Geological background of the Ulten Zone and evolution of crustal rocks
Some of the best exposures of fresh garnet peridotites in the Eastern Alps outcrop is in an area located about 40 km west of Bolzano (Fig. 2), on the mountain range (Le Maddalene) that divides the upper Non valley (Val di Non, or Nonsberg) from the Ulten valley (Val d’Ultimo, or Ultental). This area, known as Nonsberg, or Ultental, or Ulten (zone or unit, hereafter Ulten Zone), is part of the Upper Austroalpine domain and has been the subject of intensive geological research (see Martin et al., 1998; Morten et al., 2004 for reviews). The Upper Austroalpine system consists of a metasedimentary cover and upper-to-lower crustal slices derived from the Mesozoic passive margin of the Adria microplate (Dal Piaz, 1993). According to Flügel (1990) and Neubauer and von Raumer (1993), the Austroalpine system of the Eastern Alps represents a remnant of the Variscan belt. North of the Tonale and Giudicarie lines, the Austroalpine system comprises a northern, cover-bearing nappe (Ortler nappe) and a southern, overlying, cover-free nappe (Tonale nappe) (Thöni, 1981).
The Ulten Zone belongs to the Tonale nappe that is separated to the North from the Ortler nappe by the Late Cretaceous Peio Line (Andreatta, 1948). To the south, the Giudicarie Line (a segment of the Periadriatic fault system) tectonically divides the Tonale nappe from the Southalpine domain. The Tonale nappe is subdivided into Tonale and Ulten Zones, which are divided by the Val Clapa and Rumo Lines (Morten et al., 1976-1977). The Tonale Zone is mainly composed of sillimanite-bearing metasediments, metagranitoids with subordinated marbles, calc-silicates and mafic layers affected by retrograde Variscan metamorphism. In addition, an Alpine greenschist-facies overprint locally affects these rocks (Thöni, 1981).
Figure 2. Geological sketch of the Central-Eastern Italian Alps and localization of the Ulten Zone (square).
The Ulten Zone is known as a site of interest for mineral collectors since the mid-nineteenth century (Doblicka, 1852). Before the XXth century, the Ulten Zone was the subject of intense field and petrographic works (Sandberger, 1866; Stache, 1880-1891; Ploner, 1881; Hammer, 1899). A major advance in geological knowledge of the UZ was provided by Andreatta in 1936 who published the geological and structural map (Fig. 3) that already reported the major gneiss and kinzigite that compose the Ulten Zone basement, and showed the metre- to hundred meters-long peridotite bodies (in full red, named oliviniti) disposed along a major structural arrangement. Since 1970, several studies marked a new interest on the Ulten Zone basement (Amthauer et al., 1971; Brenneis, 1971; Morten et al., 1976; Herzberg et al., 1977; Rost and Brenneis, 1978, see also the historical review by Tumiati and Martin, 2003) and culminated into the paper by Obata and Morten (1987) that provided the first modern petrologic account on the spinel- to garnet-facies transition of the Ulten Zone peridotites and the metasomatic reactions governing the formation of garnet-amphibole peridotite. The crustal Ulten Zone lithologies have been successively subdivided into migmatites, garnet-kyanite gneisses and subordinate metagranitoids: an updated map is shown in Figure 4. The barrel-shaped ultramafic lenses have been subdivided into spinel and amphibole-bearing garnet peridotites: these have been recognized to be located between the garnet-kyanite gneiss and the overlying migmatites. Boudins of mafic amphibolites and rare retrogressed eclogites also occur as lenses in the gneiss and migmatites. Different from the Tonale Zone, the Ulten Zone retains well-preserved Variscan high-pressure (eclogite to granulite facies) metamorphic structures and assemblages, only weakly overprinted by Alpine metamorphism (Obata and Morten, 1987; Martin et al., 1994; Godard et al., 1996).
Figure 3. Geological map of the Ulten Zone.
The Ulten zone gneisses and migmatites are strongly foliated (Fig. 5A). The gneisses show mylonitic textures (Martin et al., 1994, 1998; Godard et al., 1996) and are composed by mm- to cm-sized bands of alternating layers with garnet + kyanite + biotite + rutile (Fig. 5B) and with quartz + plagioclase ± K-feldspar, transposed along an early (S1) mylonitic foliation. Such banded structure derive from S1 deformation of former melanosomes and leucosomes related to a pre-S1 partial melting event (Godard et al., 1996). The gneisses grade upward into stromatic and nebulitic migmatites, which mostly show the mineral assemblage quartz + plagioclase + muscovite + biotite + garnet ± kyanite. Strongly residual layers in the gneiss and migmatite domains are almost entirely made of garnet + kyanite (rocks named ultenite, Fig. 5C, D): coexistence of garnet and kyanite in the restitic layers is strong indication that the early partial melting event(s) took place at depth during evolution of the Ulten Zone basement. In the least migmatized domains, amphibolitized mafic lenses still preserve the relicts of a former eclogite paragenesis (Benciolini and Poli, 1993; Hauzenberger et al., 1996; Godard et al., 1996; Del Moro et al., 1999). The migmatites and the enclosed peridotites are cut by trondhjemitic veins (plagioclase ± quartz ± biotite; Martin et al., 1994, 1998; Godard et al., 1996), which have been interpreted as the product of crystallization of calcalkaline deep-crust partial melts mixed with in-situ melts (Del Moro et al., 1999).
Figure 4. Geological map of the Ulten Zone area.
Figure 5. Migmatites and gneiss from the crustal Ulten Zone unit.
The metamorphic-melting history of the Ulten Zone basement can be summarized as follows (Del Moro et al., 1999): i) pre-S1 (Godard et al., 1996) or syn-S1 (Hauzenberger et al. 1996) migmatization by dehydration melting; ii) S1 mylonitic deformation at eclogite-facies conditions; iii) post-S1 injection by exotic melts and further migmatization at eclogite-granulite conditions during decompressional uplift; iv) S2 shearing and retrogression.
Figure 6. Summary of available pressure-temperature paths reconstructed for the Ulten Zone crust.
In the Ulten Zone crust, thermobarometry of the peak high-pressure stage is hampered by extensive retrogression during uplift and post-S1 migmatization. The available estimates define a metamorphic evolution characterized by a clockwise P-T path, with a HP peak followed by thermal relaxation up to maximum temperature and, finally, retrogression under amphibolite- to greenschist-facies conditions. A summary of P-T paths drawn so far for the Ulten crustal rocks is reported in Figure 6. Using multi-equilibrium geothermobarometry, Godard et al. (1996) calculated P-T conditions of 1.0-2.0 GPa, 600-900 °C for the garnet-kyanite gneisses. The large spread of P-T data depends on different assumptions on the water activity (XH2O) in the fluid phase occurring during the metamorphic evolution. In the same study, Godard et al. (1996) estimated pressure conditions below 1 GPa for the migmatites. Based on cation exchange and net-transfer geothermobarometry coupled with multi-equilibrium calculations, Hauzenberger et al. (1996) suggested a metamorphic peak of at least 1.5 GPa and 750°C. According to Tumiati et al. (2003), the prograde metamorphic evolution of the Ulten Zone crust reached a pressure peak near or within the coesite stability field (> 2.5 GPa, T ≈ 800°C), and eclogites and gneisses experienced peak high-pressure conditions similar to those of garnet peridotites. Mineral inclusions in large kyanite porphyroblasts along with calculated P-T pseudosections presented by Braga et al. (2007) allowed reconstruction of a complete P-T path characterized by a prograde epidote-amphibole-facies stage (P-T ≈ 0.85 GPa, 600 °C), a pressure peak at about 1.1-1.2 GPa and a subsequent temperature increase up to ~ 750 °C. Braga and Massonne (2008) suggested that the occurrence of P-T paths characterized by different peak conditions in a single tectonic unit may be the result of incomplete HP re-equilibration of large crustal blocks or, alternatively, the tectonic assemblage of crustal slivers suffering different P-T paths and finally amalgamated within a subduction channel. Ranalli et al. (2005) modelled the retrograde metamorphic evolution of the Ulten crust as a two-stage process. First, a near-isothermal decompression P-T path brought the crustal and mantle rocks association to mid-crustal levels (P = 0.7 GPa and T ~ 500 °C) in late Variscan times (~300 Ma), constrained on the basis of Rb-Sr white-mica geochronology (Hauzenberger et al., 1993). Subsequently, the Ulten Zone basement underwent a slow decompression cooling path that ended in Permian-Jurassic times as indicated by Rb-Sr ages of biotite (Del Moro et al., 1999). Sm–Nd dating of peridotites, country migmatites and eclogites yielded late Variscan (about 330 Ma) garnet–whole rock and garnet–clinopyroxene ages (Tumiati et al. 2003). These converging ages have suggested to Tumiati et al. (2003) that crystallization of the HP garnet peridotites, partial melting and eclogitization of crustal rocks took place simultaneously at eclogite-facies, and points to coupling of crust and mantle during subduction. The possible crust-mantle coupling based on this reconstruction is reported in Figure 7. After this event, all lithologies underwent exhumation along a common retrograde P-T path (Godard et al., 1996; Tumiati et al., 2003).
Figure 7. P-T evolution of Ulten crust and mantle after Nimis and Morten (2001), Tumiati et al. (2003).