thissequence provided upper Famennian (Strunian) age.

The Upper VSC Sequence is composed of the GrandaçosFormation (dark shales with phosphate nodules, cherts andfine-grained volcanogenic sediments), the “Borra de Vinho”

Fig. 1 Geological map of the Neves Corvo antiform at the Rosário area (LNEG for LundinMining, 2010). Inset shows the location of the Iberian PyriteBelt and the study area is outlined in black

Miner Deposita

Formation (purple shales), the Godinho Formation (shales,siliceous shales and felsic volcaniclastic rocks), and theBrancanes Formation (composed of dark shales and thin-bedded greywackes), which makes the transition to theMértola Formation, both units having an upper Visean agegiven by palynomorphs and ammonoids dating. The uppersequence has a total thickness in excess of 300 m. At theNeves Corvo mine, this lithostratigraphic sequence has beendated as lower upper Visean (Oliveira et al. 2004; Pereira etal. 2008). Direct evidence for the inferred hiatus betweenLower and Upper VSC sequences was obtained at the roadcut north of Rosário village.

The accurate stratigraphic position of the newly definedRibeira de Cobres Formation (shales, siltstones and finevolcaniclastic rocks, up to 150 m thick) is uncertain sinceit is bounded by faults and no positive age determinationwas possible. This unit occurs below the Grandaços Forma-tion (see Fig. 1) and is interpreted as representing a lateralfacies variation to the grey siliceous and black shales of theGraça Formation (only identified in drill hole cores) whichyielded palynomorphs of lower Visean age (Oliveira et al.2004; Pereira et al. 2008).

From a structural point of view, the Rosário–NevesCorvo antiform is a gently SE-dipping Variscan struc-ture with a SW vergence. NW trending folds embodythe structure, with associated cleavage dipping 60–70°to NE. This deformation is related to the second episodeof Variscan folding, occurring throughout the IberianPyrite Belt. Underground, in the Neves Corvo mine,thrust sheets with a piggyback style also folded in largeopen folds during the second episode of folding wererecognised (Oliveira et al. 2004). Some of these thrustplanes crop out at the SE termination of the NE limb ofthe antiform. Also in the NE limb and central part of theantiform, the Borra de Vinho and the Godinho formations areanot represented. This gap appears to correspond to a discon-formity between the Mértola Formation flysch succession andthe underlying Grandaços Formation. NE–SW trending faultsare very common and in the mine region they are interpretedby mine geologists as growth faults that were reactivated inlate Variscan time.

Petrography of volcanic rocks

Samples were collected from outcrops and from the explo-ration boreholes indicated in Fig. 1. Thin sections of 130samples were subsequently prepared and inspected underthe petrographic and metallographic microscopes. Fourtypes of volcanic rocks basalt, dacite porphyry, rhyoliteporphyry and rhyodacite microporphyry were distinguishedon the basis of their mineralogy, textures and geochemicalsignatures (see next section).

The basalt displays an intersertal texture with randomlyoriented plagioclase microliths in a chloritized groundmass,rich in iron oxides. This rock type contains mainlycarbonate-filled vesicles. The three porphyry types are mi-crocrystalline rocks with euhedral to subhedral feldspar andquartz phenocrysts, and can be distinguished on the basis ofproportion and size of the phenocrysts. Samples with phe-nocrysts up to 7–10 % of the rock volume and with feldspar(1–3 mm) predominating over quartz (1–2 mm) were clas-sified as dacites. In the rhyolite porphyry, the phenocrystpopulation accounts for 15 % of the rock volume, withquartz predominating over feldspar phenocrysts, both inthe 1–4 mm range size. Finally, in the rhyodacite micro-porphyry phenocrysts constitute 7–10 % of the rock volume,with quartz and feldspar phenocrysts occurring in similarproportions and size (1–2 mm). The feldspar phenocrystsare often altered to carbonates, epidote and sericite. Thegroundmass of the three porphyry types consists of fine-grained sericite, quartz and feldspar and accessory chlorite,carbonate and epidote. Occasionally, the three types ofporphyry are cut by quartz veins.

Geochemistry of volcanic rocks

Seventy samples were crushed in a tungsten alloy jawcrusher, pulverised in an agate ring mill and subsequentlyanalysed for major elements and a suite of trace elements byinductively coupled plasma mass spectrometry (ICP/MS)following lithium metaborate/tetraborate fusion, at Actlabs(Ontario, Canada). Lithogeochemistry data for the morerelevant elements are presented in Electronic supplementarymaterial (ESM Table 2). These data allow the classificationof the volcanic rocks within the calc–alkaline series, accord-ing to the Irvine and Baragar (1971) triangular diagram(ESM Fig. 1), as is common in the IPB. The analysedsamples plot mostly in the rhyolite and rhyodacite/dacitefields as indicated in the alkali–silica diagram of Le Bas etal. (1986; ESM Fig. 2) and in the immobile element diagramof Winchester and Floyd (1977; ESM Fig. 3). Only samplesMD2-314.5, MD2-326, CP2-65 and MD3A classify asbasalts. The bimodal nature of the volcanism, with predom-inance of felsic rocks over mafic rocks is consistent withwhat is generally reported for the IPB (Munhá 1983;Mitjavila et al. 1997).

As previously reported for other areas of the IPB (e.g.Aljustrel (Dawson et al. 2001), Albernôa (Rosa et al. 2004),and Serra Branca (Rosa et al. 2006)), binary plots of thefelsic rocks analysed, using high-field strength elementsshow alteration lines that converge towards the origin, con-sistent with their relatively immobile behaviour (Fig. 2).Based on the variation of the immobile element concentra-tions along the alteration lines, mass changes are estimated

Miner Deposita

to have reached up to 40 %, likely as the result of theaddition of silica and other mobile elements along fracturenetworks, or its addition to or removal from the vitreousmatrix of hyaloclastic rocks. Despite the dispersion causedby fractionation trends, covariation between element pairs ineach felsic rock type can be used to broadly discriminate thefelsic rock types based on their element ratios (Fig. 3). Theleast evolved felsic rock, the dacite porphyry has ratios of70<Al2O3/TiO2<90 and 600<Zr/TiO2<900, while the

rhyodacitic microporphyry and rhyolite porphyry have higherAl2O3/TiO2 (>90) and Zr/TiO2 (>900, often >1,000) ratios,respectively.

Most felsic rock samples plot in the within-plate field ofPearce et al. (1984) diagrams (Fig. 4), which is consistentwith the currently accepted model of IPB magmatism in acontext of attenuated continental lithosphere within pull-apart basins resulting from an oblique collision (Silva etal. 1990; Quesada 1991; Tornos et al. 2002). In theRosário–Neves Corvo antiform area, there is therefore anoverall agreement between the petrographic features and thegeochemical signatures of the felsic volcanic rocks ana-lysed, and the geotectonic setting admitted for the IPB. Thissuggests that magmatic heat flow should have been com-paratively higher than in other sectors of the belt, where lowtemperatures of crustal fusion have been proposed as anexplanation for unexpectedly low concentrations of high-field strength elements in the felsic volcanic rocks formedthere (Rosa et al. 2004, 2006).

The behaviour of mobile elements and how it relates tothe alteration minerals formed was assessed using the alter-ation box plot of Large et al. (2001). This graphic represen-tation is based on the combination of the alteration indexproposed by Ishikawa et al. (1976) and the chlorite–carbon-ate–pyrite index proposed by Large et al. (2001). On plot-ting our data in this diagram, both a diagenetic and ahydrothermal trend are defined (Fig. 5). The diagenetictrend, directed towards albite, is, according to these authors,typical of the interaction with seawater at low temperature(spilites and keratophyres). The hydrothermal trend, direct-ed towards sericite, is typical of distal ore-related hydrother-mal alteration. Hydrothermal trends indicative of proximalhydrothermal alteration patterns were not detected, in accor-dance with the general absence of mineralogical, petro-graphic or other geochemical evidence of proximal ore-forming hydrothermal activity in the study area.

Physical volcanology

The reconstruction of the volcanic centres of the study areawas carried out both resorting to facies analysis along creeksand roads where the VSC crops out with great continuity, andto relogging core from old exploration holes drilled in theRosário–Neves Corvo antifom. The VSC comprises abundantfelsic volcanic units, subordinate mafic units and relativelythick volcaniclastic units that are more abundant in the centraland SE zones of the study area. The low metamorphic grade(prehnite–pumpellyite facies) and the thin-skinned tectonicstyle of the IPB allow the preservation of primary volcanictextures away from the principal shear zones.

In the Rosário–Neves Corvo antiform, the felsic unitshave rhyolitic, rhyodacitic and dacitic composition and are

Fig. 2 Zr–TiO2 diagram displaying an alteration line for each of thefelsic volcanic rock types

Fig. 3 Al2O3/TiO2–Zr/TiO2 diagram discrimination diagram for felsicvolcanic rocks

Miner Deposita

largely predominant over the mafic units. They were distin-guished in the field, according to their phenocryst contentand contact relationships with other volcanic and sedimen-tary units, and subsequently confirmed on the basis of theirgeochemical signatures (previous section). The felsic unitscomprise dominant coherent facies with minor autoclasticbreccia at their margins, and occur as lavas and domes. This

is the most common type of volcanoes throughout the IPBand may also comprise abundant and thick pyroclastic units(Rosa et al. 2010).

The coherent facies typically define thick and laterallyextensive intervals. Typical textures include even, continuousand planar flow bands (Fig. 6a) that are parallel to the contactsof the volcanic units but locally show tight flow folds.

Fig. 4 Geotectonic settingdiagram of Pearce et al. (1984).Only felsic rock samples

Fig. 5 Alteration box plot(Large et al. 2001)

Miner Deposita

Spherulites, lithophysaes and elongated quartz-filled amyg-dales are also abundant in the volcanic groundmass. Thespherulites can be up to 2 cm across and are aligned alongthe flow bands (Fig. 6b), whereas the lithophysaes can be up to20 cm across and have a nucleus made of silica. Spherulitesand lithophysaes are characteristic of high-temperature devit-rification of glass (Lofgren 1971) and are typical of coherentfacies that occur in the inner portions of lavas and domes.

The autoclastic facies are dominated by breccia intervalswith variable thickness that occur at the contacts of the felsicunits or more rarely in their inner parts, and grade to thecoherent facies through intervals of discrete fractures. Theclasts are angular, have planar and parallel margins, andrange from 1 to 50 cm across (Fig. 6c). The breccia intervalsare typically clast-supported with few or no matrix betweenthe angular clasts. The clast characteristics and the abundantflow foliation are consistent with clasts having formed bybrittle fragmentation during flow of lavas or domes due toviscosity contrasts within the flows, and correspond to auto-breccia. Other clasts are irregular, and have planar andcurviplanar margins, which is typical of quench fragmenta-tion and hyaloclastite (Pichler 1965).

The lavas and domes are dominantly rhyodacitic. However,some rhyolitic and dacitic terms were also recognised, havingdistinct distribution and facies association within the studyarea. Dacitic units were identified in the Cobres creek and incore from boreholes MD2 (Fig. 7) and MT1. These units arerelatively thin and have small volume, but their mode ofemplacement was not reliably determined as the top contactsare poorly exposed. The rhyodacitic units are the most abun-dant and occur throughout the study area. These units are thick(up to 500 m), have great lateral extent and are dominated bycoherent facies. Their top contacts with sedimentary units aregenerally sharp, or consist of a sediment–matrix igneousbreccia (Rosa et al. 2013) with the spaces between the rhyo-dacitic clasts occupied by sediments. The rhyolitic units arethinner and less abundant than the rhyodacitic units and scatterthroughout the study area. They typically have irregular anddiscordant contacts with enclosing rhyodacitic and sedimen-tary units that indicates they are intrusive.

In addition to the lavas and domes, abundant thick (up to30 m), massive to normally graded and poorly sorted poly-mictic breccias occur throughout the area. These componentsare typically angular, vary from 0.5 to 20 cm long and havevolcanic or sedimentary origin. The volcanic clasts have sim-ilar composition and textures to the rhyodacitic and rhyoliticunits; however, aphyric clasts and fragments of quartz andfeldspar phenocrysts are also present. Small black wispyelements are part of these units as well. They occur parallelto bedding and have uncertain composition due to their smallsize and strong flattening. These characteristics suggest theymay have been originally dense glassy clasts, or pumice clasts,which in any case would favour their strong compactioncompared with other clasts of crystalline groundmass. Thesedimentary clasts are more abundant at the coarser lower partof the volcaniclastic units and consist of chert and shale.

The large volume of coherent facies and in situ autoclas-tic breccia of felsic units indicate a proximal setting to thevolcanic centres. The fact that the volcanic clasts in thepolymictic breccias have similar composition to the rhyoda-citic and rhyolitic units suggests that the former weresourced from the felsic units. The dominantly massive andpoorly sorted nature of the polymictic breccia units indicatesthey have not been subject to density sorting, suggestingthey have been emplaced proximal to their source, definingvolcaniclastic aprons at the margins of the volcanic centres(e.g. Rosa et al. 2010).

In the Rosário–Neves Corvo antiform, the basalts are lessabundant than the felsic units but were recognised through-out the area, locally with significant thickness and lateralextent. The basalts typically define coherent intervals withminor clastic zones at the contacts. They occur at the base ofthe VCS sequence, immediately above the PQ Group, or atthe top of the volcanic sequence, above the youngest felsicunits (e.g., Maria Delgada Creek; see Fig. 1). In some cases,the basaltic units have irregular and discordant top contactswith the bedding of sedimentary units, which indicates theyare intrusive, whereas in other cases they have a breccia atthe contact, which grades to the overlying volcaniclasticunits and indicates they are extrusive.

Fig. 6 aMacro-spherulites, up to 2 cm in diameter, aligned parallel to flowbanding in a rhyodacitic unit. b Flow-banded rhyodacite, showing flowfolds. c Jigsaw fit autoclastic rhyodacitic breccia

Miner Deposita

Isotope geochronology

Analytical methods

Zircon grains were separated from selected samples usingstandard density and magnetic procedures at the Universityof Algarve (Portugal) and at the University of Oslo(Norway). These procedures included the crushing of sam-ples in a jaw crusher, followed by sieving of the <425 μmfractions. The heavy minerals from these fine fractions werethen separated using heteropolytungstate solution. Subse-quently, the heavy mineral fraction was separated into frac-tions with different magnetic susceptibilities using a Frantzisodynamic separator. Finally, zircons were handpicked,

under a binocular microscope, from the nonmagnetic frac-tion and mounted in epoxy and polished. Detailed imagingof each grain was carried out using cathodoluminescenceand backscattered electron imaging using a JEOL JSM-6460LV scanning electron microscope. Special care wasput into identifying inherited cores and limits between dif-ferent generations of magmatic zircons to ensure that anal-ysis did not include different domains.

U–Pb isotope analyses were performed using a NU Plas-ma HR multicollector ICP-MS at the Department of Geo-sciences, University of Oslo, which is equipped with a U–Pbcollector block (Simonetti et al. 2005). A New Wave/Mer-chantek LUV213 Nd:Yag laser microprobe was used. Sam-ples were ablated in He (gas flow=1.0 l/min) in a specially

Fig. 7 Simplified descriptivelog of drill hole MD2, showingthe location of the samplesselected for U–Pb isotopegeochronology

Miner Deposita

built ablation cell. The He aerosol was mixed with Ar (gasflow=0.7 l/min) in a Teflon mixing cell prior to entry intothe plasma. This gas mixture was optimised daily for max-imum sensitivity. The laser beam was focussed in apertureimaging mode with a circular spot geometry; all analyseswere made in static ablation mode. Masses 204, 206 and 207were measured in secondary electron multipliers and 238 inthe extra-high Faraday collector of the Nu Plasma U–Pbcollector block. 235U was calculated from the signal at mass238 using a natural 238U/235U=137.88. Laser conditionswere: beam diameter, 40 μm; pulse frequency, 10 Hz; beamenergy density, ca. 0.06 J/cm2.

A measurement procedure included a 30 s on peak back-ground measurement with the laser turned off, followed by60 s of ablation. Standard zircons GJ-01 (609±1 Ma;Jackson et al. 2004) and 91500 (1065±1 Ma; Wiedenbecket al. 1995) were used for calibration. These calibrationstandards were analysed in duplicate at the start and end ofeach analytical session, and at regular intervals throughoutthe day. Tl correction for Pb isotope mass discrimination inthe plasma was not used. Mass number 204 was used as amonitor for common 204Pb. In an ICP-MS analysis, 204Hgoriginating from the argon supply contaminates mass 204,giving a background counting rate of ca. 1,000 cps. Thecontribution of 204Hg from the plasma was eliminated byon-mass background measurement prior to each analysis. Atthe low laser energy used, there was no excess ionisation of204Hg from the gas supply during ablation, so that the on-mass background measurement is representative for theconditions during analysis. When necessary, the observedsignals at masses 206 and 207 were corrected for common206Pb and 207Pb after integration of the signal, using ob-served 204Pb and average common-lead composition givenby the global lead evolution curve of Stacey and Kramers(1975) at the uncorrected 206Pb/238U age.

Raw data from the Nu Plasma time-resolved analysisprogramme were imported into an in-house Microsoft Excel/VBA spreadsheet programme (NuAge.xlt, written by T.Andersen) for interactive selection of isotopically homoge-neous integration intervals, background correction, calibrationto standards and calculation of ages. To minimise laser-induced fractionation of U from Pb, we followed Jacksonet al. (2004) in combining data for isotopically homogeneousintervals of the time-resolved signal of an unknown with thecorresponding time (=depth) intervals on the standards, as-suming similar ablation rates for standards and unknowns.Background-corrected signals for mass numbers 204, 206,207, 235 and 238 and the 207Pb/206Pb, 206Pb/238U and 207Pb/235U isotope ratios were plotted as traces of observed voltageand voltage ratios against ablation time, these allowing homo-geneous parts of the run to be interactively selected for inte-gration. Isotopically homogeneous segments of the time-resolved traces were calibrated against the corresponding time

interval for each mass in the reference zircon. The raw datafrom the calibration standards were fitted to a relationship

y ¼ x aþ bxþ ctð Þwhere y is the true isotope ratio; x is the observed ratio for therelevant depth interval; t is time since the first standard anal-ysis in a series; and a, b and c are coefficients determined bythe built-in regression algorithms of Microsoft Excel 2003.The non-linear, 3D calibration curve was used to compensatefor an observed deviation from linearity of the ion counters athigh counting rates. Two reference samples were used ascalibration standards. Observed errors in background and sig-nal for reference zircon and unknown, and the uncertainty ofthe published standard composition were propagated through,using normal error propagation formulas (e.g. Taylor 1997).

The long-term (>2 years) precision is <1 % for 206Pb/238Uand 207Pb/206Pb (2 SD, Isoplot 4.0; Ludwig 2003) was used toplot probability density plots and identify distinct age frac-tions which can be present in one sample. In case these distinctage fractions were identified, grains belonging to each ofthem were discriminated. Subsequently, concordia diagramswere plotted to calculate concordia ages for each of the agefractions.

Results

Zircons from eight samples were analysed. Additionally, theTemora-2 (TIMS-ID U–Pb age: 416.8±1.3 Ma; Black et al.2004) reference zircon was run as an unknown. The mostrelevant results are included in ESM Table 3, with a sum-mary of obtained U/Pb ages compiled in ESM Table 1.Samples R14, C6, MD2-106 and CP2-366 contain a mixtureof zircons with different, yet Variscan, ages. Figure 8 dis-plays three concordia diagrams that yield concordia ages(Ludwig 1998) of 359±3, 366±1 and 373±2 Ma, based on4, 29 and 8 grains of sample R14 (a rhyodacite), respective-ly. Figure 9 displays three concordia diagrams that yieldconcordia ages of 359±1, 366±1 and 373±1 Ma, based on8, 19 and 8 grains of sample C6 (a rhyodacite), respectively.Another rhyodacite (Fig. 10), sample MD2-106, providedthree grains that correspond to an age of 358±2 Ma , sevengrains that correspond to an age of 365±2 Ma, three grainsthat correspond to an age of 373±2 Ma and finally, fourgrains that correspond to an age of 385±2 Ma. A rhyolite,sample CP2-366, yielded ages of 354±2, 361±1 Ma, 373±2and 382±4 Ma, based on 6, 29, 7 and 2 zircon grains,respectively (Fig. 11). For all these samples, the youngestage fraction is interpreted to be the emplacement age, withthe older grains reflecting inherited components.

For samples R1, C3, MD13, and MD2-308 (rhyodacites)only one age fraction has been identified (Fig. 12). Insample R1, 15 zircon grains yielded a concordia age of364±2 Ma; in sample C3, 17 grains gave a concordia age

Miner Deposita

of 364±2 Ma; in MD13, 8 grains provided a concordia ageof 364±3 Ma, and finally in MD2-308, 14 grains provided a

concordia age of 374±2 Ma. The identification of only oneage fraction can be the result of the small number of zircon

Fig. 8 Concordia diagrams for the different age fractions identified insample R14

Fig. 9 Concordia diagrams for the different age fractions identified insample C6

Miner Deposita

grains that were recovered and analysed for each sample. Inthis case, the obtained age can be either interpreted as themost recent age, inferred as the emplacement age, andinherited fractions that might be present could not have beenrecovered. Or, the obtained age can be of an inheritedfraction and the youngest fraction, corresponding to theemplacement age, could not have been recovered. Therefore,these results need to be interpreted with caution. However, itseems most likely that, similar to the other rhyodacites, thesefour rocks were also emplaced at approximately 359 Ma. Inthis case, the identified age fractions would correspond only toinherited components.

Palynostratigraphic studies

Materials and methods

For palynology studies, a total of 212 samples of sed-imentary rocks were collected, mostly from drill cores.Drill cores selected in this research include MD2 (17

samples), CP1 (15 samples), MT1 (36 samples), MT2(17 samples), A6-1 (12 samples), CP2 (20 samples),NC16 (13 samples) and NC20 (9 samples). In addition,73 samples from selected outcrops were also investigat-ed. Simplified logs of the studied drill cores are presented inFig. 13.

The biostratigraphic research undertaken was based onpalynomorphs. Standard palynological laboratory proce-dures were used in the extraction and concentration of thepalynomorphs from the host sediments (Wood et al. 1996).The slides were examined in transmitted light, using a BX40Olympus microscope equipped with an Olympus C5050digital camera. All samples, residues and slides are storedin the LNEG-LGM at S. Mamede Infesta, Portugal. Themiospore biozonal scheme used follows the standard West-ern Europe Miospore Zonations (after Clayton et al. 1977;Streel et al. 1987; Higgs et al. 1988, Clayton 1996; Pereira1999; Pereira et al. 2007) and is presented in Fig. 14,together with the ranges of selected miospore taxa recov-ered. Important stratigraphic taxa are illustrated in ESMFigs. 4 and 5.

Fig. 10 Concordia diagrams for the different age fractions identified in sample MD2-106

Miner Deposita

Results

This study showed that surface samples are very poor in termsof preservation of organic matter, rendering them useless forcollecting stratigraphic information. Although some drill coresamples were extremely pyritised, which does not contributeto the preservation of miospores and acritarchs, most samplesprovided organic matter in suitable conditions for the study ofpalynomorphs. These positive results are those that have beenincluded in the simplified logs of Fig. 13, their host units andformations being described from base to top. In Fig. 14, achronostratigraphic chart for the Rosário–Neves Corvo anti-form, is tentatively proposed, based on the palynostratigraphicdata obtained coupled with our interpretation of the geologicrelationship between the various units and their correlationwith the sequence previously set at the Neves Corvo mine area(Oliveira et al. 2004).

Phyllite–Quartzite Group

At the core of the Rosário–Neves Corvo antiform, the blackshales of Phyllite–Quartzite Formation occurring in boreholes

NC16, CP1, CP2 and NC20 provided moderately preservedmiospore assemblages, of the LN Biozone, of upper Strunianage (uppermost Famennian; Streel et al. 2006), characterisedby the presence of abundant specimens of Apiculiretusisporasp., Auroraspora macra, Diducites sp., Indotriradites sp.,Geminospora lemurata,Grandispora echinata,Knoxisporitesliteratus, Punctatisporites irrasus, Retispora lepidophyta,Retusotriletes incohatus, Rugospora flexuosa, Vallatisporitespusillites, and Vallatisporites verrucosus together with veryrare specimens of the guide species Verrucosisporites niditus(ESM Fig. 4). Prasinophytes of the genusMaranhites spp. andacritarchs are also common in the recovered assemblages, inparticular the species Gorgonisphaeridium sp., Gorgoni-sphaeridium plerispinosum, Navifusa bacilla, Veryhachiumsp., and Umbellasphaeridium saharicum (ESM Fig. 5). Thelatter is also a key species for the Strunian age. An interestingfeature observed in these studied assemblages, in particularthose recovered from boreholes NC16 and NC20, is thecommon presence of the species Aneurospora greggsii, Che-linospora sp., Cristatisporites triangulatus, Cymbosporitessp.,Dictyotidium sp., Lophozonotriletes sp., Verrucosisporitesbulliferus, Verrucosisporites premnus and Verrucosisporites

Fig. 11 Concordia diagrams for the different age fractions identified in sample CP2-366

Miner Deposita

scurrus. These species are interpreted as reworked material ofFrasnian age. In other areas, mainly to the SW of the Cobrescreek (Fig. 1; samples X2–X6), the Barrancão member iscomposed of laminated dark shales with siliceous lenses andnodules. These facies were sampled in an outcrop along theBarrancão creek. Four samples yielded a moderate preservedmiospore assemblage, assigned to the VH Biozone of Famen-nian age (Fig. 14). The assemblages contain Apiculiretusis-pora sp., G. echinata and Punctatisporites sp. and aredominated by specimens of Maranhites spp. These determi-nations confirm the setting of the Barrancão member as a newunit of the PQG.

Lower volcanic sedimentary complex

The black to grey shales intercalated in the volcaniclasticrocks belonging to the Neves Formation (see boreholes A6-1 and MD2), provided moderately to poorly preserved mio-spore assemblages, assigned to the LN Biozone, of upperStrunian age. Specimens of Punctatisporites sp., R. lepido-phyta, Retusotriletes sp., R. flexuosa, Vallatisporites sp. and

V. verrucosus are present in the assemblages. Maranhitesspp. and rare acritarchs are also part of these assemblages.

Upper volcanic sedimentary complex

In the upper VSC units, only two units allowed biostrati-graphic results, the Grandaços Formation and the God-inho Formation. The Grandaços Formation provided verypreliminary biostratigraphic information in the study ar-ea. Data from outcrop samples R3 and R4 (Fig. 1) reg-ister a poor determination of miospores. In this section,the Grandaços Formation is in direct contact with theMértola Formation. Samples at 349.5 and 342.0 m fromborehole MD2 provided a poorly preserved miosporeassemblage of lower upper Visean age, based on thepresence of rare specimens of Lycospora sp. and Denso-sporites sp. In the Neves Corvo mine region, this unitwas dated as lower Visean (Oliveira et al. 2004; Pereiraet al. 2008), whereas the Godinho Formation shalesyielded miospores ascribed to the NM Biozone, indicat-ing a lower upper Visean age.

Fig. 12 Concordia diagrams for samples yielding only one age fraction: R1, C3, MD13, and MD2-308

Miner Deposita

Baixo Alentejo Flysch Group (Mértola Formation)

The shales and greywackes of the Mértola Formation weresampled in borehole MT1 providing miospores speciesassigned to the NM Biozone, indicating a lower upperVisean age. This unit is correlative to the Mt1 beds of theNeves Corvo mine (Oliveira et al. 2004; Fig. 14).

Discussion

The isotope geochronology study indicates that in theRosário area there were, at least, five discrete magmaticevents, at approximately 354, 359, 365, 373 and 384Ma. The rhyolite surrounded by Strunian (360.7±2.7Ma; Kaufmann 2006) sedimentary rocks was emplacedat ∼354 Ma and is interpreted to be locally intrusive, con-firming the results of the physical volcanology study. Therhyodacites are mostly extrusive and were emplaced at ap-proximately 359 Ma, but possibly also at ∼365 and ∼373 Ma.While these two latter events may or may not have beenexpressed by the emplacement of dated volcanic rocks, theoldest event (at ∼384 Ma) was certainly not. It is possible thatthese older magma batches were emplaced at deeper levels

and therefore the equivalent volcanic rocks have not beensampled. In any case, these results are noteworthy in the sensethat, unlike what has been generally reported to date in otherareas of the Iberian Pyrite Belt (Rosa et al. 2009); the studiedvolcanic rocks register the successive melting of pre-existing,but still Variscan, volcanic rocks, as indicated by the signifi-cant fraction of inherited zircon grains. This suggests that, inaddition to a relatively high heat flow, as is interpreted fromthe geochemical signature, this heat flow remained for asignificant time. Therefore, at least locally, the volcanic activ-ity must have extended for about 30 Ma, which would havefavoured vigorous and long-lasting hydrothermal activity,favourable to the formation of massive sulphide deposits.Also, this volcanism was active during the Strunian, whichhas been repeatedly noticed as a favourable time hori-zon for VHMS mineralization in the IPB (Matos et al.2011). Furthermore, the focussed discharge of largevolumes of ore-forming fluids into a deep and quiet basinwith anoxic conditions, as documented by its sedimentaryrecord (shales rich in organic matter), is also auspiciousfor VHMS mineralization. However, despite this set offavourable conditions, massive sulphide mineralization, aspresent in the Neves Corvo mine, is still to be recognised inthe Rosário area.

Fig. 13 Simplified stratigraphic logs of boreholes NC16, CP1, CP2,MD2, NC20 and MT1 drilled by Billiton-Minaport, A6-1 drilled byRiofinex, and MT2 drilled by Somincor. The locations of positivelystudied samples are shown and the main palynostratigraphic

correlations are displayed. The red box, in borehole MD2, correspondsto the interval logged in detail in Fig. 11.Mt Mértola Fm., s GrandaçosFormation, g Godinho Formation, n Neves Formation, PQ PhylliteQuartzite Formation

Miner Deposita

Both the Neves Corvo and Rosário volcanic centreshave characteristics of lava–cryptodome–pumice conevolcanoes, which are a typical volcano type of theIPB (Rosa et al. 2010). Nevertheless, the results of thisinvestigation provide plausible reasons for the contrast-ing resource endowment between these two volcaniccentres. The absence of proximal facies of significanthydrothermal systems indicated by the mineral assemb-lages and the geochemical signatures identified in thevolcanic rocks studied can be the result of obliterationof the upper part of the volcanic edifice as recorded byan important hiatus in the local stratigraphic succession.At the road cut north of Rosário village, the top contactof the rhyodacitic unit (tentatively dated at 364±2 Ma—sample R1) consists of a breccia with sedimentarymatrix (Fig. 15), in which sediment fills the spacesbetween clast-supported breccia (Rosa et al. 2013).The sedimentary component of the breccia consists ofsiliceous shales with Chert nodules, and is in lateralcontinuity with outcrops of the Grandaços Formationthat yielded a spore assemblage of lower upper Viseanage (∼340 Ma). Therefore, there is a hiatus of ∼24 Mabetween the emplacement of the volcanic rocks and

shale sedimentation. Even if the tentative date of therhyodacitic unit is considered to be ∼359 Ma, asdiscussed above, the stratigraphic hiatus corresponds toa ∼19 Ma time span. In any case, this time gap isindicative of the erosion of the upper part of the volca-nic edifice and, consequently, of any possible massivesulphide mineralization that might have formed. This isfurther confirmed by abundant volcanic textures that areindicative of coherent facies occurring in the inner partof lavas and domes. This set of evidence is in accor-dance with the biostratigraphic results obtained for theNeves Corvo mine (Oliveira et al. 2004). Anotherpossible reason for the lack of massive sulphide miner-alization in the Rosário area (exception, stockworkmineralization in Algaré; see Fig. 1) is the absence ofa thick and continuous pyroclastic substrate, overlain byrhyolitic lavas and black shales of Strunian age asidentified at the Neves Corvo mine (Rosa et al. 2008,2010). This permeable pyroclastic substrate mighthave been a critical factor in favouring hydrothermalcirculation and, ultimately, ore formation associatedto the Neves Corvo volcanic centre (Rosa et al.2008).

Fig. 14 Chronostratigraphic chart for the Rosário area based onthe palynostratigraphic data obtained in this research, and corre-lation with the Neves Corvo mine sequence (Oliveira et al. 2004).

Absolute ages following Korn and Kaufmann (2009) for theTournaisian and Visean (Carboniferous) and Kaufmann (2006)for the Devonian

Miner Deposita

Concluding remarks

The structure of the Rosário–Neves Corvo antiform is com-plex, resulting from the thin-skinned tectonic style typical ofthe IPB and of late Variscan fragile segmentation. The VSCis made up of a lower and an upper sequence. The lowersequence, from oldest to youngest, consists of mafic andfelsic volcanic rocks, and the Neves Formation of upperStrunian age. The upper volcanic sequence comprises fromoldest to youngest, the Ribeira de Cobres Formation, theGrandaços Formation, the “Xistos Borra de Vinho” Forma-tion with jaspers, the Godinho Formation and the BrancanesFormation. The Grandaços Formation was tentatively datedas mid-upper Visean, and the Godinho and Mértola Forma-tions were dated as mid upper Visean. Two new lithostrati-graphic units are suggested: the Barrancão member (fromPQG) of upper Famennian age, consisting of laminated darkshales with siliceous lenses and nodules, and the Ribeira deCobres Formation (from VSC), characterised by shales, silt-stones and fine-grained volcaniclastic rocks. The latter for-mation is considered as a possible lateral equivalent to theGraça Formation hosting a felsic volcanic unit not identifiedin outcrops in the Rosário area, but only recognised in drillcore at the Neves Corvo mine (Oliveira et al. 2004). Thegeochemical and geochronological results obtained in thisstudy indicate that a long lasting (∼30 Ma) and large volca-nic system operated in the Rosário area. Such a system wasthe key stone for starting and maintaining the circulation ofmineralising fluids, possibly leading to the formation ofVHMS deposits and associated hydrothermal alterationhaloes, of which the neighbouring Neves Corvo deposit isa representative. This important hydrothermal system wouldhave been active during the Strunian, a favourable geolog-ical time for ore formation in the IPB, coeval with sedimen-tation in a quiet and anoxic environment, attested by blackshale deposition. Also, similar to what happens at Neves

Corvo, the study area corresponds to a proximal setting to aneffusive volcanic centre.

However, no evidence for hydrothermal alteration indic-ative of close proximity to massive sulphide mineralizationhas been found in the study area. Rather, only hydrothermalalteration typical of distal settings has been identified andonly small pyrite disseminations were found. One possibil-ity is that this may reflect the lack of explosive volcanicactivity sourcing thick and continuous pyroclastic unitsbelow the felsic flows, a factor that is envisaged as havingplayed an important role in the formation of the NevesCorvo massive sulphide ores. Furthermore, the age con-straints and volcanic facies analysis addressed in this studyindicate a significant stratigraphic hiatus between the volca-nic units and the immediately overlying sediments of theGrandaços Formation, implying significant erosion of thetop of the volcanic centre, which would have obliterated anypossible massive sulphide mineralization from the geologi-cal record. This hiatus extends over the entire Tournaisianand the base of the Visean. Nevertheless, good potential forVHMS mineralization exists in marginal or lateral areas ofthe Rosário volcanic centre. Those areas have potential tohost massive sulphide mineralization at depth, and are worthbeing tested through drilling.

Acknowledgments This work was sponsored by Lundin Miningthrough project VOLCROSARIO—Physical Volcanology and Petro-chemical Studies of Volcanic Rocks in the Neves Corvo ExplorationArea, in the Pyrite Belt. We would also like to thank Paulo Fernandesand Bruno Rodrigues (University of Algarve) for their help with zirconseparation, and acknowledge Berit Løken Berg and Siri Simonsen(University of Oslo) for their assistance with sample characterisationand assistance during LA-ICPMS work. The authors would also like tothank José Garras Leal (LNEG) for field work during geologicalmapping of the area. Jaime Máximo (LNEG) helped the team as forthe GIS system associated with the Rosário mapping. This is contri-bution no. 33 from the Isotope Geology Laboratory of the Departmentof Geosciences, University of Oslo.

Fig. 15 a Road cut, north of the Rosário village, showing the contactbetween the rhyodacites (left) and the shales (right). b Enlarged areamarked in A with a square, showing the infiltration breccia at the

contact (solid black line) between the clast-supported rhyodacitic brec-cia (R) and the shales (S). White-coloured sediment filling the spacebetween clasts in the rhyodacite breccia is indicated by arrows

Miner Deposita

References

Black LP, Kamo SL, Allen CM, Davis DW, Aleinikoff JN, Valley JW,Mundil R, Campbell IH, Korscha RJ, Williams IS, Foudoulis C(2004) Improved 206Pb/238U microprobe geochronology by themonitoring of a trace-element-related matrix effect; SHRIMP, ID–TIMS, ELA–ICP–MS and oxygen isotope documentation for aseries of zircon standards. Chem Geol 205:115–140

Carvalho D, Barriga FJAS, Munhá J (1999) Bimodal siliciclasticsystems—the case of the Iberian Pyrite Belt. In: Barrie CT,Hannington MD (eds), Volcanic-associated massive sulfidedeposits: processes and examples in modern and ancient settings.Rev Econ Geol vol 8. pp 375–408

Clayton G (1996) Mississippian miospores. In: Jansonius J, McgregorDC (eds), Palynology: principles and applications. AmericanAssociation of Stratigraphic Palynologist Found 2:589–596

Clayton G, Coquel R, Doubinger J, Gueinn KJ, Loboziak S, Owens B,Streel M (1977) Carboniferous miospores of Western Europe:illustration and zonation. Meded Rijks Geol Dienst 29:1–71

Dawson GL, Barret TJ, Caessa P, Alverca R (2001) The Feitais poly-metallic massive sulphide deposit, Southern Portugal. In: TornosF, Pascual E, Sáez R, Hidalgo R (eds) Massive sulphide depositsin the Iberian Pyrite Belt: new advances and comparison withequivalent systems: proceedings of GEODE workshop, Aracena,Spain, October 2001. Geode, Univ. Huelva, p 15

Higgs K, Clayton G, Keegan BJ (1988) Stratigraphic and systematicpalynology of the Tournaisian rocks of Ireland. Geological SurveyIreland Spec Pap 7:1–93

Irvine TN, Baragar WRA (1971) A guide to the chemical classificationof the common volcanic rocks. Can J Earth Sci 8:523–548

Ishikawa Y, Sawaguchi T, Iwaya S, Horiuchi M (1976) Delineation ofprospecting targets for Kuroko deposits based on modes of vol-canism of underlying dacite and alteration halos. Min Geol26:105–117

Jackson SE, Pearson NJ, Griffin WL, Belousova EA (2004) The appli-cation of laser ablation-inductively coupled plasma-mass spectrom-etry to in-situ U–Pb zircon geochronology. Chem Geol 211:47–69

Kaufmann B (2006) Calibrating the devonian time scale: a synthesis ofU–Pb ID-TIMS ages and conodont stratigraphy. Earth Sci Rev76:175–190

Korn D, Kaufmann B (2009) A high-resolution relative time scale forthe Visean Stage (Carboniferous) of the Kulm Basin (RhenishMountains, Germany). Geol J 44:306–321

Large RR, Gemmell JB, Paulick H, Huston DL (2001) The alteration boxplot: a simple approach to understanding the relationship betweenalteration mineralogy and litogeochemistry associated withvolcanic-hosted massive sulfide deposits. Econ Geol 96:957–971

Le Bas MJ, Le Maitre RW, Streckeisen A, Zanettin B (1986) Achemical classification of volcanic rocks based on the total alka-li–silica diagram. J Petrol 27:745–750

Leca X, Ribeiro A, Oliveira JT, Silva JB, Albouy L, Carvalho P,Merino H (1983) Cadre géologique des minéralisations de NevesCorvo (Baixo-Alentejo, Portugal)—Lithostratigraphie, paléo-géographie et tectonique. Mém BRGM 121–1983. Éditions duBRGM, Orléans, France, p 79

Lofgren G (1971) Experimentally produced devitrification textures innatural rhyolitic glass. Geol Soc Am Bull 82:111–124

Ludwig KR (1998) On the treatment of concordant uranium–lead ages.Geochim Cosmochim Acta 62:665–676

Ludwig KR (2003) Isoplot 3.0—A geochronological toolkit for Micro-soft Excel: Berkeley Geochronology Center Special Publicationno. 4, p 70

Matos JX, Pereira Z, Rosa CJP, Rosa DRN, Oliveira JT, Relvas JMRS(2011) Late Strunian age: a key time frame for VMS depositexploration in the Iberian Pyrite Belt. 11TH SGA biennial meeting,4th edn. Abs. Book, Antofagasta, Chile, pp 790–792

Mitjavila J, Martí J, Soriano C (1997) Magmatic evolution and tectonicsetting of the Iberian Pyrite Belt volcanism. J Petrol 38:727–755

Munhá J (1983) Hercynian magmatism in the Iberian Pyrite Belt. In:Sousa MJL, Oliveira JT (eds), The Carboniferous of Portugal,Mem Serv Geol Portugal 29:39–81

Oliveira J, Pereira Z, Carvalho P, Pacheco N, Korn D (2004) Stratigraphyof the tectonically imbricated lithological succession of the NevesCorvo mine area, IPB, Portugal. Miner Deposita 39:422–436

Pearce JA, Harris NBW, Tindle AG (1984) Trace element discrimina-tion diagrams for the tectonic interpretation of granitic rocks. JPetrol 25:956–983

Pereira Z, Pacheco N, Oliveira JT (2003) A case of applied palynology:dating the lithological succession of the Neves–Corvo Mine, IberianPyrite Belt, Portugal. In: Wong TE (ed) Proceedings of the XVthInternational Congress on Carboniferous and Permian Stratigraphy.R. D. Academy Arts and Sciences, Utrecht, pp 345–354

Pereira Z (1999) Palinoestratigrafia do Sector Sudoeste da Zona SulPortuguesa. Comun IGM, Portugal 86:25–57

Pereira Z, Matos JMX, Fernandes P, Oliveira JT (2007) Devonian andcarboniferous palynostratigraphy of the South Portuguese Zone,Portugal. Comun Geol 94:53–79

Pereira Z, Matos, JMX, Fernandes P, Oliveira JT (2008) Palynostratig-raphy and systematic palynology of the Devonian and carbonif-erous successions of the South Portuguese zone, Portugal. Mem34 INETI, pp 1–176

Pichler H (1965) Acid hyaloclastites. Bull Volcanol 28:293–310Quesada C, Fonseca P, Munhá J, Oliveira J, Ribeiro A (1994) The

Bela-Acebuches Ophiolite: geological characterization and geo-dynamic significance. Bol Inst Geol Mineir Madrid 105:3–49

Quesada C (1991) Geological constraints on the Paleozoic tectonicevolution of tectonostratigraphic terranes in the Iberian massif.Tectonophysics 185:225–245

Relvas JMRS (2000) Geology and metallogenesis at the Neves Corvodeposit, Portugal. Ph.D. thesis, Universidade de Lisboa, pp 1–319

Relvas J, Tassinari C, Munhá J, Barriga FJAS (2001) Multiple sourcesfor ore-forming fluids in the Neves Corvo VMS deposit of theIberian Pyrite belt (Portugal): strontium, neodymium and leadisotope evidence. Miner Deposita 36:416–427

Relvas JMRS, Barriga FJAS, Pinto AMM, Ferreira A, Pacheco N,Noiva PC, Barriga G, Baptista R, Carvalho D, Oliveira V, MunháJ, Hutchinson R (2002) The Neves-Corvo deposit, Iberian PyriteBelt, Portugal: impacts and future, 25 years after the discovery.Soc Econ Geol Spec Publ 9:155–176

Relvas JMRS, Barriga FJAS, Ferreira AMM, Noiva PC, Pacheco N,Barriga G (2006a) Hydrothermal alteration and mineralization inthe Neves–Corvo volcanic-hosted massive sulfide deposit, Portu-gal. I. Geology, mineralogy, and geochemistry. Econ Geol101:753–790

Relvas JMRS, Barriga FJAS, Longstaffe F (2006b) Hydrothermalalteration and mineralization in the Neves–Corvo volcanic-hosted massive sulfide deposit, Portugal. II. Oxygen, hydrogen,and carbon isotopes. Econ Geol 101:791–804

Rosa C, McPhie J, Relvas JMRS, Pereira Z, Oliveira T, Pacheco N(2008) Facies analyses and volcanic setting of the giant NevesCorvo massive sulfide deposit, Iberian Pyrite Belt, Portugal. MinerDeposita 43:449–466

Rosa C, McPhie J, Relvas JMRS (2010) Type of volcanoes hosting themassive sulfide deposits of the Iberian Pyrite Belt. J VolcanolGeotherm Res 194:107–126

Miner Deposita

Rosa C, McPhie J, Relvas JMRS (2013) Distinguishing peperite fromother sediment-matrix igneous breccias: lessons from the IberianPyrite Belt. Bull Volcanol (in press)

Rosa D, Inverno C, Oliveira V, Rosa C (2004) Geochemistry ofvolcanic rocks, Albernoa area, Iberian Pyrite Belt, Portugal. IntGeol Rev 46:366–383

Rosa D, Inverno C, Oliveira V, Rosa C (2006) Geochemistry andgeothermometry of volcanic rocks from Serra Branca, IberianPyrite Belt, Portugal. Gondwana Res 10:328–339

Rosa D, Finch A, Andersen T, Inverno C (2009) U–Pb geochronologyand Hf isotope ratios of magmatic zircons from the Iberian PyriteBelt. Mineral Petrol 95:47–69

Schermerhorn LJG (1971) An outline stratigraphy of the Iberian PyriteBelt. Bol Geol Min Madrid 82:239–268

Silva JB, Oliveira JT, Ribeiro A (1990) South Portuguese Zone. Struc-tural outline. In: Dallmeyer RD,Martinez Garcia E (eds) Pre-Mesozoic geology of Iberia. New York: Springer, pp348–362

Simonetti A, Heaman LM, Hartlaub RP, Creaser RA, MacHattie TG,Böhm C (2005) U–Pb zircon dating by laser ablation-MC-ICP-MS using a new multiple ion counting Faraday collector array. JAnalyt At Spectrom 20:677–686

Stacey JS, Kramers JD (1975) Approximation of terrestrial lead isotopeevolution by a two-stage model. Earth Planet Sci Lett 26:207–221

Streel M, Higgs K, Loboziak S, Riegel W, Steemans PH (1987) Sporestratigraphy and correlation with faunas and floras in the type

marine Devonian of the Ardenne-Rhenish regions. Rev PalaeobPalynol 50:211–229

Streel M, Brice D, Mistiaen B, (2006) Strunian. Geologica Belgica 9 1-2:105–109

Taylor JR (1997) An introduction to error analysis, 2nd edn. UniversityScience Books, California

Tornos F (2006) Environment of formation and styles of volcanogenicmassive sulfides: the Iberian Pyrite Belt. Ore Geol Rev 28:259–307

Tornos F, Casquet C, Relvas JMRS, Barriga FJAS, Sáez R (2002) Therelationship between ore deposits and oblique tectonics: the SWIberian Pyrite Belt. In: Blundell DJ, Neubauer F, von Quadt A(eds). The timing and location of major ore deposits in an evolv-ing orogen. Geol Soc London, Special Publ 204:179–198

Wiedenbeck M, Alle P, Corfu F, Griffin WL, Meier M, Oberli F, VonQuadt A, Roddick JC, Spiegel W (1995) Three natural zirconstandards for U-Th-Pb, Lu-Hf, trace element and REE analyses.Geostand Newsl 19:1–23

Winchester JA, Floyd PA (1977) Geochemical discrimination of dif-ferent magma series and their differentiation products using im-mobile elements. Chem Geol 20:325–343

Wood GD, Gabriel AM, Lawson JC (1996) Palynological techniques-processing and microscopy. In: Jansonius J, Mcgregor DC (eds),Palynology: principles and applications. American Association ofStratigraphic Palynologist, Found 1:29–50

Miner Deposita