-
Subduction related metamorphism in the Alps: Review of isotopic ages based on petrology and their geodynamic consequences
by Alfons Berger(1) and Romain Bousquet(2)
Citation of this article:
Bousquet R., Oberhänsli R., Goffé B., Wiederkehr M., Koller F., Schmid S. M., Schuster R., Engi M., Berger A. & Martinotti G., 2008 Metamorphism of metasediments in the scale of an orogen: A key to the Tertiary geodynamic evolution of the Alps. in Tectonic Aspects of the Alpine-Dinaride-Carpathian System Eds edited by Siegesmund, S., Fügenschuh, B. & Froitzheim, N. Geological Society, London, Special Publications, 298, 393-412
Abstract
We summarize ages of the high pressure/low temperature (HP/LT) metamorphic evolution of the Central and the Western Alps. The individual isotopic mineral ages are interpreted to represent either: (1) early growth of metamorphic minerals on the prograde path, (2) timing close to peak metamorphism or (3) retrograde resetting of the chronometers at still elevated pressures. Therefore, each individual age cannot easily be transferred to a geodynamic setting at a certain time. These different data indicate a subduction related metamorphism between 62 and 35 Ma in different units (e.g. Voltri Massif, Schistes Lustrés of the Western Alps, Tauern window). Oceanic and continental basement units show isotope ages related to eclogitic or blueschist facies metamorphism between 75 and 40 Ma. Most of these ages may record equilibration along the retrograde path, except of some Lu/Hf garnet ages and some zircon SHRIMP ages, which provide information on the prograde path. These different isotope ages are interpreted as different steps along pressure-time paths and so may provide some information on the geodynamic evolution. The data record a continuous subduction, which is ongoing for several tens of millions years. In a large-scale picture, we have to assume fragmentation of the down-going plate in order to explain the available P-T and t data. This interpretation questions the ongoing driving force for subduction during the disappearance of the Alpine Tethys.
Introduction
The Alpine belt formed during a process of convergence, subduction and collision between the European and Adria continental plate during Mesozoic and Cenozoic times (e.g. Ernst 1971; Dal Piaz et al. 1972; 2003; Trümpy 1975; Le Pichon et al. 1988; Dewey et al. 1989). The intervening oceanic crust progressively deformed and partially accreted to the continental margins, which are now sandwiched between the overlying Austroalpine nappe units and the underlying rocks of the European domains. The occurrence of high pressure-low temperature metamorphism (HP/LT) of the oceanic and some adjacent continental units is now well known from many parts of the Alps (see for review Niggli 1978; Goffé & Chopin 1986; Frey et al. 1999; Oberhänsli et al. 2004). Nevertheless, there are different opinions about the age and paleogeographic position of some HP/LT units (e.g. Cretaceous vs. Paleogene, number of subduction sites; see for example Polino et al. 1990),which are partly due to conflicting determinations of the timing of metamorphism (e.g. Chopin & Maluski 1980; Hunziker et al. 1992; Gebauer 1999) and partly due to a fragmentary view of the Alpine belt.
The identification of ophiolites and their position in the Alpine edifice are crucial for the reconstructions of the Alps (e.g. Dal Piaz 1974; Dietrich 1980; Bigi et al. 1990; Polino et al. 1990; Höck & Koller 1989; Froitzheim et al. 1996). Furthermore, it is well accepted that high-pressure/low-temperature (HP-LT) metamorphism of such ophiolites is associated to subduction related processes, whereas high-temperature/medium-pressure (HT-MP) metamorphism is associated to collision related processes. A better understanding can be gained through integration of structural analysis at various scales, petrological estimates, and isotope dating (e.g. Compagnoni et al. 1977; Gosso et al. 1979; Polino et al. 1990; Oberhänsli 1994; Froitzheim et al. 1996; Handy & Oberhänsli 2004). The identification of ophiolites is well established in the Alps, whereas the metamorphic evolution in time and space is less clear. Another large difficulty arises by defining units (rock masses), which underwent the same geodynamic evolution (see discussion in Bousquet 2007). In other words, we have to combine information from local samples towards tectonic units or paleo-plates to understand geodynamic processes in time and space.
Understanding geodynamic evolution during subduction requires the combination of different datasets. In this context, accurate constraints on pressure (P), temperature (T) and time (t) are key elements (e.g. Compagnoni et al. 1977; Ernst & Dal Piaz 1978). Recently, new isotopic ages and P-T conditions were published on the high-pressure/low-temperature metamorphic evolution using eclogites and blueschists (e.g. Gebauer 1999; Agard et al. 2002; Rubatto & Hermann 2003; Liati et al. 2005; Brouwer et al. 2005). Compiling all of the available data result in relatively high density of data, which make the Alps suitable to investigate subduction processes in time and space. When compiled isotopic ages are combined with petrological information, the resulting information includes substantial uncertainties. These uncertainties will be summarized into two groups: (1) interpretation (geological meaning) of isotopic data ages per se, and (2) correlation of units in a geodynamic context.
With regards to point 1) Individual isotopic mineral ages from high-pressure (HP) metamorphic samples may be interpreted as:
- Early metamorphic growth along a prograde path
- Peak metamorphic conditions
- Strong overprint along the retrograde path
- Diffusional resetting of the chronometer
- Relict ages of the protolith
With regards to point 2) The knowledge of a certain age has to be translated into a space-time movement of tectonic units or to fragments within mélange units. This includes the reconstruction of coherent unit for a certain part of the P-T-t evolution. In this paper, we combine such different information and discuss possible models for the geodynamic evolution of the Alps. The Alps are starting to have enough information, in order to compare data from nature with numerical modeling. Such geodynamic modeling is available since several decades, and published models using different techniques and aims (e.g. Oxburg & Turcotte 1974; Davy & Gillet 1986; Escher & Beaumont 1997). Modern geodynamic modeling proposed P-T-t paths, which can be compared with data from nature (e.g. Bousquet et al. 1997; Burov et al. 2001; Roselle et al. 2002; Goffé et al. 2003; Stöckert & Gerya 2005). Therefore, we have to understand the meaning of individual P, T, and t data and link these data into a geodynamic context, which is the aim of this contribution.
The tectonic base, as used in this contribution, is summarized from the tectonic map of the Alps (Schmid et al. 2004a) and the geodynamic map of the Alps (Oberhänsli et al. 2004). The latter both maps base on the structural model of Italy (Bigi et al. 1991). In general, the tectonic units have to be subdivided into units belong to the Adria margin, the Piemonte-Liguria Ocean, the Briançonnais microcontinent, the Valais Ocean and the European margin. In addition smaller portions of the Meliatta Ocean are preserved in the eastern part of the Alps. The main exhumed parts of the Western- and Central Alps are related to Late Cretaceous-Paleogene orogen including subduction of the Alpine Tethys and collision of Adria and Europe. The tectonic evolution of the Alps has been already discussed from different points of view (e.g. Tapponier 1977; Frisch 1979; Trümpy 1982; Froitzheim et al. 1996; Stampfli et al. 1998; Dal Piaz 1999; Schmid & Kissling 2000; Dal Piaz et al. 2003; Schmid et al. 1996, 2004a, 2004b).
This contribution summarizes data related to subduction and early evolution of the Alps. For simplicity we do not deal with the later and skip later important tectono-metamorphic evolution, although we are aware of the complex evolution following subduction, which include the exhumation of the HP/LT rocks.
The pressure-temperature-time evolution during subduction
The main published results will be summarized in the following sections. This compilation does not give a complete summary of isotope ages and P-T data, rather it examines data that are directly relevant to the subduction history. For simplicity, data are mostly presented without the analytical error. The following data are organized into the major tectonic units from south(east) to north(west) (Sesia-Lanzo zone, Piemonte-Liguria Zone, Briançonnais and Valais).
Sesia Zone and Dent Blanche nappe
The Sesia Zone of the western Austroalpine is a portion of Adriatic continental crust recording Alpine eclogite-facies assemblages (Fig. 1). The recognition that granites underwent eclogite-facies conditions within this zone provided the first demonstration of subduction of continental crust (Dal Piaz et al. 1972; Compagnoni & Maffeo 1973; Compagnoni et al. 1977; Lardeaux et al. 1982; Oberhänsli et al. 1982, 1985; Lardeaux & Spalla 1991; Venturini 1995; Spalla et al. 1996). The Alpine evolution is characterized by relics of blueschist- to eclogite-facies prograde metamorphism, followed by a HT blueschist-facies re-equilibration during decompression (e.g. Castelli 1991; Pognante 1991 and references therein), and then by a greenschist facies overprint (Dal Piaz et al. 1972; Compagnoni et al. 1977; Oberhänsli et al. 1985), which is generally associated with mylonitic textures (Stünitz 1989; Spalla et al. 1991).
The more external Dent Blanche nappe s.l. is located at the same structural level as the Sesia-Lanzo Zone, namely on top of the entire ophiolitic Piemonte-Ligurian Ocean (Fig. 1). It consists of the Dent Blanche, M. Mary and Pillonet basement and cover units, which display relics of a blueschist facies imprint and a pervasive greenschist facies overprint (Ballevre et al. 1986; Dal Piaz 1999). Other basement slices (Mt. Emilius, Glacier-Rafray, Etirol-Levaz) preserve relics of an eclogitic imprint but are located at a lower structural level of the nappe stack (i.e. within the underlying ophiolitic Piemonte-Ligurian Ocean) and are reported as lower Austroalpine outliers (Dal Piaz 1999; Dal Piaz et al. 2001).
More recently, the Sesia Zone and the Lower Austroalpine Outliers have been interpreted as one or more extensional allochtons within the Piemonte-Liguria Ocean (Froitzheim et al. 1996; Dal Piaz 1999; Dal Piaz et al. 2001; Schmid et al. 2004a). The original position of these units within an oceanic plate is of special importance for geodynamic and tectonic models (see discussion). The subduction direction of this unit is towards the southeast, which is consistent with Tertiary subduction of the Piemonte-Liguria Ocean.
SHRIMP zircon ages in the Sesia Zone indicate that the metamorphic evolution began at 76 Ma and subduction related metamorphism at ~65 Ma (Fig. 2; Rubatto et al. 1999). A late Cretaceous subducting related metamorphism in eclogite has been suggested based on Lu/Hf ages on garnet (69 Ma; Duchêne et al. 1997). While the Sesia evolution is well constrained between 60 and 80 Ma (see Frey et al. 1999 and Handy & Oberhänsli 2004 for reviews), the similar eclogitic imprint of the lower Austroalpine outliers (Mt. Emilius-, Glacier-, Rafray-, Etirol-Levaz- units) are less clear. These units display Eocene Rb-Sr white mica ages (49-45 Ma; Dal Piaz et al. 2001).
click on the picture to zoom or click here to download Figure 1: Simplified metamorphic map of the Alps modified from Oberhänsli et al. (2004). The colours of the map represent the various metamorphic conditions preserved in the Alps (see colour code in the P/T diagram in the middle right). P-T paths are shown together with the isotope ages of subduction related metamorphism for various localities across the Alps (the ages are colour coded as shown in the upper left). Data source and references for P/T and t data are given in Table 1 and 2 |
 Figure 2: Compiled isotope and petrological data from selected continental and oceanic units. In the middle column are exemplary P-T path for some HP rocks. Note that in the Central Alps a large variety of P-T path are estimated and only a few are presented. The temperatures in the Central Alps are systematic higher compared to surrounding units. The right column presents the pressure-time paths using the isotope and P-T data (compare also Figs. 1 and 3). Numbers refer to Table 1. |
Piemonte-Liguria Ocean
 Table 1: Selected isotope age data relevant to the early tectonic history of the units. Units involved in the subduction processes, but without evidence of HP/LT metamorphism are shown in grey. The area-code will be find in Figures 2, 3 and 5. These data are presented graphically in Figure 1.Download as pdf |
The Piemonte-Liguria Ocean (or Zone) is a structurally composite nappe system extending along the entire arc of the Western Alps (e.g. Bigi et al. 1990). In the north-western Alps it is classically divided into two principal units, the upper unit (e.g. Tsaté nappe or Combin zone in the north) and the lower unit (e.g. Zermatt-Saas Fee nappe in the north; e.g Sartori & Thélin 1987; Ballèvre & Merle 1993; Dal Piaz 1999 and references therein). The distinction between these ophiolitic units was based both on lithostratigraphic and metamorphic differences (Bearth 1962, 1967; Kienast 1973; Dal Piaz 1965, 1974; Caby et al. 1978; Marthaler 1984; Marthaler & Stampfli 1989) and structural data (e.g. extensional faults; Ballèvre & Merle 1993; Reddy et al. 2003). These ophiolitic units extend south of the Aosta valley, through the Cottian to the Ligurian Alps; these are generally reported as external and internal units of the Piemonte-Liguria Zone, corresponding to the Combin and Zermatt-Sass units, respectively.
The Zermatt-Saas Fee unit and southern homologues (Monviso, Voltri Group) are composed mainly of metamorphic ophiolites (metabasalt, metagabbro, serpentinite) with minor cover sequences. These units are well known for their high-pressure mineral assemblages. The discovery of coesite inclusions within garnet in Lago di Cignana suggests that some pieces of the Piemonte-Liguria Ocean were deeply subducted (~2.8 GPa; Reinecke 1991). The spatial distribution of possible ultra high pressure rocks is not clear at the moment. The Combin zone (northwestern Alps) is characterized by blueschist facies relics scatterly preserved below a pervasive greenschist facies overprint (Dal Piaz & Ernst 1978; Ayrton et al. 1982; Baldelli et al. 1983; Sperlich 1988; Martin et al. 1994) in the metabasites while garnet-chloritoid-phengite assemblages in surrounding metapelites indicate higher metamorphic conditions up to 1.4 GPa and 450°C (Bousquet, 2007). In the South the Schistes Lustrés zone of the Cottian Alps is characterized by occurrences of Fe, Mg-carpholite bearing assemblages in the western part and chloritoid-bearing assemblages in the eastern part with Fe, Mg-carpholite only as relics in quartz veins (Goffé & Chopin 1986; Agard et al. 2001). On the basis of metapelite and metabasite mineralogy, P-T estimates for the upper most unit at maximum depth increase from west to east from ~1.0-1.2 GPa / 300-350°C to 1.4-1.5 GPa / 450-500°C (Agard et al. 2001; Messiga et al. 1999; Schwartz et al. 2000a).
Evidences of HP⁄LT metamorphism have also been found in the Avers nappe of the Central Alps (eastern Switzerland). Mineral assemblages formed of glaucophane-garnet ± chloritoid in metabasalts (Oberhänsli 1978), glaucophane-phengite in marbles, and garnet-chloritoid in calcschists (Wiederkehr 2004). The lower unit in the east is represented by the Platta-, Lizun, Malenco and Arosa units (e.g. Schmid et al. 2004). The Platta Nappe shows no HP⁄LT metamorphism, but a very-low grade, Alpine metamorphism in the north and a greenschist facies metamorphism in the south (e.g., Müntener et al. 2000; Frey & Ferreiro-Mählmann 1999). The metamorphic evolution of the metasediments (e.g. Tsaté in the western Alps and Avers nappe in eastern Switzerland) is in contrast to the greenschist facies metamorphism of some of the mentioned ophiolitic units. These are Balagne, Nebbio and Pineto-Tribbio units in the southwestern Alps and Corsica and the Platta-Lizun units in the eastern part of the central Alps.
The eclogites of the Zermatt-Saas Fee unit show Lu⁄Hf and Sm⁄Nd ages of 49 and 41 Ma, respectively (Table 1, Fig. 2; Amato et al. 1999; Lapen et al. 2003). The Lu⁄Hf data are interpreted as the time of blueschist-eclogite transition (Lapen et al. 2003). The zircons record SHRIMP age of 44 Ma, which is interpreted as the HP⁄LT metamorphism (Rubatto et al. 1998). The Monviso eclogites show spread in Sm⁄Nd, Lu⁄Hf and SHRIMP ages, which may in part related to analytical uncertainties. The Sm⁄Nd garnet isochrones result in ages between 60 ± 12 and 61 ± 9 Ma (Cliff et al. 1998), whereas the Lu⁄Hf garnet ages, SHRIMP zircon data are significant younger (49 and 45 Ma; Duchêne et al. 1997; Rubatto & Hermann 2003; Table 1). These younger ages correspond somehow with Rb⁄Sr phengite isochron of Cliff et al. (1998). HP-phengites of Fe, Mg-carpholite bearing metasediments of the external unit are dated between 60 and 55Ma by Ar⁄Ar for peak pressure (Agard et al. 2002) while the exhumation down to greenschist facies conditions took place from 48 to 40 Ma (Hunziker et al. 1992; Markley et al. 1998; Agard et al. 2002, Bucher 2003). The greenschist facies metamorphism in the southern Platta- and Malenco units is Cretaceous in age. These are recorded in K⁄Ar amphibole ages, which are 90-69 Ma (Deutsch 1983) and detail Ar⁄Ar data of amphiboles in the Malenco, which spread between 80-90 Ma (Table 2, Villa et al. 2000). Similar to these data, the data in Balagne unit indicate a greenschist facies metamorphism at ~80 Ma (Maluski 1977; see also discussion in Handy & Oberhänsli 2004).
Briançonnais
 Table 2: Review of selected PT data relevant for subduction-related metamorphism in the Alps. For localities and paleogeographic subdivision see Table 1.Download as pdf |
The paleogeographic significance and the metamorphic evolution of the Briançonnais microcontinent have been under debate for a long time (e.g. Stampfli 1993; Monié 1990). Despite the relatively small size of this microcontinent, it displays a complex Alpine metamorphic pattern. According to the metamorphic history, three areas can be defined within the Briançonnais microcontinent.
- The so-called internal massifs (Dora Maira-, Gran Paradiso-, Monte Rosa-units) show eclogite facies metamorphism. These are often metagranitoid rocks with interlayered thin layers of metasediments and/or metabasic rocks. These units are pervasively overprinted by alpine tectono-metamorphic reworking (eclogitic and greenschist to amphibolite facies; Dal Piaz & Lombardo, 1986; Chopin et al. 1991; Frey et al 1999, Dal Piaz 2001). The occurrence of coesite in the Dora Maira nappe gives a minimum pressure of ~3 GPa (Chopin et al. 1991; Chopin & Schertl 1999; Simon & Chopin 2001). The estimated pressures in the Monte Rosa nappe differ between 1.8 to 2.4 GPa (Dal Piaz & Lombardo 1986; Le Bayon et al. 2006a) at temperatures around 500°C. Maximum pressures are 1.8 - 2.0 GPa at 500°C in the Gran Paradiso massif (cf. Benciolini et al. 1984; Brouwer et al. 2002; Le Bayon et al. 2006b). U/Pb of rutile in a vein of the Monte Rosa indicate HP/LT metamorphism at 42.6 Ma (Lapen et al. 2007). In addition, zircon SHRIMP data of 35 Ma in metasediments are published (Rubatto & Gebauer 1999), although the position of these metasediments inside the Monte Rosa unit has been questioned (see Dal Piaz 2001). Furthermore, Th/Pb monazite ages of 34-32 Ma in the Monte Rosa nappe indicate decompression down to ~1.0 GPa (Engi et al. 2001b), which clearly are in conflict with with an age for peak pressure at 35 Ma. The Gran Paradiso is dated by Rb/Sr phengite-apatite isochron, which indicates an equilibration near maximum pressures at 43 Ma (Meffan-Main et al. 2004). The Dora Maira unit is dated by U/Pb in a range of minerals (Tilton et al. 1991), Lu/Hf in eclogite (Duchêne et al. 1997) and zircon SHRIMP age (Gebauer et al. 1997; Rubatto & Hermann 2001), Ar/Ar and Rb/Sr dating (Di Vincenzo et al. 2006). These data show a spread between 38 and 33 Ma, which have been correlated to subduction related metamorphism. Chemical age data in monazite indicate two metamorphic events in the Dora Maira unit (Vagelli et al. 2006). Monazite cores record a metamorphic stage at 60 Ma, and a second monazite forming event at 37 Ma. The latter age clearly overlap with other data presented above (Table 1). The meaning of the 60 Ma monazite age is not clear at the moment.
- The internal Briançonnais (Ambin-, Vanoise-, Ruitor-, Mt. Fort-, Tambo/Suretta-units) consists of combined basement and metasedimentary units, which shows a clear HP/LT imprint. Fe, Mg-carpholite (Goffé 1977, 1984; Goffé et al. 1973; Goffé & Chopin 1986), aragonite (Goffé & Velde 1984), jadeite (Cigolini 1995) in metasediments and occurrences of lawsonite and jadeite in metabasites (Lefèvre & Michard 1976; Schwartz et al. 2000b) indicate an evolution at the blueschist-eclogite transition (see Oberhänsli et al. 2004; Fig. 1). An increase in metamorphic grade from blueschist in the west towards blueschist-eclogite facies in the east has been documented in post-Hercynian metasediments (Bucher & Bousquet 2007). Paradoxically, in the northwestern Alps, only the uppermost units of the Briançonnais domain (the Mont Fort nappe) display HP metamorphic evolution (Schaer 1959; Bearth 1963; Bousquet et al. 2004).
The age of the HP metamorphism in the internal Briançonnais was debated for a while (see review in Caby 1996), despite the occurrences of glaucophane in Eocene sediments (Ellenberger 1958). In the western Alps, Ar/Ar data on HP-phengites in post-Hercynian metasediments indicate that these units reached peak pressure between 50 and 43 Ma while exhumation and nappe stacking took place between 43 and 35 Ma (Bucher 2003). Ar/Ar ages in the Tambo/Suretta nappes (Central Alps) indicate that the pressure-dominated low temperature metamorphism occurred in the range of ~45 Ma (Challandes et al. 2003). Similar ages (~ 45 Ma) have been obtained for the Tenda massif in Corsica (Brunet et al. 2000), which has a similar tectonic position.
- The Briançonnais further to the west are composed only of sedimentary rocks (summary in Schmid et al. 2004a). These units display only greenschist metamorphic assemblages (see review in Goffé et al. 2004; Bousquet et al. 2004).
Valais Ocean
The rocks of the Valais ocean can be followed from the western Alps to the Engadine window (see Fig. 1 and Schmid et al. 2004a). The sedimentation in the Valais Ocean is mainly Cretaceous in age (e.g. Steinmann 1994), but continues also in the Paleogene. The sediments are often stripped off from their substratum and were incorporated into an accretion wedge during subduction. Ophiolites and magmatic rocks associated with the metasediments are rare (Versoyen at the French/Italy boundary; Vals- Engadine- and Chiavenna area in the Central and in the Eastern Alps). The metasediments as well as some of the magmatic rocks of the Valais Ocean show clear evidence for blueschist to eclogite facies metamorphism (e.g. Goffé & Oberhänsli 1992; Goffé & Bousquet 1997; Bousquet et al. 1998).Metamorphic pressures in the accretion wedge range from 1.2-1.4 GPa in the east and up to 1.7 GPa in the west at low temperatures (Bousquet et al. 2002). Eclogite facies rocks are also connected to the Valais Ocean (Antrona ophiolite; Oberhänsli 1994; Schmid et al. 2004a). In addition, some authors have recently related the Balma unit to the Valais Ocean (Pleuger et al. 2005; Liati & Froitzheim 2006) instead of the Zermatt-Saas Fee unit (Gosso et al. 1979).
Unfortunately, the timing of the HP metamorphic event of the Valais metapelites remains poorly resolved. An upper limit of the high-pressure metamorphism at 40-35 Ma is given by a Paleocene- Eocene Radiolaria in carpholite-chloritoid-bearing metasediments in the Central Alps (Bousquet et al. 2002). From isotope dating (Schürch 1987; Cannic et al. 1999) and the fossils in the HP-metasediment, it is reasonable to assume a high-pressure metamorphism around 40-35 Ma for the Valaisan domain. This age is in agreement with 39-40 Ma old zircon SHRIMP ages published for the Antrona and Balma units (Liati et al. 2005, Liati and Froitzheim 2006).
The oceanic part in the eastern Alps are mainly visible in the Tauern window (with several exceptions). The units in the Tauern window are subdivided into a lower unit (Venediger nappe) and an upper unit (Glockner nappe). The lower unit is related to the European margin with its sedimentary cover (Schmid et al. 2004a). The upper unit include magmatic rocks, schist and calcschists, which are correlated with the eastern continuation of the Valais ocean (Schmid et al. 2004a). However, some sediments of the lower units have similarities to the Piemonte-Liguria and the Valais Ocean. However, these units are not separated by a microcontinent, and therefore, we will consider ”one“ ocean in the Eastern Alps (see Stampfli et al. 2002). The magmatic and metasedimentary units of this ocean suffered blueschist to eclogite facies metamorphism (e.g. Miller 1974; Spear & Franz 1986; Zimmermann et al. 1994; Hoschek 2001). Maximum pressures reached in the eclogite zone are ~2 GPa. These rocks underwent subduction-related metamorphism in the time between 55-45 Ma and exhumation from this time on (Zimmermann et al. 1994; Ratschbacher et al. 2004). The eclogite zone became coupled with the blueschist facies metasediments at around 35 Ma and both continued exhumation to 1.0 and 0.5 GPa (Table 1, Fig. 1; Ratschbacher et al. 2004).
Assimilated units at the European margin (Central Alps)
A former plate boundary is preserved as a Tectonic Accretion Channel in the Central Alps (TAC; Engi et al. 2001a). This represents a deep-seated part of the plate boundary including fragments of the mantle, meta-basalts, metagabbros and metasediments. The continental basement units can be related to material above and below, but the source of the incorporated oceanic material is unclear. These units are treated as the trace either of the Valais Ocean (e.g. Froitzheim et al. 1996; Schmid et al. 1996, 2004a), or of the Piemonte-Liguria Ocean (Stucki et al. 2003). Studies in the mélange units of the Central Alps indicate that peak pressure of eclogites varies from ~1.0 - ~4.0 GPa at variable temperatures (e.g. Ernst 1981; Heinrich 1982; Pfiffner 1999; Paquin & Altherr 2002; Nimis & Trommsdorff 2002; Dale & Holland 2003; Brouwer et al. 2005; Fig. 2). The lowest pressure is reported from the northern part of the Adula nappe complex and the highest pressure from different lenses in the Southern Steep Belt (e.g. Heinrich 1982). The temperature of the eclogite stage is in the range of 700-800°C, which is higher than in the subducted units of the Western Alps (e.g. Zermatt-Saas Fee unit; Fig. 2).
The available ages are concentrated on Alpe Arami- and Cima di Gagnone-bodies (Table 1, Fig. 1; Gebauer 1996, 1999; Becker 1993). The garnet Sm/Nd isochrones of these localities indicate ages between 38 and 42 Ma (Becker 1993). Zircon SHRIMP data indicate HP ages between 43 and 35 Ma (Gebauer 1996). Hermann et al. (2006) show that zircons developed at still elevated pressure, but on the retrograde path at 34 Ma (Monte Duria body). The interpretations of the Sm/Nd ages in the Central Alps are under debate. Brouwer et al. (2005) conclude that these ages are reset by the high temperature history of these bodies, which is supported by the same Lu/Hf ages in these localities (Fig. 2). In addition, Lu/Hf ages in one locality (Alpe Repiano) show growth of garnet at ~61 Ma (Brouwer et al. 2005; Table 1). Using the Lu and Hf contents of the separates in combination with laser ablation ICP-MS data of the Lu and Hf content of garnet, it is possible to combine growth evolution with Lu concentrations in different garnet separates. The preserved zonation indicates that these ages represent garnet growth (see also discussion in Skora et al. 2006). The core of the garnets with high Lu content indicate growth at the garnet in position along the prograde path at ~61 Ma.
Discussion
 Figure 3: Pressure-time ”path“ for some selected samples/units. The key information include the depth of subduction inferred from petrological estimates and related exhumation (compare also to Figure 2).(a) pressure-time path, where dating of the prograde path is inferred. (b) pressure-time path, where only the exhumation is documented. (c) Youngest sediments in the different units. Note that subduction must have been active, while sedimentation was occurring. References: R98: Rubatto et al. 1998; R04: Ratschbacher et al., 2004; B05: Brouwer et al., 2005; A02: Agard et al. 2002; RH03: Rubatto & Hermann 2003; L03: Lapen et al. 2003; A99: Amato et al. 1999; MM04: Meffein Main et al. 2004 (for further information see Table 1 and 2). |
The presented ages show, that the timing of subduction related metamorphism is mainly spread between ~75 and 40, with some local event at 33 Ma (Figs. 2 and 3). Some of these ages represent the prograde path (i.e. Lu/Hf: Lapen et al. 2003; Ar/Ar in phengites: Agard et al. 2002) and other may be related to retrograde path at elevated pressures (e.g. Hermann et al. 2006). The isotopic HP age information, P-T data and geodynamic interpretation include a large number of uncertainties. These uncertainties and the consequences by combining the data can be summarized into following questions:
- What is the meaning of each measured age per se?
- What is the paleogeography before and during subduction?
- What is the type and amount of fragmentation during subduction?
- Can we reconstruct subduction in time and space?
- How many suture zones are hidden in the Central Alps?
- What is the paleogeographic position of the Malenco-Platta units (and their southwestern equivalents)?
In the following, we will mainly formulate these questions, rather than present already solutions.
What is the meaning of each measured age?
Dating of rock forming minerals has the advantage that we directly combine petrological information with isotope dating. However, accessory phases (zircon, monazite, etc) are often rich in U and Th, and they are robust in respect to isotopic resetting and hence are well suited for dating. However, in the case of accessory phases the relation to pressure and temperature has to be evaluated. Datasets from accessory phases versus rock forming minerals have both advantages and disadvantages.
In order to discuss the relation between isotopic age and P/T data, we have to discuss the ”closure-temperature“ as introduced by Dodson (1973; see also Ganguely & Tirone 1999). However, the role of temperature-induced diffusion is of minor importance for most geochronometer at most grain sizes, and thermally induced diffusion is the key element of the ”closure-temperature“ concept (see discussion in Villa 1998). In addition, other processes influence the ”resetting“ of an isotope system in a certain mineral. These processes are mainly deformation, mineral reactions and interactions with fluids (see for example Federico et al. 2005).
Unfortunately, the respective contributions from volume diffusion, mineral reactions, deformation or fluids on the resetting of a certain isotope system is unknown, so such processes can not be generalized.
Despite these problems, the large number of individual data and several detail studies on the meaning of the age indicate some processes are more important in one or the other mineral. For example, diffusional resetting is rarely a problem for U/Pb zircon ages. Therefore, many SHRIMP ages give important insights in the evolution of zircons. New zircon can be produced during HP-events (e.g. Rubatto & Hermann 2003), but the relation to the P-T evolution is not always clear. The zircon SHRIMP ages are most useful if these data can be combined with trace element data and/or inclusion relationships (e.g. Rubatto 2002; Rubatto & Hermann 2003; Hermann et al. 2006). Ar/Ar mica ages in blueschist-facies units often record mineral growth along the prograde path (e.g. Federico et al. 2005; Agard et al. 2002). However, the role of excess Ar in high-pressure rocks has to be considered (e.g. Scaillet 1996; Brunet et al. 2000, but see also Gouzu et al. 2006). At higher temperatures and/or stronger deformation, resetting of this isotope systems along the retrograde path often occurs, so in the higher temperature and strongly deformed units we exclude the Ar/Ar ages for our discussion. Lu/Hf garnet ages are most useful because the distribution coefficient of the parent element is large between garnet and other minerals and diffusion of REE and Hf is slow in garnets (Scherer et al. 2000). Therefore, Lu/Hf in garnet has a huge potential for dating prograde metamorphic evolution (e.g. Lapen et al. 2003; Skora et al. 2006). In contrast to the accessory phases, garnet stability is well investigated. Diffusional resetting by volume diffusion in garnet is in most cases of minor importance, because of the low temperatures and slow element diffusion coefficients. Garnet isochron dating (especially Lu/Hf) has less problems with mineral reactions or volume-diffusion, but has other problems with REE bearing mineral inclusions inside the garnet, which may disturb age information (Scherer et al. 2000). Another problem of porphyroblastic garnet is zonation of REE, which are either solely related to petrological processes or related to age zonation (Bouwer et al. in prep; Lapen et al. 2003; Skora et al. 2006).
In four localities, the zircon SHRIMP data are slightly younger than the Lu/Hf age (Zermatt-Saas Fee-, Sesia-, Arami-, Monviso units), whereas the Sm/Nd garnet ages show no systematics with respect to other isotope ages. Lu-Hf ages record in most cases the early growth of garnet (e.g. Lapen et al. 2003, Skora et al. 2006), which occurs during the prograde path at the blueschist-eclogite transition. The meaning of the zircon U/Pb age in respect to P and T may be different, but the effect of fluids seems important in many cases. Taking this information in account, the Lu/Hf garnet ages may represent part of the prograde path except some HT samples in the Central Alps. Zircon SHRIMP data may relate to a range of events, depending on the metamorphic reactions and availability of fluids; all of which influence the growth of new zircon shells.
The paleogeographic situation
In order to combine P-T-t information with the geodynamic evolution of an orogen, we require a palinspastic reconstruction (e.g. Laubscher 1971a, 1971b; Schmid & Kissling 2000; Schmid et al. 2004a, 2004b). The Alpine tectonics can be subdivided into a Cretaceous and a Tertiary orogen, which are separated by the extensional Gosau event (e.g. Froitzheim et al. 1994, 1996). The Cretaceous orogen includes westward thrusting in the Austroalpine, which has been related to closing of the western Meliata Ocean. HP/LT rocks related to closing of the Meliata Ocean occur, which will not be discussed here (see Thöni & Miller 1996; Janak et al. 2004; Thöni 2006 for this theme). We discuss the Upper Cretaceous to Paleogene subduction of the aforementioned oceans and continental fragments.
The subducted units can subdivided into two basins and the related continental units. The basins are the Valais-Ocean in the north and the Piemonte-Liguria Ocean in the south (Figs. 1 and 4; e.g. Stampfli & Borel 2004). The northern continent is Europe and the southern one, as summarized in this contribution, is Adria. The forementioned oceans are separated by the Briançonnais-unit, which represent a topographic high between the oceans, including a continental basement and its sedimentary cover. The oceanic metasediments in the eastern part (Tauern window) show characteristics of the Piemonte-Liguria Ocean as well as the Valais Ocean, but these units are not separated by a continental domain (Frisch 1979; Stampfli & Borel 2002; Schmid et al. 2004a). Therefore, we will consider ”one“ ocean in the Eastern Alps (Fig. 4, see also Stampfli et al. 2002). The different oceanic basins are part of the Alpine Tethys. This subdivision is accepted by most authors, but the palinspastic reconstruction and position of each tectonic unit is under debate. However, the inferred shape of these Oceans require a special paleogeographic situation at the triple-point of the Valais- and Piemonte-Liguria Ocean and the lateral end of the Briançonnais in the central part of the present day Alps (Fig. 4).
 Figure 4: Schematic palaeogeographic sketch for late Cretaceous (reconstruction is inspired from Stampfli & Marchand 1997; Stampfli et al. 2002; Stampfli & Borel 2002; Schmid et al. 2004a). Numbers of present day units refer to Table 1. |
It will be important for the discussion to constrain the size of these oceans in order to discuss subduction process (see below). Such data are obtained from palinspastic reconstructions and are in the range of 400-500 km for the combined Piemonte-Liguria- and Valais Ocean in the west (Stampfli et al. 2002). The size of the ocean in the east (Tauern window) must be somewhat smaller (Laubscher 1971a, 1975).
What is the type and scale of fragmentation during subduction?
In the former section, we concentrate on the prograde evolution during subduction. The difference in P-T path requires also differences in tectonic evolution, which most likely include off-scraping and fragmentation along the destructive plate boundary. However, the preserved information also includes the exhumation part of the investigated units, boudins or clasts. However, compiling P-t evolution will give information on the time of fragmentation.
In general, subduction is considered to be a continuous process, which requires a coherent lithospheric plate (scale of kilometers to hundreds of kilometers) in order to maintain the driving force for subduction. In any scenario of a coherent subducting plate, HP/LT metamorphism occurs over a certain time interval (depending on the size of the ocean and the subduction rate). The present day record of this long lasting process and related metamorphism may be different, and two end-member scenarios can be discussed: (1) different exhumed parts of this process record the time interval of metamorphism or (2) the record of this metamorphism is related to a single exhumation event. The latter scenario is not related to the time interval of metamorphism during subduction. However, both scenarios required fragmentation of the subducting plate at different scales (m and/or km). In this context, fragmentation means off-scraping of sections of the plate. In the case of large sheared-off proportions (km size) of a subducting plate, large ophiolitic sequences are transferred to the orogenic wedge (e.g. Zermatt Saas Fee-, Platta units). In the example of small fragments (m-scale), we find complex tectonic mélange units, which contain HP/LT fragments of oceanic origin. The latter process was active in the Central Alps and possibly in the Tauern. The mechanism for separating pieces of the subducting oceanic crust and mantle slivers is not well understood. The understandings of such fragmentation processes are further obscured by later deformation related to exhumation. However, cutting out a large piece of the crust during subduction will change the geometry and driving force during subduction. In contrast, development of a tectonic accretion channel operates over long time intervals and subduction may be continued.
In terms of subduction and the record of HP metamorphic rocks, we may subdivide the units in the Alps into following groups:
1) Coherent continental basement units (e.g. Dora Maira-, Gran Paradiso, Monte Rosa- and Sesia units)
2) Stripped-off metasediments (e.g. Schistes Lustrés)
3) Small HP fragments within mélange zones (e.g. Entrelor area, Central Alps)
4) Coherent sections of oceanic lithosphere (e.g. Monviso-, Zermatt-Saas Fee, Platta units)
Although some coherent continental basement units contain minor sedimentary cover units, most of the cover units were lost before reaching HP metamorphism. This is well visible in the Briançonnais of the Western Alps (e.g. Bucher et al. 2003).
The oceanic metasediments are related to different oceans (Valais and Piemonte-Liguria) and have different P-T-t evolution. Some of them are included in an accretion wedge along the upper part of the destructive plate boundary. These units show a strong internal deformation, which is evident from their structures and their isotopic ages (Agard et al. 2002). Similar is true for tectonic mélange zones in the Central Alps, where different oceanic fragments show different P-T-t paths (Brouwer et al. 2005). The aforementioned metasediments and the deeper parts of the subduction channel are the possible locations of small-scale fragmentation. Small-scale fragmentation along a plate boundary may operate at times during ongoing subduction. In contrast, the separation and exhumation of complete ophiolite sequences is most likely to drastically change the geometry and driving force of subduction.
Can we reconstruct subduction in space and time?
 Figure 5: Interpretation of the palaeogeographic reconstructions and available HP/LT data as given in Table 1 and 2: (a) Palaeogeographic sketch map of the Alpine Tethys based on the supposed existence of two oceans and the Brianconnais microcontinent. The location of profiles in b-d are shown. (b) Schematic profile A through the western Alps: (c) Schematic profile B through the central Alps; (d) schematic profile C through the area of the Tauern. Numbers in the profile and the pressure-time path correspond to units or fragments as used in Table 1 (see also Figs. 1 and 2). |
As discussed in the last sections, there are many uncertainties related to the significance of age determinations and of the location of each unit in its former paleogeographic position. Furthermore, there appears to be a large time span (ca. 35 Ma) for HP/LT metamorphism. By combining the presented data (Figs. 2, 3), we have a tool to check the consistency of some tectonic models in terms of comparing map and section view (Fig. 5). In general, we have to consider coherent plates to maintain the driving force for subduction. One model indicates that subduction simply propagated to the northwest (e.g. Ernst, 1971, Dal Piaz et al. 1972, Polino et al. 1990, Schmid & Kissling 2000 and literature therein). Some data fit such a propagation model well, but others do not. Such a model would also require similar ages along strike of the same paleogeographic unit. For example, the basement of the Briançonnais (Dora Maira, Gran Paradiso, Monte Rosa) should have similar HP/LT ages at similar depth. In contrast to this prediction, there is significant variation in the HP/LT ages (35 Ma (60 Ma), 43 Ma, 45 Ma, see Table 1).
Testing possible scenarios, we may discuss ages perpendicular to the destructive plate boundary at different positions (Fig. 5). A distance between units before subduction can be calculated by combining the timing of HP metamorphism (at similar depth) and an assumed subduction rate (Fig. 5). In the last section we present available P-T and t information for some units of interest (Figs. 3, 5) To examine space evolution we employ pressure-time path, using calculated metamorphic pressures and interpolation of available isotope ages (Fig. 3). In order to have comparable ages from burial to exhumation, we can calculate the distances of these units at given subduction rate. Using the subduction-related metamorphism of the Sesia Zone as one point and calculating the distance to the metasediments in the Western Alps or the Zermatt-Saas Fee unit, we get distances of 150 km and 300 km, respectively (using subduction rate of 1.5 cm/a; see plate moving data in Schmid et al. 1996; Dewey et al. 1989). These distances fit roughly into the size of the oceans from independent plate reconstructions (Stampfli & Marchand 1997). Accepting the overall palaeogeographic configuration of the different subducted units, the differences in pressure-time path have a major consequence. It shows, that large-scale units (km sized) are cut off from the subducting plate, but the overall subduction continues (see also section on fragmentation). These age depth data can be also used the other way around. The timing and a certain subduction depth combined with a palinspastic reconstruction result in a certain subduction rate (1-2 cm/a). These subduction rates indicate relative slow subduction process compared with modern subduction rates worldwide.
How many oceans are hidden in the Central Alps?
In the Central Alps, we mainly find three units containing relics of oceanic crust: (1) the Tectonic Accretion Channel (TAC), (2) Chiavenna ophiolite, and (3) metasediments of the Northern Steep Belt. In the TAC units, there is one body indicating early HP metamorphism (~61 Ma), but most fragments record a pressure dominated retrograde equilibration between 42 and 34 Ma (Table 1, 2; Gebauer 1996; Brouwer et al. 2005, Hermann et al. 2006). The rocks in the Southern Steep Belt are already incorporated in the nappe edifice and suffered Barrovian overprint at ~32 Ma, which was contemperanuous with the ascent and emplacement of the Bergell pluton (Gebauer 1996; Berger et al. 1996; Oberli et al. 2004). Barrovian overprinting occurs much later in the northern part of the Lepontine. The metasediments of the Northern Steep Belt show relics of a blueschist facies metamorphism, although the age is not known. The Chiavenna ophiolite shows no evidence for any subduction related metamorphism. The metamorphic ages (Barrovian?) in the Chiavenna ophiolite varies between 42-35 Ma (Talerino 2002; Liati et al. 2003). Geometrical consideration indicates, that the main suture in the Central Alps is part of the Valais Ocean (Froitzheim et al. 1996; Schmid et al. 1996, 2004a). In this view, the trace of the Piemonte-Liguria Ocean is represented by the Avers and Platta nappes. This hypothesis is further supported by the occurrence of Briançonnais units in the collisional nappe stack between the Adula nappe complex and the Avers Nappe (Piemonte-Liguria Ocean).
However, relics of the Piemonte-Liguria Ocean have also been proposed to occur in the Southern Steep Belt (Stucki et al. 2003). The Piemonte-Liguria Ocean would be missing in a cross section through the Southern Steep Belt by assuming the Adula and other ophiolites are only the relics of the Valais Ocean. This situation can be explained by later deformation along the Periadriatic Lineament.
Alternatively, the Adriatic plate may have served as an upper plate for a long time, with fragments of different paleogeographic units accreting to this upper plate. This would require strong tectonic transport to have occurred early in the tectonic history, whereas later tectonic movements were more coherent. This would allow juxtaposition of different fragments within the tectonic mélange units. In this view the prograde Lu/Hf age at ~61 Ma can be related to an early fragment along this upper plate.
What is the position of the Malenco-Platta system?
In eastern Switzerland, a part of the Piemonte-Liguria Ocean is preserved with its passive continental margin between the Adria continent and the ocean (e.g. Manatschal & Bernoulli 1999; Müntener et al. 2000). These passive continental margin and oceanic crust include the Platta-, Malenco-, Forno-, Lizun and Arosa units. The Platta-Malenco-system was dominated by west directed thrusting, followed by greenschist facies overprinting during Cretaceous times (e.g. Deutsch 1984; Ring 1992; Handy et al. 1996; Villa et al. 2000), whereas other parts of the Piemonte-Liguria Ocean (e.g. Voltri, Monviso-, Zermatt-Saas Fee units) were subducted during early Tertiary times. The stacking of the Platta Nappe is towards the West, which is incompatible with a south-directed subduction in the Paleogene (in the central and eastern Alps). All these data are in contrast to a joint evolution for the Tauern, Zermatt-Saas Fee and Platta units along the same plate boundary. Froitzheim et al. (1996) proposed that the Austroalpine units and the Platta unit were both subject to Cretaceous thrusting towards the west, with a migrating deformation front from east to west. However, the Platta-Malenco system is part of the Piemonte-Liguria Ocean and not part of the Meliata Ocean (as indicated by the age of the sediments).
This interpretation includes major thrusting in Cretaceous times when the orogenic evolution is related to closing of the Meliata Ocean and the thickening of the Adriatic plate. The thrusting of the Platta unit requires maintaining open oceans in order to deposit the preserved younger sediments. This requires an extraordinary shape and position of ocean-continent transition at that time and a large transfer zone somewhere in the northern part of the eastern Alps (note the questionmarks in Fig. 6a). One possible solution would involve a large-scale plate boundary with curvilinear geometry (Fig. 6) or, alternatively, a series of large transform faults (Beccaluva et al. 1994, Stampfli et al. 1998, 2002). The first scenario would allow for west directed thrusting of part of the Piemonte-Liguria Ocean and preservation of a open Ocean with their magmatic crust, which were subsequently subducted elsewhere.
In dependent from such proposals, the Platta-Err and Malenco units are already in the Paleogene are slice of the hanging plate (orogenic lid: Schmid et al. 1990). This does not exclude incorporation of metasediments of the Piemonte-Liguria Ocean at younger times (see relations of the Avers Nappe to the underlying Briançonnais units). At present the appears no clear solution to this problem, but large changes in plate movement directions have to be assumed at the Cretaceous/Tertiary boundary (Le Pichon et al. 1988; Schmid & Kissling 2000).
 Figure 6: Sketch of the possible position of the Platta-Malenco units in Upper Cretacous times. (a) Situation before the Platta greenschist-facies metamorphism. Note the transfer zone located in the north of the Alps (located today at the north of the Northern Calcareous Alps). (b) Situation after nappe stacking of the Platta-Err system. Note the change in convergence direction. |
Summary and Conclusion
There are an increasing number of isotope ages on the HP/LT metamorphism in the Alps. Taking all ages in account, a 35 million-year history of HP/LT metamorphism is preserved in the western and central Alps (we exclude the middle Cretaceous HP history in the Saualpe/Koralpe and related areas; see Thöni 2006). The compiled data show that preserved ages of HP/LT metamorphism are spread over the entire time interval, which is assumed for the subduction process. This is in contrast to the common view of single exhumation events for any one unit and the preservation of only ”one“ HP event. The preserved stages of subduction in the Alps present a problem of possible ongoing subduction while exhumation is occuring. In examples of internal deformation in the HP/LT metamorphism leads to variation in ages depending on the individual position of the fragment (see for example Schiste Lustré of the Cottian Alps, Agard et al. 2002; Central Alps, Brouwer et al. 2005). In contrast, the mechanisms and dynamic of removal of large pieces of the oceanic crust during subduction is not understand at this point of research. The dilemma of the completely different P-T-t evolution of the Platta nappe and the Piemont-Liguria units in the Western Alps require a different geodynamic settings along the same plate boundary. We proposed a scenario, which requires complex plate boundary geometries or a series of transfer faults.
The summary of HP/LT data indicate ongoing subduction at contemperanous accretion of fragments along the destructive plate boundary. However, there are fundamental differences recorded in the size and type of fragmentation during subduction. Small-scale fragmentation along a plate boundary may be active during ongoing subduction. This process produced mélange zones, but the driving force for subduction remained. Therefore, it may be possible to find a number of different time steps of an ongoing subduction history preserved in a mélange. In contrast, the incorporation of large, oceanic slices in the nappe pile may predominately occur during a change in plate arrangement.
Acknowledgments
We thank I. Mercolli and O. Müntener for comments on a first draft of the paper. Schweizerischer Nationalfonds has supported our research over several years (2000-055306.98, 20-63593.00, 20020-101826, and 200020-109637). Giorgio Dal Piaz provide a extended review and additional literature data and Gerhard Franz a careful review. Carl Spandler improved the english. We thank all of them very much.