The main contributions towards a better reconstruction of P-T paths for the whole northern part of Sardinia are graphically reported in Figure 12. Reference papers are those by Franceschelli et al. (1989, 1990, 1998, 2002), Ricci (1992), Carosi et al. (2004) and Giacomini et al. (2005a).
Ricci (1992, Figure 3) has also shown that P-T loops, varying in shape and time evolution, characterise the different portions of the metamorphic basement in northern Sardinia, probably owing to diachronous metamorphic evolution of the chain in different areas. However, the general P-T-t path for Variscan metamorphism in Sardinia is that of collisional belts, showing the typical clockwise trend attributed by England and Thompson (1984) to late thermal relaxation following the initial thickening stage. The P-T-t path in Nurra (NW Sardinia) is described by Franceschelli et al. (1990). In southern Nurra, the P-T path (Figure 12a) is poorly constrained but seems to indicate P=0.7-0.8 GPa for peak pressure at 350°C (end of D1) and T = 420°C for peak temperature at P=0.6 GPa (end of the early D2 phase). For northern Nurra (Figure 12b), the end of the D1 phase is marked by the attainment of peak pressure at 0.8 GPa at a temperature of 400°C, while during early decompression of the D2 phase, northern Nurra basement rocks reached peak temperatures of 450°- 500°C for pressures of 0.6-0.7 GPa. The late stage of the D2 phase is characterised by a decrease in both pressure (from 0.6 to 0.4 GPa) and temperature (from 480°C to 420°C). During D3, the P-T decrease was: T from 420°C to 300°C, P from 0.4 to 0.2 GPa.
Figure 12. Pressure and temperature evolution
Pressure and temperature evolution in the northern Sardinia basement: shapes of P-T paths are redrawn from: a, b (Nurra) Franceschelli et al. (1990); c, d (Asinara Island) Carosi et al. (2004); e, l, m (Western Gallura, Anglona) Ricci (1992); f (granulitic rocks from Montiggiu Nieddu) Franceschelli et al. (2002); g (E: retrogressed eclogites and M: migmatite near Golfo Aranci) Giacomini et al. (2005a); h (eclogitic rocks from P.ta de li Tulchi) Franceschelli et al. (1998); i (metapelitic rocks from garnet to sillimanite+ K-feldspar zones, NE Sardinia) Franceschelli et al. (1989), Ricci et al. (2004). U.Grt = upper garnet zone; L.Grt = lower garnet zone. Dashed line: Al2SiO5 triple point after Holdaway (1971).
Very few, fragmentary data are available in literature for a precise reconstruction of the P-T path followed by the medium- and high-grade rocks on Asinara Island (Figure 12c,d), perhaps owing to the strong thermal overprint of late HT/LP metamorphism, able to obliterate any evidence of the D1 phase (early part of the P-T-t path). The highest P-T values calculated with thermobarometers for the D2 phase in L-MGMC rocks are P= 0.8-1.0 GPa and T=560°C (Carosi et al., 2004, Figure 4) and T=600°-650°C (Ricci, 1992).
The first part of the P-T path shows an increase in temperature of about 100°C and a decrease in pressure of up to P=0.2-0.3 GPa. The final part of this path shows isobaric cooling, with a T decrease down to T=400°-500°C. For migmatite and high-grade rocks, Di Pisa et al. (1993) calculated a pressure decrease from 0.7-0.8 GPa at temperatures of 720°-740°C (migmatite stage) to 0.3-0.4 GPa at temperatures of 500°-600°C (amphibolite stage). However, more recently, in melanosomes from migmatites at Punta Scorno, Oggiano and Di Pisa (1998) have found kyanite relics containing sillimanite prisms evolving into acicular aggregates. The attainment of kyanite isograd force pressure at values higher than 0.8-0.9 GPa for high-grade rocks on Asinara.
According to Ricci (1992), and similar to the Asinara path, the Anglona and western Gallura areas (Figure 12e,l,m) (northern coast facing Corsica) show a P-T-t path crosscutting the maximally relaxed geotherm due to a post-thickening heat supply (Thompson and England, 1984). Ricci (1992) attributed this behaviour to reheating of a previously-cooled and uplifted block by late intrusion (Anglona) or to extensional tectonics caused by shallow emplacement of syntectonic, peraluminous granitoids (Palau-Santa Teresa di Gallura).
Giacomimi et al. (2005a) calculated T=680-750°C P=0.8 GPa up to maximum values of T= 788°C and 1.02 GPa for the migmatite formation at Golfo Aranci (Figure 12g, path M). For the amphibolite stage, the authors obtained temperatures of 582°C and 663°C and pressures of 0.6-0.7 GPa.
P-T-t paths of metabasite with granulite (Figure 12f) and eclogite (Figure 12g,h) facies relics in NE Sardinia have been defined by Franceschelli et al. (1998) and Giacomini et al. (2005a).
At Punta de li Tulchi, after the first papers by Miller et al. (1976) and Ghezzo et al. (1982), Franceschelli et al. (1998) reconstructed the P-T-t path (Figure 12h) of HP metamorphism by calibrating different thermobarometers applied to relict eclogite and granulite assemblages still preserved within amphibolitic metabasites.
The following P-T conditions were obtained. For the eclogite stage, a temperature range of 564°-669°C (Grt-Cpx geothermometer using calibration by Ellis and Green, 1979) and a minimum pressure of 1.3 GPa at 700°C, based on jadeite content of relict omphacite, were calculated for a sample crystallised at the beginning of the eclogite-granulite transition. For the granulite stage, two temperature ranges of 765°-899°C and 789°-829°C were calculated with a Cpx-Opx thermometer; pressure estimates yielded 1.0-1.2 GPa. For the amphibolite stage, calculations provided temperatures of 550°-650°C and pressures of 0.3-0.7 GPa; temperatures of 300°-400°C and pressures of 0.2-0.3 GPa were calculated for the greenschist stage.
A similar reconstruction of the P-T path (Figure 12g, path E) is now proposed for the metabasites with HP metamorphic relics at Golfo Aranci by Giacomini et al. (2005a, see Figure 12). The occurrence of the edenite-andesine pair as a small inclusion within a kyanite porphyroblast is interpreted by the authors as testifying to an early prograde amphibolite stage with T=580°-605°C and P=0.8-1.0 GPa. For the subsequent eclogite stage at temperatures of 550°-700°C, maximum pressures of 1.32-1.63 GPa and of 1.40-1.72 GPa were calculated on the basis of jadeite contents of 36% and 44% respectively, yielded by omphacite microprobe analyses. The clinopyroxene-garnet thermometer indicates a temperature range of 580°-690°C and a pressure range of 1.3-1.8 GPa. The decompression part of the P-T path is characterised by omphacite destabilisation and the appearance of clinopyroxene-plagioclase symplectites, typical of the granulite stage, for which geothermobarometry yielded T=650°-790°C and P=0.9-1.2 GPa. A further pressure decrease led to the almost complete transformation of symplectite-rich granulites into garnet amphibolites (HT amphibolites), for which temperatures of 650°-790°C and pressures of 0.9-1.2 GPa have been calculated. The final, complete amphibolitisation of the original granulitic rock generated amphibolites with lobate grain boundaries yielding equilibration temperatures of 550°-650°C at 0.5-0.7 GPa.
For the whole evolution of Montiggiu Nieddu metabasites, P-T conditions (Figure 12f) were estimated as follows. Granulite stage equilibrium conditions were achieved only on a domain scale (Bethume and Davidson, 1997, with references). For two corona microdomains, the garnet-clinopyroxene thermometer, using calibrations by Ellis and Green (1979) and Powell (1985), yielded temperatures in the 700°-740°C range, while the garnet-orthopyroxene thermometer gave a range of 650°-700°C. Using the composition of garnet cores and orthopyroxene and plagioclase enclosed in porphyroblastic garnet, Franceschelli et al. (2002) obtained, with various calibrations, T= 680°-750°C and P=0.8-0.9 GPa. Comparable values of T =700°-750°C and P= 0.8-1.0 GPa were calculated for the coronas of ultramafic amphibolites by Franceschelli et al. (2002), substantially agreeing with the previous estimates of 750°C and 1.0 GPa proposed by Ghezzo et al. (1979). For the amphibolite stage, the garnet +amphibole + plagioclase +quartz assemblage suggests temperatures in the 540°-640°C range and the garnet-amphibolite-plagioclase-quartz barometer gives pressures of 0.4-0.6 GPa. The late appearance of chlorite, tremolite and epidote indicates re-equilibration at temperatures of 330°-400°C and pressures of 0.2-0.3 GPa in the final part of the P-T path.
For NE Sardinia, from Siniscola to Punta de li Tulchi, Franceschelli et al. (1989) and Ricci (1992) proposed the following P-T-t path (Figure 12i). The ascending limb of the path corresponds to maximum thickening of the crust, with attainment of a peak pressure of 1.2 GPa at temperatures of 550°-600°C: this limb is characterised by prograde metamorphism and static post-D1 crystallisation of key minerals from biotite up to the sillimanite+ K-feldspar zone. The highest pressure part of this path is calculated precisely only in the garnet zone, still maintaining minerals crystallised during the M1 event (D1 phase). From the staurolite + biotite zone to the sillimanite + K-feldspar one, rocks recorded only decompression and temperature decrease after the thermal climax, as strong D2 deformation and peak temperatures had almost completely erased any evidence of the previous D1 phase. This explains why, for medium- and high-grade zones, the peak pressure part of the path is hypothetical and backward-extrapolated from the peak temperature of the downgoing limb. Backward extrapolation from peak temperature of high-grade zones is partially constrained by the finding of relict minerals such as kyanite.
The folllowing estimates of P-T conditions have been published: Di Vincenzo et al. (2004), using various geothermobarometers, calculated T = 497°-544°C, P=1.0-1.1GPa for the D1 phase and T= 521°-560°C, P=0.7-0.9 GPa for the D2 phase in minerals of the garnet zone.
For the staurolite + biotite zone, Franceschelli et al. (1989) calculated T =610°C, P= 0.8-1.0 GPa; Di Vincenzo et al. (2004), T=588°-624°C, P= 0.6- 0.9 GPa. As regards the kyanite + biotite zone, Carosi and Palmeri (2002) estimated T < 650°C and P>0.9 GPa for the D1 phase and T= 650-700°C, P=0.7-0.9 for the D2 phase. For the sillimanite + K-feldspar zone (near Punta de Li Tulchi), Cruciani et al. (2001) found evidence that melting started in the kyanite field, migmatite later attained peak conditions in the sillimanite field at T= 700°-720°C, P= 0.6-0.8 GPa. The downgoing part of the P-T-t path ended at T 630°-670°C, P=0.4-0.6 GPa.
Summarising all previously-reported data for the P-T-t path of Variscan metamorphic rocks from Sardinia and taking many recent radiometric data into account (Table 2), the following sequence of events may be proposed. An early phase of HP metamorphism characterised the beginning of the Variscan orogenic cycle in Sardinia. Evidence of this event has thus far been found only in some amphibolitic metabasites containing relics of eclogitic mineral assemblages enclosed in the HGMC or L-MGMC. The age of protoliths seems to be Middle Ordovician, as suggested by three U-Pb zircon ages of 453 ± 14 Ma, 457 ± 2 Ma and 460 ± 5 Ma, (Table 1) respectively found for metabasites with eclogite relics by Palmeri et al. (2004), Cortesogno et al. (2004) and Giacomini et al. (2005a).
Moreover, Palmeri et al. (2004), on the basis of SHRIMP zircon U-Pb data, put forward the hypothesis that 400 ± 10 Ma might be the age of the eclogite formation in NE Sardinia. For eclogites (eclogite A) included in the HGMC, Cortesogno et al. (2004) also obtained a zircon age of 403 ± 4 Ma, interpreted as dating the high-grade event.
According to Cortesogno et al. (2004), eclogites B and the hosted Gneiss Complex must be attributed to a Neoproterozoic (?) Cambro-Ordovician rifting event, giving rise to bimodal tholeiitic-rhyolitic volcanic rocks interbedded within a pelitic-psammitic sedimentary sequence, later subducted and intruded by granitic melts during the Early-Middle Ordovician.
For eclogites embedded within the Sardinian HGMC and Savona Crystalline Massif, Giacomini et al. (2005b) propose a Middle Ordovician protolith age, an Early Visean age for eclogite facies metamorphism and a Late Visean age for amphibolite facies metamorphism, i.e. respective ages of 460-450 Ma, 345 Ma and 320 Ma.
A question now arises: did migmatite in fact experience the HP event revealed by enclosed eclogite lenses? Except for kyanite relics found in plagioclase, no UHP mineral such as coesite or diamond has thus far been found in migmatite.
The following chronological sequence may be proposed for Variscan metamorphic events. According to Di Vincenzo et al. (2004), the early thickening stage of the D1 -M1 event did not start before ~360 Ma and probably took place at 345-340 Ma, in agreement with the Tournaisian age (355-345 Ma) of deformed metamorphic rocks from SE Sardinia, and with an 40Ar-39Ar age of 345 ± 4 Ma on actinolite from a metagabbro (Del Moro et al., 1991). Moreover, the time span from 360 to 345 Ma is consistent with the minimum theoretical time lapse of 20 Ma indispensable for the change from a passive to active collisional margin and for the beginning of exhumation (Thompson et al., 1997). More specifically, Di Vincenzo et al. (2004) found apparent 40Ar-39Ar ages of 340-315 Ma for muscovites in the garnet zone, with the oldest ages (340-335 Ma) for syn-D1 white mica, and age clustering at 320-315 Ma for most syn-D2 white mica.
Giacomini et al. (2005a), taking into account the petrological features of a 344 ± 7 Ma-old leucosome from a NE Sardinia migmatite (Ferrara et al., 1978), proposed an age of 345 Ma for the muscovite dehydration melting event that took place in Golfo Aranci gneisses.
Pervasive fluid infiltration into metabasite lenses from enveloping migmatites transformed anhydrous granulite facies assemblages into hydrated upper amphibolite facies assemblages at 352 ± 3 Ma (zircon resetting age by Giacomini et al., 2005a). The age values of 344 and 352 Ma seem to indicate that migmatisation, partial melting and related metasomatic processes took place soon after the end of the D1-M1 phase, and that peak temperatures in the basement were attained very close to the D1-D2 boundary. Similarly, according to Elter et al. (1999), age values around 350 Ma, 345 ± 4 (Del Moro et al., 1991) and 344 ± 7Ma (Ferrara et al., 1978) mark the boundary in the axial zone between the end of collisional tectonics, when peak pressure is reached (D1 phase), and the beginning of the extensional tectonics responsible for exhumation and uplift (D2 phase). Ricci et al. (2004) also stated that "the 345 Ma age could therefore be close to the collisonal stage or represent the beginning of the exhumation".
In Anglona and on Asinara Island, Barrovian metamorphic mineral assemblages were overprinted by late Variscan (Rb-Sr age of 303 ± 6 Ma on muscovite, Del Moro et al., 1991) HT/LP parageneses linked to gravitative collapse, exhumation of the chain and shallow emplacement of intrusive granitoids (high-K calc-alkaline type, 310-290 Ma, Di Vincenzo et al., 2004).
Ar-Ar and Rb-Sr ages from 305 to 298 Ma were obtained by Laurenzi et al. (1991) on muscovites from the synkinematic leucogranites of the Monte Grighini SZ, deformed by shear movements. An age of 300 Ma could reasonably be attributed to the Sardinian LSZ.