evolution of the ligurian tethys in the western alps: sm/nd and u/pb geochronology and rare-earth...

Ž .Chemical Geology 175 2001 449–466www.elsevier.comrlocaterchemgeo

Evolution of the Ligurian Tethys in the Western Alps: SmrNdand UrPb geochronology and rare-earth element geochemistry of

ž /the Montgenevre ophiolite France`Sylvie Costaa,), Renaud Cabyb

a Max-Planck Institut fuer Chemistry, Abteilung Geochemie, Postfach 3060, 55020 Mainz, Germanyb Laboratoire de Tectonophysique, USTL, place E. Bataillon, 34095 Montpellier Cedex 05, France

Received 18 May 1999; accepted 9 June 2000

Abstract

We provide geochemical and geochronological data for gabbro, diorite and albitite samples from the Montgenevre`ophiolite in the Western Alps. This well-preserved remnant of the Piemont–Ligurian oceanic basin shows evidence ofintra-oceanic deformation and metamorphism, but has suffered minor ductile deformation and metamorphism during theAlpine orogeny. The gabbros have geochemical features and initial Nd isotopic signatures similar to that of Mid-OceanicRidge gabbros, indicating that they were extracted from a depleted mantle source with no evidence of continental

Ž Ž . .contamination ´ T )q8 . Cumulation played an important role in the genesis of these gabbros, and their rare-earthNdŽ .element REE patterns are controlled by the major cumulus phases. Modelling the REE data for the gabbros and diorites

Ž .provides support for the hypothesis that the dioritic magmas have been derived from small-degreeF5% partial melting ofŽ .the surrounding gabbros during shearing at high-temperature high-T .

Zircons from a leucodioritic vein within sheared gabbros are poorly discordant and cluster the Concordia at 156"3 Ma,while zircons from an albitite lens within the mantle-rocks display concordant ages at 148"2 Ma. In both cases, the zirconsshow no evidence of inheritance, and these values are interpreted as crystallisation ages. The;160–150 Ma age bracketvery likely records a late stage in the magmatic history of the Montgenevre ophiolite, since the emplacement of diorite`clearly post-dates the layered gabbros. These radiometric ages correlate well with the Late Bathonian to Early Kimmeridgian

Ž .stratigraphic age;160–140 Ma of the earliest post-ophiolitic radiolarian sediments. The whole-rock Sm–Nd system ofŽ Ž . .the gabbros yielded an isochron age of 198"22 Ma ´ T of q8.8 . This value is significantly older than the;165 MaNd

age of Western Alps ophiolitic gabbros published elsewhere, but is similar within error to the;180 Ma age of eclogitisedgabbros from the Ligurides and Corsica. This early Jurassic age is interpreted as the emplacement age of the gabbroic meltsinto the mantle lherzolites and could be associated with the early stage of rifting between the European and Austro-Alpinecontinental margins.

These ages and their interpretation are consistent with the model of asymmetric mantle-denudation by an obliqueŽ .detachment fault, proposed by Lemoine et al. 1987 . This model implies that the newly formed lherzolite–gabbro oceanic

) Corresponding author. Currently at Department of Earth Sciences, Monash University, Clayton Victoria 3168, Australia.Ž .E-mail address: [email protected] S. Costa .

0009-2541r01r$ - see front matterq2001 Elsevier Science B.V. All rights reserved.Ž .PII: S0009-2541 00 00334-X

( )S. Costa, R. CabyrChemical Geology 175 2001 449–466450

domain probably remained close to the spreading centre and therefore experienced slow cooling and low spreading rates.This appears to be correlated by our radiometric results that suggest a life span of at least 30 Ma for the formation of theoceanic crust in this part of the Piemont–Ligurian ocean.q2001 Elsevier Science B.V. All rights reserved.

Keywords: Western Alps; Ophiolite; UrPb; SmrNd; Geochemistry

1. Introduction

Ophiolite massifs of the Western Alps, the Lig-urides, the Northern Apennines, and Corsica repre-sent remnants of the Piemont–Ligurian oceanic basin,a branch of the Mesozoic Tethys which formedbetween the divergent European and Austro-Alpine

Žpassive margins e.g. Weissert and Bernouilli, 1985;.Lemoine et al., 1987 . Located in southeastern

France, the Montgenevre ophiolite is one of the best`preserved ophiolite occurrences in the external

Ž .Piemont zone of the Western Alps Fig. 1 . Unlikemost ophiolitic remnants of the Alps, this massif hassuffered minor ductile deformation and metamor-phism related to the Alpine orogenesis. Therefore,the ophiolite was the subject of numerous studiesoutlining the primary structural, petrological andgeochemical features of the oceanic floor, in order toprovide comprehensive genetic models for the

ŽPiemont–Ligurian ocean Mevel et al., 1978; Lewisánd Snewing, 1980; Bertrand et al., 1987; Caby,

.1995 .These investigations have shown that the Mont-

genevre ophiolite consists primarily of serpentinised`lherzolites, gabbros and pillowed lavas, in addition

Ž .to veins and dykes of diorite and albitite Fig. 1 .The gabbro sequence presents all the petrologicaland geochemical characteristics of modern oceanicridges, and shows evidence of pervasive intra-oc-eanic deformation and metamorphism prior to theintrusion of basaltic dykes and the development of

Žthe Alpine tectono-metamorphic events Mevel et al.,´.1978; Caby, 1995 . The dioritic veins essentially

outcrop within sheared gabbros in the upper part ofthe sequence. In contrast, the albititic veins, whichusually crosscut lherzolites and mylonitic gabbros,can locally be intimately associated with deformeddioritic veins. Although this ophiolite massif appearsto be an ideal candidate for dating the magmatic,metamorphic and deformational events related to

oceanisation in the Alps, very few chronologicalstudies have been undertaken so far. Available datasuggests a long time span between the emplacement

Ž .of the gabbros Carpena and Caby, 1984 and the´deposition of an oceanic sedimentary cover in the

Ž .Late Jurassic De Wever and Caby, 1981 . This mayimply that this section of oceanic crust was emplaced

Ž .in a low-spreading ridge Caby, 1995 . However, nodirect geochronological information is available onthe magmatic components of the ophiolitic suite sofar.

The present work aims to provide such age con-straints, by applying the SmrNd and UrPb datingtechniques to different types of magmatic rocksŽ .gabbro, diorite, albitite of the oceanic crust pre-served at Montgenevre. In addition, rare-earth ele-`

Ž .ment REE data are used to establish the geneticrelationships between sheared gabbros and dioriticveins.

2. Geology of the Montgenevre ophiolite`

The ophiolite forms a thin klippe, which rests onan oceanic cover sequence with blueschist facies

Ž .mineral assemblages Replatte-Lago Negro unit ; it-self overlying Triassic dolomites of the European

Ž . Ž .continental margin Briançonnais Fig. 1B . In con-trast to other ophiolites of the Piemont–Ligurianzone that were affected by early Alpine metamor-phism of blueschist to eclogite facies conditions, theMontgenevre ophiolite has only suffered a weakÀlpine overprint, characterised by albite, prehnite,

Žpumpellyite, actinolite and epidote Mevel et al.,´.1978; Lewis and Snewing, 1980 . The effects of

Alpine ductile deformation are also limited. TheMontgenevre ophiolite represents a remnant of a`higher nappe, which is now exposed as a down-

Ž .faulted block Caby, 1995, and refs. therein . It haspreserved most of the primary signatures of oceanic

( )S. Costa, R. CabyrChemical Geology 175 2001 449–466 451

Ž . Ž .Fig. 1. A Tectonic setting of the Montgenevre ophiolite within the Western Alps. B Simplified geological map of the Montgenevre` `Ž . Ž .ophiolite. Both maps modified after Bertrand et al., 1987. C Simplified stratigraphic section of the Montgenevre ophiolite. 1 Crystalline`

Ž . Ž . Ž . Ž . Ž .massifs; 2 Ultra-Dauphinois; 3 Sub-Brianc¸onnais; 4 Brianc¸onnais; 5 Continental margin series; 6ASchistes lustresB with ophiolites;´Ž . Ž . Ž .7 Helminthoide-flysch; Ophiolite sequences: 8 Undifferentiated pillowed basalts in white when hidden below Recent deposits , dolerite

Ž . Ž . Ž . Ždykes are not shown; 9 Ultramafic arenites and conglomerates; 10 Gabbros and dioritic rocks; 11 Serpentinised peridotites in grey. Ž .when hidden below Recent deposits ; Oceanic cover sequences: 12 Undifferentiated Upper Jurassic to CretaceousAschistes lustresB with´

Ž .blueschist facies metamorphic assemblages; 13 Replatte-Lago Nero unit with lower grade metamorphic assemblages; Passive TethyanŽ . Ž . Ž . Ž .paleo-margin: 14 Upper Triassic dolomites and Lower Jurassic sediments; 15 Main Jurassic Fault; 16 Alpine thrust; 17 Alpine

Ž .backthrust; 18 Late Alpine fault.

crust, as well as evidence for intra-oceanic, high-Ž .temperature high-T metamorphism and ductile de-

Ž .formation Mevel et al., 1978; Caby, 1995 .´The base of the ophiolite is comprised mainly of

spinel harzburgites and lherzolites, less plagioclase

lherzolites, and minor cumulate rocks such as pyrox-Ž .enites, dunites and wehrlites Bertrand et al., 1987

Ž .Fig. 1C . The mantle peridotites show a metamor-phic foliation, occasionally oblique to the composi-tional layering and parallel to the inferred paleo-


Moho. The latter was reworked by high-T shearŽ .zones Caby, 1995 .

Above these highly serpentinised rocks and lo-cally preserved paleo-Moho outcrops, a thin se-quence of layered troctolites and olivine-bearinggabbros, essentially free of post-magmatic, high-Tdeformation. These layered rocks are cut by mafic

Ž .ferrogabbro to pyroxenite sills Fig. 1C , which arehighly altered by a low-temperature ocean-floormetamorphism, characterised by the presence ofchlorite, serpentine, sphene and epidote. A poorlylayered sequence of coarse-grained and fine-grainedgabbros, with subordinate pegmatitic gabbros andpockets of microgabbro overlie these basal rocks.The gabbros consist of plagioclase and clinopyrox-ene, with minor olivine and opaques, and theclinopyroxenes show variable degrees of replace-

Žment by late-magmatic amphiboles Mevel et al.,´.1978 . They are affected by a system of anastomos-

ing, high-T shear zones consisting of mylonites andŽ .ultramylonites Caby, 1995 .

Dioritic rocks form veins, dykes and diffuse im-pregnation within sheared gabbros, mostly in the

Ž .upper part of the gabbro sequence Fig. 1C . Withprogressive deformation, these differentiated rocksgrade into high-T protomylonites and mylonitesŽ .Caby, 1995 . Dykes of pegmatitic diorite mainly cutthe less deformed gabbros. The primary magmaticmineralogy of the diorites includes plagioclase,clinopyroxene, pargasitic amphibole, and abundantaccessory minerals such as ilmenite, sphene, apatiteand zircon. Based on textural relationships, it hasbeen proposed that the dioritic magma originatedfrom partial melting of the already consolidated gab-bros during their shearing at temperatures ofG8508CŽ .Caby, 1995 . The dioritic melts percolated repeat-edly through active mylonite zones towards the lessdeformed upper section of the gabbroic sequenceŽ .Caby, 1995 .

A 400-m thick sequence of pillowed lavas, show-ing mineralogical and geochemical characters similar

Žto that of Mid-Oceanic Ridge basalts Lewis and.Snewing, 1980; Bertrand et al., 1987 , rests with a

tectonic contact on top of the gabbros. Basalt dykescut across deformed gabbros and mantle peridotitesŽ .Caby, 1995 .

Dykes and sills of albitite were emplaced withinŽ .peridotites and gabbros Caby, 1995 , and according

Ž .to Chapelle 1990 within basalts as well. However,this last important point has never been confirmed byour field observations. The primary mineralogy ofthese fine-grained rocks typically consists of ran-

Ž .domly oriented grains of albiteG90% , magmaticamphibole pseudomorphosed into tremolite–actino-lite, and abundant accessory minerals such as allan-ite, apatite and zircon. The origin and significance ofthese rocks is controversial. Considering their geo-chemical characteristics and the typological classifi-

Ž .cation of their zircons, Chapelle 1990 proposed thatthey represent fractionated mantle-derived magmasof tholeiitic character. On the basis of textural rela-

Ž .tionships, Caby 1995 suggested that they couldrepresent more evolved liquids differentiated from adioritic magma. Albitites and albite–granites, oftenassociated with ferrodiorites, are also very commoncomponents of the ophiolitic remnants of the Apen-

Žnines, the Ligurian Alps and Corsica Ohnenstetterand Ohnenstetter, 1980; Ohnenstetter et al., 1981;

.Borsi et al., 1996 . Based on petrological and geo-chemical evidence, these authors proposed that thealbitites represent the final differentiation products ofa MORB-type mantle source.

3. Age constraints on Piemont–Ligurian ophio-lites

Radiometric data providing direct age constraintson the formation of the Piemont–Ligurian oceanicbasin are scarce, and most of them were acquiredrecently. In the Western Alps, lenses of pegmatiticgabbros within the basal wildflysch of the Gets

Ž .nappe Pre-Alps were dated using the UrPb andŽ . Ž .ArrAr techniques Bill et al., 1997 . Zircons UrPb

Ž .and amphiboles ArrAr provided identical values of166"2 Ma, interpreted as the crystallisation age ofthe gabbros. Further east in the Zermatt–Saas ophio-lite, eclogitic metagabbros yielded similar UrPb zir-

Žcon ages at 163"2 Ma Rubatto and Gebauer,.1996 . In the Ligurian Alps, eclogitised gabbros of

the Voltri group provided a whole-rock SmrNdŽ .isochron age of 177"23 Ma Borsi, 1995 , whereas

UrPb zircon ages clustering at 155–150 Ma wereobtained for plagiogranites and the host ferrodioriteŽ .Borsi et al., 1996 . In the Northern Apennines,


plagioclase–clinopyroxene pairs of spinel–lherzo-lites from the External Ligurides units yieldedSmrNd isochron ages of;165 Ma, which wereinterpreted as the age of metamorphic re-equilibra-tion from spinel- to plagioclase-facies conditionsŽ .Rampone et al., 1995 . In the Internal Liguridesunits, an olivine gabbro gave a similar Sm–Nd inter-

Ž .nal isochron of 164"14 Ma Rampone et al., 1998 .In the same area, a plagiogranite provided an UrPb

Ž .zircon age of 153"1 Ma Borsi et al., 1996 , whileArrAr datings undertaken on amphiboles from sev-eral diorite and plagiogranite samples returned simi-

Ž .lar ages at;158 Ma Bortolotti et al., 1995 . In theCorsican ophiolites, metagabbros provided amphi-

Žbole KrAr ages of 181"6 Ma Beccaluva et al.,.1981 , while two albitite samples yielded zircon

ŽUrPb ages of 161"3 Ma Ohnenstetter et al.,.1981 . As a whole, available geochronological data

of magmatic rocks from different ophiolites rangefrom 180 to 150 Ma, with gabbroic rocks recording

Ž .relatively older ages 180–160 Ma than dioriticŽ .intrusives, albitites and plagiogranites 160–150 Ma

Ž .Fig. 2 .

Ž .Fig. 2. Summary of isotopic ages with 2s errors available formagmatic rocks in Alpine ophiolites. Squares refer to gabbroicrocks, circles to diorites, diamonds to albitites and plagiogranites,and stars to mantle rocks. Black symbols refer to UrPb ages,white symbols to ArrAr ages and crossed symbols to SmrNdages. The stratigraphic age of the first post-ophiolitic radiolarian

Ž .sediments is given for comparison. Ages are from: 1 RubattoŽ . Ž . Ž . Ž . Ž .and Gebauer 1996 , 2 Bill et al. 1997 , 3 this work, 4 Borsi

Ž . Ž . Ž . Ž . Ž . Ž .1995 , 5 Borsi et al. 1996 , 6 Rampone et al. 1995 , 7Ž . Ž . Ž . Ž .Rampone et al. 1998 , 8 Bortolotti et al. 1995 , 9 Beccaluva

Ž . Ž . Ž .et al. 1981 , and 10 Ohnenstetter et al. 1981 .

Indirect age constraints, based on the chronos-tratigraphy of the first post-ophiolitic sediments, pro-vide a younger limit for the formational age of theoceanic basin. Radiolarites and associated deep-seasediments were deposited over either pillow lavasand gabbros, or mantle peridotites, of manyPiemont–Ligurian ophiolites with the exception of

ŽMontgenevre e.g. De Wever and Caby, 1981; Schaaf`et al., 1985; Marcucci and Conti, 1995; De Wever

.and Danelian, 1995 . De Wever and BaumgartnerŽ .1995 , who have recently re-assessed the age ofsupra-ophiolitic radiolarites in the Western Alps,suggested that they may be diachronous, with an agedifference of up to 10 Ma in two adjacent areas.However, as no detailed chronostratigraphy is possi-ble in the overlying metaradiolarites due to isoclinalfolding, the exact position of the samples collectedby these authors is unknown. Moreover, as the pre-sent-day thickness of the radiolarites varies from afew centimeter to about 30-m, it is impossible to saywhether the difference in ages reported by De Wever

Ž .and Baumgartner 1995 simply reflects the durationof radiolarian sedimentation, or if a sedimentationdiachronism does exist. In this case, diachronism inthe formation of the underlying basalts and gabbroscould also be expected. In summary, the first radio-larian sediments of the oceanic cover which havebeen dated in the Western Alps, the Apennines andCorsica are never older than the Late Bathonian andmay be as young as Late Oxfordian. This 160–140

ŽMa age bracket according to the time scale by Odin,.1994 is in agreement with available radiometric data

Ž .for the magmatic rocks Fig. 2 .The first radiometric age constraints were based

on a number of KrAr and Fission Track dates.ŽHowever, their significance is now questioned e.g.

.Bill et al., 1997 , since they have not been corrob-orated by recently published, high-precision age data.KrAr ages were measured on whole rocks andamphiboles from ophiolites of the Western Alps and

Žthe Northern Apennines e.g. Beccaluva et al., 1981;.Frontignie et al., 1982 . These dates are widely

spread in the range 180–40 Ma, which very likelyreflects various degrees of resetting of the KrArisotopic system by subsequent tectono-metamorphicevents. Fission-track ages were measured on apatitesand zircons from different ophiolitic massifs of the

Ž .Western Alps Carpena and Caby, 1984 , the North-´


Ž .ern Apennines Bigazzi et al., 1972, 1973 and Cor-Ž .sica Carpena et al., 1979 . Ages measured on ap-´

atites always yielded values younger than 40 Ma,which can be attributed to later resetting due to

Ž .Alpine orogenic activity Carpena et al., 1979 . Zir-´cons provided older ages in the range of 185–160

ŽMa in the Northern Apennines Bigazzi et al., 1972,.1973 , and 210–190 Ma in the Western Alps, with

the oldest value being measured at Montgenevre`Ž .Carpena and Caby, 1984 . The latter authors con-´cluded that these ages could reflect cooling of theoceanic crust below;2208C after intra-oceanicmetamorphism, and that subsequent blueschist meta-morphism may have been short-lived, since it did notsignificantly affect the fission tracks in zircons.However, the major problem of all these fission-trackages is that they are much older than the radiometricdata based on UrPb, SmrNd and ArrAr isotopicsystems, which have much higher closure tempera-tures than the fission-track system. Moreover, theages were calculated with an older value of 7.03=

y17 y1 Ž .10 year for thelF constant Naeser, 1978 .This constant has since been revised, and the value

y17 y1 Žnow commonly used is 8.46=10 year Van.den Haute et al., 1988; De Corte et al., 1991 . When

recalculated with this new parameter, the fissiontrack ages are significantly younger, and fall in therange of 150–130 Ma for the zircons of the NorthernApennines and 175–160 Ma for those of the WesternAlps.

4. Analytical procedures

Preparation of whole-rock powders and mineralseparates, as well as measurement of major and traceelement concentrations were done at the University

Ž .of Montpellier France . UrPb and SmrNd isotopicanalyses, and determination of REE concentrationswere performed at the Max-Planck Institute fur¨

Ž .Chemie in Mainz Germany .Major elements were determined by standard wet

methods adapted at the University of MontpellierŽ .Dostal et al., 1986 . Rb, Sr, Ba, V, Co, Cr, Ni, Cuand Zn were analysed by atomic absorption. Preci-sion and accuracy of the trace element data wasevaluated from replicate analyses of standards. Ingeneral, precision was better than 5–10%.

REE abundance was determined by the High-Per-Ž .formance Liquid Chromatographic HPLC tech-

nique, following the procedure described by CassidyŽ .1988 . Approximately 50 mg of the ground rocksamples were dissolved in Teflonw bombs using amixture of concentrated hydrofluoric and nitric acidsat 2208C. After evaporation to dryness, the sampleswere soluble in a mixture of nitric and oxalic acids.The REEs were first separated as a group on cation-exchange resins with a combination of nitric andoxalic acids. They were then separated from oneanother by dynamic ion-exchange HPLC on a re-versed phase-column, using a gradient elution from0.05 to 0.5 M HIBA. The uncertainty for the REEconcentrations was estimated to a precision ofF5%,based on the reproducibility of standard solutions.

For the SmrNd analyses, the whole-rock powderswere spiked with a mixed149Sm–150Nd tracer, thendissolved using a mixture of concentrated hydrofluo-ric and nitric acids in Teflonw bombs at 2208C. Smand Nd were separated using a chemical proceduresimilar to that described by White and PatchettŽ .1984 . Total procedural blanks were less than 60 pgfor both Sm and Nd. Isotopic measurements wereperformed on a MAT 261 mass spectrometer,equipped with multiple collectors operating in staticmode. 143Ndr144Nd ratios were normalised to146Ndr144Nds0.7219. In the course of our analy-ses, repeated measurements of the La Jolla Ndstandard provided a mean143Ndr144Nd ratio of

Ž .0.511836"0.000015 2s ; ns20 . For isochronm

calculations, a minimum uncertainty of 0.003% wasassumed for the143Ndr144Nd ratios, based on thereproducibility of the standard. If the error of aparticular analysis was higher than this value, thehigher value was used. The uncertainty for the147Smr144Nd ratios was estimated to a precision of0.3%.

UrPb isotopic analyses of zircon fractions wereperformed following the conventional technique. Zir-cons isolated from bulk rock samples were sortedaccording to morphology, size and colour under abinocular microscope. Before dissolution, the zirconfractions were washed with cold 2 N hydrochloricacid, then with warm 7 N nitric acid. After spikingwith a mixed 205Pb–233U tracer, the zircons weredissolved using concentrated hydrofluoric acid inTeflonw bombs lodged inside Krogh-style digestion


vessels at 2208C. U and Pb were separated andpurified on small separation columns using the

Ž .chemical procedures recommended by Krogh 1973 .Isotope compositions were measured on a MAT 261mass spectrometer equipped with multiple Faradaycollectors and a SEM. Pb was run in static multicol-lector mode and U by peak jumping in the SEM.Analytical data were corrected from processingblanks, common Pb contribution and mass discrimi-nation. Total Pb blanks were less than 150 pg andhad the following isotopic composition208Pb: 207Pb:206Pb: 204Pbs37.346: 15.464: 18.185: 1. Correctionfor common Pb was made using the Stacey and

Ž .Kramers 1975 values corresponding to an age ofŽ .160"10 Ma sample MG3511 or 150"10 Ma

Ž .sample MG542 . Fractionation factors of 0.900‰and 0.991‰ per atomic mass unit were used for Pband U, respectively, based on repeated analyses ofNBS 982 and natural U standards.

5. Geochemistry

Major and trace element analyses of Montgenevre`ophiolitic rocks were discussed in detail by Bertrand

Ž .et al. 1982, 1987 . We present additional major andtrace element data for the seven gabbroic samples,two dioritic veins and the albitite lens, that we havedated using the SmrNd and UrPb techniques. A

brief petrographic description of the samples is pre-sented in Table 1. Major and trace element data isgiven in Table 2. The REE results listed in Table 3are presented graphically in Fig. 3.

5.1. Major and trace elements

One olivine-bearing cumulate was sampled fromthe base of the layered gabbroic sequence, and fiveclinopyroxene-bearing gabbros were taken from the

Žoverlying, poorly layered sequence Fig. 1C and.Table 1 . The chemical composition of gabbroic

cumulate MG36, with 38 wt.% silica, 31 wt.% mag-nesia, a total iron content of 10 wt.%, and its enrich-

Ž . Ž .ment in Cr 7280 ppm and Ni 2040 ppm reflectŽ .essentially olivine cumulation Table 2 . The

clinopyroxene-bearing gabbros show ultrabasic tobasic compositions similar to that described by

Ž .Bertrand et al. 1987 , with SiO contents between2

41 and 53 wt.%, MgO ranging from 2 to 8 wt.%, andTiO from 0.2 to 6.6 wt.%. Sr, Cr and Ni concentra-2

tions vary sympathetically with the silica content ofthe samples.

Ž .The two ferrogabbros MG535 and MG537 con-tain more clinopyroxene than plagioclase, and ac-cordingly have high Fe O , TiO and V contents.2 3 2

The other studied gabbros contain either more pla-Ž .gioclase than clinopyroxene MG25 , or roughly

Žequal amounts of these two minerals MG534 and.MG536 . The mineralogical composition of these

Table 1Ž .Petrographic characteristics of ophiolitic samples studied in the Montgenevre ophiolite. Mineral abbreviations after Kretz 1983`

Sample Rock type Primary magmatic mineralogy Texture, deformation, etc.

MG36 Cumulate PlqOlqAm"Cpx"Spl"Mag Coarse-grained, undeformed rock from base of layered sequence,with serpentinised Ol, chloritised Pl and Amp–Cpx rims replacedby Tr–Chl–Agt

MG537 Ferrogabbro PlqCpxqAmq Amphibole- and apatite-rich, protomyloniteŽ .Apq IlmqSpn

Ž .MG535 Ferrogabbro PlqCpxqAmq IlmqSpn Ilmenite- and sphene-rich, high-T myloniteMG566 Anorthosite Pl"Cpx"Ol Coarse-grained rock containing 95% epidotised anothiteMG25 Leucogabbro Pl"Cpx"Am"Ol Pegmatitic gabbro, containing 85% epidotised labradoriteMG534 Gabbro PlqCpx"Ol Fine-grained gabbro, showing a random fabric

Ž .MG536 Gabbro PlqCpx"Amq Spn Protomylonite with clinopyroxene clasts replaced by amphiboleŽ .MG3511 Diorite PlqCpx"Amq SpnqApqZrn Fine-grained vein within sheared gabbrosŽ .MG15 Leucodiorite Pl"Cpx"Amq SpnqApqZrn Coarse-grained vein with 80% oligoclase, within sheared gabbros

Ž .MG542 Albitite Pl"Amq AlnqApqZrn Fine-grained lens with 90% albite, which cuts across lherzolites


Table 2Ž . Ž . ŽMajor wt.% and trace element ppm concentrations for gabbro, diorite and albitite samples of the Montgenevre ophiolite unpublished`

.data from C. Dupuy, 1985

MG36, MG537, MG535, MG566, MG25, MG534, MG536, MG15, MG542,Cumulate Ferrogabbro Ferrogabbro Anorthosite Leucogabbro Gabbro Gabbro Leucodiorite Albitite

( )Major elements wt.%SiO 38.03 41.40 44.72 49.17 50.85 50.98 52.57 60.50 62.602

TiO 0.22 4.05 6.60 0.11 0.17 1.19 0.32 1.85 0.312

Al O 5.10 12.72 12.66 25.46 21.74 15.16 17.60 17.00 16.252 3Ž .Fe O t 10.00 17.50 14.00 1.82 1.94 6.54 4.18 2.58 1.452 3

MnO 0.16 0.29 0.24 0.02 0.05 0.14 0.09 0.07 0.03MgO 31.05 5.84 7.00 2.36 4.02 8.33 6.58 1.70 3.56CaO 3.32 9.12 9.78 10.50 12.27 10.54 10.02 6.05 3.27Na O 0.03 3.65 3.40 4.72 4.06 4.10 4.52 7.82 9.102

K O 0.01 0.05 0.07 0.12 0.39 0.11 0.22 0.01 0.012

P O 0.04 2.40 0.05 0.05 0.05 0.05 0.06 0.37 0.102 5

LOI 10.19 2.10 1.63 5.42 4.30 2.71 3.17 0.55 3.10Total 98.15 99.12 100.15 99.75 99.84 99.85 99.33 98.30 99.78

( )Trace elements ppmRb 1 1 1 1 2 1 1 2 1Sr 23 192 197 261 317 208 348 260 1298Ba 18 -50 -50 110 55 -50 -50 -50 137V 78 490 733 25 65 184 101 26 22Co 122 45 47 11 11 35 33 8 8Cr 7280 6 49 49 709 67 484 9 12Ni 2040 26 45 76 92 66 103 18 59Cu 12 56 76 4 5 29 12 3 1Zn 107 21 75 16 16 35 34 12 8

Table 3Ž . Ž .REE concentrations ppm and elementrelement ratios for gabbros and diorites of the Montgenevre ophiolite nd: not determined`

MG537, MG535, MG566, MG25, MG534, MG536, MG3511, MG15,Ferrogabbro Ferrogabbro Anorthosite Leucogabbro Gabbro Gabbro Diorite Leucodiorite

( )Rare-earth elements ppmLa 17.37 1.04 0.60 0.82 0.99 5.19 8.99 23.39Ce 61.50 3.75 1.50 2.27 3.16 14.89 27.85 68.00Pr 11.03 0.80 0.24 0.35 0.60 2.14 4.11 10.02Nd 62.72 5.12 1.02 1.68 3.37 9.53 18.95 45.26Sm 18.95 2.11 0.21 0.45 1.09 2.46 4.97 12.45Eu 5.01 1.06 0.37 0.32 0.60 0.66 1.21 2.61Gd 21.17 3.35 0.23 0.64 1.78 3.11 6.22 15.93Tb 3.72 0.61 0.02 0.11 0.35 0.60 1.17 3.25Ho 4.45 0.91 0.02 0.17 0.54 0.97 1.83 5.18Er 11.73 2.59 nd 0.48 1.55 3.05 5.75 16.93Tm 1.41 0.35 nd 0.07 0.21 0.48 0.82 2.73Yb 8.14 2.32 0.09 0.49 1.31 3.24 5.60 18.45Lu 1.06 0.35 nd nd 0.18 0.47 0.73 2.49Ž .LarSm N 0.58 0.31 1.80 1.15 0.57 1.33 1.14 1.18Ž .LarYb N 1.44 0.30 4.50 1.13 0.51 1.08 1.08 0.86

)EurEu 0.76 1.22 5.15 1.82 1.32 0.73 0.67 0.57


ŽFig. 3. Chondrite-normalised REE patterns for gabbros squares. Ž .and diamonds and diorites circles sampled in the Montgenevre`

Ž .ophiolite. Normalisation is based on Evensen et al. 1978 .

samples controls their bulk chemistry, with MG25being alumina-rich, and MG534 and MG536 having

Ž .high magnesium and iron-contents Tables 1 and 2 .Felsic samples are represented by diorite MG3511,

leucodiorite MG15 and albitite MG542. The twolatter only were analysed for major and trace ele-ments. The leucodiorite is an oligoclase-rich,coarse-grained vein which was emplaced withinsheared gabbros of the clinopyroxene-rich sequence,whereas the albitite is a fine-grained lens which cutsacross the lherzolites, and shows no magmatic or

Ž .post-magmatic ductile deformation Table 1 . BothŽsamples have similar concentrations of SiO 60–632

. Ž .wt.% and Al O 16–17 wt.% , and their Na O and2 3 2

CaO contents reflect cumulation of oligoclase andŽ .albite, respectively Table 2 . In addition, the albititeŽ . Ž .is enriched in Sr 1298 ppm and Ba 137 ppm

relative to the leucodiorite.

5.2. Rare-earth elements

With the exception of apatite-rich sample MG537,all gabbros have REE contents less than 20 times

Ž .that of Chondrites Fig. 3 . Varying amounts ofclinopyroxene and plagioclase in the samples influ-ence the Chondrite-normalised patterns. For exam-ple, clinopyroxene-rich ferrogabbro MG535 has a

wŽ . xslightly LREE-depleted pattern LarSm s0.31NŽ ) .with a small positive Eu anomaly EurEu s1.22 ,

indicative of cumulitic clinopyroxene. In contrast,the pattern of anorthosite MG566, which contains95% anorthite, is characterised by an enrichment in

wŽ . xLREE LarYb s4.5 and a marked positive EuNŽ ) .anomaly EurEu s 5.2 . Gabbro MG534 and

leucogabbro MG25 show intermediate patterns be-tween these two types. Ferrogabbro MG535 is anapatite- and amphibole-rich protomylonite, which isconsistently REE-enriched, compared to the othergabbros. In addition, its normalised pattern is charac-terised by a relative enrichment in the middle REEs

Ž ) .and a negative Eu anomaly EurEu s0.76 .Protomylonitic gabbro MG536 has distinct REE-

features, which are in fact similar to that of the twodiorite veins MG3511 and MG15. The three samples

Žhave higher REE concentrations 20–100=.Chondrite , and their normalised patterns resemble

each other with small negative Eu anomaliesŽ ) .EurEu ;0.6 and otherwise essentially flat pat-

wŽ . xterns LarYb ;1 . Such characteristics can re-N

flect derivation by crystallisation from an evolvedliquid that has already crystallised clinopyroxene,plagioclase and olivine. This may apply to the caseof protomylonitic gabbro MG536. Alternatively,these features can indicate derivation by partial melt-ing with clinopyroxene, plagioclase and olivine re-tained in the source. This hypothesis may apply tothe case of the two diorite samples and is consistentwith textural relationships, which favour derivationof the diorite melts from partial melting of thealready consolidated clinopyroxene-rich gabbros dur-

Ž .ing their progressive shearing Caby, 1995 .In order to test the reliability of the latter hypothe-

sis, we have modelled the REE data of these rocks.We have chosen a model of modal batch partialmelting, in which the partial melt is continuallyreacting in situ with the solid residue, until it can

Žescape as a single batch of magma e.g.; Rollinson,


.1993, and refs. therein . This is particularly appropri-ate in a situation where small pockets of melt areextracted from their source during deformation, andare subsequently injected in veins, dikes or shear-zones as a single batch of melt. The concentration ofa trace element in the meltC is related to itsL

concentration in the unmelted sourceC by the0

expression:

C rC sD qF 1yPŽ .0 L 0

where D is the bulk partition coefficient of the0

minerals in the source rock,P is the bulk partitioncoefficient of the minerals which make up the melt,and F is the weight fraction of melt produced. In thecase of modal melting, the minerals in the sourcerock contribute to the melt in proportion to theirinitial concentration, and therefore this equation sim-plifies to:

C rC sD qF 1yD .Ž .0 L 0 0

In our model, the source rock was assumed to bean unaltered gabbro of the clinopyroxene-rich se-quence, with the REE composition and mineral as-

Žsemblage 50% plagioclase, 45% clinopyroxene, 5%.olivine of sample MG534. This sample represents

the rock type commonly affected by deformation andintimately associated with diorite melts in theclinopyroxene–gabbro sequence. The mineralrmelt

Ž .partition coefficients given by Fujimaki et al. 1984were used in the calculation. The modelling indicates

Fig. 4. Chondrite-normalised REE-patterns corresponding to aŽ .model of modal batch partial melting applied to gabbro square

Ž .and diorite circle samples.

Ž .that the small-degreeF5% partial melting of thisgabbro sample gives rise to a melt with an REE-composition very similar to that of diorite vein

Ž .MG3511 Fig. 4 . Such small percentages of meltingare compatible with figures observed in natural ex-amples of mafic and felsic melt generation. Indeed,measurement of U-series activity ratios in plume-de-rived and mid-oceanic ridge basalts show that meltcan separate from its source region when the melt

Žfraction exceeds a few tenths of a percent Mc-.Kenzie, 2000 . Studies of dioritic magmatic systems

Žalso show that small percentages of meltF1% to.10% , migration and segregation are possible pro-

cesses under conditions of tectonic stress, in particu-Žlar in shear zones Vigneresse et al., 1996; John and

.Stunitz, 1997 .¨

6. Geochronology

6.1. SmrNd isotope systematics

The olivine cumulate, five clinopyroxene gabbrosŽ .and two dioritic veins Table 1 were analysed for

their Sm and Nd isotopic compositions. The analyti-cal results are given in Table 4 and shown on aconventional isochron diagram in Fig. 5. Duplicateanalyses were carried out for some samples, usingdifferent aliquots of the same rock powders, and

Ž .identical results were obtained Table 4 .In an isochron diagram, the whole-rock SmrNd

system of the gabbro samples forms a linear array,and the best-fit line to the data points corresponds toan age of 198"22 Ma, with an MSWD value of

Ž . Ž .0.61 and an initial T of q8.8 Fig. 5 . The dataNd

points for the two diorite veins fall slightly off thelinear array and were not included in the age calcula-tion. This is justified by the fact that these samplesare clearly younger, since they crosscut the deformedgabbros, as discussed above. The 156 Ma referenceisochron, which corresponds to the zircon UrPb age

Ž .of one of the diorite veins see below , is shown forcomparison.

Ž .Individual ´ T values calculated for the gab-Nd

bros that define the linear array vary fromq8.5 toq9.4, and the initial value calculated for the isochron


Table 4SmrNd isotopic results for gabbro and diorite whole-rock samples of the Montgenevre ophiolite´

a bŽ . Ž .Sample Concentrations Isotope ratios ´ T T MaNd dm

147 144 143 144Ž . Ž .Sm ppm Nd ppm Smr Nd Ndr Nd "2s

GabbrosMG36, Cumulate 0.25 0.79 0.1914 0.513069 15 q8.5 225MG537, Ferrogabbro 19.50 64.31 0.1834 0.513087 10 q9.1 107repeat 19.35 63.88 0.1832 0.513080 20 q9.0 134MG535, Ferrogabbro 2.15 4.96 0.2624 0.513173 50 q8.8 223repeat 2.14 4.93 0.2621 0.513179 10 q8.9 248MG25, Leucogabbro 0.60 1.86 0.1934 0.513073 17 q8.6 219MG534, Gabbro 1.30 3.27 0.2402 0.513152 10 q8.9 319MG536, Gabbro 2.46 8.76 0.1697 0.513084 72 q8.2 88

DioritesMG3511, Diorite 5.54 19.58 0.1710 0.513049 10 q8.5 195MG15, Leucodiorite 13.70 48.29 0.1715 0.513016 27 q8.0 296

a Ž143 144 . Ž147 144 . Ž . Ž .Referenced to Ndr Nd churs0.512638 and Smr Nd churs0.1966 Goldstein et al., 1984 ; T values wereNd

calculated atTs198 Ma for gabbros, andTs156 Ma for diorites.b Ž143 144 . Ž147 144 . Ž .Referenced to Ndr Nd dms0.513144 and Smr Nd dms0.2220 Michard et al., 1985 .

falls within this range. These values are indicative ofa depleted mantle-source and show no evidence ofcontamination by continental material. Depleted-mantle model ages calculated for the gabbro samplesare all younger than 275 Ma. For diorite samplesMG3511 and MG15, the initial values wereNd

calculated at 156 Ma, which is the UrPb crystallisa-tion age of MG3511, as discussed in the next sec-tion. The corresponding values ofq8.5 andNd

q8.0 do not show any evidence for continental

Fig. 5. 143Ndr144Nd vs. 147Smr144Nd isochron diagram forŽ . Ž .gabbro squares and diorite circles whole rock samples of the

Montgenevre ophiolite. 2s errors on ratios are not shown when`they are smaller than size of symbols.

contamination, and are in agreement with a model ofderivation of these samples from partial melting ofthe gabbros.

6.2. Zircon UrPb ages

Zircons were isolated from diorite vein MG3511and albitite lens MG542. The UrPb isotopic resultsare presented in Table 5 and shown on a Concordiadiagram in Fig. 6.

Ž .Essentially large 50–200mm , clear and trans-parent zircons were recovered from the dioritic veinMG3511. The crystals were sorted in three different

Ž . Ž .groups: a short and prismatic grains, b large andŽ .elongated grains, and c large, elongated andŽ .rounded grains, with type a being the most abun-

dant. Most of the grains in the different groupsŽ .contain inclusions. Three fractions of type a and

one fraction of each of the two other types wereanalysed. All zircon fractions are characterised byvery low Pb and U contents of 0.2–0.5 and 9–19ppm, respectively, and they all present varying de-

Žgrees of discordance ranging from 4% to 25% Table.5 . However, considering the relatively young ages

of the zircons analysed, their UrPb ages are prefer-able to the PbrPb ages. Four of the five analysed

Ž .fractions 1–5 have apparent UrPb ages ranging

()

S.Costa,R

.Cabyr

Chem

icalGeology

1752001

449–

466460

Table 5UrPb isotopic results for zircon fractions from diorite MG3511 and albitite MG542. Pb) refers to radiogenic lead. Errors on isotopic ratios and apparent ages are 2s

206 204)Ž . Ž .Sample Weight mg Concentrations Pbr Pb %Pb Isotope ratios Apparent ages Ma

206 238 207 235 207 206 206 238 207 235 207 206Ž .measuredŽ . Ž .U ppm Pb ppm Pbr U Pbr U Pbr Pb Pbr U Pbr U Pbr Pb

Diorite MG3511Ž .Fraction a : short, prismatic, bipyramidal

1 2.0 10 0.2 107 72 0.02428"24 0.1659"55 0.04955"150 156"2 156"5 174"702 4.4 12 0.3 317 91 0.02419"15 0.1643"20 0.04927"43 154"1 155"2 161"203 4.0 9 0.2 251 91 0.02432"16 0.1675"25 0.04996"50 155"1 157"2 193"23

Ž .Fraction b : large, elongated, eroded4 2.7 11 0.3 308 99 0.02448"17 0.1673"28 0.04956"47 156"1 157"2 175"22

Ž .Fraction c : elongated, prismatic, bipyramidal5 2.4 11 0.3 112 68 0.02643"17 0.1832"45 0.05026"114 168"1 171"4 207"52

Albitite MG542Ž .Fraction d : short, prismatic, milky

6 2.6 167 4.3 245 81 0.02320"19 0.1569"20 0.04906"49 148"1 148"2 151"237 2.7 175 4.5 260 82 0.02322"20 0.1569"18 0.04901"31 148"2 148"2 149"148 0.9 179 4.7 230 82 0.02353"16 0.1612"17 0.04969"23 150"1 152"1 181"11

Ž .Fraction e : short, prismatic, bipyramidal, transparent9 1.2 170 4.3 217 80 0.02323"18 0.1568"24 0.04897"58 148"1 148"2 148"27

Ž .Fraction f : large, prismatic, bipyramidal, transparent10 0.9 140 3.7 149 73 0.02338"16 0.1581"20 0.04905"42 149"1 149"2 151"20


Fig. 6. UrPb Concordia diagram for zircons from dioritic veinŽ . Ž .MG3511 ellipses 1–5 and albitite lens MG542 ellipses 6–10 .

Quoted ages correspond to averaged values of zircon fractions 1,2, 3, 4 for sample MG3511 and 6, 7, 9, 10 for sample MG542.

Ž .between 154"1 and 157"2 Ma Table 5 andŽprovide an average UrPb age of 156"3 Ma Fig.

.6 . In the absence of later disturbances related to theAlpine orogeny, this value is interpreted as the crys-tallisation age of the dioritic magma. The last grain

Ž .fraction 5 displays older UrPb ages at;170 MaŽ .Table 5, Fig. 6 , which could possibly reflect inheri-tance.

All the zircons extracted from the albitite lens areŽ .small 20–40mm prismatic grains that occur in

Ž .three different types: a short and milky grainsŽ .containing numerous inclusions; b short and trans-

Ž .parent grains with few or no inclusions, and clarger and elongated grains, with few or no inclu-

Ž .sions, type a being the most abundant. Three frac-Ž .tions of type a and one fraction of each of the two

other types were analysed. All zircon fractions havelow total Pb concentrations of 3–5 ppm and signifi-

Ž .cantly higher U values of 140–180 ppm Table 5 .With the exception of zircon fraction 8, which is20% discordant, all zircon fractions yield identicalconcordant ages ranging from 148"2 to 149"2

Ž .Ma Table 5 . In the absence of syn- or post-mag-matic deformation in this rock, the average value of148"2 Ma calculated for the concordant fractionsonly is interpreted as the crystallisation age of the

Ž . Ž .albitite Fig. 6 . The discordant grain fraction 8displays slightly older UrPb ages at;150–152Ma.

7. Results significance and implications

The Montgenevre gabbros represent shallow in-`trusions in the spinel peridotites that form the base ofthe ophiolite sequence. They clearly display geo-

Ž .chemical features Bertrand et al., 1987; this studyand initial Nd isotopic signatures attesting that theyare tholeiitic melts extracted from a depleted mantlesource with no insight of continental contaminationŽ Ž . .´ T Gq8 . Cumulation played an important roleNd

in the genesis of the clinopyroxene gabbros, whoseREE patterns are controlled by the major cumulusphases, namely clinopyroxene and plagioclase.

Although the Zr content of the clinopyroxeneŽ .gabbros is not negligible Bertrand et al., 1987 ,

these rocks are virtually free of zircon, and weanalysed them using the SmrNd technique. Wholerock samples, including five clinopyroxene gabbrosand one olivine gabbro from the base of the gabbropile, define a SmrNd isochron age of 198"22 Ma

Ž .with an initial ´ T of q8.8. This date is inter-Nd

preted as the emplacement age of the gabbroic meltsinto the mantle peridotites, an event that clearlyrepresents an early stage in the development of theMontgenevre ophiolite. Within its large error, this`Lower Jurassic value is similar to the;180 Ma age

Žof eclogitised gabbros from the Ligurides Borsi,. Ž .1995 and Corsica Beccaluva et al., 1981 , but is

significantly older than the;165 Ma age of otherŽophiolitic gabbros from the Western Alps Rubatto

. Ž .and Gebauer, 1996; Bill et al., 1997 Fig. 2 . Thisindicates that intrusion of gabbroic melts into themantle peridotites was spread over a time period ofG30 Ma throughout the Piemont–Ligurian oceanicbasin. The Montgenevre ophiolite could therefore`represent a much earlier remnant of this ocean,relative to neighbouring ophiolites.

The Montgenevre gabbroic sequence was affected`by high-T shearing at;8508C, which did not de-form its basal section, dominantly composed ofolivine gabbros. Minor amounts of diorite melt wereemplaced in veins and dykes that mainly crosscut the

Ž .sheared clinopyroxene gabbros Fig. 1C . Accordingto field evidence, these melts may have been partlyderived from synkinematic partial melting of the host

Ž .gabbros Caby, 1995 . Modelling the REE data forour gabbro and diorite samples strongly supports thisinterpretation, as this shows that the diorites can be


Ž .generated from low degree 5% or less partial melt-ing of the surrounding clinopyroxene gabbros duringtheir high-T deformation. The values ofq8.0Nd

andq8.5 calculated for the two diorite samples atthe time of their crystallisation are also in agreementwith the model of derivation of these rocks by partialmelting of the gabbros.

In contrast to the gabbros, the diorites contain anunusual high amount of euhedral zircons, some ofthem poikilitic. Zircons extracted from one dioritevein provided a nearly concordant UrPb age at156"3 Ma, which, in the absence of later distur-bances related to the Alpine orogeny, is interpretedas the crystallisation age of the dioritic melt. Somediorite dykes, nearly free of late-magmatic deforma-tion, cut across already deformed amphibole gab-bros, but were in turn progressively deformed. This

Ž . Ž .suggests that 1 the extraction of dioritic melts, 2Ž .their crystallisation in pockets and dykes, and 3

their high-T deformation, were very closely relatedin space and time. As a consequence, the MiddleJurassic age of the diorite vein, which is identical tothat of other diorites from the Ligurides and the

Ž .Apennines Bortolotti et al., 1995; Borsi et al., 1996 ,records a late stage in the formation of theMontgenevre ophiolite.`

The albitite melts, which may represent moreevolved liquids differentiated from a gabbroic or

Ž .dioritic magma Bertrand et al., 1987; Caby, 1995 ,were emplaced slightly later as shown by the concor-dant UrPb age at 148"2 Ma of albitite sampleMG542. The sample, which crosscuts spinel lherzo-lites, shows no evidence of high-T deformation.However, Montgenevre albitites also form large`composite veins that are intimately associated withdeformed diorite veins and crosscut mylonitic gab-bros. This Upper Jurassic age, which is similar withinerrors to that of albitites or plagiogranites from the

ŽLigurides, the Apennines and Corsica Ohnenstetter.et al., 1981; Borsi et al., 1996 , could therefore be

associated with the very last stage of oceanic defor-mation of the Montgenevre ophiolite. Intrusion of`young basaltic dykes that fed the overlaying pillowsequence occurred while high-T ductile deformation

Ž .of the gabbro pile was waning Caby, 1995 andradiolarian sediments were being deposited.

Over the past 30 years, different geodynamicmodels have been proposed in an attempt to explain

the peculiar features of the Piemont–Ligurian ophio-lites. These models fall in two main categories. Inthe first, the ophiolites are thought to have formed ina normal oceanic lithosphere, but to represent eitherŽ .1 peculiar oceanic regions such as transcurrent

Žfault zones e.g. Gianelli and Principi, 1977; Ishi-. Ž .watari, 1985; Weissert and Bernoulli, 1985 , or 2

Žremnants of a slow spreading oceanic ridge e.g.Barret and Spooner, 1977; Lagabrielle and Cannat,

.1990 . In the second category of models, thePiemont–Ligurian ophiolites did not develop in amature oceanic lithosphere and are viewed as rem-nants of an embryonic ocean mainly characterised bysporadic MOR-type basaltic flows directly lying on a

Žperidotite–gabbro floor e.g. Elter, 1975; Bortolottiet al., 1976; Piccardo, 1977; Lombardo and Pog-

.nante, 1982; Pognante et al., 1986 . Following simi-Žlar models proposed for the Red Sea Wernicke,

. Ž .1985 and the Galicia margin Boillot et al., 1987 ,the Piemont–Ligurian models imply the tearing apartof thinned continental crust and general denudationof sub-continental mantle by means of passive exten-

Žsion of the continental lithosphere Lemoine et al.,1987; Piccardo et al., 1994; Rampone and Piccardo,

.1999 .The structural, petrographic and geochemical fea-

tures of the Montgenevre ophiolite, the radiometric`data presented here and their interpretation are con-sistent, in particular, with the model of asymmetricmantle-uncovering by an oblique detachment fault,

Ž .proposed by Lemoine et al. 1987 for the formationŽ .of the Piemont–Ligurian oceanic floor Fig. 7 .

Indeed, the 200–180 Ma ages for the gabbros ofŽ .Montgenevre, the Ligurides Borsi, 1995 and Cor-`

Ž .sica Beccaluva et al., 1981 could be associatedwith the early stage of rifting between the Europeanand Austro-Alpine continental margins. During thisrifting episode, partial melting of the Austro-Alpinemantle may have been responsible for the earlygeneration of gabbroic liquids. As these liquids couldnot escape towards the surface, they were trapped inmagma chambers and underplated along the detach-ment normal fault beneath the European continental

Ž .margin Fig. 7A .The ;165 Ma ages for some Western Alps

Žgabbros Rubatto and Gebauer, 1996; Bill et al.,.1997 and Apennines lherzolites and gabbros

Ž .Rampone et al., 1995; 1998 could reflect the onset


Fig. 7. Model of mantle-uncovering through a detachment normalfault and formation of the Piemont–Ligurian oceanic basin, re-

Ž .drawn from Lemoine et al. 1987 . During the Lower to MiddleŽ .Jurassic ;210–160 Ma , rifting takes place between the Euro-

pean and Austro-Alpine thinned continental margins and gabbroicbodies are emplaced below the European margin along the normal

Ž .detachment fault. From the Middle to Late Jurassic;160 Ma ,spreading of the uncovered lherzolitic–gabbroic floor commencesand the uplifted gabbroic rocks are ductilely deformed at depthalong the detachment fault.

of spreading of the lherzolite–gabbro oceanic floor.During this stage, new gabbroic liquids may havebeen generated, while the gabbroic masses previ-ously emplaced in the sub-continental lherzoliticmantle were pushed away from the underplatespreading centre, in response to the activity of thedetachment fault zone and continuous domal uplift

Ž .of the underlying asthenospheric mantle Fig. 7B .The younger 160–150 Ma ages for Montgenevre`

dioritic and albititic veins and the plagiogranites ofŽthe Apennines and Ligurides Bortolotti et al., 1995;

.Borsi et al., 1996 mark the high-T deformationalhistory and late-stage melting events that affected the

Ž .uncovered lherzolite–gabbro oceanic floor Fig. 7B .Ž .These radiometric ages correlate well Fig. 2 with

the Late Bathonian to Early Kimmeridgian strati-Ž .graphic age ;160–140 Ma of the earliest post-

Žophiolitic radiolarian sediments e.g. De Wever and.Baumgartner, 1995 .

8. Concluding remarks

An underplate spreading centre developing asym-metrically with respect to the continental rifting axis

Ž .Fig. 7 , as suggested by the asymmetric mantle-un-Ž .covering model of Lemoine et al. 1987 , has impor-

tant thermal and kinematic consequences for theevolution of the Piedmont–Ligurian ocean. This mayimply in particular that the newly formedlherzolite–gabbro oceanic domain remained close tothe spreading centre and, therefore, experienced slowcooling and low spreading rates. This appears to becorroborated by the radiometric results discussedhere.

The difference of about 10 Ma between the UrPbŽ . Ž .ages of diorites 156 Ma and albitites 148 Ma

portrays only part of the magmatic history of theMontgenevre ophiolite, since the formation of diorite`veins was a late event clearly post-dating the pile oflayered gabbros. Zircon UrPb ages published else-

Ž .where Fig. 2 give a consistent time span of aboutŽ15 Ma between the crystallisation of gabbros;165

.Ma in some Western Alps ophiolites and that ofŽplagiogranites in the Ligurides and Apennines;150

. Ž .Ma . Earlier emplacement;200–180 Ma of gab-broic melts within the mantle lherzolites of Mont-genevre, the Ligurides and Corsica may indicate that`the uncovering and spreading of the oceanic floortook place at least 15 Ma after the onset of riftingbetween the Brianc¸onnais and Austro-Alpine mar-gins.

These results thus suggest a life span of at least30 Ma for the formation of the Piemont–Ligurianoceanic domain. A tentative comparison may bemade with the Zabargad Island in the Red Sea,where very little spreading has been recorded sinceemplacement of an asthenospheric mantle diapir into

Žthe Pan-African continental lithosphere e.g. Stylesand Gerdes, 1983; Nicolas et al., 1987; Boullier et

. Žal., 1997 some 22 Ma ago Bosch and Bruguier,. y11999 . At a slow-spreading rate of 1–2 cm year ,

300 to 600 km of Piemont–Ligurian oceanic crustcould have been generated, a figure in agreement

Žwith paleogeodynamic reconstructions e.g. Lemoine.et al., 1987, and refs. therein .

Acknowledgements

This investigation was partially supported by theŽ .Centre National de la Recherche Scientifique France


Ž .and the Max-Planck Gesellschaft Germany . WeŽ .wish to thank C. Dupuy USTL, Montpellier for

providing us with unpublished major and trace ele-ment data. Thanks are due to A.W. Hofmann for hisconstant interest, and to W. Abouchami, H. Feld-mann and E. Griesshaber for assistance during ana-

Žlytical work at the MPI, Mainz. A. Nicolas USTL,.Montpellier is thanked for stimulating discussions

and his interest in unraveling the emplacement modeand age of the Alpine ophiolites. The Department of

Ž .Earth Sciences Monash University is thanked forits logistic support during preparation of this paper.We also acknowledge E. Rampone and an anony-mous reviewer for their competent comments thathelped to improve the manuscript, and PubliScienceEditing and Translation for editing the English ver-sion of this text.

References

Barret, J.J., Spooner, E.F.C., 1977. Ophiolite breccias associatedwith allochtonous oceanic crustal rocks in the East LigurianApennines, Italy — a comparison with observations fromrifted oceanic ridges. Earth Planet. Sci. Lett. 35, 79–91.

Beccaluva, L., Chiesa, S., Delaloye, M., 1981. KrAr age determi-nation on some Tethyan ophiolites. Rend. Soc. Ital. Mineral.Petrol. 37, 869–880.

Bertrand, J., Courtin, B., Vuagnat, M., 1982. Elaboration d’unsecteur de lithosphere oceanique liguro-piemontais d’apres les´ ´ `

Ž .donnees de l’ophiolite du Montgenevre Hautes-Alpes, France .´ Òfioliti 2r3, 155–196.

Bertrand, J., Dietrich, V., Nievergelt, P., Vuagnat, M., 1987.Comparative major and trace element geochemistry of gab-broic and volcanic rock sequences, Montgenevre ophiolite,`Western Alps. Schweiz. Mineral. Petrogr. Mitt. 67, 147–169.

Bill, M., Bussy, F., Cosca, M., Mason, H., Hunziker, J.C., 1997.High-precision UrPb and 40Ar–39Ar dating of an Alpine

Ž .ophiolite Gets Nappe, French Alps . Eclogae Geol. Helv. 90,43–54.

Bigazzi, G., Bonadonna, F.P., Ferrara, G., Innocenti, F., 1973.Fission-track ages of zircons and apatites from Northern Apen-nines ophiolites. Fortschr. Miner. 50, 51–53.

Bigazzi, G., Ferrara, G., Innocenti, F., 1972. Fission-track ages ofgabbro from Northern Apennines ophiolites. Earth Planet. Sci.Lett. 14, 242–244.

Boillot, G., Recq, M., Winterer, E.L., 1987. Tectonic denudationof the upper mantle along passive margins: A model based on

Ž .drilling results ODP leg 103, western Galicia margin, Spain .Tectonophysics 132, 335–342.

Borsi, L., 1995. Geochemical characterization and radiometric

determination of meta-Fe-gabbros and meta-plagiogranitesfrom ophiolitic sequence of Voltri group, Sestri-Voltaggio

Ž . Ž .zone Ligurian Alps and Bracco unit Northern Apennines .Plinius 13, 44–51.

Borsi, L., Scharer, U., Gaggero, L., Crispini, L., 1996. Age, originänd geodynamic significance of plagiogranite in lherzolitesand gabbros of the Piedmont–Ligurian oceanic basin. Earth

Ž .Planet. Sci. Lett. 140 1–4 , 227–241.Bortolotti, V., Cortesogno, L., Gianelli, G., Piccardo, G.B., Serri,

G., 1976. I filoni basaltici delle ofioliti dell’ Appenninosettentrionale e il loro significato della formazione del bacinooceanico ligure. Ofioliti 3, 331–364.

Bortolotti, V., Cellai, D., Chiari, M., Vaggelli, G., Villa, I.M.,1995. 40Arr39Ar dating of Apenninic ophiolites: 3. Pla-giogranites from Sasso di Castro, Northern Tuscany, Italy.Ofioliti 20, 55–65.

Bosch, D., Bruguier, O., 1999. An early Miocene age for theŽhigh-temperature event in gneisses from Zabargad Island Red

. Ž .Sea, Egypt : mantle diapirism? Terra Nova 10 5 , 274–279.Boullier, A.M., Firdaous, K., Boudier, F., 1997. Fluid circulation

Ž .related to deformation in the Zabargad gneisses Red Sea rift .Tectonophysics 279, 281–302.

Caby, R., 1995. Plastic deformation of gabbros in a slow spread-ing Mesozoic ridge: Example of the Montgenevre ophiolite,`

Ž .Western Alps. In: Vissers, R.L.M., Nicolas, A. Eds. , Mantleand Lower Crust Exposed in Oceanic Ridges and in Ophio-lites. Kluwer Academic Publishing, pp. 123–145.

Carpena, J., Caby, R., 1984. Fission-track evidence for Late´Triassic oceanic crust in the French Occidental Alps. Geology12, 108–111.

Carpena, J., Mailhe, D., Naeser, C.W., Poupeau, G., 1979. Sur la´ ´datation par traces de fission d’une phase tectonique d’ageêocene superieur en Corse. C. R. Acad. Sci. 289D, 829–832.´ ` ´

Cassidy, R.M., 1988. Determination of rare-earth elements inrocks by liquid chromatography. Chem. Geol. 67, 185–195.

Chapelle, B., 1990. La lithosphere oceanique de la Tethys ligure.` ´ ´ŽEtude des magmatismes basiques et acides massifs ophioli-

.tiques du Montgenevre et de Haute-Ubaye . Unpublished the-`sis, Grenoble, France, 196 pp.

De Corte, F., Van den Haute, P., De Wispelaere, A., Jonckheere,R., 1991. Calibration of the fission-track dating method: is Cuuseful as an absolute thermal neutron fluence monitor? Chem.Geol. 86, 187–194.

De Wever, P., Baumgartner, P.O., 1995. Radiolarians from thebase of the supra-ophiolitic Schistes Lustres formation in the´

Ž .Alps Saint-Veran, France and Traversiera Massif, Italy . Mem.´ ´Geol. Lausanne 23, 725–730.´

De Wever, P., Caby, R., 1981. Datation de la base des SchistesŽLustres post-ophiolitiques par des radiolaires Oxfordien´

.superieur-Kimmeridgien moyen dans les Alpes Cottiennes´ ´Ž .Saint-Veran, France . C. R. Acad. Sci. Paris 292, 427–472.´

De Wever, P., Danelian, T., 1995. Supra-ophiolitic radiolaritesŽ .from Alpine Corsica France . Mem. Geol. Lausanne 23,´ ´

731–735.Dostal, J., Baragar, W.R.A., Dupuy, C., 1986. Petrogenesis of the

Natkusiak continental basalts, Victoria Island, Northwest Ter-ritories, Canada. Can. J. Earth Sci. 23, 622–632.


Elter, P., 1975. L’ensemble ligure. Bull. Soc. Geol. France 6,´984–997.

Evensen, N.M., Hamilton, P.J., O’Nions, R.K., 1978. Rare-earthabundances in chondritic meteorites. Geochim. Cosmochim.Acta 42, 1199–1212.

Frontignie, D., Dalaloye, M., Bertrand, J., 1982. Ages radiometri-´ques KrAr des elements ophiolitiques de la nappe des Gets´ ´Ž .Haute-Savoie, France . Eclogae Geol. Helv. 75, 117–226.

Fujimaki, H., Tatsumoto, M., Aoki, K., 1984. Partition coeffi-cients of Hf, Zr and REE between phenocrysts and ground-masses. J. Geophys. Res. 89, B662–B672.

Gianelli, G., Principi, G., 1977. Northern Apennines ophiolite: anancient transcurrent fault zone. Soc. Geol. Ital. 96, 53–58.

Goldstein, S.L., O’Nions, R.K., Hamilton, P.J., 1984. A Sm–Ndstudy of atmospheric dusts and particulates from major riversystems. Earth Planet. Sci. Lett. 70, 221–236.

Ishiwatari, A., 1985. Alpine ophiolites: product of low-degreemantle melting in a Mesozoic transcurrent rift zone. EarthPlanet. Sci. Lett. 76, 93–108.

John, B.E., Stunitz, H., 1997. Magmatic fracturing and small-scale¨melt segregation during pluton emplacement: evidence from

Ž . Ž .the Adamello massif Italy . In: Bouchez, J.L. Ed. , Granite:From Segregation of Melt to Emplacement Fabrics. KluwerAcademic Publishing, pp. 55–74.

Kretz, R., 1983. Symbols for rock forming minerals. Am. Mineral.58, 277–279.

Krogh, T.E., 1973. A low contamination method for the hy-drothermal decomposition of zircon and extraction of U andPb for isotopic age determination. Geochim. Cosmochim. Acta46, 637–649.

Lagabrielle, Y., Cannat, M., 1990. Alpine Jurassic ophiolitesresemble the modern central Atlantic basement. Geology 18,319–322.

Lemoine, M., Tricart, P., Boillot, G., 1987. Ultramafic and gab-broic ocean floor of the Ligurian Tethys domain of the West-ern Alps, Corsica, Apennines: in search of a genetic model.Geology 15, 622–625.

Lewis, A.D., Snewing, J.D., 1980. The Montgenevre ophiolite`Ž .Hautes-Alpes, France : metamorphism and trace element geo-chemistry of the volcanic sequence. Chem. Geol. 28, 291–306.

Lombardo, B., Pognante, U., 1982. Tectonic implications in theevolution of the Western Alps ophiolite metagabbros. Ofioliti7, 371–394.

Marcucci, M., Conti, M., 1995. Radiolarian biostratigraphy of thecherts in the sedimentary cover of the Apenninic ophiolitesŽ .Italy . Mem. Geol. Lausanne 23, 799–812.´ ´

McKenzie, D., 2000. Constraints on melt generation and transportfrom U-series activity ratios. Chem. Geol. 162, 81–94.

Mevel, C., Caby, R., Kienast, J.R., 1978. Amphibolite faciesćonditions in the oceanic crust: example of amphibolitizedflaser-gabbro and amphibolites from the Chenaillet ophiolite

Ž .massif Hautes-Alpes, France . Earth Planet. Sci. Lett. 39,98–108.

Michard, A., Gurriet, P., Soudant, M., Albarede, F., 1985. Ndìsotopes in French Phanerozoic shales: external versus internalaspects of crustal evolution. Geochim. Cosmochim. Acta 49,601–610.

Naeser, C.W., 1978. Fission Track Dating: U.S. Geological Sur-vey Open-File Report. pp. 76–190.

Nicolas, A., Boudier, F., Montigny, R., 1987. Structure of Zabar-gad Island and early rifting of the Red Sea. J. Geophys. Res.92, 461–474.

Odin, G.S., 1994. Geological time scale. C. R. Acad. Sci. ParisŽ .II 318, 59–71.

Ohnenstetter, M., Ohnenstetter, D., 1980. Comparison betweenCorsican albitites and oceanic plagiogranites. Arch. Sci.Ž . Ž .Geneva 33 2–3 , 201–220.

Ohnenstetter, M., Ohnenstetter, D., Vidal, P., Cornichet, J., Her-mitte, D., Mace, J., 1981. Crystallization and age of zirconfrom Corsican ophiolitic albitites: consequences for oceanicexpansion in Jurassic times. Earth Planet. Sci. Lett. 54, 397–408.

Piccardo, G.B., 1977. Le ofioliti dell’ areale ligure: petrologia edambiante geodinamico di formazione. Rend. Soc. Geol. Ital.Mineral. Petrol. 33, 221–252.

Piccardo, G.B., Rampone, E., Vanucci, R., Cimmino, F., 1994.Upper mantle evolution of ophiolitic peridotites from theNorthern Apennine: petrological constraints to the geodynamicprocesses. Mem. Soc. Geol. Ital. 48, 137–148.

Pognate, U., Perotto, A., Salino, C., Toscani, L., 1986. Theophiolitic peridotite of the Western Alps: record of the evolu-tion of a small oceanic-type basin in the Mesozoic Tethys.Mineral. Petrogr. Mitt. 35, 47–65.

Rampone, E., Hofmann, A.W., Piccardo, G.B., Vannuggi, R.,Bottazzi, P., Ottolini, L., 1995. Petrology, mineral and isotope

Žgeochemistry of the External Liguride peridotites Northern. Ž .Appenines, Italy . J. Petrol. 36 1 , 81–105.

Rampone, E., Hofmann, A.W., Raczek, I., 1998. Isotropic con-Ž .trasts within the Internal Liguride ophiolite N. Italy : the lack

of a genetic mantle–crust link. Earth Planet. Sci. Lett. 163,175–189.

Rampone, E., Piccardo, G.B., 1999. The ophiolite–oceanic litho-sphere analogue: new insights from the Northern ApenninesŽ .Italy . Geol. Soc. Am., Spec. Publ., in press.

Rollinson, H.R., 1993. Using geochemical data: evaluation, pre-sentation and interpretation. Longman Scientific and Technicaland Wiley, p. 352.

Rubatto, D., Gebauer, D., 1996. Timing of formation and subduc-Ž .tion of the Zermatt–Saas ophiolites Western Alps : a com-

bined cathodoluminescence–Shrimp study using zircon. Mitt.Osterr. Mineral. Ges. 141, 193–195.

Schaaf, A., Polino, R., Lagabrielle, Y., 1985. Nouvelle decouverte´de radiolaires d’age Oxfordien superieur-Kimmeridgienˆ ´ ínferieur, a la base d’une serie supra-ophiolitique des schistes´ ` ´

Žlustres piemontais Massif de Traversiera, Haut Val Maıra,´ ´ ¨. Ž .Italie . C. R. Acad. Sci. Paris II 301-14, 1079–1084.

Stacey, J.C., Kramers, J.D., 1975. Approximation of terrestrial Pbisotope evolution by a two-stage model. Earth Planet. Sci.Lett. 26, 207–221.

Ž .Styles, P., Gerdes, K.D., 1983. St John’s island Red Sea : a newgeophysical model and its implications for the emplacement ofultramafic rocks in fracture zones and at continental margins.Earth Planet. Sci. Lett. 65, 353–368.

Van den Haute, P., Jonckheere, R., De Corte, F., 1988. Thermal


neutron fluence determination for fission-track dating withmetal activation monitors: a re-investigation. Chem. Geol. 73,233–244.

Vigneresse, J.L., Barbey, P., Cuney, M., 1996. Rheological transi-tions during partial melting and crystallization with application

Ž .to felsic magma segregation and transfer. J. Petrol. 37 6 ,1579–1600.

Weissert, H.J., Bernouilli, D., 1985. A transform margin in the

Mesozoic Tethys: evidence from the Swiss Alps. Geol. Rund-sch. 74, 665–679.

Wernicke, B., 1985. Uniform sense normal simple shear of thecontinental lithosphere. Can. J. Earth Sci. 22, 108–125.

White, W.M., Patchett, P.J., 1984. Hf–Nd–Sr isotopes and incom-patible element abundances in island arcs, implications formagma origin and crust–mantle evolution. Earth Planet. Sci.Lett. 67, 167–185.

evolution of the ligurian tethys in the western alps: sm/nd and u/pb geochronology and rare-earth...

Documents