Three generations of monazite in Austroalpine basement rocks to the south of the Tauern Window: evidence for Variscan, Permian and Eo-Alpine metamorphic events

Abstract

Three monazite generations were observed in garnet-bearing micaschists from the Schobergruppe in the basement to the south of the Tauern Window, Eastern Alps. Low-Y monazite of Variscan age (321 ± 14 Ma) and high-Y monazite of Permian age (261 ± 18 Ma) are abundant in the mica-rich rock matrix and in the outer domains of large garnet crystals. Pre-Alpine monazite commonly occurs as polyphase grains with low-Y Variscan cores and high-Y Permian rims. Monazite of Eo-Alpine age (112 ± 22 Ma) is rarer and was observed as small, partly Y-enriched grains (3 wt. % Y₂O₃) in the rock matrix and within garnet. Based on monazite-xenotime thermometry, Y + HREE values in monazite indicate minimum crystallization conditions of 500 °C during the Variscan and 650 °C for the Permian and Alpine events, respectively. Garnet zoning and thermobarometric calculations with THERMOCALC 3.21 record an amphibolite facies, high-pressure stage of ~600 °C/13–16 kbar, followed by a thermal maximum at 650–700 °C and 6–9 kbar. The Eo-Alpine age for these two events is supported by inclusions of Cretaceous monazite in the garnet domains used for thermobarometric constraints and through the high growth temperatures of Eo-Alpine monazite, which is consistent with that of the thermal maximum (~700 °C). The age and growth conditions of a few Mn-rich garnet cores, sporadically present within Eo-Alpine garnet, are unclear because inclusions of monazite, plagioclase and biotite necessary for thermobarometric- and age constraints are absent. However, based on monazite thermometry, Permian and Variscan metamorphic conditions were high enough for the growth of pre-Alpine garnet. The formation of Variscan garnet and its later resorption, plus Y-release, would also explain the high Y in Permian monazite, which cannot originate from preexisting Variscan monazite only. Monazite of Variscan, Permian and/or Eo-Alpine ages were also observed in other garnet-bearing micaschists from the Schobergruppe. This suggests that the basement of the Schobergruppe was overprinted by three discrete metamorphic events at conditions of at least lower amphibolite facies. While the Variscan event affected all parts of this basement, the younger events are more pronounced in its structurally lower units.

1 Introduction

The Austroalpine basement immediate to the south of the Tauern Window within the Northern-Defereggen-Petzeck Group (the Schobergruppe as a part of it) was overprinted during the Eo-Alpine orogeny at variable metamorphic conditions (e.g., Schuster et al. 2004; Schmid et al. 2004; Schulz et al. 2005). According to some authors (Hoinkes et al. 1999; Linner et al. 2000) and as outlined in the metamorphic map of the Alps (Frey 1999a, b; Neubauer et al. 1999; Hoinkes et al. 1999; Schmid et al. 2004), even Eo-Alpine high-pressure amphibolite facies conditions were reached.

However it is difficult to assess the pre-Alpine metamorphic evolution in this part of the basement, in particular the Permian and Variscan events, since the older metamorphic assemblages are mostly overprinted and common geochronometers largely re-equilibrated. Accessory phases like monazite or zircon are usually refractory even at high-grade metamorphic overprints (e.g. Rubatto et al. 2006) and may therefore better record earlier thermal events than common minerals. However, although highly refractory, monazite shows certain reactivity as well and can crystallize at almost every P–T conditions, including very low grades (Read et al. 1987; Evans et al. 2002; Rasmussen et al. 2001; Rasmussen and Muhling 2007, 2009; Wan et al. 2007; Wilby et al. 2007; Biševac et al. 2011). As a consequence, several generations of monazite may be observed within a single sample or even a single grain (e.g., DeWolf et al. 1993; Cocherie et al. 1998; Pyle 2006; Spear et al. 2008; Petrík and Konečný 2009; Martins et al. 2009; Schulz and von Raumer 2011).

Apart from geochronological aspects, the chemical composition of monazite is a valuable petrogenetic indicator. The incorporation of Y and HREE in monazite follows a solvus and can be used as a thermometer, the so called monazite–xenotime miscibility gap thermometer (Gratz and Heinrich 1997; Heinrich et al. 1997; Pyle et al. 2001). In addition, Y-zoning in monazite may be correlated with distinct stages of garnet growth or breakdown (Pyle and Spear 2003), as garnet highly fractionates this element. Y will thus be less available for monazite when garnet is crystallizing simultaneously and the other way around when garnet breaks down (e.g. Zhu and O`Nions 1999a, b; Pyle et al. 2001; Krenn et al. 2009; Martins et al. 2009). Therefore the Y distribution in monazite can be useful in clarifying if garnet growth occurred prior to, at the same time as or after monazite.

In this study, we present data from polymetamorphic Schobergruppe rocks hosting distinct generations of monazite with specific Y contents. Three monazite generations have been identified and corroborated by comparison with other samples from this region. Variable Y contents evidence the existence of pre-Alpine metamorphic assemblages with simultaneous crystallization of garnet (e.g., Gaidies et al. 2008; Pyle and Spear 2003).

2 Geological setting

The studied samples are micaschists from the Austroalpine basement of the Schobergruppe (provinces Carinthia and Tyrol in Austria; Fig. 1a), which in turn belongs to the Northern-Defereggen-Petzeck Group (Fig. 1b). This part of the Austroalpine basement nappe is located to the south of the central to eastern Tauern Window (Schulz 1993a, b). The samples were taken close to eclogitic amphibolites (Fig. 1b, d) described by Schulz (1993a) and Schulz et al. (2005). Other micaschist samples from the Schobergruppe or from adjacent parts of the Austroalpine basement (Fig. 1b, e) have been considered for age control. Some of them were located in areas free of Alpine overprint, according to available Carboniferous K–Ar and Rb–Sr mica ages (Fig. 1c).

Fig. 1

Geological map showing the Schobergruppe (SG) within the tectonic framework of the Eastern Alps (a) and in more detail with sample location (b) (modified after Schulz et al. (2005, 2008). (c) Distribution of K–Ar and Rb–Sr mica ages to the south of the Tauern Window (compiled from Borsi et al. (1978); Hoinkes et al. (1999); Schuster et al. (2001)). Cross section through the Schobergruppe and Prijakt eclogitic amphibolites (d) and through the Austroalpine basement to the south of Hopfgarten (e). Sample locations and Th–U–Pb chemical ages of monazite are also shown

As in many crystalline complexes in the Eastern Alps, the Northern-Defereggen-Petzeck Group was overprinted by Alpine metamorphism during Cretaceous times (e.g. Schuster et al. 2004; Schulz et al. 2008). This metamorphic overprint reached greenschist to high-pressure amphibolite facies conditions (Exner 1962; Oxburgh et al. 1966; Troll and Hölzl 1974; Troll et al. 1976, 1980; Schulz 1993a; Linner et al. 2000). In high-grade zones, relicts of the pre-existing Devonian-Carboniferous Variscan metamorphism are often blurred.

So far, Variscan amphibolite facies metamorphism has been reported mainly from the basement south and southeast to the Schobergruppe (Schulz 1990, 1993b; Schuster et al. 2001; Schulz et al. 2005, 2008; Steidl et al. 2009, 2010a, b), where mica cooling ages do not record significant Alpine overprint (Fig. 1c). From a tectonic point of view, the area to the south of the so-called Defereggen-Antholz-Vals shear zone (DAV) belongs to an upper part of the Austroalpine basement, while the Schobergruppe represents a structurally lower part.

Variscan ages in metapelitic basement rocks south to the Tauern Window are generally rare. Schulz et al. (2005, 2008) for instance report Variscan monazite from the Schobergruppe and Steidl et al. (2010a, b) Variscan monazite from the Michlbach Complex, in the eastern Defereggen Alps, south to the DAV. Geochronological data to the east of the Schobergruppe range from 90 to 310 Ma (Hoke 1990; Schuster et al. 2001) and are interpreted as a mixture of Variscan and Eo-Alpine ages. Similarly, Permian ages in metapelitic rocks were also interpreted as mixed ages (Carboniferous to Cretaceous) or were ascribed to a slow Variscan cooling history (e.g. Borsi et al. 1978). So far, traces of a Permian event in the Schobergruppe seem to be restricted to the emplacement of pegmatite bodies and related HT/LP overprint nearby. Although Permian pegmatites are very common in the basement of the Schobergruppe (Bücksteeg 1999), it is not clear if this area was also affected by a Permian HT/LP event as is the case in other crystalline complexes south and southeast to the Schobergruppe (Schuster and Stüwe 2008).

3 Samples

Monazite and garnet were studied in detail in three garnet-bearing micaschist samples (520, 527 and Alk 8a) located in the vicinity of eclogitic amphibolites. The rocks are strongly foliated with modal abundances of approximately 5–10 vol. % garnet, 20–25 % quartz, 15–20 % plagioclase, 15–20 % muscovite, 10–15 % biotite, 3–5 % staurolite, 1–5 % chlorite and <3 % kyanite. Accessory phase are apatite, zircon, monazite and xenotime. The prevailing Ti-phase is ilmenite (0.5–2 vol. %).

Garnet is variable in size and shape, ranging from small (<300 μm) isometric, partly, rounded grains up to very large (several mm) eu- to subhedral crystals (Fig. 2). Under the microscope, garnet appears relatively homogeneous and not polyphase with little marginal breakdown and alteration. Inclusions of ilmenite, quartz, biotite, muscovite and plagioclase are common everywhere in the garnet. Garnet was also observed around staurolite (Fig. 2) or occurs intergrown with staurolite and/or mica-rich batches with remnants of staurolite (Fig. 2). Staurolite forms up to 5 mm large, eu- to subhedral crystals, sometimes twinned, sometimes elongated parallel to the foliation. In some places, staurolite occurs together with kyanite (see below). Kyanite is scarce and occurs as small grains or clusters (partly fibroblastic) associated with mica and/or staurolite.

Fig. 2

Microphotographs showing garnet crystals associated with staurolite from sample 520 (a) and 527 (b)

Monazite from other garnet-bearing micaschists in this area was studied by Schulz et al. (2005, 2008). The corresponding samples were taken in the vicinity and below the eclogitic amphibolites to the south of the Oligocene Defereggen-Antholz-Vals shear zone (Fig. 1d, e). The samples were situated within and above a pegmatite-rich zone and within the monotonous metapsammopelitic Defereggen Group. The age of metamorphism remained unclear and the peak P–T conditions were considered pre-Alpine in age (Schulz et al. 2005, 2008). This is why we have revisited this area and resampled metapelites with zoned monazite as well as monazite inclusions in garnet, in order to establish correlations between monazite- and garnet growth stages.

4 Monazite

4.1 Chemical Th–U–Pb monazite dates

Backscattered electron imaging (BSE), phase identification with an energy-dispersive system (EDS) as well as dating and chemical analyses using the wavelength-dispersive system (WDS) were carried out with a JEOL-JX8600 electron microprobe following the procedures described in Krenn et al. (2008) and Krenn and Finger (2004). Single monazite dates and weighted average ages (Table 1) were calculated utilizing the method of Montel et al. (1996) and isoplot 2.1 (Ludwig 2001) considering the analytical 2 sigma errors on Th, U and Pb measurements. The analytical Pb-errors range from 0.010 to 0.016 for a Pb dwell time of 160–400 s.

Table 1 Th, U, Pb concentration, Th* values, ages and 2 sigma errors of monazites

The statistical distribution of measured monazite ages shows three peaks corresponding to Carboniferous (Variscan), Permian and Cretaceous (Eo-Alpine) times, respectively (Fig. 3). Data cluster around 300–330, 250–280 and 70–135 Ma, respectively, whereas a few dates between 220 and 250 Ma and ~300 Ma are interpreted as a mixture of ages due to the polyphase nature of monazite (see below). The weighted average dates per sample are listed in Table 1. In the Th* versus total-Pb diagram after Suzuki et al. (1991), pre-Alpine and Alpine monazite analyses arrange themselves along three trendlines (Fig. 4), which provide isochron ages of 321 ± 14 Ma, 261 ± 18 Ma and 112 ± 22 Ma, respectively. The Variscan trendline is characterized by a slope of 0.01429x (±0.0006x), an interception value of ~0 and a MSWD value of 0.4; the Permian regression defines a slope of 0.0117x (±0.0008x), an intercept value of ±0.0007 and a MSWD value of 0.3 and the Alpine trendline has a slope of 0.0055x (±0.0013x), an intercept value of 0.001 and a MSWD value of 0.24.

Fig. 3

Frequency diagram of monazite ages in samples 520, 527 and Alk 8

Fig. 4

Th* vs. Pb diagram after Suzuki et al. (1991) showing the trendlines defined through Alpine, Permian and Variscan monazite. Age bars on the right site are isochrones forced through 0

Variscan, Permian and/or Eo-Alpine (Cretaceous) monazite were also observed in other samples from the Schobergruppe and from the Defereggen Group, south to the DAV (Schulz et al. 2005, 2008). Sample HPr10 shows Permian and abundant Cretaceous monazite ages (Fig. 1b, d; Table 2), samples Alk2, Sti14 and P24 preferentially Permian ages and samples Alk 8, 839, 431b and 750b Variscan (Carboniferous) monazite ages (Fig. 1d, e; Table 2). It is important to note that it was beyond the scope of these earlier studies to analyze all monazite grains in a sample. Therefore, it is likely that other monazite generations would be found as well if studied in more detail and the absence of a monazite age in these samples should not be over- or misinterpreted.

Table 2 Th-U–Pb CHIME model ages of monazite from other garnet-bearing mica schists from the Schobergruppe and adjacent Austroalpine basement, as reported in Schulz et al. (2005)

4.2 Morphology and chemistry of monazite

In all three samples investigated in this work, Variscan and Permian monazite form morphologically indistinguishable, relatively large eu- to subhedral grains (ca. 20–150 μm) (Fig. 5a, b). Some crystals show straight grain boundaries, others are irregular and embayed (Fig. 5b, c). Monazite grains were observed in the matrix as well as inclusions in staurolite and in the outermost domains of large garnet crystals (Fig. 5a), but not in the cores of the latter.

Fig. 5

Backscattered electron images showing monazite from samples 520, 527 and Alk 8. Polyphase monazite (Variscan core, Permian rim) enclosed in the outer domain of garnet (a, b). Polyphase matrix monazite with a Variscan core (labeled as analysis 1), Variscan rim (analysis 2) and an outermost Permian rim (analysis 3) (c). Polyphase matrix monazite with a Variscan core and Permian rim (d, e). Alpine monazite enclosed in garnet. (h) Alpine monazite in the staurolite-kyanite bearing matrix (f, g). Chemistry and ages of monazite are listed in Tables 1, 2, 3, and garnet compositions are shown in Figs. 10 and 11

Approximately 1/3–1/2 of all monazite grains display a characteristic zoning in BSE images with bright cores (enriched in thorium; up to 16 wt.  % ThO₂) and darker rims with lower Th contents (Table 3). Dating of these rim zones revealed that most of them are Permian in age, although both Variscan and Permian rims were locally observed (Fig. 5c). In Fig. 5c, a bright Th-rich Variscan core is surrounded by an optically darker Variscan rim, which in turn is followed by an outermost Permian rim (Fig. 5c; Table 3). Variscan domains (cores and rims) are low in Y (<1 wt.  % Y₂O₃), while Y₂O₃ values in Permian rims are systematically (0.5–1.5 wt.  %) higher than in the corresponding Variscan core (Table 3). In average, Variscan monazite shows higher Th- and lower Y contents than Permian monazite (3–16 wt.  % ThO₂ and 0.1–1wt.  % Y₂O₃ vs. 1–8 wt.  % ThO₂ and 0.2–2.5 wt.  % Y₂O₃; Figs. 6, 7). In addition, Variscan monazite provides a slight positive covariation between Th and Y, while Th and Y in Permian monazite are negatively correlated or not correlated at all (Fig. 7).

Table 3 Selected microprobe analyses of monazites

Fig. 6

Th versus Y trends of Variscan, Permian and Alpine monazite

Fig. 7

Chemical composition of Variscan and Permian monazite

Eo-Alpine monazite occurs in small and unzoned grains (<10 μm), which, in some places, arrange themselves in clusters (Fig. 5f, g). Monazite in sample Alk8 was only observed in the matrix, while monazite in samples 520 and 527 was found in the matrix as well as enclosed in small, optically homogeneous and widely unaltered garnet crystals (Fig. 5f, g). Eo-Alpine monazite is not texturally associated with allanite (as is in sample HPr10 from Schulz et al. 2005, 2008) or with pre-Alpine monazite, but was found together with or nearby xenotime (Fig. 5h). Eo-Alpine monazite yields ca. 1.5–6 wt.  % ThO₂ and shows a considerable Y-variation from ca. 1 wt. % Y₂O₃ (as is the case for monazite in garnet) up to 2.9 wt. % Y₂O₃ (Fig. 6; Table 3). The highest Y-values were observed in a matrix at the boundary of staurolite-kyanite-muscovite aggregates and in the close vicinity to xenotime crystals (Fig. 5). Xenotime is common in all samples and occurs as small grains or in clusters in the matrix and enclosed in garnet (see below).

4.3 Monazite-xenotime thermometry

Minimum growth temperatures of monazite were obtained utilizing monazite-xenotime miscibility gap thermometers. Figure 8 shows the experimentally calibrated 2 kbar thermometer curve of Gratz and Heinrich (1997), the empirical thermometer of Heinrich et al. (1997), which was obtained on low-pressure rocks and the empirical thermometer of Pyle et al. (2001) calibrated on high pressure rocks. Also shown is the experimentally 2 kbar curve for huttonite-bearing monazite with ca. 10 % huttonite (Seydoux-Guillaume et al. 2002). This thermometer curve runs sub-parallel to the 2 kbar curve of Gratz and Heinrich (1997) at ca. 30–50 °C lower temperatures. The Gratz and Heinrich (1997) thermometer was calibrated in the binary CePO₄–YPO₄ system disregarding the stronger Y-uptake of huttonite-bearing monazite.

Fig. 8

Xenotime versus temperature diagram showing the published monazite-xenotime miscibility gap thermometer curves and the minimum formation temperatures of Variscan, Permian and Alpine monazite under study

The highest xenotime value observed in Permian monazite (~8 mol %) intersects the low-pressure curves of Heinrich et al. (1997); Gratz and Heinrich (1997) and Seydoux-Guillaume et al. (2002) between ca. 650 and 700 °C (Fig. 8). The highest xenotime value observed in Eo-Alpine monazite (ca. 10 mol %) intersects the thermometer curve of Pyle et al. (2001) at temperatures of ca. 650 °C, those calibrated for low-pressure rocks at ~700 °C. Maximum xenotime contents observed in Variscan monazite (5 mol %) implies growth temperatures between 450 °C (Pyle et al. 2001) and 550 °C (Heinrich et al. 1997). Table 4 lists the (minimum) formation temperatures calculated utilizing the thermometer-functions given by Gratz and Heinrich (1997); Heinrich et al. (1997) and Pyle et al. (2001).

Table 4 Xenotime values of monazite and corresponding monazite-xenotime miscibility gap temperatures

5 Major phase composition and geothermobarometry

Most garnet crystals, including those with inclusions of Eo-Alpine and Permian monazite (Fig. 5a, f) and those surrounding staurolite (Fig. 9b), show a simple core to rim zoning with a rimward decrease of MnO and CaO (1–0.1 wt. %; 3–0.1 wt. %) and increase of MgO (2.5–4 wt. %; Fig. 10). However, a few of the larger garnet crystals also yield a Mn-enriched core (plateau) with 3–4 wt. % MnO and 1–2 wt. % CaO and MgO (Fig. 11), followed by a rim domain, which is characterized by an abrupt decrease of MnO and increase of CaO and MgO (up to ca. 3–4 wt. %). The composition of the latter rims is similar to the garnet crystals shown in Fig. 10. A narrow retrogressive rim (max. 100–200 μm) is common in all investigated garnet grains.

Fig. 9

Backscattered electron images of garnet from samples 520, 527 and Alk 8. Euhedral garnet with a simple zoning and a narrow alteration rim (a). Garnet around staurolite shown in Fig. 2b (b). Large garnet with diffuse zoning and inclusion of Permian monazite in its outer domains (c). Large garnet with diffuse zoning and inclusion of xenotime. Chemical profiles are shown in Figs. 10 and 11 (d)

Fig. 10

Chemical composition of low Mn-garnet, shown in Figs. 5a, f and 9a, b

Fig. 11

Chemical composition of high Mn-garnet grains, shown in the Fig. 9c and d

Biotite is chemically homogeneous with a Mg/(Mg + Fe) of 0.4–0.5 and Ti contents from 1.5 to 2 wt % TiO₂, whether in the matrix or enclosed in garnet. Muscovite usually shows a weak diffuse zoning with slightly varying Na- and Ti values (1.3–2.5 wt. % Na₂O and 0.6–1.5 wt. % TiO₂). Plagioclase composition ranges from An₂ in the core to An₁₂ at the rim.

An independent set of reactions obtained with the THERMOCALC software 3.21 (Holland and Powell 1998) yields P–T estimates of ca. 600–700 °C and 6–9 kbar for the assemblage biotite, muscovite, plagioclase-core and Mg-enriched garnet domains, and 550–600 °C and 14–16 kbar for the assemblage biotite, muscovite, plagioclase-rim and grossular-enriched garnet domains. We are aware that pressure estimates suffer from low An-contents of the plagioclase (e.g. Todd 1998). However similar P–T estimates of 650–700 °C/8–10 kbar and 550–650 °C/12–14 kbar were obtained by Schulz et al. (2005) for other low- and high-Ca metapelites from the Schobergruppe. These authors used the garnet-biotite thermometer of Bhattacharya et al. (1992) in combination with the garnet-muscovite-biotite-plagioclase barometer of Holland and Powell (1990) with updated activity models from Ganguly et al. (1996) and Powell and Holland (1993). The P–T estimates obtained from metapelites overlap with those obtained from eclogites and eclogitic amphibolites (Schulz 1993a; Schulz et al. 2008), which are intimately associated with the studied metapelites (see Fig. 1b and discussion below).

6 Discussion

6.1 Formation of pre-Alpine monazite

The occurrence of pre-Alpine monazite in the Schobergruppe samples is a clear evidence that parts of this area were pervasively metamorphosed during Variscan and Permian times. We conclude from the monazite-xenotime miscibility gap thermometry that the Variscan and Permian events reached temperatures of at least 500 and 650 °C, respectively (Fig. 8; Table 4). However, many Variscan monazite grains show xenotime contents of <5 mol % possibly because they crystallized at lower temperatures along the prograde Variscan metamorphic path or because they failed to attain maximum Y contents. Growth of monazite at temperatures much below 500 °C is not consistent with its high Th contents (up to 16 wt. % ThO₂) and with its large, euhedral crystals. Low-T monazite usually has much lower Th contents and/or its crystals are smaller and sub- to anhedral (e.g., Read et al. 1987; Rasmussen et al. 2001; Evans et al. 2002; Kryza et al. 2004; Wilby et al. 2007; Krenn et al. 2008; Del Rio et al. 2009; Biševac et al. 2011; Čopjaková et al. 2011). In addition monazite often occurs in greater abundances at upper greenschist to lower amphibolite facies conditions in many metapelitic rocks (Kingsbury et al. 1993; Lanzirotti and Hanson 1996; Wing et al. 2003).

It is likely that Variscan monazite with lower Y contents grew at a stage when Y was trapped in other phases like garnet. Garnet has indeed a high affinity for Y (Pyle and Spear 1999; Pyle et al. 2001) and this element will be less available for monazite if garnet growth occurs before monazite; alternatively, Y will be released and made available when garnet breaks down. Thus low Y contents in some Variscan monazite grains could be explained by its uptake in garnet porphyroblasts, which crystallized at Variscan times prior to monazite. Subsequent garnet resorption coupled to Y release, could be responsible for the much higher Y contents observed in Permian monazite (Fig. 7), which cannot come from the Variscan low-Y monazite generation alone. It is also likely that xenotime crystallized during this period of garnet resorption (e.g. Pyle et al. 2001). Possibly some of the xenotime grains observed within garnet formed together with Permian high-Y monazite.

The replacement of Variscan, low-Y monazite by Permian, high-Y monazite might be explained in terms of dissolution and reprecipitation processes caused through the high temperatures and higher Y activity during the Permian event. Retrogression of the samples during the Permian coupled with liberation of Y could be a plausible explanation for the onset of abundant Permian monazite. Variscan low-Y monazite grains probably recrystallized in order to adjust their Y contents to the higher temperatures during the Permian event. This is a recrystallization mechanism that has been reported in other studies as well (e.g. Janots et al. 2008). A high strain rate (Berger et al. 2006) and a high degree of retrogression/alteration (Lanzirotti and Hanson 1996; Poitrasson et al. 1996, 2000; Krenn and Finger 2007; Budzyń et al. 2011) are also favorable for monazite recrystallization. Indentions and protrusions observed at the contacts between distinct monazite age zones (Fig. 5) suggest that pre-existing Variscan monazites were marginally replaced through a dissolution and reprecipitation mechanism (e.g. Putnis 2002, 2009). On the other hand, the euhedral shape of Variscan monazite cores also indicate that some of them were probably overgrown by neighboring monazite substance. This could be an explanation for the lower Th-contents in some Permian monazite rims compared to the corresponding Variscan monazite cores.

6.2 Formation of Eo-Alpine monazite

Although monazite can survive high-grade metamorphism and even migmatite-grade conditions (e.g. Zhu and O`Nions 1999a, b; Martins et al. 2009), it is surprising that so many pre-Alpine monazite grains survived the Eo-Alpine amphibolite facies event. It is assumed that pre-Alpine monazites remained unaffected because their Y contents (at least in Permian monazite) were high and the degree of retrogression of the samples during the Eo-Alpine event was low. Krenn and Finger (2007) showed that the pre-existing Variscan monazite generation in polymetamorphic rocks from Crete remained largely unchanged in quartz-rich layers but barely survived in mica-rich layers of the same rock. The results showed that a lack of monazite age generation should not be misinterpreted as indicating that the degree of a metamorphic event was low or that a metamorphic event did not occur at all. Krenn and Finger (2007) also showed that the younger monazite generation likely formed in allanite-bearing domains, while allanite-absent domains hardly yielded young monazite. Indeed, in sample HPr10 from the Schobergruppe, where Eo-Alpine monazite is abundant (Schulz et al. 2005), allanite is common as well. It served as a direct precursor to Eo-Alpine monazite, as reported in other studies (e.g. Wing et al. 2003; Janots et al. 2006, 2008). No allanite formed in samples 520, 527 and Alk8, probably because of a low bulk Ca content, which is more favorable to the growth of monazite relative to that of allanite according to Janots et al. (2007) and Spear (2010). Allanite is often the prevailing REE-bearing phase in metapelites up to lower amphibolite facies conditions, where it reacts to monazite (e.g., Wing et al. 2003). The allanite to monazite transition has a slight positive slope in P–T space (Janots et al. 2007; Spear 2010) and moves towards lower temperatures with decreasing CaO whole-rock content or with increasing Al₂O₃ content. This results in a decrease or complete disappearance of the allanite stability field.

6.3 Growth of garnet

We identified three growth stages of garnet (M1–M3). M1 is observed sporadically in the cores of a few garnet crystals and is characterized by high Mn and low Mg values. M2 shows high Ca and intermediate Mg values and can be related to a high-pressure amphibolite facies stage at 550–600 °C/13–16 kbar. Garnet growth zone M3 prevails in many garnet crystals and is characterized by decreasing Ca and increasing Mg at low Mn contents resulting from a thermal maximum at 650–700 °C/6–9 kbar. The following assemblages (including quartz and muscovite) were observed together with the garnet growth zones M1–M3: M1 − garnet + chlorite; M2 − garnet + chlorite + biotite + plagioclase + monazite + xenotime ± staurolite ± kyanite; M3 − garnet + chlorite + biotite + plagioclase + staurolite + kyanite + monazite + xenotime. M2 and M3 stages were also documented in other micaschists from the Schobergruppe (Schulz et al. 2005, 2008) and match the thermobarometric results obtained from eclogitic amphibolites (Schulz 1993a; Schulz et al. 2008). Eclogitic amphibolites (Schulz 1993a; Schulz et al. 2008) contain a high-pressure assemblage of garnet plus clinopyroxene, which can be related to the M2 stage observed in metapelites and a subsequent high-temperature assemblage with Ca-amphibole plus zoisite, which can be ascribed to the M3 stage in metapelites. Eclogitic amphibolites are situated structurally below and above the micaschists with parallel planar-linear structures indicative of a common deformation history. P–T estimates for M2 and M3 are also in agreement with those obtained for other basement rocks from the Schobergruppe (Hoinkes et al. 1999; Linner et al. 2000).

However, while Hoinkes et al. (1999) ascribed the M2 and M3 stages to a Cretaceous event, Schulz (1993a) and Schulz et al. (2005, 2008) proposed a Variscan age. Results from this study, in particular the observation of Eo-Alpine monazite inclusions within M2 and M3 garnet domains, suggest that the M2 and M3 stages are Eo-Alpine rather than pre-Alpine. It is unlikely that the Eo-Alpine monazite grains shown in Fig. 5e crystallized after they were enclosed in garnet. There are no textural evidences such as pseudomorphs, recrystallization patches or overgrowths, which would argue for a later formation of monazite. Also, there are no cracks visible along which REEs could have entered garnet. In addition, REEs are relatively immobile during regional metamorphic processes and migration of REEs is expected rather at low metamorphic grade associated with diagenetic-, hydrothermal- and metasomatic processes (see Čopjaková et al. 2011 and references therein). Moreover, the high xenotime content observed in some Eo-Alpine monazite is also consistent with crystallization during the M3 thermal peak.

6.4 Pre-Alpine development of garnet

The age and growth conditions of M1-related garnet are not well known because they lack inclusions of biotite, plagioclase and monazite. Although we do not know if M1-related garnet domains represent pre-Alpine remnants, this is likely for several reasons. According to the monazite thermometry, samples 520, 527, Alk 8a were metamorphosed at temperatures of 500 and 650 °C during the Variscan and Permian events, respectively. This is high enough for the growth of garnet. In addition, growth of Variscan and Permian garnet has been documented from other metapelitic samples from the basement south of the Tauern window (Schuster et al. 2001; Steidl et al. 2010a, b). Steidl et al. (2010a, b) proposed P–T conditions of 3.8–5.8 kbar and up to 650 °C for the Permian event and assumed amphibolite facies conditions for the Variscan event. Similarly our results would argue for a high-temperature Permian event and a Variscan event at lower amphibolite facies conditions.

According to the structural position of some of the studied metapelites (Fig. 1b, d), Permian recrystallization in these rocks may be related to the intrusion of Permian pegmatites. However, Permian monazite in other metapelites, which occur at a greater distance to pegmatites (Fig. 1), would rather argue for a discrete Permian HT/LP event. This is supposed by Schuster and Stüwe (2008) for the metamorphic complexes south and southeast of the Schobergruppe (Jenig and Strieden Complexes). Moreover, the large number of pegmatites across the entire basement of the Schobergruppe (Bücksteeg 1999) is also consistent with a regional Permian HT/LP event (Schuster and Stüwe 2008), which would have been accompanied by partial melting of the lower crust.

Permian metamorphism is widespread in the Eastern Alps (Habler and Thöni 2001; Schuster et al. 2001) and is particularly pronounced in the crystalline complexes of the so-called Wölz-Koralpe nappe system (Schuster et al. 2004). Remnants of Variscan mineral assemblages are rarely observed in rocks from the Wölz-Koralpe nappe system and have been reported for instance from the Strieden and Jenig Complexes (Schuster et al. 2001, 2004), the Michelbach Complex of the Defereggen Mountains (Steidl et al. 2010a, b) or from the Rappold Complex (Gaidies et al. 2008). From these units, the Rappold Complex shows a similarly high-grade metamorphic Eo-Alpine overprint (Gaidies et al. 2008) like the studied samples from the Schobergruppe. Other crystalline complexes of the Wölz-Koralpe nappe system (e.g., Wölz Unit, Saualpe-Koralpe Complexes, Pohorjeh Mountains, Schneeberg Complex) usually contain assemblages indicative for a Permian and/or a Eo-Alpine imprint (e.g., Abart and Martinelli 1991; Schuster and Thöni 1996; Thöni and Miller 1996; Bernhard and Hoinkes 1999; Faryad and Chakraborty 2005; Gaidies et al. 2006; 2008; Schuster et al. 2001).

7 Conclusions

Petrographic and textural evidence from different generations of monazite crystals indicate a Cretaceous Eo-Alpine age for the high-pressure amphibolite facies metamorphic events M2 and M3 recorded in the Schobergruppe. In contrast, relict monazite crystals record pre-Alpine metamorphic events of amphibolite grade, whatever the intensity of the subsequent Alpine overprint. Monazite records Carboniferous, Permian and Cretaceous age groups in agreement with preexisting data in the Schobergruppe, like sample HPr10, which documents a distinct Cretaceous monazite crystallization event in the lower parts of the Northern-Defereggen-Petzeck Group. The finding of Permian monazite ages in samples with both multi and single monazite age populations implies a metamorphism event clearly distinct from those of Variscan and Alpine ages. This study also illustrates the resistance of monazite at very high-grade metamorphic conditions and its potential use as a relict in unraveling the complex history of polymetamorphic rocks.