The most impressive feature of metamorphic zoning in NE Sardinia (Figure 3) is the rapid increase in metamorphic grade in a very restricted area, from biotite to a sillimanite + K-feldspar zone in only 40 km. The metamorphic zoning of metapelitic and metapsammitic sequences outcropping in northern Sardinia has been studied by observing the regional distribution of AKFM minerals (Franceschelli et al., 1982b; Elter et al., 1986). Six zones were distinguished from south to north: 1) biotite; 2) garnet; 3) staurolite + biotite; 4) kyanite+biotite 5) sillimanite; 6) sillimanite + K-feldspar (Figure 2). The garnet zone has been further subdivided into lower and upper garnet zones (i.e. garnet + albite and garnet +albite-oligoclase zones of Franceschelli et al., 1982a). This zoning is apparently continuous and gradual from medium- to high-grade, in spite of the abrupt lithological change and the major tectonic line (Posada-Asinara Line of Cappelli et al., 1992) dividing high-grade rocks from medium-grade ones.
The biotite zone is defined by the first appearance of biotite. The mineralogy is quartz, albite, muscovite, biotite and chlorite, with minor K-feldspar, epidote, carbonate and ilmenite in variable combinations. The sedimentary bedding of the protolith is still recognisable a few kilometres south of the village of Lula (Franceschelli et al., 1982b; Ricci et al., 2004). Mica, quartz and graphite helicitic inclusions sometimes occur in albite porphyroblasts. White mica parallel to S1 schistosity is a paragonite-poor, celadonite-rich muscovite. Chlorite is an Mg-Fe chlorite and biotite shows an XMg in the 0.4-0.5 range.
The garnet zone is defined by the first appearance of almandine-rich garnet and by the coexistence of garnet, biotite and chlorite. The main mineral assemblage is quartz, albite, oligoclase, garnet, muscovite, biotite, chlorite, ilmenite, sphene, and, locally, chloritoid and incipient staurolite in varying combinations.
Diachronous development characterised the main constituents; S1 minerals, quartz, muscovite, biotite and chlorite, are recrystallised parallel to S2, while the first garnet and chloritoid are post-D1, pre-D2; that same initial garnet can have a post-D1, pre-D2 core and a syn-D2 rim. A second garnet generation represented by small subhedral to euhedral crystals appears clearly syn-D2. Helbing and Tiepolo (2005) attributed all pre-D2 minerals to a pre-Variscan metamorphic event.
Carosi and Palmeri (2002) discovered that syn- to post-D1 garnet crystals, included in albite porphyroblasts, show increasing MgO content from core to rim and lower CaO content with respect to the core of garnets growing along S2 in the same thin section. This behaviour clearly indicates the attainment of peak temperature before the beginning of the D2 phase, and subsequent decompression during this phase. The sequence peak temperature attainment - decompression is also suggested by the chemical zoning observed by Carosi and Palmeri (2002) in white mica flakes from the garnet zone: the core is still a celadonite-rich mica (Si4+ = 3.3), while the rim is pure muscovite (Si4+ = 3.01). The same result, syn-D1 celadonite-rich mica and stable syn-D2 muscovite, was found by Di Vincenzo et al. (2004). The great decrease in celadonite content may be ascribed to a temperature increase and /or pressure decrease, probably to both simultaneously. Plagioclase, which is almost pure albite (An = 1-4%) in the albite-bearing zone, appears, in the oligoclase-bearing zone, as large albite (An= 1-4%) porphyroblasts, often surrounded by a thin oligoclase rim (An= 18-22%) or as small syn-D2 oligoclase crystals (An= 16-18%). Abundant chlorite (XMg =0.4-0.5) and biotite (XMg= 0.4-0.5) occur as syn-D2 flakes (parallel to S2), wrapped around garnet and plagioclase porphyroblasts.
Garnet was generated by means of the following reactions (Franceschelli et al., 1982a): 1) chlorite + muscovite + quartz = garnet + biotite + H2O; 2) carbonate + epidote + quartz = garnet + CO2 + H2O. The latter reaction is proposed to explain the remarkable grossularite content, particularly in garnet cores.
The staurolite + biotite zone is defined by the first appearance of the staurolite + biotite association observed on contact between metapelites and granodioritic orthogneisses and augen gneisses in the Lodé Antiform. Fractured staurolite and garnet porphyroblasts, flattened and rotated by D2, are often embedded in a mylonitic matrix consisting of sheet silicates and quartz. Plagioclase, staurolite, garnet and biotite have syn-D1 pre-D2 cores and sometimes syn-D2 rims (Franceschelli et al., 1982a, b; Carosi and Palmeri, 2002). Garnet porphyroblasts display bell-shaped Mn zoning and a gradual Mg, Fe increase and Ca decrease from core to rim. Staurolite has an XMg ratio ≈ 0.17, biotite XMg ratio= 0.50-0.60 and muscovite low celadonite content. The staurolite-producing reaction in the KFMASH system (Spear and Cheney, 1989) is: garnet + chlorite + muscovite = staurolite + biotite + quartz +H2O.
The kyanite + biotite zone is characterised by the mineral assemblage kyanite, biotite, garnet, white mica and plagioclase. Staurolite relics are rimmed by new biotite. Garnet principally occurs as 0.1-0.2 mm sized idioblasts with cloudy cores rich in optically-undetectable inclusions. The average core and rim compositions are respectively Alm76, Prp11, Sps7, Grs6 and Alm75, Prp10, Sps9 and Grs6. Limpid garnet crystals contain microinclusions of idiomorphic anorthite, epidote and margarite and show the following average composition: Alm 62, Grs18, Prp7, Sps3, for the cores and Alm 67, Grs17, Prp10, Sps6 for the rims (Connolly et al., 1994). Staurolite, biotite, kyanite, garnet, white mica and plagioclase are all syn-D1, pre-D2 (Franceschelli et al., 1982b), at least in the cores, perhaps syn-D2 in the rims (Carosi and Palmeri, 2002).
Plagioclase inclusions in limpid garnet are strongly calcic (An = 67-99%), while the same inclusions in cloudy garnet show An = 22-59%. The XMg of biotite is 0.4-0.5. White mica is a Na-and Fe-, Mg-poor muscovite. The following KFMASH discontinuous reaction (Spear and Cheney, 1989) staurolite + chlorite + muscovite = kyanite +biotite + quartz +H2O, accounts for the appearance of kyanite + biotite.
The beginning of the sillimanite zone is marked by the entry of sillimanite and the disappearance of kyanite. Near the boundary with the kyanite + biotite zone, sillimanite-bearing rocks show mylonitic texture and, northwards, compositional layering due to the alternation of 0.1-0.3 mm thick mafic and quartz-feldspathic layers aligned along the main S2 schistosity. The mineral assemblage includes quartz, plagioclase, garnet, sillimanite, muscovite, biotite and rare K-feldspar. Very fine-grained sillimanite needles, sometimes aggregated in nodules, are associated with small anhedral biotite flakes and/or included in quartz, plagioclase and muscovite crystals. Garnet occurs as fine-grained unzoned euhedral crystals with high spessartine content (up to 22%) and grossularite in the 3-7 % range, XMg ratio ≈ 0.16. Matrix biotite shows XMg in the 0.35-0.43 range and TiO2 up to 2.6 %, muscovite with an Na/(Na+K) ratio of 0.06-0.09. Plagioclase has a uniform oligoclase-andesine composition.
The sillimanite + K-feldspar zone consists mainly of gneiss, nebulitic and layered migmatites. Mesosomes are medium-grained, with biotite flakes parallel to S2 schistosity. They consist of quartz, plagioclase, garnet, fibrolitic sillimanite, biotite, K-feldspar and sporadic relics of kyanite in variable combinations. Retrograde overgrowth of muscovite on fibrolite has been observed locally. Leucosomes are medium- to coarse-grained poorly foliated rocks, with granitic to trondhjemitic composition. Plagioclase with An 25-32 sometimes shows myrmekitic intergrowth with K-feldspar. Garnet composition is Alm 59-67, Prp 0-11, Sps 10-25, Grs <4. Muscovite has low celadonite (0.06-0.09 a.p.f.u.) and paragonite content (0.06-0.10). K-feldspar is a frequently-zoned microcline with low anorthite content and a moderate Na/(Na+K) ratio (0.10-0.12). Biotite shows an XMg in the 0.30-0.50 range. Leucosomes may be regarded as having mainly derived from the muscovite dehydration melting reaction (Franceschelli et al., 1989; Cruciani et al., 2001).
Cruciani (2003) described two varieties of migmatite found at Punta Sirenella (north of Olbia): 1) amphibole-bearing migmatites; 2) kyanite + sillimanite bearing migmatites. Amphibole-bearing migmatites consist of close alternation of coarse- to fine/medium-grained leucosomes made up of quartz, plagioclase, ± K-feldspar, biotite and amphibole, and grey mesosomes (Figure 4a). Kyanite + sillimanite migmatites have trondhjemitic leucosomes consisting of quartz, plagioclase, biotite, medium-grained kyanite crystals, sillimanite, garnet and retrograde muscovite (Figure 4b).
Figure 4. NE Sardinia
A: Field photography of amphibole migmatites along the coast of NE Sardinia. Arrows indicate the coarse-grained amphibole (ferroan pargasite) on the D2-folded trondhjemitic leucosome.
B: Photomicrograph of kyanite-sillimanite migmatites along the coast of NE Sardinia. One polar.
There are three major outcrops of low- to high-grade metamorphic rocks in north-central Sardinia: western Gallura, Anglona (Lago Coghinas) and Goceano (Figure 1, 3). Studies on metamophic rocks from western Gallura and Anglona were carried out by Oggiano and Di Pisa (1992) and Ricci (1992).
In the western Gallura region, the L-MGMC is separated from the HGMC by an ENE-verging mylonitic belt. The L-MGMC is principally made up of micaschists with subordinate paragneiss bodies. The micaschists are strongly foliated rocks consisting mainly of quartz, plagioclase, biotite, muscovite, staurolite, kyanite and garnet. Along the shear zone, a phyllonitic belt a few hundred metres thick was formed from the original micaschists.
Migmatites of the HGMC outcrop northwards from the mylonitic belt. Although the migmatites show great textural and compositional variety, two main groups may be distinguished: metatexites and diatexites (Oggiano and Di Pisa, 1992). Metatexites are stromatic migmatites defined by the discontinuous alternation of leucosomes, melanosomes and mesosomes. Leucosomes, mainly trondhjemitic in composition, are granoblastic rocks characterised by vague foliation. These migmatites are very common in northern Sardinia, and their origin has been interpreted as the result of subsolidus differentiation (Ferrara et al., 1978; Oggiano and Di Pisa, 1992; Palmeri, 1992).
Diatexite structures vary greatly (agmatitic, nebulitic, schlieren, stromatic, etc.), revealing that these rocks were generated by in situ partial melting followed by melt segregation and mobilisation. Diatexites consist of plagioclase, quartz and K-feldspar in modal proportions similar to those of the minimum melting point. According to Del Moro et al. (1991), these rocks were probably generated by dehydration melting of muscovite and/or biotite deriving from crustal protoliths rich in arenaceous components. Diatexites show both trondhjemitic leucosomes, similar to those of metatexites, and granitic leucosomes. According to Oggiano and Di Pisa (1988), trondhjemitic leucosomes were generated before granitic ones. Migmatitic layering and trondhjemitic leucosomes are folded by D2.
The metamorphic rocks in the Anglona region outcrop in the area of the Coghinas lake, between Limbara leucogranites to the north and eastern and western volcano-sedimentary successions of Oligo-Miocene age. Metamorphic rocks from Anglona are pelitic to arenaceous metasediments with minor calc-silicates, amphibolite lenses, quartzites and granitoid bodies of varying composition.
Three main deformation phases (D1, D2, D3) have been recognised in this area (Oggiano and Di Pisa, 1992). In this region between Nurra and the Posada valley, the metamorphic basement shows LP/HT syn-late D2 to pre-D3 metamorphism that perhaps identifies a thermal dome with sillimanite in the central part, andalusite + sillimanite in adjacent NE and SW areas and only andalusite in the extreme SW corner of the region (Oggiano and Di Pisa, 1992; Ricci, 1992). Al2SiO5 polymorphs are associated to K-feldspar and cordierite. This LP/HT metamorphism is a late overprint produced by decompression on a previous Barrovian metamorphic event, testified by frequent relics of staurolite, garnet and plagioclase. According to Oggiano and Di Pisa (1992), similar evolution characterises Baronie and southern Gallura.
In the Anglona rocks, inclusion trails parallel to S1 schistosity are often recognisable in plagioclase and staurolite porphyroblasts, suggesting post-D1, pre- to syn-D2 growth. The helicitic and sigmoidal texture of garnet indicates syn-D2 growth and destral shear movement (Oggiano and Di Pisa, 1992, Figure 4). High-T low-P minerals developed during the late stages of D2 and before D3 as suggested by: a) fibrolitic sillimanite, parallel to D2 and deformed by D3; b) andalusite porphyroblasts, including poorly-crenulated S2 schistosity, preceeding the highly-crenulated S2 schistosity of the matrix; c) andalusite porphyroclasts deformed by D3 and enveloped by the composite foliations of the Anglona shear zone; d) cordierite growth coeval with that of andalusite.
Detailed studies of metamorphic zoning in the Nurra region have been carried out by Carmignani et al. (1979, 1982b) and Franceschelli et al. (1990). The major difference with respect to NE Sardinia is represented by the existence of an extensive chlorite zone (Figure 2, 5).
Figure 5. Nurra Paleozoic basement
Structural geological sketch map of the Nurra Paleozoic basement, showing metamorphic zonation and traces of S1 and S2 schistosity. Graphical representation of changing folding style is also shown (after Franceschelli et al., 1990, modified).
The chlorite zone shows close alternation of granoblastic quartzitic and micaceous layers, consisting of white mica (muscovite ± paragonite), albite, chlorite, and chloritoid in varying combinations. Accessory minerals are carbonate, epidote, Fe-Ti oxides and graphite. Slaty cleavage is replaced northwards by strain-slip S1 schistosity. The increasing importance of S2 schistosity, observed moving northwards, explains why sheet silicates are oriented parallel to S2 in the northern part of the chlorite zone (Figure 5). Here, subrounded albite porphyroblasts still preserve inclusion trails parallel to relict S1 schistosity, now oriented discordantly with respect to the enveloping S2 one.
In the northernmost part of the chlorite zone, crenulations of S2, mainly defined by opaque minerals, begin to appear on a regional scale. Albite porphyroblasts (An 1-4) occur as subrounded crystals of 0.1 mm. S1 muscovite shows a broad range of Si a.p.f.u., with maximum values of 6.65-6.70 in metarhyolite samples and minimum values of 6.07-6.20 in Al-rich samples. Paragonite yielded MgO <0.20 wt%, FeO <0.40 wt% and a K/(K-Na) ratio ranging from 0.06 to 0.10. Chlorite shows the following composition: (a.p.f.u.) AlVI= 2.76-3.14; Fe= 5.46-6.00 Mg=2.90-3.42.
The biotite zone - Carmignani et al. (1882b) discovered that the bulk composition of rocks may influence the first appearance of biotite. Early appearance of biotite was observed in metabasites at Monte Rugginosu, outcropping in the chlorite zone: low Al content anticipates while high Al content delays biotite entry into a prograde sequence.
Simpson (1998) observed that the D2 phase produced open folds with interlimb angles of 40°-140° (mean 80°) in the chlorite zone and tight N-vergent isoclinal folds with interlimb angles of 25° at the onset of the biotite isograd. The abrupt change in interlimb angles defines the transition from dehydration-absent to dehydration-active metamorphic environments.
In the biotite zone, S1 still survives in lens-shaped micaceous layers, where its presence is defined by mica blades sub-orthogonal to enveloping S2 surfaces. Albite porphyroblasts show a xenoblastic core, including trails of opaque, quartz and phyllosilicate grains and a limpid rim devoid of inclusions. The trails, generally curved, denote growth, mostly subsequent to the beginning of the D2 phase. The inclusion trails of biotite porphyroblasts depicted S1 orientation and form wide angles with S2 schistosity. Biotite (AlVI=0.76 a.p.f.u.; Fe =2.65 a.p.f.u.; Mg =2.11 a.p.f.u.) is Al-rich annite (Carmignani et al., 1982b).
Garnet zone - This zone is defined by the first appearance of almandine- rich garnet in pelitic rocks. The analogy between NE Sardinia and Nurra as regards the presence of upper and lower garnet zones is worth noting. The garnet zone is characterised by an increase in mineral grain-size from 0.1-0.2 to 1-2 mm, and by the mineral assemblage albite, oligoclase, garnet, muscovite, biotite, chlorite, and minor chloritoid. S2 is the main schistosity and S3 becomes more and more evident towards the north. Inclusion trails in albite porphyroblasts may be rectilinear, gently or strongly folded and sometimes characterised by two orientations. These microstructures suggest a gradual deformation of S1 schistosity during the D2 phase. Garnet porphyroblasts of 0.1-0.2 mm occur in the matrix and as fresh idioblastic crystals within albite porphyroblasts. Garnet is Alm 59-82 with subordinate Grs 9-25, Sps 1-12 and Prp 5-9. Mn content and a Ca/(Ca+Mg+Fe) ratio decrease from core to rim. The opposite trend characterises Fe and Mg. A slight increase in Mg/(Mg+Fe) ratio was observed from core to rim. Biotite composition is as follows: AlVI= 0.79-0.91 a.p.f.u.; XMg = 0.39-0.48. In the upper garnet zone, albite porphyroblasts are mantled by oligoclase rims (An20-24).
The first description of Asinara metamorphic rocks (Figure 6) can be found in Ricci (1972), who reported albite-oligoclase, garnet, K-feldspar, sillimanite and relict staurolite as characteristic metamorphic minerals. The occurrence of these minerals suggested that Asinara was affected, during the first Variscan phase, by the same Barrovian metamorphism well-known all over northern Sardinia. The "Barrovian" hypothesis was confirmed by the discovery of kyanite relics replaced by sillimanite within melanosomes of Punta Scorno migmatites (Oggiano and Di Pisa, 1998).
Figure 6. Asinara Island
Geological sketch map of Asinara Island (after Carosi et al., 2004, modified).
Two metamorphic complexes, juxtaposed along narrow belts of high-strain concentration, have been distinguished by Carosi et al. (2004) on Asinara Island: a low- to medium-grade metamorphic complex (L-MMC) and a high-grade metamorphic complex (HGMC). The L-MGMC consists mainly of (Figure 6) : 1) fine-grained and porphyroblastic paragneisses; 2) andalusite-and sillimanite-bearing paragneisses and micaschists; 3) massive amphibolites; 4) quartzites; 5) La Reale orthogneisses; 6) augen gneisses,7) mylonitic micaschists.
Fine-grained and porphyroblastic paragneisses consist of quartz, oligoclase, garnet, biotite and white mica; accessory minerals are epidote, monazite, zircon and oxides. Andalusite- and sillimanite-bearing paragneisses and micaschists show very high modal contents of andalusite, abundant relict garnet and staurolite replaced by an andalusite + biotite +oxide paragenesis. Accessory minerals are tourmaline, ilmenite, apatite, epidote and zircon. Massive amphibolites consist of hornblende and plagioclase, associated with minor amounts of biotite, chlorite and opaque minerals.
The HGMC consists of 1) migmatites, 2) Cala d'Olivo and Punta Scorno orthogneisses; 3) Punta Scorno banded amphibolites.
Diatexites and metatexites have been distinguished among the migmatites. Diatexites consist of centimetre-thick quartz-feldspar leucosomes, biotite schlieren and restitic polydeformed amphibolite bodies. Metatexites have mesosomes made up of biotite + sillimanite + quartz + cordierite ± K-feldspar and scattered neosome patches. Amphibolites occur as banded rocks made up of alternating centimetre- to decimetre-thick amphibolite and leptynite layers or as massive hornblende-plagioclase rocks. Relics of Ca-clinopyroxene and garnet testify to a previous granulite stage.
Asinara metamorphic rocks were affected by three deformation phases related to crustal thickening in a compressive and transpressive tectonic regime, followed by a later phase of extensional deformation (Carosi et al., 2004). In spite of a general HT/LP metamorphic overprint, Barrovian assemblages, pre- to syn-kinematic with respect to the D2 deformation phase, are still recognizable.
These authors, mapping the distribution pattern of HT-LT mineral assemblages, hypothesized the existence of a regional Abukuma-type HT/LP metamorphism preceding the last deformation event.
Very few detailed studies have thus far been done on the tectono-metamorphic evolution of the Nappe and External zones (Figure 7) during Variscan orogeny. Metamorphism within the Nappe zone in central Sardinia took place during the complex event of Nappe emplacement and stacking and was prograde from low- to medium-grade P-T conditions. Two metamorphic events, M1 and M2 (D1 and D2 phases) mark the attainment of intermediate Barrovian P-T conditions (Carosi et al., 1992a). During the M1 event, low-grade metamorphism was attained with blastesis of muscovite + chlorite + albite.
Figure 7. Paleozoic basement
Geological sketch map of the Paleozoic basement around the Flumendosa antiform, east-central Sardinia (after Carmignani et al. 2001, modified).
Further data on M1 events were supplied by illite crystallinity (IC) measurements (Δ°2θ values) on the six tectonic units of the Nappe pile. From bottom to top they are: Riu Gruppa, Gerrei, Meana Sardo, Arburese and the Barbagia metamorphic complex (Figure 2). Franceschelli et al. (1992) obtained the following IC values: Riu Gruppa, 0.18-0.26; Gerrei Unit, 0.24-0.32 at the front, 0.20-0.26 at the root; Arburese Unit, 0.20-0.26; Arburese unit, on a regional scale, 0.17-0.28 with clustering values in the 0.18-0.22 range (Eltrudis et al., 1995); Genn'Argiolas, 0.20-0.28; Meana Sardo, 0.20-0.32, with slightly decreasing values from front to roots; Barbagia Metamorphic Complex, front values of 0.24-0.34 and root ones of 0.18-0.22.
Eltrudis and Franceschelli (1995) supplied further IC data for the Mulargia Lake area. From bottom to top, IC values of 0.18-0.27 for the Mulargia Unit, 0.26-0.42 for the Gerrei Unit and 0.34-0.43 for the Meana Sardo Unit reveal metamorphic zonation from the deepest epizonal to the shallowest anchizonal units.
During the M2 event, the low-grade metamorphic complex of Barbagia shows persistence of low-grade conditions, with syn-S2 blastesis of chlorite + muscovite + albite (Dessau et al., 1983). Along the axis of the Flumendosa Antiform (Figure 7), prograde metamorphism was observed from SE (Riu Gruppa Unit) to NW (Castello Medusa Unit), applying the calcite-dolomite geothermometer to dolostone levels from Siluro-Devonian marbles (Carosi et al., 1990). The values obtained confirm prograde metamorphism from the Riu Gruppa Unit (average ≈ 300°C).
A higher metamorphic grade was attained only in the deepest units of Castello Medusa and Monte Grighini and in the metamorphic complexes of Sa Lilla, Mandas, Monte Trempu and Asuni.
The Monte Grighini (Figure 8) is a NW-SE trending metamorphic complex consisting of an upper Gerrei Unit and a lower Monte Grighini Unit (Musumeci, 1992). The Gerrei Unit, made up of Middle Ordovician to Siluro-Devonian sedimentary sequences, is characterised by low-grade metamorphism (chlorite zone). The Monte Grighini Unit, considered by Carosi et al. (1990) the deepest and most metamorphic unit in the whole Nappe zone, is divided by Musumeci (1992) into two subunits (Figure 8) : 1) one, consisting of low grade phyllites and acid metavolcanic rocks (biotite zone) and 2) the other, made up of paragneisses and micaschists reaching garnet to sillimanite zones. Along the NE corner of Monte Grighini, the Monte Grighini Unit is locally overlain by the Castello Medusa Unit consisting of low grade (biotite zone) phyllites, metavolcanic rocks and calc-schists. The Monte Grighini Unit is intruded by two NW-SE elongated granite plutons, one calc-alkaline monzogranitic to tonalitic, the other peraluminous leucogranitic. According to Musumeci (1992), P-T peak conditions at the end of the D2 -M2 event were T= 600° ± 50°C and P= 0.6 ± 0.1 GPa. The following M3 event, linked to the intrusion of plutonic rocks, gave rise to the crystallisation of fibrolitic sillimanite, andalusite and cordierite and subsequent generation of biotite and garnet. Thermobarometric calibrations (Carosi et al., 1992a) yielded values of 470°C for the garnet +biotite zone and 560°C for the staurolite+ biotite zone, with pressure values of 0.5 ± 0.1 GPa. The latter stage marks the initial exhumation of the Variscan Belt by means of extensional tectonics caused by the late gravitational collapse of the thickened Variscan crust.
Figure 8. Monte Grighini Unit
Metamorphic zonation of the Monte Grighini Unit, west-central Sardinia (after Carosi et al., 1992a, modified). See text for explanation.
In the tectonic windows of Sa Lilla, Mandas, Monte Trempu, and Asuni medium- to high-grade rocks show rare S1 relics, largely prevailing S2 schistosity and the following prograde sequence of metamorphic zones: 1) chlorite; 2) biotite; 3) cordierite + andalusite; 4) cordierite + sillimanite. The coexistence of Al-silicates with cordierite occurs within the muscovite + cordierite + quartz stability field, yielding a nearly isobaric P-T path for higher-grade assemblages, corresponding to P=0.3 GPa and T= 500°-600°C, with a gradient of 60°C/km (Cappelli, 1991).
A detailed picture of illite crystallinity regional distribution (Figure 9) in the external zone of the Variscan chain in Sardinia is given by Eltrudis et al. (1995, Figure 1). For the whole sequence, from Puddinga to early Carboniferous metasediments, they found a IC range of 0.26-0.45, with a cluster of values from 0.32 to 0.38 (anchizone). Going northwards from the area around Gonnesa, the Cabitza Fm yielded IC values of 0.20-0.30, (average 0.26) with six values > 0.23 and only two < 0.23. Going southwards from the area near Gonnesa, the Cabitza Fm displays opposite behaviour, yielding IC values of 0.19-0.29 (average 0.22). The Nebida group is characterised by IC values in the 0.17-0.29 range, with a regional distribution similar to that described for the Cabitza Fm: moving northwards from the Gonnesa area, the Nebida Fm shows a IC range of 0.22-0.29, with an average of 0.26; going southwards from the Gonnesa area, results obtained for the Nebida group are: IC range of 0.17-0.23, average value of 0.20.
Figure 9. Average IC values
Distribution of average IC values (Δ°2θ x 100) in Paleozoic units from SW Sardinia; after Eltrudis et al. (1995), modified.
The following observations are suggested by the above-reported data: 1) There is a significant jump in metamorphic grade between the anchimetamorphism of Puddinga and the epimetamorphism of the underlying Cambrian to early Ordovician formations (see also Conti et al., 1978); 2) Metamorphic zonation of the Cabitza and Nebida formations, showing increasing metamoprhic grade from north (average IC values of 0.26) to south (average IC values of 0.20-0.22), is opposite to Variscan zonation, characterised by a northward increase in metamorphic grade.
Capo Spartivento metamorphic rocks (Figure 9) were described by Cocozza et al. (1977) and Mazzoli and Visonà (1992), who considered the Monte Filau gneissic body an original granitic intrusion and Settiballas micaschists as its original country rock envelope. The medium- to coarse-grained Monte Filau orthogneisses consist of quartz, plagioclase, K-feldspar, biotite, muscovite and andalusite, sillimanite and garnet. Monte Settiballas micaschists are polymetamorphic, well-foliated rocks characterised by an alternation of granoblastic and lepidoblastic millimetric layers. According to Sassi and Visonà (1989), two main tectono-metamorphic events have been recognised in the micaschists: the first one is documented by relics of andalusite, cordierite, garnet and muscovite, suggesting a high metamorphic gradient. The second one led to the development of compositional layering and the growth of new biotite and muscovite flakes along foliation planes. The Bithia Fm consists of a sequence of phyllites, quartzites, quartzitic metasandstones and metapelites, with minor lenses of marbles and metabasites. The relationships between the Monte Filau orthogneiss and the Bithia Fm support the idea that D1 deformation was pre-Variscan and contemporaneous to the emplacement of the orthogneiss protolith during the Ordovician (Carosi et al., 1995). The D2 phase was probably related to a compressional phase, with maximum shortening in the WSW-ENE direction. This phase can be correlated with the main Variscan deformation of the Iglesiente-Sulcis region. Carmignani et al. (1992, 1994) interpreted the Capo Spartivento structure as a metamorphic core complex generated during the extensional collapse of the chain.