Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
The Cerro del Almirez massif (Nevado–Filábride Complex, Betic Cordillera, SE Spain) is composed of antigorite (Atg-) serpentinite and chlorite (Chl-) harzburgite separated by a thin reaction front formed in a palaeo-subduction setting. These ultramafic rocks preserve unique prograde structures related to the pre- and peak high-pressure event (1.6–1.9 GPa and 680–710 °C). The oldest subduction-related structures are preserved in Atg-serpentinites: a penetrative S1 foliation and associated L1 stretching lineation that formed in a non-coaxial regime with a top-to-the-W sense of shear. This planar–linear fabric is crosscut by olivine ± Ti-clinohumite veins, formed during the prograde breakdown of brucite and pre-metamorphic clinopyroxene, which form a system of veins decimetres long. They record embrittlement processes due to the release of 6 vol.% of H2O associated with brucite dehydration. The growth of prograde olivine and/or tremolite porphyroblasts is syn- to post-kinematic in relation to the S1 foliation. Further reactions at higher temperature related to the complete breakdown of the Atg (i.e. Atg-out) are post-kinematic to S1. Prograde Chl-harzburgite is crosscut by sets of conjugate zones associated with grain-size reduction of olivine grains. These grain-size reduction zones are interpreted as brittle structures generated by hydrofracturing due to overpressure fluids in a compressional setting at low-differential stresses.Structures related to the exhumation process are mainly preserved in the metasedimentary host rock, where an S2/L2 planar–linear fabric developed within a shear zone dominated by a non-coaxial regime with a top-to-the-W sense of movement in a transpressional regime. Peak metamorphic conditions deduced for the schists are similar in pressure (1.3–1.9 GPa) but lower in temperature (560–590 °C) compared to the ultramafic rocks in contact with them, suggesting a major shear zone at the base of the ultramafic massif during D2 deformation in a right-handed transpressional regime.
ISSN: 0040-1951

hal-01172303
https://hal.archives-ouvertes.fr/hal-01172303
DOI : 10.1016/j.tecto.2014.12.022

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
We present a new set of brittle microtectonic measurements carried out in the Pliocene and Quaternary rocks outcropping in several key sectors of the western Betic and Rif orogen, the so-called Gibraltar orogenic arc. This data set, along with available earthquake focal mechanisms and borehole breakouts, allowed us to compile the Pliocene and Quaternary stress map of this area. This map provides new constraints for tectonic models and the present-day tectonic activity of the proposed active eastward subduction of oceanic lithosphere beneath the Gibraltar Arc and roll-back. The horizontal maximum compressive stress (S-Hmax) is NW-SE in the Betic Orogen and N-S/NNW-SSE in the southern Rif Cordillera. There is a significant consistency between S-Hmax and the displacement field deduced from GPS measurements with respect to the African plate: both appear to reflect the NW-SE convergence between the African and the European plates that is perturbed in the Rif. We propose that part of the eastern Rif behaves as a quasi-rigid block welded to the stable African plate. This block is bounded by important faults that localized most of the deformation disturbing the stress and surface displacement field. Pliocene to Quaternary N-S to NW-SE Africa-Europe plate convergence seem to be associated to the reorganization of the remnant Early Miocene subduction system in a continental-continental collision framework. Three-dimensional reconstruction of available seismic tomography plotted against the intermediate seismicity shows that only part of the old subduction system, whose orientation ranges from N20 degrees E to N100 degrees E, remains active: the portion ranging from N30 degrees E to N40 degrees E, orthogonal to the regional convergence. (C) 2011 Elsevier Ltd. All rights reserved.
ISSN: 0264-3707

hal-00671033
https://hal.archives-ouvertes.fr/hal-00671033
DOI : 10.1016/j.jog.2010.12.003

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Serpentinites form by hydration of mantle peridotite and constitute the largest potential reservoir of fluid-mobile elements entering subduction zones. Isotope ratios of one such element, boron, distinguish fluid contributions from crustal versus serpentinite sources. Despite 85% of boron hosted within abyssal peridotite being lost at the onset of subduction at the lizardite-to-antigorite transition, a sufficient cargo of boron to account for the composition of island arc magma is retained (c. 7 μg g− 1, with a δ11B of + 22‰) until the down-going slab reaches the antigorite-out isograd. At this point a 11B-rich fluid, capable of providing the distinctive δ11B signature of island arc basalts, is released. Beyond the uniquely preserved antigorite-out isograd in serpentinites from Cerro del Almirez, Betic Cordillera, Spain, the prograde lithologies (antigorite–chlorite–orthopyroxene–olivine serpentinite, granofels-texture chlorite-harzburgite and spinifex-texture chlorite-harzburgite) have very different boron isotope signatures (δ11B = − 3 to + 6‰), but with no significant difference in boron concentration compared to the antigorite-serpentinite on the low P–T side of the isograd. 11B-rich fluid, which at least partly equilibrated with pelagic sediments, is implicated in the composition of these prograde lithologies, which dehydrated under open-system conditions. Serpentinite-hosted boron lost during the early stages of dehydration is readily incorporated into forearc peridotite. This, in turn, may be dragged to sub-arc depths as a result of subduction erosion and incorporated in a mélange comprising forearc serpentinite, altered oceanic crust and pelagic sediment. At the antigorite-out isograd it dehydrates, thus potentially providing an additional source of 11B-rich fluids.
ISSN: 0009-2541

hal-01172276
https://hal.archives-ouvertes.fr/hal-01172276
DOI : 10.1016/j.chemgeo.2014.03.015

Éditeur(s) : HAL CCSD Wiley-Blackwell
Résumé : International audience
Garnet growth in high-pressure, mafic garnet granulites formed by dehydration melting of hornblende-gabbronorite protoliths in the Jijal complex (Kohistan palaeo-island arc complex, north Pakistan) was investigated through a microstructural EBSD-SEM and HRTEM study. Composite samples preserve a sharp transition in which the low-pressure precursor is replaced by garnet through a millimetre-sized reaction front. A magmatic foliation in the gabbronorite is defined by mafic-rich layering, with an associated magmatic lineation defined by the shape-preferred orientation (SPO) of mafic clusters composed of orthopyroxene (Opx), clinopyroxene (Cpx), amphibole (Amp) and oxides. The shape of the reaction front is convoluted and oblique to the magmatic layering. Opx, Amp and, to a lesser extent, Cpx show a strong lattice-preferred orientation (LPO) characterized by an alignment of [001] axes parallel to the magmatic lineation in the precursor hornblende-gabbronorite. Product garnet (Grt) also displays a strong LPO. Two of the four < 111 > axes are within the magmatic foliation plane and the density maximum is subparallel to the precursor magmatic lineation. The crystallographic relationship < 111 >(Grt)//[001](Opx,Cpx,Amp) deduced from the LPO was confirmed by TEM observations. The sharp and discontinuous modal and compositional variations observed at the reaction front attest to the kinetic inhibition of prograde solid-state reactions predicted by equilibrium-phase diagrams. The P-T field for the equilibration of Jijal garnet granulites shows that the reaction affinities are 5-10 kJ mol.(-1) for the Grt-in reaction and 0-5 kJ mol.(-1) for the Opx-out reaction. Petrographic and textural observations indicate that garnet first nucleated on amphibole at the rims of mafic clusters; this topotactic replacement resulted in a strong LPO of garnet. Once the amphibole was consumed in the reaction, the parallelism of [001] axes of the mafic-phase reactants favoured the growth of garnet crystals with similar orientations over a pyroxene substrate. These aggregates eventually sintered into single-crystal garnet. In the absence of deformation, the orientation of mafic precursor phases conditioned the nucleation site and the crystallographic orientation of garnet because of topotaxial transformation reactions and homoepitaxial growth of garnet during the formation of high-pressure, mafic garnet-granulite after low-pressure mafic protoliths.
ISSN: 0263-4929

hal-00411323
https://hal.archives-ouvertes.fr/hal-00411323
DOI : 10.1111/J.1525-1314.2008.00789.x

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
We combined microstructural observations and high-resolution crystallographic preferred orientation (CPO) mapping to unravel the active deformation mechanisms in garnet clinopyroxenites, garnet-spinel websterites, and spinel websterites from the Beni Bousera peridotite massif. All pyroxenites display microstructures recording plastic deformation by dislocation creep. Pyroxene CPOs are consistent with dominant slip on [001]{110} in clinopyroxene and on [001](100) or [001](010) in orthopyroxene. Garnet clinopyroxenites have however high recrystallized fractions and finer grain sizes than spinel websterites. Recrystallization mechanisms also differ: subgrain rotation dominates in garnet clinopyroxenites, whereas in spinel websterites nucleation and growth also contribute. Elongated shapes and strong intracrystalline misorientations suggest plastic deformation of garnet, but CPOs are weak. Clinopyroxene porphyroclasts in spinel websterites show deformation twins underlined by orthopyroxene exsolutions. Thermodynamic calculations indicate that garnet clinopyroxenites deformed at 2.0 GPa and 950-1000 °C and spinel pyroxenites at 1.8 GPa and 1100-1150 °C. The lower temperatures may explain the faster work rates implied by the finer grained microstructures in garnet clinopyroxenites. Greater stresses may have also reduced the competence contrast between garnet and pyroxene in the garnet pyroxenites and, at the outcrop scale, lowered the competence contrast between pyroxenites and peridotites, favoring mechanical dispersion of pyroxenites in the cooler lithospheric mantle.
ISSN: 0191-8141

hal-00750322
https://hal.archives-ouvertes.fr/hal-00750322
DOI : 10.1016/j.jsg.2012.02.019

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Before attaining the mantle wedge, where they trigger partial melting, volatiles released from dehydration reactions in the slab have to migrate across a relatively cold (<750 degrees C), peridotite-layer above the incoming slab. In order to unravel the mechanisms allowing for this initial stage of fluid transport, we performed a detailed field and microstructural study of metamorphic prograde peridotites in the Cerro del Almirez ultramafic massif (Betic Cordillera, Spain), where evidences of one of the most important dehydration reactions in subduction zones, the high-pressure antigorite breakdown (P = 1.6-1.9 GPa and T approximate to 680 degrees C), can be mapped in the field. This reaction led to arborescent growth of centimeter-size olivine and orthopyroxene, producing a chlorite-harzburgite with a spinifex-like texture. Microstructural observations and crystal preferred (CPO) mapping show no evidences of solid-state deformation during the prograde growth of olivine and orthopyroxene at the expenses of antigorite. However, a few tens to a hundred meters away from the reaction front, the metamorphic texture is partially obliterated by grain-size reduction in roughly planar conjugate zones, a few mm to meters wide. Grain size reduction zones (GSRZ) are characterized by (1) sharp contacts with undeformed spinifex-like texture domains, (2) important reduction of the olivine grain size (60-250 mu m), (3) olivine color change from brownish to colorless, (4) decrease in the modal amount of orthopyroxene, and (5) at the mm- to cm-scale, irregular shapes and abrupt terminations. Field and microstructural observations exclude that relative displacement took place across these GSRZ. Changes in modal composition imply reactions with fluids undersaturated in silica. Analysis of olivine crystal-preferred orientations (CPO) in GSRZ shows patterns similar, but more dispersed, than those in neighboring spinifex-like domains. It also reveals mm- to cm-scale discrete domains with rather homogeneous crystallographic orientations suggesting inheritance from the preexisting spinifex-like olivines in the host peridotite. Misorientation angles between neighboring grains in the GSRZ show peaks at similar to 5-10 degrees and similar to 20 degrees, but rotations are not crystallographically controlled. Based on these observations, we rule out the formation of the GSRZ by dynamic recrystallization during dislocation creep and propose that they record brittle deformation (microcraking) of the spinifex-like chlorite-harzburgite, probably induced by hydrofracturing at high pressure and relative low temperature conditions (680-710 degrees C). High-pressure hydrofracturing can, thus, be invoked as an efficient mechanism for fluid flow across the cold top-slab mantle layer, hence allowing the slab-derived fluids to ingress in the wedge.
ISSN: 0012-821X

hal-00534035
https://hal.archives-ouvertes.fr/hal-00534035
DOI : 10.1016/j.epsl.2010.06.029

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Geochronological studies of samples from continuous crustal sections of fossil intra-oceanic arcs are paramount for determining the precise timing of major processes that took place during the building of an oceanic island-arc crust. In this study, laser ablation ICP-MS U-Pb zircon (and conventional Rb-Sr) analyses pinpoint the timing of major events responsible for crustal growth of the Kohistan paleo-island arc (Northern Pakistan). Inheritance in magmatic rocks is evidence for recycling of intra-arc material and indicates that the Kohistan arc may have formed as early as 135 +/- 4 Ma. One older inherited grain indicates that the arc developed on a young oceanic lithosphere whose remnants include a c. 175 Ma old component. An interesting consequence is that intra-oceanic arc magmas can yield inherited zircons, which can be much older than the arc system itself without requiring recycling back into the mantle by subduction processes. In the case of the Kohistan arc, ante-arc Neotethys oceanic lithosphere was tapped by arc magmas during their way upward into the arc crust. The oldest magmatic ages measured in this study fall in the range 101-102 Ma (100.9 +/- 0.6 and 102.1 +/- 0.4 Ma). This period corresponds to arc build-up and thickening of the arc crust. Leucogranitic melts dated at 89.9 +/- 0.4 Ma and 90.9 +/- 1.0 Ma are considered as produced by dehydration/melting reaction accompanying granulitisation of the thickened lower arc crust. A maximum age for this event is 97.7 +/- 0.7 Ma, the age of emplacement of a garnet meta-tonalite affected by granulite facies metamorphism. Magmatic activity was still ongoing during and after the granulitisation as testified by emplacement of a diorite and a gabbro dated at 89.1 +/- 0.6 and 88.2 +/- 0.9 Ma respectively. The upper part of the metaplutonic sequence contains diorite samples dated at 84.6 +/- 0.5 Ma and 84.3 +/- 0.5 Ma and as young as 81.1 +/- 0.7 Ma. During this period, and based on the present day outcropping sequences, the mean crustal growth rate varied from 32 to 65 km(3)/km/Ma (volume per unit width along the strike of the arc) which is comparable to the range for present day arcs of the Western Pacific region. The correspondence of crustal growth rates observed for present-day island arcs and the good preservation of the crustal section in the Kohistan arc, make this area an exceptional natural laboratory to study arc related processes and to check models of continental crust formation by arc accretion.
ISSN: 0012-821X

hal-00644555
https://hal.archives-ouvertes.fr/hal-00644555
DOI : 10.1016/j.epsl.2011.07.016

Éditeur(s) : HAL CCSD American Geophysical Union (AGU)
Résumé : International audience
The Ronda peridotite supplies one of the best objects to document subcontinental mantle deformation, but its internal deformation and exhumation mechanisms remain controversial. Here we provide new structural data and numerical results that constrain the Oligocene-Miocene deformation history of the Ronda massif. We first describe a mantle shear zone in the northern massif that developed subcrustal strain localization during decompression and related partial melting. The deformation regime of this mantle shear zone evolved from penetrative NE-SW stretching to sinistral shear highlighted by discrete shear bands. We then show structural observations that document a viscous deformation in the southern massif occurring prior to the thrust-assisted emplacement of the peridotite during large decompression. Finally, we performed numerical investigations that quantify a high temperature of 980°C for the basal peridotite lens at the time of its crustal emplacement. Our numerical results constrain the timing of ductile deformation of the peridotite just before 22 Ma, probably between 30 and 22 Ma. Altogether, these features led us to conclude that the deformation and exhumation of the Ronda peridotite results from lithosphere thinning subsequently inverted in the course of the Oligocene-Miocene. Among available models, our findings support the hypothesis of peridotite exhumation by the inversion of a thinned back-arc continental lithosphere during westward slab rollback through the Alboran region.
ISSN: 0278-7407

insu-00924180
https://hal-insu.archives-ouvertes.fr/insu-00924180
https://hal-insu.archives-ouvertes.fr/insu-00924180/document
https://hal-insu.archives-ouvertes.fr/insu-00924180/file/tect20062.pdf
DOI : 10.1002/tect.20062

Éditeur(s) : HAL CCSD Springer Verlag (Germany) Springer Verlag
Résumé : International audience
The Cerro del Almirez massif (Spain) represents a unique fragment of serpentinized oceanic lithosphere that has been first equilibrated in the antigorite stability field (Atg-serpentinites) and then dehydrated into chlorite–olivine–orthopyroxene (Chl-harzburgites) at eclogite facies conditions during subduction. The massif preserves a dehydration front between Atg-serpentinites and Chl-harzburgites. It constitutes a suitable place to study redox changes in serpentinites and the nature of the released fluids during their dehydration. Relative to abyssal serpentinites, Atg-serpentinites display a low Fe3+/FeTotal(BR) (=0.55) and magnetite modal content (=2.8–4.3 wt%). Micro-X-ray absorption near-edge structure (μ-XANES) spectroscopy measurements of serpentines at the Fe–K edge show that antigorite has a lower Fe3+/FeTotal ratio (=0.48) than oceanic lizardite/chrysotile assemblages. The onset of Atg-serpentinites dehydration is marked by the crystallization of a Fe3+-rich antigorite (Fe3+/FeTotal = 0.6–0.75) in equilibrium with secondary olivine and by a decrease in magnetite amount (=1.6–2.2 wt%). This suggests a preferential partitioning of Fe3+ into serpentine rather than into olivine. The Atg-breakdown is marked by a decrease in Fe3+/FeTotal(BR) (=0.34–0.41), the crystallization of Fe2+-rich phases and the quasi-disappearance of magnetite (=0.6–1.4 wt.%). The observation of Fe3+-rich hematite and ilmenite intergrowths suggests that the O2 released by the crystallization of Fe2+-rich phases could promote hematite crystallization and a subsequent increase in fo2 inside the portion of the subducted mantle. Serpentinite dehydration could thus produce highly oxidized fluids in subduction zones and contribute to the oxidization of the sub-arc mantle wedge.
ISSN: 0010-7999

hal-01172310
https://hal.archives-ouvertes.fr/hal-01172310
DOI : 10.1007/s00410-015-1130-y

Éditeur(s) : HAL CCSD American Geophysical Union (AGU)
Résumé : International audience
The Atlas Mountains are characterized by high elevations and Quaternary volcanism. Long period magnetotelluric data acquired along a NNW-SSE transect reveal the presence of a conductive anomalous mantle below the High Atlas. Data dimensionality analyses show a preferent N80 degrees E strike of the deep resistivity structure in agreement with the induction vector alignment at long periods. Accordingly, a 2D inversion of the data set was carried out. Large resistive bodies at the crustal basement most likely correspond to batholiths emplaced in more conductive metapelites. They are covered by outcropping conductive sedimentary detritic and carbonate rocks. Lithospheric thinning producing anomalous mantle and basin development in the Atlas probably started during Triassic-Jurassic rifting. Inversion tectonics since the Oligocene produced low shortening on previous lithospheric weak zones, with thrusting of the Atlas above the stable African plate. Melting at the top of the anomalous mantle is connected with Quaternary basaltic volcanism in the Middle Atlas.
ISSN: 0278-7407

hal-00671017
https://hal.archives-ouvertes.fr/hal-00671017
DOI : 10.1029/2010TC002859

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
The Massif du Sud is a large ophiolitic complex that crops out in the southern region of New Caledonia (SW Pacific). It is dominated by harzburgite tectonite that locally shows a transitional gradation to massive dunite up section. Clinopyroxene, orthopyroxene and plagioclase progressively appear in dunite up to the transition to layered wehrlite and orthopyroxene-gabbro. The dunite-wehrlite and wehrlite-gabbro contacts are parallel and the latter defines the paleo-Moho.;Highly depleted modal, mineral and bulk rock compositions indicate that harzburgites are residues after high degrees (20-30%) of partial melting mainly in the spinet-stability field. Their relative enrichment in HFSE, LREE and MREE is due to re-equilibration of melting residues with percolating melts. Dunite formed in the Moho transition zone by reaction between residual mantle harzburgite and olivine-saturated melts that led to pyroxene dissolution and olivine precipitation. Rare clinopyroxene and plagioclase crystallized in interstitial melt pores of dunite from primitive, low-TiO2, ultra-depleted liquids with a geochemical signature transitional between those of island arc tholeiites and boninites.;Ascending batches of relatively high-SiO2, ultra-depleted melts migrated through the Moho transition zone and generated wehrlite by olivine dissolution and crystallization of clinopyroxene, orthopyroxene and plagioclase in variable amounts. These liquids were more evolved and were produced by higher degrees of melting or from a more depleted source compared with melts that locally crystallized clinopyroxene in dunite. Ultra-depleted magmas, non-cogenetic with those that formed the Moho transition zone, ascended to the lower crust and generated gabbroic cumulates with subduction-related affinity. Thus, the ultramafic and mafic rocks in the Moho transition zone and lower crust of the Massif du Sud ophiolite are not products of fractional crystallization from a single magma-type but are the result of migration and accumulation of different melts in a multi-stage evolution. The record of high partial melting in the mantle section, and migration and accumulation of ultra-depleted subduction-related melts in the Moho transition zone and lower crust support that the Massif du Sud ophiolite is a portion of forearc lithosphere generated in an extensional regime during the early phases of the subduction zone evolution. Our results show the existence of different types of ultra-depleted melt compositions arriving at the Moho transition zone and lower crust of an infant intraoceanic paleo-arc. Ultra-depleted melts may thus be a significant component of the melt budget generated in oceanic spreading forearcs prior to aggregation and mixing of a large range of melt compositions in the crust.
ISSN: 0009-2541

hal-00429333
https://hal.archives-ouvertes.fr/hal-00429333
DOI : 10.1016/j.chemgeo.2009.06.004

Éditeur(s) : HAL CCSD Oxford University Press (OUP)
Résumé : International audience
The processes active in the deep crust above an oceanic subduction zone during its evolution have been constrained through a detailed geochemical study (major and trace elements and Sr, Nd and Pb isotopes) of representative samples through an 30 km thick exhumed crustal section of the Cretaceous Kohistan oceanic island arc (Northern Pakistan). The use of both trace elements and radiogenic isotopes reveals two distinct geochemical suites (suites A and B) within the JijalPatanKiruKamila (JPKK) complex. Suite A is characterized by a progressive enrichment in Pb-207 and a decrease in Nd-143/Nd-144 with increasing La-N/Sm-N. Suite B has higher Pb-207/Pb-204 and lower Nd-143/Nd-144 ratios with approximately constant La-N/Sm-N. By combining trace elements with different partitioning behaviour it is demonstrated that there is an increasing contribution of the subduction component in the magmas with time. It is also possible to distinguish a slab component imprint carried by aqueous fluids from one corresponding to sediment melts. Intrusive granites are abundant in the upper levels of the JPKK section. All were generated at the arc root level (Jijal crustal section) during dehydration-melting of hornblende-rich plutonic rocks. A three-stage geodynamic model is proposed for the evolution of the arc over a period of 30 Myr. The first stage (117105 Ma) starts with the onset of subduction, which was followed by the building of the volcanic arc. The second stage (10599 Ma to 9691 Ma) corresponds to a major igneous event, which was characterized by abundant magma underplating and granulite-facies metamorphism at the arc base. Recycling of the residualcumulative lower crust into the convective asthenospheric mantle was efficient during this stage, and was probably related to thermo-mechanical erosion of the base of the crust. The last stage (9585 Ma) corresponds to a period of low magmatic activity, which marked the end of the intra-oceanic subduction. This is related to the formation of a cold blanket above the slab surface as a result of thermo-mechanical erosion of the cold walls of the subduction zone (i.e. the upper part of the slab and the base of the overriding plate), and corner flow dragging the cold material into the zone of melt generation. Ultimately, a voluminous magmatic pulse occurred around 85 Ma (forming the Chilas complex), before arccontinent collision.
ISSN: 0022-3530

hal-00420068
https://hal.archives-ouvertes.fr/hal-00420068
DOI : 10.1093/petrology/egp010

Éditeur(s) : HAL CCSD Springer Verlag
Résumé : International audience
The western part of the Ronda peridotite massif (Southern Spain) consists mainly of highly foliated spinel-peridotite tectonites and undeformed granular peridotites that are separated by a recrystallization front. The spinel tectonites are interpreted as volumes of ancient subcontinental lithospheric mantle and the granular peridotites as a portion of subcontinental lithospheric mantle that underwent partial melting and pervasive percolation of basaltic melts induced by Cenozoic asthenospheric upwelling. The Re-Os isotopic signature of sulfides from the granular domain and the recrystallization front mostly coincides with that of grains in the spinel tectonites. This indicates that the Re-Os radiometric system in sulfides was highly resistant to partial melting and percolation of melts induced by Cenozoic lithospheric thermal erosion. The Re-Os isotopic systematics of sulfides in the Ronda peridotites thus mostly conserve the geochemical memory of ancient magmatic events in the subcontinental lithospheric mantle. Os model ages record two Proterozoic melting episodes at similar to 1.6 to 1.8 and 1.2-1.4 Ga, respectively. The emplacement of the massif into the subcontinental lithospheric mantle probably coincided with one of these depletion events. A later metasomatic episode caused the precipitation of a new generation of sulfides at similar to 0.7 to 0.9 Ga. These Proterozoic Os model ages are consistent with results obtained for several mantle suites in Central/Western Europe and Northern Africa as well as with the Nd model ages of the continental crust of these regions. This suggests that the events recorded in mantle sulfides of the Ronda peridotites reflect different stages of generation of the continental crust in the ancient Gondwana supercontinent.
ISSN: 0010-7999

hal-00564508
https://hal.archives-ouvertes.fr/hal-00564508
DOI : 10.1007/s00410-009-0429-y

Éditeur(s) : HAL CCSD Springer Verlag
Résumé : International audience
Chromitite pods in the Mayarí-Cristal ophiolitic massif (eastern Cuba) were formed in the Late Cretaceous when island arc tholeiites and MORB-like back-arc basin basalts reacted with residual mantle peridotites and generated chromite-rich bodies enclosed in dunite envelopes. Platinum-group minerals (PGM) in the podiform chromitites exhibit important Os-isotope heterogeneities at the kilometric, hand sample and thin section scales. 187Os/188Os calculated at the time of chromitite crystallization (~90 Ma) ranges between 0.1185 and 0.1295 (γOs = −7.1 to +1.6, relative to enstatite chondrite), and all but one PGM have subchondritic 187Os/188Os. Grains in a single hand sample have initial 187Os/188Os that spans from 0.1185 to 0.1274, and in one thin section it varies between 0.1185 and 0.1232 in two PGM included in chromite which are only several millimeters apart. As the Os budget of a single micrometric grain derives from a mantle region that was at least several m3 in size, the variable Os isotopic composition of PGM in the Mayarí-Cristal chromitites probably reflects the heterogeneity of their mantle sources on the 10-100 m scale. Our results show that this heterogeneity was not erased by pooling and mingling of individual melt batches during chromitite crystallization but was transferred to the ore deposits on mineral scale. The distribution of the Os model ages calculated for PGM shows four main peaks, at ~100, 500, 750 and 1,000 Ma. These variable Os model ages reflect the presence of different depleted domains in the oceanic (Pacific-related) upper mantle of the Greater Antilles paleo-subduction zone. The concordance between the age of crystallization of the Mayarí-Cristal chromitites and the most recent peak of the Os model age distribution in PGM supports that Os in several grains was derived from fertile domains of the upper mantle, whose bulk Os isotopic composition is best approximated by that of enstatite chondrites; on the other hand, most PGM are crystallized by melts that tapped highly refractory mantle sources.
ISSN: 0010-7999

hal-00822742
https://hal.archives-ouvertes.fr/hal-00822742
DOI : 10.1007/s00410-010-0575-2

Éditeur(s) : HAL CCSD Elsevier
Résumé : The ultramafic–mafic complexes located at the base of obducted paleo-island arcs are commonly interpreted as evidence for intra-crustal fractionation of primitive (high-Mg#) mantle melts. The present study, realized on the Jijal mafic–ultramafic basal section of the Cretaceous Kohistan oceanic arc (N Pakistan), discards a crystal fractionation model from a single parental magma to take into account the geochemical and isotope variations observed between the plutonic crust and the underlying ultramafic section. The basal ultramafic rocks, i.e. dunites, wehrlites and clinopyroxenites, show high Mg#, extremely depleted REE patterns, no marked HFSE anomalies and high LILE enrichment. Cpx and leached whole rocks from this unit yield restricted εNd values (+ 5.5 to + 6.8), a large εSr range (− 5.9 to + 8.8) and mainly radiogenic 207Pb/204Pb ratios. The overlying mafic section, i.e. gabbroic rocks, displays low Mg# (46 < Mg# < 54), enriched REE patterns and marked HFSE anomalies. These plutonic rocks have also very homogeneous Sr and Pb isotopic compositions defining a restricted domain, which does not overlap that of the ultramafic samples. Moreover, Cpx from the basal ultramafics yield an Sm–Nd isochron at 117 ± 7 Ma interpreted as dating crystallization of the ultramafic section. Together, these results indicate that the mafic rocks on one hand and the ultramafic rocks on the other hand, originate from separate sources with specific emplacement underneath the subduction zone. A lherzolite lens, collected from the top of the mafic section, plots on the early Cretaceous Sm–Nd isochron and therefore yields the same 143Nd/144Nd initial ratio. This suggests that the ultramafic part of the lower Kohistan arc complex sampled the initiation stages of the subduction with production of magmas with boninite-like features. At this stage, the mantle wedge has been already modified by percolation of the first slab-derived fluids. On such a scenario the lherzolite lens would correspond to the lithospheric mantle underneath the arc before its initiation. A multi-stage history is proposed for the evolution for the Kohistan arc through not, vert, similar 30 Ma. Stage #1 reflects the spontaneous initiation of subduction associated with extensive boninitic affinity magmatism, stages #2 and #3 are related to arc building by tholeiitic melts and, a fourth step (stage #4) corresponds to the intra-crustal differentiation.
ISSN: 0012-821X

hal-00407206
https://hal.archives-ouvertes.fr/hal-00407206
DOI : 10.1016/j.epsl.2007.06.026