-
Metamorphic heterogeneities within a single HP unit:
overprint effect or metamorphic mix?
by Romain Bousquet
Citation of this article:
Bousquet R., 2008, Metamorphic heterogeneities within a same HP unit: overprint effect or metamorphic mix? Lithos, 103, 46-69
Abstract
Eclogites and eclogites-facies rocks in mountain belts provide evidence that material in subduction zones material can return from depths of more than 100 km to the surface in the early stages of orogenic processes. Their relationship with lower metamorphic grade rock of i.e. blueschist and greenschist facies can provide information on the late stages of orogenic processes, but are often matter of debate. Petrography of metamorphic rocks in mountain belts has mainly focused on mafic rocks systems. Nevertheless, metamorphic domains of recent mountain belts like the Alps are not only constituted by mafic rocks but also by metasediments that continuously outcrop over very large areas. Such mountain belts are made of a large part of low-temperature metasediments devoid of index minerals classically observed in mafic and quartzo-feldspathic rocks systems, allowing a direct comparison to be made. These metasediments also have various chemical and mineralogical compositions that represent an important geothermobarometric potential.
Thus we propose to study relationships between eclogites and less metamorphosed rocks for a significant case that has been long debate (Entrelor area, Western Alps). Despite metasediments continuously outcropping over the whole area, all previous metamorphic studies were carried out on dismembered sequences of mafic rocks. They have evidenced two kind of metamorphic evolution.
Three explanations are generally proposed to interpret this feature: incertainties of the thermodynamic data, the loss of HP mineralogy during the exhumation and a late-stage tectonic juxtaposition. By studying simultaneously both eclogites and metasedimentary contry-rocks, we show that rocks association in the Entrelor area can be interpreted as a metamorphic mix. This area consists of eclogites rocks embedded in a blueschist facies matrix mainly made of metapelites. Exhumation of HP metamorphic rocks reveals that different pressure peaks (1.2 GPa at 450°C vs. 2.3 GPa at 550°C) were contemporaneous. The different types of rocks have been juxtaposed at a shallow crustal level within a subduction channel.
Due to the fact the Western Alps do not reach the mature stage of a colliding belt, as the Central or as the Eastern Alps, the rocks of the Entrelor area can be viewed as an exhumed part of a frozen subduction channel attributed to a metamorphic mixing of rocks having different metamorphic evolution and accreted at great depths.
1. Introduction
Subduction and collision zones are the sites of recycling of oceanic and continental lithospheres either into the mantle or crust. While the general position of the downgoing slab is revealed by Benioff zone seismicity (e.g., Kirby et al., 1996, England et al., 2004) and seismic tomography (Spakman, 1990) and the temperature field of subduction and collision zones has been simulated based on various assumptions for steady state subduction (e.g., Oxburgh and Turcotte, 1974; England and Thompson, 1984; Thompson and England, 1984; Davy and Gillet, 1986; Peacock, 1990; Bousquet al., 1997; Goffé et al., 2003; Stöckert and Gerya, 2005), the fine-scale structure of a subduction zone is at the limit of or beyond the spatial resolution of geophysical methods. However, an ever increasing number of high-pressure and ultrahigh-pressure metamorphic rocks have been discovered in the last two decades (e. g., Chopin, 2003; O′Brien and Rötzler, 2003). These rocks prove that material returns to shallow crustal levels, and to the surface, from depths of more than 100 km. This happens either continuously or during certain episodes in the lifetime of a subduction zone, with rates corresponding to plate velocity (Duchêne et al., 1997; Gebauer et al., 1996). The record of these rocks reveals information on the pressure-temperature-time paths followed (Engi et al., 2001a; Parra et al., 2002a) and on the state of stress and the mechanisms of deformation (Stöckhert et al., 1999) at depths and on length scales which are otherwise not accessible for direct analysis.
In many mountain belts representing active and ancient convergence regions, such as subduction and collision zones, despite the fact that eclogites and eclogites-facies rocks are common and well studied, their relationship with lower metamorphic grade rocks of i.e. blueschist and greenschist facies is not always well understood. In particular associations of eclogites, blueschists and greenschists are also commonly observed (Bearth, 1973; Oberhänsli, 1980; El-Shazly, 2001). Three main reasons are generally cited to explain this: (i) differences in P-T conditions of equilibration in the rocks and tectonic juxtaposition (Platt, 1992), (ii) different protolith compositions (Oberhänsli, 1982) or (iii) different retrograde overprinting eclogites-facies rocks (Ballèvre, 1988).
If a good characterization of P-T paths is performed, eclogites and associated rock of lesser metamorphic grade in mountain belts can provide significant constraints on the different stages of orogenic evolution. However blueschist- and eclogite-facies rocks are generally described in basic protholiths, largely because of their spectacular field appearance and because of historical description of metamorphic facies (Eskola, 1929). In particular, the metamorphic evolution of mostly low-temperature metamorphic belts is determined for mafic rocks, which often occur in dismembered sequences. In such cases, relationships between the different types of metamorphic rocks are subject to numerous assumptions depending on the regional evolution or particular model.
On the other hand, metasediments outcrop continuously over very large areas. Without carefully considering metamorphic mineralogy of the metasediments, particularly in HP-LT conditions, metamorphic conditions can be misinterpreted (see discussion in Oberhänsli et al., 2003) and thus their geodynamic evolution would be misunderstood (Bousquet et al., 2002). Low-grade metapelites in high-pressure belts that reveal a specific mineralogy are the most emblematic example to illustrate this problem. The recognition of Fe-Mg carpholite in such rocks (Goffé et al., 1973) over huge areas changed the interpretation of geodynamic evolution of many mountain belts (Goffé & Chopin, 1986; Goffé et al., 1988, 1989; Goffé & Bousquet, 1997; Theye et al., 1997; Bousquet et al., 1998; Oberhänsli et al., 2001; Rimmelé et al., 2003).
In this paper, we present a study of the Entrelor area in the Western Alps where relationships between eclogites and lower grade metamorphosed mafic rocks have been discussed for a long time. We carried out detailed work on the metamorphic evolution of metasediments in comparison to the evolution of the mafic rocks to understand their relationships and exhumation processes.
2. Regional geology
2.1 The Alpine context of the field area
In the internal western Alpine nappes, eclogite and blueschist facies imprint led to the recognition of a fossil subduction zone (Ernst 1971; Dal Piaz et al. 1972). Subsequently, it had been shown that the Alpine orogen formed during a long process of convergence and collision between the European and Adria continental domains during Mesozoic and Cenozoic times (Trümpy, 1975; Le Pichon et al., 1988; Escher et al., 1997; Schmid et al., 1996, 2004; Schmid and Kissling, 2000). The intervening oceanic crust progressively deformed and partially accreted to the continental margins, and is now part of the Penninic nappes, which are sandwiched between units of the overlying Austroalpine and the underlying European domains (e.g. Dal Piaz, 1999). The occurrence of high pressure-low temperature metamorphism (HP-LT) in major parts of the oceanic and adjacent continental units is now well known from many parts of the Alps (for review, see Droop et al., 1990; Goffé et al., 2004; Bousquet et al., 2004). Nevertheless, different opinions exist about the age and paleogeographic position of the HP-LT units (i.e., Cretaceous vs. Tertiary, one or two sites of subduction, east- or westward vergence) partly due to conflicting age determinations of HP metamorphism (see Chopin and Maluski, 1980; Hunziker et al., 1992; Duchêne et al., 1997; Gebauer, 1999) and due to a fragmentary view of each paleogeographic domain (see discussion in Bousquet et al., 2002).
The Piemont-Ligurian zone in the Western Alps is classically divided into two units (Fig. 1), the Tsaté nappe (or Combin zone) and the Zermatt-Saas nappe (e.g., Sartori & Thélin, 1987; Droop et al., 1990; Dal Piaz, 1999 and references therein), separated by a major extensional fault (Philippot, 1990; Ballèvre & Merle, 1993; Reddy et al., 2003). The distinction between both units was based both on lithostratigraphic (Bearth, 1962; Marthaler, 1984; Marthaler & Stampfli, 1989) and metamorphic differences (Dal Piaz, 1965; Kienast, 1973; Caby et al., 1978; Michard et al., 1996: Schwartz et al., 2000).
The lowermost unit, the Zermatt-Saas nappe, is composed mainly of mafic and ultramafic ophiolites displaying an oceanic affinity. Since Bearth′s famous work, this nappe is well known for its high-pressure mineral assemblages (Bearth, 1967; Ernst & Dal Piaz, 1978; Chinner & Dixon, 1973; Lombardo et al., 1978; Oberhänsli et al., 1980; Pognante & Kienast, 1987). The discovery of coesite inclusions within garnet in some Mn-bearing metasediments in Lago di Cignana suggests that some parts of the Piemont-Ligurian were deeply subducted up to 2.8 GPa at 600°C (Reinecke, 1991).
The uppermost unit, the Tsaté nappe (Marthaler and Stampfli, 1989), is an oceanic unit dominated by carbonate and terrigeneous calcschists, alternating with tholeiitic metabasalts. The lack of eclogites and the occurrence of relic sodic amphiboles in metabasites (Dal Piaz & Ernst, 1978; Ayrton et al., 1982; Sperlich, 1988) as well as Mn-rich quartzitic schists associated with Mn-rich garnets (Dal Piaz, 1979; Caby, 1981) have led workers to consider the Tsaté nappe to have an overall greenschist-blueschist metamorphic evolution.
2.2 The Entrelor question
 Figure 2: Detailed view of the Entrelor area. a) Detailed geological map after Dal Piaz (1928), Cigolini (1992), Ballèvre (1988) and this study. b) View from the Entrelor pass of the Piemont-Ligurian area pinched between two parts of the Briançonnais domain represented by the Internal Briançonnais (left) and the Gran Paradiso massif (right) |
The problem of the relationship between the different types of metamorphic rocks is particularly acute in the Entrelor area. This area represents a small piece (~1-2 km wide) of the Piemont-Ligurian domain, pinched between two continental basements (Fig. 2). The underlying massif, the Gran Paradiso unit, consists mainly of abundant augengneisses derived from porphyritic granitoids of Late-Variscan age and metasediments with relics of high-temperature metamorphism of pre-Alpine age (Callegari et al., 1969; Le Bayon & Ballèvre, 2004). The overlying massif, the internal zone of the Briançonnais nappe, is made up of paragneisses and micaschists with a polymetamorphic history (Boquet 1974, Cigolini 1995), and of a mono-metamorphic sequence consisting of lower Permian to Mesozoic formations (Ellenberger, 1958; Elter, 1972). While the Gran Paradiso massif displays a metamorphic evolution into eclogite facies conditions (Compagnoni & Lombardo, 1974; Ballèvre, 1988; Borghi et al., 1996), the rocks of internal Briançonnais have only reached upper blueschist facies conditions (Cigolini, 1995; Bucher and Bousquet, 2006).
 Figure 3: Different models proposed for the Entrelor area. a) Metamorphic maps based on studies of the metabasites rocks: a1- The Entrelor area is interpreted as a main contact zone between eclogite and a less metamorphosed unit (Droop et al., 1990; Dal Piaz, 1999). a2- All rocks of the Entrelor area were buried at the same depth, reaching eclogite facies conditions (Ballèvre and Merle, 1993, Frey et al., 1999). b) Tectonic interpretations for the Entrelor shear zone, (EZS) which is separating the Piemont-Ligurian domain from the Internal Briançonnais. b1- The EZS is a backthrust acting top-to-the east as a major shear zone at the end of the evolution of the Western Alps and thus preserves HP- assemblages (Buttler and Freeman, 1996). b2- The EZS is a top-to-the northwest extensional ductile fault that assisted exhumation of HP rocks (Caby, 1996). b3- The EZS, showing apparent top-to-the northwest sense of shear, is interpreted as a refolded tectonic contact during a large-scale refolding event in the Western Alps (Bucher et al., 2003, Bucher and Bousquet, 2007). |
Although the metamorphic evolution of both surrounding basement units is relatively well constrained, evolution of the Piemont-Ligurian nappe is subject to different interpretations. This nappe is made up of carbonates and terrigenous metasediments (calcschists) interleaved with slices (at different scales) of metabasalts. Detailed mapping (Dal Piaz 1928; Cigolini, 1992) in this area reveals several types of basic rocks (Fig. 2): eclogites and greenstones that are spatially separated. This led to two main interpretations (Fig 3a). The first one assumes that greenstones and eclogites resulted from different metamorphic evolution and considers the Entrelor area as a contact zone between an HP unit (eclogite conditions facies) and a less metamorphosed unit (blueschist-greenschist conditions facies) (Elter, 1972, Droop et al., 1990; Dal Piaz, 1999). The second interpretation considers the greenstones as old eclogitic rocks totally retrogressed; in this case, the Entrelor area must be considered as part of the Zermatt-Saas unit (Ballèvre and Merle, 1993, Frey et al., 1999; Le Bayon & Ballèvre, 2004).
The contact between the Piemont-Ligurian nappe and the internal Briançonnais is also subject to debate (Fig. 3b). For some authors, this contact places greenschist facies Briançonnais rocks onto an old subduction complex of oceanic (Piemont-Ligurian ocean) and continental (Gran Paradiso) rocks and thus preserves eclogitic assemblages (b1, Buttler and Freeman, 1996; Freeman et al., 1997). However, Caby (1996) described this contact as an extensional ductile fault that assisted exhumation of blueschist rocks of the Piemont-Ligurian realm (b2). Recently, Bucher et al. (2003) demonstrated that this contact, which displays an apparent top-to-the-west sense of shear, is a refolded thrust (b3). The refolding around a NE-SW-oriented axis, approximately perpendicular to stretching lineation, that is associated with top-NW shear-senses led to the preservation of thrust-related kinematic indicators, even where former nappe contacts are overturned but the original metamorphic contacts are preserved (Bucher et al., 2004).
To decipher the metamorphic evolution of this key area of the Western Alps and to evaluate the different models proposed, we carried out detailed petrological work on all rock types.
3. Petrology and PT estimates
3.1 Methods of investigation
3.1.1 Mineral and whole rock chemistry
 Table 1: Bulk rocks analyses of some representative rocks of the Entrelor area |
The mineral compositions were determined with a JEOL JXA-8600 electron microprobe at the University of Basel (15 kV, 10 nA, PROZA correction procedure) using wollastonite (Si,Ca), albite (Na, Al), graftonite (Mn, Fe), rutile (Ti), albite (Na), orthoclase (K), Olivine (Mg) as standards.
Bulk rock compositions (Table 1) were determined on melted pellets by XRF using a Brucker AXS SRS-3400 at the Geochemical Laboratory in Basel. Bulk rock chemistries were measured on rocks in which no major layering was observed in thin sections.
3.1.2 P-T estimates methods
In order to understand the metamorphic history of the different rock types, we used two thermodynamic approaches. The first one is based on the notion of ″local equilibrium″ and the second method is based on the notion of ″bulk rock equilibrium″.
The first calculation procedure is based on the method of Berman (1991) that allows simultaneous estimates of P and T using phases that are present in the studied thin sections to check for equilibrium. The mineral assemblages used to perform the calculations were first selected using classical micro-textural criteria suggesting equilibrium state (Trotet et al., 2001; Parra et al., 2002a, Vidal et al., 2006). Assuming that the standard state thermodynamic data as well as the activity-composition relationships are well calibrated, all the reactions computed for a given mineral assemblages should intersect at a single point in the P-T field if equilibrium is achieved. Knowing that reactions may be incomplete at the rock scale, this method allows deciphering different metamorphic stages in one sample.
The second method computes stable assemblages, including mode and composition of solution phases, for specific bulk rock compositions using the THERIAK-DOMINO software (De Capitani, 1994, http://titan.minpet.unibas.ch/minpet/theriak/theruser.html). This software allows to calculate and plot thermodynamic functions, equilibrium assemblages and rock-specific equilibrium assemblage diagrams (also called pseudo-sections). Based on its approach to equilibrium by means of Gibbs free energy minimization (c.f. de Capitani & Brown, 1987) rather than solving complex and large equation systems, the Theriak-Domino software computes assemblage diagram comparable to the others approaches (see Hoschek, 2004). In such equilibrium phase diagrams, all phases are considered for each point of the P-T diagram assuming a complete thermodynamic equilibrium for the whole rock. Isochemical P-T phase-diagram sections provide important information on mineral-assemblage stability in P-T space. In all studied rocks, except for chemical zoning of some garnets, no significant srelic of the burying evolution has been observed. Thus we assume that the rocks were fully equilibrated at the pressure peak of their history. Only in such cases, we use the ″bulk rock equilibrium″ method to constrain the pressure-peak.
The updated JAN92.RGB thermodynamic database of Berman (1988) was used for all calculations, completed by the following thermodynamic data: Mg-chloritoid data of B. Patrick (listed in Goffé and Bousquet, 1997), Fe-chloritoid data of Vidal et al. (1994), chlorite data of Hunziker (2003), glaucophane data of Potel et al. (2002) and alumino-celadonite data from Massonne and Szpurka (1997). Mineral activities were calculated according to the model of Evans (1990) for glaucophane, according to the solid solution model from Parra et al. (2002b) for phengite and from Hunziker (2003) for Chlorite.
3.2 Mafic rocks
 Figure 4: Main mineralogical assemblages occurring in the mafic rocks of the Entrelor area. |
A complex set of heterogeneous and dismembered mafic and ultramafic rocks (Dal Piaz, 1928, Elter, 1972) displaying a close oceanic affinity (Dal Piaz, 1974) occurs within a matrix of metasediment in the Entrelor area. This complex is composed of serpentinites and serpentine schists, some rare amphibole-bearing metagabbros, retrogressed glaucophanites and eclogites, (garnet)- amphibolites and green- and talc-schists. Retrogressed eclogites and garnet amphibolites seem to be spatially separated from greenstone rocks (Fig. 4). Retrogressed eclogites occur mostly in the structurally lower part just above the HP gneisses of the Gran Paradiso unit. Greenstone rocks occur structurally above the amphibolite and below the contact with continental rocks of the Briançonnais domain (Internal Zone). While any greenstone has been limited the lower structural part, boudins of well-preserved eclogite have been found all over the area from the contact with the Gran Paradiso massif to the contact with the Internal Zone (Fig. 4).
3.2.1 Petrology of greenstone
 Table 2: Representative mineral assemblages occurring in the greenstones The structural formulae were calculated for Na-amphibole on 13 cations and on 23 oxygens (Robinson et al., 1982), for amphiboles on 15 oxygens and for clinozoisite on 12.5 oxygens (Deer et al., 1992). *for Clz, all iron is Fe2O3 |
Greenstones in the Entrelor area are prasinites and ovardites, with some intercalations of talc-schists. They are mainly composed of green minerals such as chlorite and actinolite and of albite, quartz, clinozoisite, epidote, titanite and some fine needles of biotite. These rocks occur as weakly deformed rocks metamorphosed in greenschist facies conditions. Amphibole and epidote form the main schistosity, while biotite associated with some feldspath grew in fine shear-bands cross-cutting the main schistosity underlined by clinozoisite and amphiboles.
Evidences of high-pressure metamorphic conditions are scare. Some blue amphiboles were found as inclusions in epidote or albite (Fig. 5a, Table 2). These Na-amphiboles have an intermediate composition (0.30 ≤ XFe3+ ≤ 0.55 and 0.18 ≤ XFe3+ ≤ 0.30), a typical crossite-like composition (Fig. 6a).
 Figure 5: HP-mineral occurring in mafic rocks. a) Only scarce inclusions of glaucophane are witness of an earlier HP-evolution in greenstones. b) Some mafic rocks preserve uncommon assemblages in chlorite-epidote rich-rocks. Such assemblage, composed of chloritoid, glaucophane, garnet, omphacite and talcclearly predates the main schistosity. |
 Figure 6: Chemistry of the HP-index minerals observed in mafic rocks. Star symbols represent analyses for the greenstones, black diamonds for eclogite type I, grey squares for eclogite type II, and open circles for eclogite type III. c) Chemistry of Na-amphiboles. We note that the chemical composition of Na-amphiboles in eclogite differs noticeably from those of the greenstone. Na-amphiboles from greenstones have a crossite-type composition and Na-amphiboles from eclogites have a glaucophane-type composition. Composition of glaucophane from eclogite-type II is not homogeneous, contrary to glaucophane from eclogite-type III. a) Chemistry of clinopyronexes of the different eclogites-types. All composition show a jadeite-content around 50%, with slight variations according to the eclogite type. In eclogite-type I, omphacites contain less than 50% of jadeite while in eclogite-type II & III they contain slightly more jadeite. b) Chemistry of garnet of the different eclogites-types. All garnets are almandine-rich, without significant variation in their composition. |
3.2.2 Petrology of (retro)eclogites and amphibolites
All eclogites and garnet amphibolites occur as variably sized blocks or lenses within calcschists and form the most important mass of mafic rocks. Based on mineral assemblages, three types of eclogites were recognized. Type I, representing the main part, is eclogite with a peak-stage Grt + Omp + Rt assemblage. The composition of clinopyroxene is Fe3+-rich (around 40% of acmite component, Fig. 6b) and the jadeite-content is less than 50%. Garnet composition does not show chemical zoning, and is almost constant (Fig. 6c). These rocks are strongly overprinted by three retrogression stages. During the first stage, glaucophane and clinozoisite grew at the expense of garnet and clinopyroxene, likely at epidote-blueschist facies conditions. Chemical composition of such glaucophane shows slight variation of Fe2+ and Fe3+ content (Fig. 6b). Glaucophane and clinozoite are oriented along the main foliation. Assemblages of actinolite and paragonite surrounding glaucophane minerals mark the second stage of retrogression. The breakdown of glaucophane suggests a pressure decrease. The third retrogression stage is indicated by some occurrences of chlorite-epidote-albite assemblages at greenschist facies conditions.
 Table 3: Representative mineral assemblages occurring in the eclogites The structural formulae were calculated for chloritoid on 12 oxygens and Fe3+ is calculated following Chopin et al. (1992), for Na-amphibole on 13 cations and on 23 oxygens (Robinson et al., 1982), for garnet on 12 oxygens, and Fe3+ is calculated from the deficit in Al into octahedral site, for clinopyroxene on 4 cations and on 6 oxygens and for talc on 22 oxygens (Deer et al., 1992). |
Type II is well-preserved eclogite, and mainly made up of clinopyroxene, garnet, zoisite, rutile and oxydes (ilmenite, magnetite). No preferred mineral orientations have been observed in these rocks. Some of the eclogites are gabbros occurring as massive, coarse-grained rocks with garnet, omphacite and relics of magmatic pyroxene. The subophitic magmatic texture is well preserved. Prograde reactions such as magmatic plagioclase and clinopyroxene replaced by an assemblage of garnet and omphacite can also be observed. Such structures have also been described east of the Gran Paradiso (Benciolini et al., 1984).
The composition of clinopyroxene indicated a slight increase of the jadeite-content by nearly 50% (Fig. 6b) in comparison with type I. Garnets have an almandine-rich composition (Fig. 6c). Any chemical zoning within garnet, omphacite have been evidenced. Veins up to 30 cm large and containing quartz, clinozoisite, tremolite, phengite, calcite and dolomite are crosscutting well-preserved eclogite bodies.
Type III is talc-eclogite that occurs as garnet and chlorite-rich lenses within eclogites, mainly of type I. Less intensively deformed rocks contain unusual mineral assemblages formed by chloritoid, talc, omphacite, garnet, and glaucophane (Fig. 6, Table 3). Compositions of omphacite and of garnet are similar to those of type II (Fig. 6b &c). Na-amphiboles seem to have a constant composition, slightly more glaucophane-rich. These minerals occur as relics in a matrix made of paragonite, chlorite and clinozoisite. It seems that these rocks were altered prior to HP-metamorphism. They thus favour recrystallization of Al-rich phases (Pawlig & Baumgartner, 2001).
Despite several studies carried out on metabasites of the Piemont-Ligurian ocean north of Gran Paradiso (e.g. Tartarotti et al., 1986, Martin and Kienast, 1987; Martin and Tartarotti, 1989), PT estimates from these rocks are scare. Benciolini et al. (1984), used the garnet-pyroxene equilibrium to estimate temperatures between 466 and 530°C for pressures between 0.87 and 1.17 GPa. Knowing that to the north (Reinecke, 1991) as in the south (Lombardo and Pognante, 1982), rocks from the Piemont-Ligurian ocean were buried at greater depth, we reinvestigated the PT conditions of mafic rocks from the Entrelor area, using mineralogical data described above.
 Figure 7: P-T conditions estimated for the mafic rocks. a) Estimates for the eclogites. The ″local equilibrium″ method [a1] and the ″bulk rock equilibrium″ method [a2] indicate that assemblages occurring in the eclogites type III were formed at around 2.5 GPa and 550 °C. Veins containing newly formed calcite and dolomite clearly post-dating the HP event by cross-cutting eclogites type II indicate PT conditions around 1.5-1.3 GPa and 550-500°C, depending on the XCO2. Eclogites were exhumed by isothermal decompression. b) The peak pressure for the greenstones was estimated at about 1.0 GPa and 400°C (″local equilibrium″ method [b1], ″bulk rock equilibrium″ [b2]) and a re-heating up to 500°C during exhumation can be observed. Both methods indicate similar results. |
The local equilibrium method of Berman (1991) applied to the talc-chloritoid bearing rocks yielded pressure estimates around 2.5 GPa for temperature around 550°C (Fig. 7a1), while the ″bulk rock″ method of De Capitani (1994) for the same rock describes a small stability field for the Ctd-Tc-Gt-Omph-Pg assemblage, at 2.3 GPa for the same temperature (Fig. 7a2). Using the mineralogy of the veins crosscutting the eclogites, from local equilibrium calculations we obtain conditions around 1.3-1.5 GPa for temperature ranging between 500-550°C (Fig. 7a1). Thus eclogites were buried to great depth along a cold gradient (less than 8°C/km) and were isothermally exhumed to crustal depths.
PT conditions estimated for the greenstones with both methods differ significantly from those of the eclogites and they indicate that the greenstones were buried at shallower depths (around 1.0 GPa) and colder temperatures (400°C, Fig 6b1 & 6b2). Occurrences of biotite with feldspath, clinozoisite and chlorite constrain the late evolution of these rocks. We note heating up to 500°C during exhumation along a clockwise P-T path (Fig. 7b1).
Knowing that both rock types (eclogites and greenstones) are today exposed in the same area, the significance of the PT estimations can be questioned. Did both types of rocks experience two different metamorphic evolutions? Or are greenstones representing fully retrogressed old eclogites? In order to get more information and to answer these questions, we will present a detailed study of the metasediments in which these metabasites occur.
Metasediments
 Figure 8: Main mineralogical assemblages occurring in the metasediments of the Entrelor area. |
The metasediments form the country-rocks of the Entrelor area between the Gran Paradiso massif and the internal Briançonnais (Fig. 2). They are mainly composed of calcschist, sometimes interlayered with metapelitic or metamarly beds, and they represent the typical Schistes Lustrés described all over the Western Alps (Marthalet & Stampfli, 1989; Deville et al., 1992).
While no metamorphic studies were carried out on the basic ″Schistes Lustrés″ north of the Gran Paradiso, those occurring in the south display a metamorphic evolution at blueschist conditions as indicated by the occurrence of (Fe, Mg)-carpholite (Goffé and Chopin, 1986, Agard et al., 2001, Goffé et al., 2004) in calcschists and of garnet-lawsonite-glaucophane assemblage in marbles (Ballèvre and Lagabrielle, 1994).
4.1 Petrology
We focus our study on calcschist that continuously occur over the whole area. As usual in such calcschist, metamorphic mineralogy is poor compared to metabasites. The rocks are mainly composed of calcite, quartz, phengite, and variable amounts of chlorite, while occurrences of graphite are restricted to the metapelitic layers. The preferred orientation of phengite and quartz in the calcite-quartz matrix defines a penetrative foliation in calcschist. Quartz shows undulose extinction and grain boundary migration. Despite the lack of index minerals indicating HP conditions such as (Fe, Mg)-carpholite, glaucophane or jadeite, occurrences of garnet, chloritoid, phengite and paragonite testify to a HP metamorphic evolution for the metasediments (Fig. 8, Table 4). While no geographic correlation of mineral assemblages is observed (Fig. 8), it has been noticed for microstructures: while the main schistosity is expressed by phengite, chlorite, clinozoisite, quartz, plagioclase (mainly albite) and small garnet, chloritoid and big garnet associated with phengite occur as porphyroblasts (Fig. 9). Chloritoid associated with phengite, quartz and paragonite occurs also as inclusions in big garnets. Such inclusions are sometimes forming an old schistosity (Bucher et al., 2004), or they are rotated.
 Table 2: Representative mineral assemblages occurring in the greenstones The structural formulae were calculated for Na-amphibole on 13 cations and on 23 oxygens (Robinson et al., 1982), for amphiboles on 15 oxygens and for clinozoisite on 12.5 oxygens (Deer et al., 1992). *for Clz, all iron is Fe2O3 |
 Figure 9: HP-mineral occurring in calcschist. a) Example of microstructures commonly observed in metasediments (Vaud991): Porphyroblasts of chloritoid and garnet surrounded by phengite, chlorite and clinozoisite in a strongly deformed rock. We note occurrences of different generations of garnet and phengite, according to their texture and size. b) Chemical profile across chloritoid and garnet. Chloritoid and garnet II do not show significant chemical zoning in their composition. |
In contrast to chloritoids that lack chemical zoning, we note variable chemical compositions for garnet. According to their microstructural site, garnets show different chemical patterns (Fig. 10). Large garnets (100 μm) that occur as porphyroblasts are zoned, with a spessartine component decreasing from core to rim and almandine content increasing from core to rim (type I ″normal zoning″, Fig. 10a). Zoning in the grossular and pyrope component is weak. This garnet zoning is interpreted to result from primary growth at increasing temperatures (Spear, 1997; Tinkham and Ghent, 2005). The external rim of large garnets contains only inclusions of quartz, while the core contains inclusions of chloritoid, phengite and quartz.
 Figure 10: Chemical composition of the garnet found in metasediments. Three types of garnet have been observed according to their chemistry and the microstructural site. Type 1 is composed of big garnets (~125 μm) occurring as porphyroblast and showing ″normal″ chemical zoning; the spessartine component decreases from core to rim, and almandine content increasing from core to rim. Type 2 is composed of relatively small garnet (50μm) displaying no chemical zoning. The smallest garnets (20μm), type 3, occurring in young shear zones, have an ″inverse″ chemical zoning: almandine component decreasing from core to rim, and spessartine content increasing from core to rim. |
The second type of garnet does not show chemical zoning (type II ″no zoning″, Fig. 10b). These garnets are small (30 to 50 μm), and occur within the main schistosity. Their chemical composition is the same as the chemical composition of the external rim of the large garnets. The third type of garnet occurs associated with chlorite and phengite in shear zones that post-date the main schistosity. These garnets are chemically zoned with almandine component decreasing from core to rim, and spessartine content increasing from core to rim (type III ″inverse zoning″, Fig. 10c). We interpret this garnet zoning as resulting from growth at decreasing temperatures.
Margarite is observed in samples lacking garnet (Fig. 8, Table 4). In such samples, the HP mineral assemblage is composed of chloritoid, zoisite, phengite and opaque minerals. Margarite is growing around chloritoid in association with phengite and chlorite.
4.2 P-T estimates
We used mineral assemblages with garnet to constrain the metamorphic evolution of the metasediments. We present results from multi-equilibrium calculations performed on samples representative of each metamorphic stage (Fig. 11).
The local equilibrium method is applied on garnet type I (core), chloritoid, paragonite, phengite I and quartz assemblage documents a HP stage around 1.4 GPa and 450°C (Fig. 11a). ″Bulk rock″ calculation also estimates typical blueschist facies conditions (1.2-1.5 GPa, 450-530°C) for chloritoid, garnet, phengite I assemblages (Fig. 11b).
The assemblage composed of garnet type II, chlorite, phengite II, plagioclase, and clinozoisite is stable at HP amphibolite conditions around 0.7-0.8 GPa and 530°C (Fig. 11c). These conditions indicate significant heating during exhumation of metasediments. This heating is also confirmed by type III garnet occurring in late shear zones. Associated with phengite and chlorite, they are stable at LP amphibolites conditions around 0.4-0.5 GPa and 500°C (Fig. 11c).
 Figure 11: P-T conditions estimated from the mineral assemblages observed in calcschists. All investigated rocks were graphite-free. a) PT estimates using the multi equilibrium technique. a1) PT estimates (1.4 GPa, 450°C) for the assemblage garnet (type 1)-chloritoid-phengite-quartz-paragonite. a2) PT estimates for post pressure-peak assemblages containing garnets type II and III and associated minerals. b) P-T estimate using the ″bulk rock″ method. In this case, only the pressure-peak can be computed due to the fact that many HP-relics are always stable in the rock. Both methods indicate similar results. |
Discussion
5.1 Eclogite facies in metasediments?
Results of our calculations indicate that eclogites and metasedimentary rocks did not record similar peak P-T conditions. Lower PT conditions (blueschist facies conditions) have been recorded within country rocks (metasediments) relatively to the eclogitic slices. Three explanations can be suggested to interpret these observations: uncertainty of the thermodynamic data, the loss of HP mineralogy during the exhumation and a late-stage tectonic juxtaposition of metapelites and eclogites.
5.1.1 Thermodynamic?
The same thermodynamic database was used for all PT estimates. Difference in pressure of around 1 GPa cannot only be a consequence of thermodynamic inaccuracies. On the other hand these inaccuracies can eventually affect the pressure-peak estimate, but certainly not the pressure gap.
5.1.2 Retrogression?
Some workers suggest that metapelites dehydrate during exhumation (e.g. Heinrich, 1982, Ballèvre 1988) and then fluids will destroy HP mineral assemblages. However, occurrences of chloritoid as porphyroblast or as inclusion in garnet indicate that some HP imprint can be preserved despite the exhumation. Field observations (Vuichard and Ballèvre, 1988; Agard et al., 2001), experiments (Chopin and Schreyer, 1983; Vidal et al., 1994) and thermodynamic calculations (Proyer, 2003) indicate that chloritoid-garnet is the main assemblage at HP conditions for temperatures exceeding 450°C in Na-poor metasediments. Despite an important overprint during the exhumation, metasediments of the Entrelor preserved some petrological information of the prograde evolution. We are convinced as it has been recently demonstrated (Brosse et al., 2002; Song et al. 2007) that Na-poor metasediments can also display a specific mineralogy comparable to that of the eclogites if the need arises.
5.1.3 Tectonic evolution?
PT conditions estimated for the rocks of the Entrelor area reveal an important tectonic dilemma. How can rocks having the same protolith (greestones and eclogites) displaying different metamorphic evolution occur in the same area? For the Entrelor area and its neighbour regions, this dilemma was interpreted either as two metamorphic units separated by a shear zone (Dal Piaz, 1999) or as one unit in which greenstones are representing fully retrogressed eclogites (Ballèvre, and Merle, 1993). Finding of well-preserved eclogites near the greenstones invalidates the first model, while lack of minerals or pseudomorphs of garnet or lawsonite as well thermodynamic calculations argue against the second interpretation.
In the Entrelor area, metasediments are forming the country-rocks for the different bodies of metabasite. While the size of greenstones bodies does not exceed 20-30 m, the size of eclogitic bodies is varying from some meters to about a hundred meters (see figure 2, the Grivola mountain). The metamorphic evolution of the metasediments is slightly different from the evolution of eclogites and the greenstones, accrediting the assumption of different metamorphic evolutions for the rocks of the Entrelor area. While further to the south, in the ″Schistes Lustrés″ and associated basic rocks, blueschists (in the west) and eclogites (in the east), are spatially separated (Goffé et al., 2004), blueschist and eclogites in the Entrelor area are mixed. This suggests a common prograde evolution of the Schistes Lustrés and associated basic rocks all over the Western Alps as an accretionary wedge (Agard et al., 2001), but mechanisms of exhumation seem to be different north and south of the Gran Paradiso.
5.2 Metamorphic mix?
 Figure 12: New metamorphic map of the Western Alps (after Oberhänsli et al., 2004). The contact between rocks buried at eclogite facies conditions and those buried only at blueschist facies conditions is gradual. We note that in addition to the Entrelor area, another area (north of Acceglio massif) is displaying lower PT conditions recorded within country rocks (metasediments) relative to the mafic slices (Goffé et al., 2004). Despite the debate on the age of the peak pressure for the rocks of the Piemont-Ligurian domain, the ages obtained on phengite for the early exhumation stage are coherent for all types of rocks (from a- Duchêne et al., 1997; b- Agard et al., 2002; c- Meffan-Main et al., 2004; d- Dal Piaz et al., 2001; e- Bucher, 2003). PT paths: A- this study. B- Agard et al., 2001; C- Ballèvre and Lagabrielle, 1994; D- Lombardo et al., 1978; Schwartz et al., 2000. |
The metamorphic pattern described in the Entrelor area is generally interpreted as a mélange zone (e.g. Cloos, 1984; Parkinson, 1996), described as ″mappable bodies of fragmented and mixed blocks in a matrix″ (Silver and Beutner, 1980). However, the mechanisms of the exhumation of HP rocks in a mélange zone remain a matter of debate and a number of feasible models have been proposed (e.g., Cloos, 1982; Platt, 1993; Ring et al., 1999). In the corner flow model, originally proposed by Hsu (1971), and further developed (Cloos, 1982; Cloos and Shreve, 1988a, 1988b; Shreve and Cloos, 1986; Gerya et al., 2002a), exhumation of high-pressure metamorphic crustal slices at rates on the order of plate velocity is driven by forced flow in a wedge-shaped subduction channel. The observation that many exhumed HP rocks can remain poorly deformed, implying a low bulk viscosity of the material in the subduction channel, is taken to support the concept of forced flow (Burov et al., 2001; Engi et al., 2001b). In several models serpentinite is proposed as a low viscosity layer allowing exhumation of HP rocks (Coleman, 1961; Schwartz et al., 2001) in subduction channels by hydration of the mantle wedge (Guillot et al. 2001; Gerya et al., 2002b; Arcay et al., 2005, 2006). In the Entrelor area, although few serpentinites are present the low viscosity of the metasediments relative to the others rocks is consistent with a subduction channel model. The intensity of mixing in a channel is a function of viscosity contrast between the channel and the surrounding rocks: the higher the viscosity contrast, the more important the mixing (Gerya et al., 2002b).
While the age of the HP event in the Western Alps is matter of debate and seems to be dependent on the rock types (Duchêne et al., 1997; Cliff et al., 1998; Perchuk and Philippot, 2000; Agard et al., 2002), the exhumation history is better constrained. The early stage of exhumation, still at blueschist facies conditions was dated around 45 Ma (Dal Piaz et al., 2001; Agard et al., 2002; Meffan-Main et al., 2004; Fig. 12), the greenschist overprint around 35 Ma (Agard et al., 2002; Bucher, 2003) and the latest exhumation stages documented by fission-track ages on zircon and apatite have been dated at 30 Ma and 20 Ma respectively (Fügenschuh and Schmid, 2004; Malusà et al., 2005).
 Figure 12: New metamorphic map of the Western Alps (after Oberhänsli et al., 2004). The contact between rocks buried at eclogite facies conditions and those buried only at blueschist facies conditions is gradual. We note that in addition to the Entrelor area, another area (north of Acceglio massif) is displaying lower PT conditions recorded within country rocks (metasediments) relative to the mafic slices (Goffé et al., 2004). Despite the debate on the age of the peak pressure for the rocks of the Piemont-Ligurian domain, the ages obtained on phengite for the early exhumation stage are coherent for all types of rocks (from a- Duchêne et al., 1997; b- Agard et al., 2002; c- Meffan-Main et al., 2004; d- Dal Piaz et al., 2001; e- Bucher, 2003). PT paths: A- this study. B- Agard et al., 2001; C- Ballèvre and Lagabrielle, 1994; D- Lombardo et al., 1978; Schwartz et al., 2000. |
We propose that our findings (different P-T paths, heating and cooling events at moderately decreasing pressures, similar ages for the exhumation history) can be explained best by mass flow in a subduction channel environment. This environment implies that the assemblage of HP rocks of the Entrelor area, eclogites, greenstones, and metasediments, cannot represent a crustal section that was already coherent at HP conditions as commonly believed. The coherency was attained during significant exhumation of these HP rocks possibly at a depth close to 30 km corresponding to 1 GPa at about 45 Ma (Fig. 13). The mainspring for the exhumation of different rock types could be the extrusion of a buoyant body (Gran Paradiso) composed of poorly deformed granitic rocks and gneiss. This rigid body acts as a bulldozer pushing up many types of rocks to the surface during its exhumation. Such a mechanism allows exhumation of HP-rocks at relatively low subduction rates (around 2 cm/yr, Gorczyk et al., 2006) as in the Alps (Schmid et al., 1997).
A similar mechanism of exhumation of eclogites, related to the extrusion of continental crust was already proposed for the Central Alps (Engi et al., 2001b; Berger et al., 2005), and seems to play an important role in the dynamics of the Alpine belt.
Conclusions
The assemblage of rocks in the Entrelor area is interpreted as a metamorphic mixture, consisting of eclogites rocks embedded in a blueschist facies matrix consisting of metapelites and greenstones. Exhumation of HP metamorphic rocks revealing different peak pressures (1.2 GPa at 450°C vs. 2.3 GPa at 550°C) was contemporaneous. They have been finally juxtaposed at a shallow crustal level within a subduction channel. Subduction processes cannot be considered as a single pass process; instead, return flow of a considerable portion of crustal and upper mantle material must be accounted for (Gerya et al., 2002a) and the exhumation of different types of rocks can not be considered independently of each other (Engi et al., 2001b). Because the Western Alps did not reach the mature stage of a colliding belt, as the Central and Eastern Alps, the rocks of the Entrelor area can be viewed as an exhumed part of a frozen subduction channel described as a metamorphic mixture of rocks with different metamorphic evolution, and accreted at depths.
Acknowledgments
S.M. Schmid is thanked for fruitful discussions and for its enthusiasm for the Alpine geology. Geological and petrological knowledge of C. Cigolini were useful to improve this work. I feel grateful to him for his help. This work has benefited from many comments by B. Fügenschuh, S. Bucher, R. Caby, R. Polino and A. Loprieno. Despite our difference of opinion, I would thank G. V Dal Piaz for his remarks and his criticisms that forced me to improve my thinking. The patience and encouragements of Taras Gerya to write this paper were really helpful. Reviews of A. Perchuk and T. Gerya significantly improve the manuscript. Most of the funding was supplied by the Swiss National Science Foundation (projects 20-55559.98 and 2000-63391.00). All this work would not be possible without permit to work in the Gran Paradiso national park. I thank all the personal of the park for their support.
References
Download references as file for bibliographic reference manager: endnote or Bibtex