4.a. The basement and the cover sequence

The polycyclic basement of the Dora-Maira ‘Basement Complex’, with the exception of the meta-intrusive bodies, is interpreted as deriving from a pelitic to psammitic metasedimentary sequence spanning from Precambrian to Devonian (e.g. Nosenzo et al. Reference Nosenzo, Manzotti, Krona, Ballèvre and Poujol2024; Carosi et al. Reference Carosi, Montomoli, Iaccarino, Dana, Corno, De Cesari and Spina2025; Dana et al. Reference Dana, De Cesari, Montomoli, Iaccarino, Corno, Rubatto and Carosi2025), which was metamorphosed into garnet-and chloritoid-bearing micaschist and paragneiss. Minor carbonate and organic rich-siliceous sedimentary intercalations are preserved as polycyclic marble (Cadoppi et al. Reference Cadoppi, Camanni, Balestro and Perrone2016; Vola et al. Reference Vola, Ardit, Frijia, Cavallo, Natali, Mion, Lugli and Primavori2022) and black-quartzite. Notably, garnets in some micaschist and meta-mafic rocks of the polycyclic basement has recorded more than one orogenic cycle (Sandrone and Borghi, Reference Sandrone and Borghi1992, Nosenzo et al. Reference Nosenzo, Manzotti, Poujol, Ballèvre and Langlade2022, Reference Nosenzo, Manzotti and Robyr2023). The metasedimentary monocyclic cover overlying the basement in the study area (Figure 2b) is intensely deformed and highly dismembered, showing clear evidence of superposed folding and multiple levels of décollement along weak horizons.

This monocyclic sequence was described in detail by Caron (Reference Caron1977), who referred to it as the Séries de Sellerie–Fontane after the two-type localities (Figure 2a). Notably, calcschists and impure marbles cropping out in the Subiasco–Pellice Valley (Figure 2a) were previously attributed to the Piemonte-Ligurian units by some authors (e.g. Deville et al. Reference Deville, Fudral, Lagabrielle, Marthaler and Sartori1992; Angiboust and Glodny, Reference Angiboust and Glodny2020). However, our observations support the interpretation originally proposed by Wheeler (Reference Wheeler1991), namely that these rocks belong to the Mesozoic cover of the Dora-Maira Massif. Unlike the calcschists of the Piemonte-Ligurian units, they are neither associated with meta-ophiolitic rocks nor contain ophiolitic-derived detritus. Instead, the calcschists are characterized by continental-derived detritus, including quartz and rare dolostone pebbles. Where associated with marbles, the latter contain detrital components and display lithological features comparable to the Upper Jurassic Briançonnais marbles (Lemoine, Reference Lemoine1961; Lemoine et al. Reference Bourbon, de Graciansky, Rudkiewicz, Megard-Galli, Lemoine, Bas, Arnaud-Vanneau, Arnaud, Dumont, Gidon and Tricart1986; Gidon et al. Reference Gidon, Kerckhove, Michard, Tricart, Gotteland, Gout, Leblanc, Lefèvre, Le Gernic, Mégard-Galli and Michel-Noël1994; Michard and Henry, Reference Michard and Henry1988). Altogether, these lithological characteristics indicate a Briançonnais affinity and support their attribution to the Mesozoic cover of the Dora-Maira Massif rather than to the Piemonte-Ligurian units. This interpretation is further supported by comparison with the calcschists exposed in the mixed Cima Lubin Shear Zone of the southern Dora-Maira, which are intimately associated with mafic and ophiolitic rocks (Michard et al. Reference Michard, Schmid, Lahfid, Ballèvre, Manzotti, Chopin, Iaccarino and Dana2022; Dana et al. Reference Dana, De Cesari, Montomoli, Iaccarino, Corno, Rubatto and Carosi2025; Figure 2a). Due to the strong deformation, reconstructing the original thicknesses and true lithostratigraphic relationships among the various lithologies is challenging. Nevertheless, by comparison with similar sequences in the neighbouring valleys, the following tentative, lithostratigraphic succession can be proposed, from the bottom to the top (Figure 2b):

1. (i) The felsic Luserna orthogneiss (dated as Permian in the area around Col Clapier by Nosenzo et al. Reference Nosenzo, Manzotti, Krona, Ballèvre and Poujol2024; at Colle Giulian by Dana et al. Reference Dana, De Cesari, Montomoli, Iaccarino, Corno, Rubatto and Carosi2025 and in the Susa Valley by Balestro et al. Reference Balestro, Roà, Bertok, Gattiglio, Ghignone, Groppo, van Schijndel and Festa2026) is usually the first rock-type occurring at the top of the polycyclic basement. Based on their frequent association with metasediments and distinctive structural position, they have been interpreted as the metamorphosed and deformed remnants of an original volcanic and/or sub-volcanic sequence associated with very shallow intrusions (Vialon, Reference Vialon1966; Wheeler, Reference Wheeler1991; Nosenzo et al. Reference Nosenzo, Manzotti, Krona, Ballèvre and Poujol2024; Dana et al. Reference Dana, De Cesari, Montomoli, Iaccarino, Corno, Rubatto and Carosi2025). They are associated with quartz-phengite-bearing micaschist (‘silvery micaschist’; Bussy and Cadoppi, Reference Bussy and Cadoppi1996) interpreted as the product of shearing and metasomatism of the orthogneiss (Cadoppi, Reference Cadoppi1990; Compagnoni et al. Reference Compagnoni, Crisci and Sandrone1983) or as original sedimentary levels (Vialon, Reference Vialon1966).

2. (ii) Quartzite: they are rare, generally thin (<5 m) and have been sporadically observed only in a few localities (e.g. Col Clapier). Quartzite usually overlies the Luserna orthogneiss.

3. (iii) Micaschist with sigmoidal lenses, rich in chloritoid, white mica and ±garnet (and commonly associated with intercalated layers and pods of ankerite), here referred to as Fontane micaschist. It frequently occurs in spatial association with the Luserna orthogneiss and calcschist. However, the stratigraphic position of this lithology relative to the Luserna orthogneiss remains uncertain (Figure 3a–c).

   Figure 3.

   (a) Fontane micaschist (sample DD20), characterized by sigmoidal lenses composed of chloritoid ± white mica and ankerite pods. (b) Detail of a Fontane micaschist with small orange ankerite pods. (c) Association between Fontane micaschist and Luserna orthogneiss; garnet-chloritoid-bearing micaschist are folded together with the orthogneiss in thin, centimetre-scale layers. (d) Typical meta-dolostone showing alternating whitish and bluish layers. (e) Grey-blue sandy marble containing deformed blue clasts (arrows) within a meta-calcarenitic matrix, wrapped by the main foliation. (f) Folded and deformed cover sequence in the Subiasco area (Pellice Valley area), preserved as a lens pinched within the core of a synformal structure. (g) Omission of Triassic–Jurassic carbonate-bearing rocks, with calcschist directly in contact with Fontane micaschist.

   

   Figure 3. Long description

   Panel A: A photo of a rock formation labeled Fontane micaschist with sigmoidal lenses composed of chloritoid and white mica, and ankerite pods. Panel B: A close-up photo of Fontane micaschist showing small orange ankerite pods. Panel C: A photo showing the association between Fontane micaschist and Luserna orthogneiss, with garnet-chloritoid-bearing micaschist folded together with the orthogneiss in thin, centimeter-scale layers. Panel D: A photo of typical meta-dolostone displaying alternating whitish and bluish layers. Panel E: A photo of grey-blue sandy marble containing deformed blue clasts within a meta-calcarenitic matrix, wrapped by the main foliation. Panel F: A photo of a folded and deformed cover sequence in the Subiasco area, preserved as a lens pinched within the core of a synformal structure. Panel G: A photo showing the omission of Triassic-Jurassic carbonate-bearing rocks, with calcschist directly in contact with Fontane micaschist.

4. (iv) Meta-dolostone and dolomitic marble: they commonly exhibit a banding of whitish and blueish layers (Figure 3d), with variable proportions of dolomite and calcite (Vialon, Reference Vialon1966; Nosenzo et al. Reference Nosenzo, Manzotti, Krona, Ballèvre and Poujol2024). Their thickness can be up to 15–20 m where in contact with quartzite or Fontane micaschist. The equivalent of these rocks in the Susa Valley (northern Dora-Maira) has been dated by Franchi (Reference Franchi1911) to the Middle-Upper Triassic on the basis of its fossils content.

5. (v) Grey-blue sandy marble (up to 20–30 m thick): this lithotype is characterized by a lower portion containing centimetre-scale bluish dolomite lenses and deformed clasts within a meta-calcarenitic matrix with detrital quartz grains, all wrapped by the main foliation (Figure 3e). It can be tentatively attributed to the Middle–Late Jurassic based on correlation with similar rocks from the Briançonnais units (Lemoine, Reference Lemoine1967; Dana et al. Reference Dana, Iaccarino, Schmid, Petroccia and Michard2023; Pantet et al. Reference Pantet, Epard, Masson, Baumgartner-Mora, Baumgartner and Baumgartner2023; Nosenzo et al. Reference Nosenzo, Manzotti, Krona, Ballèvre and Poujol2024).

6. (vi) Calcschist and grey impure marble: these rocks are generally richer in siliciclastic components and may contain meta-dolostone clasts. They occur in contact with all the previously described lithologies, whereas marble and meta-dolostone progressively thin southwards and disappear towards the Pellice Valley, except for a few metre-thick slices. In the Susa Valley, equivalent lithologies were dated to the Upper Cretaceous by Marthaler et al. (Reference Marthaler, Fudral, Deville and Rampnoux1986) based on foraminifera, whereas palynological data from our study area indicate a Lower Cretaceous (Aptian–Albian) to Upper Cretaceous age (Spina A., pers. comm.).

The Permo–Mesozoic sequence is deformed and dismembered (Figure 3f, g), due to multiple décollement levels (Figure 2b) and/or an original unconformity between younger (calcschists) and older rocks. In some localities, carbonate-bearing formations are partially or entirely absent (Figure 3g). Because the age of the Fontane micaschist from the cover sequence of the Dora-Maira Massif remains unconstrained, two samples (23PIN43, FONT2; Figure 2a) were selected for detrital zircon geochronology. In addition, one garnet-chloritoid-bearing micaschist sample from the basement (DD97) was analysed for comparison. Two samples of Fontane micaschist (23PIN43, DD20) were selected for thermodynamic modelling using PERPLE_X, whereas one sample (FONT2) was used to estimate average P–T conditions with THERMOCALC. Finally, six calcschist and marble samples (Table 1) from the cover sequence were collected for RSCM analyses following the approach of Beyssac et al. (Reference Beyssac, Goffé, Chopin and Rouzaud2002 a, b).

4.b. Structural analysis

Deformation in poly-and monocyclic rocks is strongly heterogeneous, and different fabrics are preserved in different areas depending on lithology and structural level. Overall, four distinct deformation phases have been recognized affecting both the basement and the cover. The earliest Alpine deformation event (D1) is represented by relics of a continuous S1 foliation, locally preserved within microlithons of the S2 foliation in garnet-chloritoid-bearing micaschist (Section 5) and orthogneiss of both the basement and cover. The D2 deformation phase is associated with strong strain partitioning. In some domains, D2 produced E-W-trending tight to isoclinal folds (F2), generally parallel to the stretching lineation. These folds affecting both cover and basement rocks are only locally preserved (Figure 4a) and are associated with the development of a spaced S2 foliation. In other domains, D2 is characterized by the development of a mylonitic fabric associated with the S2 foliation and characterized by a dominant top-to-the-SW sense of shear (in present day orientation). This mylonitic fabric is well developed in garnet-chloritoid-bearing micaschist (including both the Fontane micaschist and lithologically similar, ankerite-free, basement micaschist), orthogneiss and locally in marble (Figure 4b). A third deformation phase (D3) generated open to isoclinal NW-SE-trending folds (Figure 4c, d). D3 is associated with the development of an S3 crenulation cleavage, accompanied by recrystallization of chlorite, muscovite and albite. In most of the micaschist and paragneiss of the basement, as well as in impure marble and calcschist, S3 commonly represents the dominant penetrative fabric (Figure 4c). Type III interference patterns (Ramsay, Reference Ramsay1967) resulting from superposed D2 and D3 folding are commonly observed. D3 structures were subsequently overprinted by a fourth deformation phase (D4), interpreted as a doming event. D4 produced open, upright N-S-trending F4 folds at scales ranging from centimetric to hectometric. This late event caused progressive steepening and bending of the main foliation towards the tectonic contact separating the Dora-Maira Massif from the overlying Piemonte-Ligurian units.

Figure 4.

Meso-scale structures in the study area. (a) F2 fold affecting the contact between Fontane micaschist and dolomitic marble (S1), with S2 as axial planar foliation. (b) Top-to-the-SW, σ-type rotated porphyroclast associated with the S2 main foliation in a mylonitic sandy marble of the cover sequence. (c) F3 folds in impure marble; a local S3 crenulation cleavage is visible. (d) Decameter-scale F3 folds affecting the Luserna orthogneiss and adjacent calcschists in the Subiasco area (Pellice valley in Figure 2a). (e) Cargneules/tectonic breccia layer underlying the Triassic marble sequence. (f) Basement–cover (represented by Luserna orthogneiss and calcschist) interface in the Subiasco Valley, a tributary of the Pellice river, transposed towards parallelism to S2/S3.

Figure 4. Long description

Panel A: A photo showing a geological formation with layers of Fontane micaschist and meta-dolostone. A hammer is placed for scale. Panel B: A photo displaying a mylonitic sandy marble with top-to-the-SW sigma-type rotated porphyroclasts. Panel C: A close-up photo of F3 folds in impure marble, with a local S3 crenulation cleavage visible. Panel D: A landscape photo showing decameter-scale F3 folds affecting the Luserna orthogneiss and adjacent calcschists in the Subiasco area. Panel E: A photo of a cargneules/tectonic breccia layer underlying the Triassic marble sequence. Panel F: A photo of the basement-cover interface in the Subiasco Valley, with Luserna orthogneiss and calcschist transposed towards parallelism to S2/S3.

4.c. Basement–cover relationships

The Permo–Mesozoic cover of the Dora-Maira Massif is commonly detached from the underlying polycyclic basement and are therefore regarded as ‘para-autochthonous’ (Borghi et al. Reference Borghi, Cadoppi, Porro, Sacchi and Sandrone1984; Fudral, Reference Fudral1998; Nosenzo et al. Reference Nosenzo, Manzotti, Krona, Ballèvre and Poujol2024; Balestro et al. Reference Balestro, Gattiglio, Roà and Festa2025). In the Germanasca and Chisone valleys, the basement–cover interface is locally marked by high-strain zones formed under ductile-to-brittle transition conditions, or by levels of cargneules and tectonic breccias (Figure 4e). Where exposed, this contact appears as a relatively young shear zone that acted as a detachment level after D3 folding, as it cross-cuts pre-existing D3 structures. However, available data do not allow us to precisely constrain the magnitude of displacement along this interface, nor to determine whether it represents a late disharmonic feature or the reactivation of an older shear zone that originally accommodated significant earlier movements. Within this framework, Nosenzo et al. (Reference Nosenzo, Manzotti, Krona, Ballèvre and Poujol2024) identified what they referred to as ‘La Fracho Shear Zone’, a ductile-to-brittle top-NW shear zone interpreted as the basal tectonic contact separating the Permo–Mesozoic cover (their Serre Unit) from the basement (Muret Unit) in the Germanasca valley (Figure 2a). Our field mapping (Figure 2a), however, significantly differs from previous studies because cover rocks occur both in the footwall and in the hangingwall of this structure (Salza di Pinerolo area, Figure 2a). Moreover, our observations show that the cover rocks found in the Serre Unit of Nosenzo et al. (Reference Nosenzo, Manzotti, Krona, Ballèvre and Poujol2024) also crop out within the Muret Unit in the area further south. Nosenzo et al. (Reference Nosenzo, Manzotti, Krona, Ballèvre and Poujol2024) extended their mapping of the Serre Unit only as far as Villa di Prali (Figure 2a), but sequences of Permo–Mesozoic cover rocks are present also further south, where no such clear tectonic contact as the ‘La Fracho Shear Zone’ between cover and basement is observed. These observations suggest that La Fracho Shear Zone only locally coincides with the basement–cover interface, implying that it represents a relatively late structure that cross-cuts a pre-existing tectonic pile that even includes the Piemonte-Ligurian units according to our observation. Therefore, it does not represent a nappe divider between a Muret and Serre Unit.

In the southern part of the study area (Pellice valley, southern part of Figure 2a), remnants of the meta-sedimentary cover, mainly calcschist, marble and Fontane micaschist, are tightly folded together with polycyclic basement lithologies attributed to the Muret Unit, into which the Permo–Mesozoic sequence is folded by D2 and D3 structures (Figures 2a, 3f, 4d). In this southern area, the basement–cover interface is parallel to the S2 foliation (Figure 4f) and shows no evidence of significant strain localization, excluding any substantial late reactivation of the contact. This S2 foliation is folded by metric to decametric F3 folds-oriented NW-SE, representing the dominant structural style of the area (Figure 4d). Therefore, as the main planar elements seen in this southern area of the basement–cover interface (i.e., both S2 and S3 foliations) are recognized in basement and cover without any discontinuity of fabric, the study area records a polyphase Alpine deformation history of both basement and cover, with local detachment occurring only during very late ductile-to-brittle stages such as observed in the north by Nosenzo et al. (Reference Nosenzo, Manzotti, Krona, Ballèvre and Poujol2024).

5.a. Sample 23PIN43 (Colle Giulian, Germanasca valley)

In sample 23PIN43 (Table 1; Figure 5a; modal proportions in Supplementary Material 1), two distinct foliations are recognized. An early S1 foliation defined by quartz, rutile, chloritoid (Cld 1) and white mica (Wm 1) is locally preserved within microlithons (Figure 5b) of the main S2 foliation. Chloritoid occurring within the microlithons locally contains inclusions of graphite, rutile and white mica, suggesting growth during the early deformation stages.

Figure 5.

Mineral phases and petrography of sample 23PIN43. (a) Processed X-ray map showing the mineral assemblage (h/G = hole/graphite). (b) Relict of S1 foliation in a microlithon of the S2 foliation defined by white mica (Wm 1) (XPL = Crossed Polarized Light). (c) Aggregate of chloritoid (Cld 2) and white mica (Wm 2) defining the main S2 foliation, wrapping around S1 chloritoid (Cld 1) (PPL= Plane Polarized Light). (d) Particular growth zoning in garnet highlighted by inclusions. (e) Aggregates of chlorite and albite, interpreted as pseudomorph after sodic-amphibole (PPL).

Figure 5. Long description

Panel A: Processed X-ray map displaying various mineral phases including garnet, quartz, plagioclase, albite, chlorite, chloritoid, rutile, ilmenite, titanite, and holes/graphite. Panel B: Microlithon showing relict S1 foliation defined by white mica under cross-polarized light. Panel C: Aggregate of chloritoid and white mica defining the main S2 foliation, wrapping around S1 chloritoid under plane polarized light. Panel D: Garnet with particular growth zoning highlighted by inclusions. Panel E: Aggregates of chlorite and albite, interpreted as pseudomorph after sodic-amphibole under plane polarized light.

The S2 foliation is a mylonitic fabric defined by millimetre-thick layers of either chloritoid (Cld 2) + rutile + white mica (Wm 2) or white mica (Wm 2) + paragonite + rutile + graphite, which are parallel to stretched quartz-rich lenses (Figure 5a, c). The S2 foliation wraps around garnet porphyroblasts. Microstructural relationships indicate that only the outer rims of the garnet are syn-kinematic with respect to D2 and grew contemporaneously with the development of the S2 foliation, whereas the inner portions predate D2. Garnet porphyroblasts range from 0.5 to 5 mm in diameter (Figure 5a). Inclusions consist of quartz, coesite, rutile, zircon, chloritoid, paragonite ± epidote (possibly replacing lawsonite) and white mica. Their spatial distribution within garnet records a protracted growth history (Figure 5d). The garnet cores (Grt 1a) represent the earliest growth stage and contain mainly quartz, paragonite (possibly replacing lawsonite) and white mica inclusions, indicating growth under prograde metamorphic conditions during the initial stages of D1. Mantle domains (Grt 1b) are characterized by the occurrence of coesite, white mica, rutile and rare chloritoid inclusions, recording continued garnet growth during D1 and the attainment of peak UHP conditions. In contrast, the outer rims (Grt 2) mainly containing quartz and white mica are interpreted to have formed during D2, syn-kinematically with the development of the S2 foliation. Accordingly, coesite inclusions are preferentially restricted to the Grt 1b mantle domains (see also below), whereas quartz, white mica, rutile and zircon occur across multiple garnet domains, reflecting their stability or recrystallization during different stages of the metamorphic evolution. Rutile occurs both in the matrix and as inclusions in garnet and chloritoid. Rutile grains in the matrix can display ilmenite coronas, suggesting partial retrograde re-equilibration. Quartz is recrystallized by grain–boundary migration overprinted by late bulgingReference Stipp, Stünitz, Heilbronner and Schmid.

Post-D2 mineral growth is associated predominantly to static mineral replacement and to brittle deformation. Chlorite and albite locally replace the rim of the garnet porphyroblasts, whose fractures are locally filled by chlorite ± white mica. Chlorite commonly grows statically overprinting the S2 foliation. Rhomboidal domains, composed of post-D2 fine-grained chlorite and albite aggregates, are interpreted as pseudomorphs replacing sodic amphibole that grew along the S2 foliation (Figure 5e). Some of the albite may also derive from the replacement of primary Na-rich clinopyroxene during decompression and subsequent retrograde metamorphism. Although Na-rich clinopyroxene is not preserved in the studied sample, its absence is not unexpected, as this phase is rarely preserved in high-pressure metapelites despite being commonly predicted by phase equilibria modelling (see Manzotti et al. Reference Manzotti, Millonig, Gerdes, Whitehouse, Jeon, Poujol and Ballèvre2025 a). In this sample, a well-developed S3 foliation does not occur.

5.b. Sample DD20 (Serre, Chisone valley)

In sample DD20 (Table 1; Figure 6a; modal proportions in Supplementary Material 1), two distinct foliations can be identified. The S1 foliation, defined by chloritoid (Cld 1) and white mica (Wm 1), is again locally observed within microlithons of the main S2 foliation (Figure 6b). The D2 is associated with the development of the main mylonitic foliation (S2) defined by the shape preferred orientation of chloritoid (Cld 2), white mica (Wm 2), paragonite and millimetre-thick layers of either chloritoid + graphite + rutile or white mica + rutile + graphite, parallel to stretched quartz-rich lenses (Figure 6d). As in the previous sample, S2 wraps around garnet porphyroblasts with strain shadows containing mainly quartz (Figure 6e). Garnet inclusions consist of quartz, white mica, paragonite, zircon and rutile. Paragonite ± epidote inclusions, possibly replacing former lawsonite, are preferentially observed in the garnet core (Grt 1a). In contrast, zircon, rutile and white mica inclusions are distributed throughout the garnet domains. Quartz inclusions show a more heterogeneous distribution as they are particularly concentrated at the onset of the mantle domain (Grt 1b), where they form interconnected clusters along a network of fractures (former coesite?). Chloritoid ranges from 0.1 to 1 mm in size. Most chloritoid crystals are zoned, with a lighter-blue core and a darker-blue rim, suggesting a Mg-richer core (Cld 1) and a Mg-poorer rim (Cld 2). Inclusions in chloritoid consist of quartz, white mica, rutile and graphite. Quartz is recrystallized by grain–boundary migration overprinted by late bulgingReference Stipp, Stünitz, Heilbronner and Schmid. Post-D2 mineral growth is characterized by predominantly static mineral replacement by chlorite and albite. As in sample 23PIN43, albite may also derive from the breakdown of Na-rich clinopyroxene during decompression and retrograde evolution, although this phase is not preserved in the studied sample. Chloritoid crystals occur often as fractured and partly dissolved at rims. Fractures in chloritoid are locally filled by chlorite ± white mica. Garnet porphyroblasts are replaced by chlorite along their outer edges and (together with quartz) along fractures. Rutile in the matrix displays dark ilmenite coronas.

Figure 6.

Mineral phases and petrography of sample DD20A. (a) Processed X-ray map showing the mineral assemblage (h/G = hole or graphite). (b) Relict of S1 foliation within a microlithon of the S2 foliation (XPL). (c) Relicts of chloritoid (Cld 1) and white mica (Wm 1) defining the S1 foliation, in a microlithon of the S2 foliation (PPL). (d) Aggregates of chloritoid (Cld 2) and white mica (Wm 2), defining the main S2 foliation in a quartz-rich domain (PPL). (e) Garnet wrapped by the S2 main foliation, with strain shadows mainly made of quartz (PPL).

6.a. Sample 23PIN43 (Colle Giulian, Germanasca valley)

Garnet in this sample is compositionally zoned (Figure 7), with distinct core (Grt 1a), mantle (Grt 1b) and rim (Grt 2) domains. From core to mantle, garnet shows a slight increase in X_Prp (0.05–0.06 to 0.07–0.08) and X_Alm (0.79 to 0.80–0.83), followed by a decrease in the rim (Figure 7, Table S1 in Supplementary Material 2, Figure S2 in Supplementary Material 1). X_Grs decreases slightly from core to mantle (0.10–0.09 to 0.07–0.08) and X_Sps from 0.05–0.06 to 0.03–0.02, whereas the rim is strongly enriched in grossular (X_Grs = 0.12–0.35; Figure 8; Figure S2). White mica also displays clear zoning, particularly evident in X-ray compositional maps (Figure 8a, b). Si⁴⁺, Fe and Mg decrease towards the crystal edges (Figure 8a, Table 2), whereas Ti and Na increase (Figure 8b) in the same direction. White mica cores (Wm 1) are characterized by Si⁴⁺ a.p.f.u. contents of 3.50–3.65 (Figure 8c). The rims (Wm 2), in equilibrium with the main foliation, have slightly lower Si⁴⁺ contents (3.45–3.50 a.p.f.u.). Plagioclase is close to pure albite (X_An = 0.01). Chloritoid aligned along the main S2 foliation (Cld 2) has Mg-rich cores (X_Mg = 0.19–0.20) and Fe-rich rims (X_Mg = 0.10–0.17). Monocrystalline inclusions of chloritoid within garnet core (Cld 1) have X_Mg = 0.19–0.20. Paragonite, frequently associated with chloritoid, has X_Na = 0.90–0.95. Chlorite has X_Mg = 0.34–0.38. It is worth noting that, although white mica and chloritoid occur in different microstructural positions (e.g. related to S1 or S2), their compositional zoning patterns are generally similar. This reflects the fact that many crystals were reoriented during the subsequent deformation event (D2) and continued to grow from the same pre-existing cores. Consequently, minerals associated with S2 commonly preserve inherited D1 cores, while developing thicker syn-D2 rims.

Figure 7.

X-ray map (WDS = Wavelength-Dispersive Spectroscopy) of garnet grains from samples 23PIN43, DD20 and FONT2, showing zoning for Fe, Ca, Mn and Mg. The interface between garnet core and rim is regular, suggesting limited or no dissolution of garnet core before the growth of garnet rim. Mineral inclusions are shown with different symbols in the Mg map. Abbreviations: R = Rim (Grt 2), M = Mantle (Grt 1b), C = Core (Grt 1a). Electron Probe Micro-Analyzer (EPMA) spot analyses are reported in Supplementary Material 2.

Figure 7. Long description

Panel 23PIN43: Four X-ray maps display the distribution of Fe, Ca, Mn, and Mg in garnet grains. The Fe map shows a clear distinction between the core and rim regions. The Ca map highlights areas with higher calcium concentration. The Mn map indicates manganese distribution, and the Mg map shows magnesium distribution with mineral inclusions marked by different symbols. Panel DD20: Similar X-ray maps for Fe, Ca, Mn, and Mg in another sample. The Fe map shows a regular interface between the core and rim. The Ca map highlights calcium-rich areas. The Mn map shows manganese distribution, and the Mg map indicates magnesium distribution with marked mineral inclusions. Panel FONT2: X-ray maps for Fe, Ca, Mn, and Mg in the third sample. The Fe map shows a distinct core and rim. The Ca map highlights calcium concentration. The Mn map shows manganese distribution, and the Mg map indicates magnesium distribution with mineral inclusions marked by different symbols.

Figure 8.

Mineral chemistry and zoning for white mica in samples 23PIN43, DD20 and FONT2. X-ray map (WDS) showing zoning in white mica for Mg and Na in samples 23PIN43 (a, b), DD20 (d, e) and FONT2 (g, h). White mica compositions for samples 23PIN43 (c), DD20 (f) and FONT2 (i). EPMA spot analyses are reported in Supplementary Material 2.

6.b. Sample DD20A (Serre, Chisone Valley)

Garnet displays growth zoning (Figure 7, Table S2 in Supplementary Material 2) characterized by a slight decrease in X_Alm from core (Grt 1a) to rim (Grt 2) (0.76–0.75 to 0.73–0.74) and in X_Sps (0.12–0.13 to 0.05), accompanied by an increase in X_Grs (0.05 to 0.11–0.14) and X_Prp (0.08 to 0.09). Grossular content increases markedly at the outer rim, reaching values up to X_Grs = 0.15 (Figure 7; Figure S2 in Supplementary Material 1). White mica also exhibits growth zoning (Figure 8d, e), with Fe, Si⁴⁺ and Mg decreasing towards the crystal edges, and Ti and Na showing a slight increase (Figure 8d). The crystal core (Wm 1) is characterized by Si⁴⁺contents of 3.50–3.65 a.p.f.u. (Table 3), whereas the rims (Wm 2), in equilibrium with the main foliation, have slightly lower Si⁴⁺contents (3.45–3.50 a.p.f.u.). Chloritoid has slightly Mg-rich cores (Cld 1, X_Mg = 0.19–0.20) and more Fe-rich rims (Cld 2, X_Mg = 0.17–0.18). Paragonite has X_Na = 0.90–0.95, and chlorite has X_Mg = 0.48–0.50.

6.c. Sample FONT2 (Colletto di Fontane, Germanasca Valley)

This sample shows microstructural and petrographic features comparable with those observed in sample 23PIN43. Garnet displays growth zoning (Figure 7, Table S3 in Supplementary Material 2) characterized by a slight decrease in X_Alm (0.81–0.73 to 0.71–0.67) and in X_Sps (0.06–0.05 to 0.03) from core (Grt 1a) to mantle (Grt 1b), accompanied by a slight increase in X_Grs (0.06 to 0.11–0.12) and X_Prp (0.06 to 0.09). Grossular content increases markedly at the rim (Grt 2), reaching values up to X_Grs = 0.23 (Figure S2 in Supplementary Material 1; Table S3 in Supplementary Material 2). White mica also exhibits growth zoning (Figure 8g, h), with Fe, Si⁴⁺ and Mg decreasing towards the crystal edges, and Ti and Na showing a slight increase (Figure 8h). The crystal cores (Wm 1) are characterized by Si⁴⁺contents of 3.50–3.65 a.p.f.u., whereas the rims (Wm 2), in equilibrium with the main foliation, have slightly lower Si⁴⁺contents (3.45–3.50 a.p.f.u.). Chloritoid has slightly Mg-rich cores (Cld 1, X_Mg = 0.19–0.20) and Fe-rich rims (Cld 2, X_Mg = 0.17–0.18). Paragonite has X_Na = 0.90–0.95, and chlorite has X_Mg = 0.48–0.50.

9.a. Bulk compositions

The protoliths of samples 23PIN43 and DD20 classify as shale following the bulk chemical classification SandClass3 of Herron (Reference Herron1988), compositionally close to the Worldwide Average Pelite of Forshaw and Pattison (Reference Forshaw and Pattison2023) (Figure S4¸ bulk compositions in Supplementary Material 1, Tables S3 and S4).

9.b. P–T results for metapelite 23PIN43

Three equilibrium assemblage diagrams were calculated for sample 23PIN43 using progressive garnet fractionation of the effective bulk composition. The first diagram (Figure 10a) computed using the whole-rock bulk composition (Bulk I; Table S3 in Supplementary Material 1) is dominated by three-and four-variant fields, with minor two-and five-variant fields. Garnet is stable over most of the P–T space above 480°C. Biotite appears only at high temperatures and pressures below 1.4 GPa. Rutile is the main Ti-bearing phase, with ilmenite becoming stable at T > 500°C and P < 0.7–0.8 GPa. Glaucophane is predicted at T < 550°C and P between 1.6 and 2.0 GPa (Figure S7; Supplementary Material 1). Chloritoid remains stable up to 3.1 GPa at around 550°C. Kyanite appears for T ≥ 550°C at P > 0.7–1.2 GPa. The measured composition of the garnet core (X_Prp = 0.06, X_Sps = 0.06, X_Alm = 0.79, X_Grs = 0.09) and the corresponding isopleths indicates crystallization at c. 500–550°C and 2.6–2.8 GPa, within the Car–Grt–Cld–Wm–Omp–Lws–Rt–Qz field, consistent with quartz rather than coesite occurring as inclusions in the garnet core. The measured Si⁴⁺ content in white mica (3.45–3.50 a.p.f.u.) and the X_Mg of chloritoid inclusions in the garnet core (0.19–0.20) are consistent with the P–T estimates shown in Figure 10b (black ellipsoid).

Figure 10.

(a) Equilibrium assemblage diagram calculated using Bulk I (garnet core composition) for sample 23PIN43. (b) Overlap of selected garnet compositional isopleths (X_Prp = 0.06, X_Sps = 0.06, X_Alm = 0.79, X_Grs = 0.09), together with X_Mg in chloritoid and Si⁴⁺ (a.p.f.u.) in white mica; the intersection is highlighted by the black ellipse. (c) Overlap of garnet compositional isopleths for the garnet mantle (Bulk II; X_Prp = 0.08, X_Sps = 0.02, X_Alm = 0.82, X_Grs = 0.08), combined with X_Mg in chloritoid and Si (a.p.f.u.) in white mica. (d) Overlap of garnet compositional isopleths for the garnet rim (Bulk III; X_Prp = 0.03, X_Sps = 0.02, X_Alm = 0.70, X_Grs = 0.25) and Si (a.p.f.u.) in white mica. Equilibrium assemblage diagrams calculated using Bulk II and Bulk III are provided in Supplementary Material 1 (Figure S5).

A second diagram (Figure 10c; Figure S5a) was computed using Bulk II (Table S2), obtained by subtracting the garnet core composition from Bulk I. Garnet fractionation shifts the garnet-in line towards higher temperatures by c. 50°C, whereas the overall topology and isopleth positions remain largely unchanged. Mantle crystallization corresponds to P–T conditions of c. 2.7–3.1 GPa and 540–560°C, based on the measured composition of garnet close to coesite inclusions (X_Prp = 0.08, X_Sps = 0.02, X_Alm = 0.82, X_Grs = 0.08). The modelled Si⁴⁺ content of white mica in this field (c. 3.45–3.50 a.p.f.u.) matches the measured values. The X_Mg of chloritoid rims and inclusions in the outer core of garnet (0.19) is also consistent with the inferred P–T range. Calculated modal abundances for this stage are around ∼4 vol% for chloritoid and ∼4 vol% for garnet, consistent with the estimated modal proportion of garnet core + mantle (Supplementary Material 1; Figure S6). Kyanite is predicted at ∼8 vol% but was not observed, possibly due to later replacement or bulk composition uncertainties (Figure S7). Lawsonite (<0.5 vol%) and carpholite (<2 vol%) are predicted in minor amounts and were not directly observed, whereas omphacite (∼7–8 vol%) was likely replaced during decompression (Figure S6–7 in Supplementary Material 1). Although Na-pyroxene is not preserved in our samples, its predicted stability at peak HP–UHP conditions is consistent with previous thermodynamic modelling of Dora-Maira metapelites. Its complete replacement by albite during decompression likely explains its absence, as suggested by the rare jadeite inclusions in rutile reported by Manzotti et al. (Reference Manzotti, Millonig, Gerdes, Whitehouse, Jeon, Poujol and Ballèvre2025 a). The peak temperature estimated from the modelling is within uncertainties, comparable with the temperature estimated by RSCM (505°C ± 15°C) for this sample.

A third equilibrium assemblage diagram (Figure S5b, Figure 10d) was calculated using Bulk III (removing garnet core + mantle, i.e. Grt 1a and Grt 1b, from Bulk II) to model the garnet rim, which is characterized by a markedly higher grossular content (X_Grs = 0.25), implying crystallization at significantly lower pressures. Because decompression may lead to H₂O undersaturation (Le Bayon et al. Reference Le Bayon, Pitra, Ballèvre and Bohn2006; Manzotti et al. Reference Manzotti, Rubatto, Zucali, Korh, Cenki-Tok, Ballevre and Engi2018), this equilibrium assemblage diagram was computed with a fixed H₂O content corresponding to that stored in the solid phases stable at peak P–T conditions (Bulk III composition; Table S3). The garnet rim composition (X_Prp = 0.03, X_Sps = 0.02, X_Alm = 0.70, X_Grs = 0.25), plots within the Grt–Cld–Wm–Chl–Pg–Rt–Qtz field at c. 1.6–2.0 GPa and 525–570°C (Figure 10d). The measured Si⁴⁺ in white mica rims (3.25–3.35 a.p.f.u.) is consistent with this P–T range. Glaucophane likely grew during early decompression, in association to the S2 foliation, between ∼2.4 and 2.0 GPa, and was later on pseudomorphosed by chlorite + albite aggregates at lower pressure conditions, in agreement with petrographic observations.

9.c. P–T results for metapelite DD20

The equilibrium assemblage diagram for DD20 (Bulk I; Table S4 in Supplementary Material 1; Figure 11a) shows a broadly similar topology to that of 23PIN43. The main differences are that glaucophane is absent and kyanite appears at temperatures as low as 480°C above 2.7 GPa. The measured composition of the garnet core (X_Prp = 0.08, X_Sps = 0.12, X_Alm = 0.75, X_Grs = 0.05) and the corresponding isopleths indicate crystallization at c. 540–570°C and 2.5–2.7 GPa, within the Car–Grt–Cld–Wm–Omp–Ky–Lws–Rt + Qtz field. The measured Si content in white mica (3.45–3.50 a.p.f.u.) and the X_Mg of chloritoid inclusions in the garnet core (0.14–0.15) are consistent with the P–T range inferred from garnet composition.

Figure 11.

(a) Equilibrium assemblage diagram calculated using Bulk I for sample DD20. (b) Overlap of selected garnet core compositional isopleths (X_Prp = 0.08, X_Sps = 0.12, X_Alm = 0.75, X_Grs = 0.05), together with X_Mg in chloritoid and Si (a.p.f.u.) in white mica; the intersection is highlighted by the black ellipse. (c) Overlap of garnet compositional isopleths for the garnet mantle (Bulk II; X_Prp = 0.09, X_Sps = 0.09, X_Alm = 0.75, X_Grs = 0.06), combined with X_Mg in chloritoid and Si (a.p.f.u.) in white mica. (d) Overlap of garnet compositional isopleths for the garnet rim (Bulk III; X_Prp = 0.10, X_Sps = 0.05, X_Alm = 0.72, X_Grs = 0.13) and Si⁴⁺ (a.p.f.u.) in white mica. Equilibrium assemblage diagrams calculated using Bulk II and Bulk III are provided in Supplementary Material 1 (Figure S8).

Figure 11. Long description

Panel A: A phase diagram shows equilibrium assemblage for Bulk I in sample DD20. The x-axis represents temperature in degrees Celsius, and the y-axis represents pressure in gigapascals. Various mineral assemblages are labeled within the diagram. Panel B: A phase diagram overlays selected garnet core compositional isopleths (XPrp=0.08, XSps=0.12, XAlm=0.75, XGrs=0.05) with XMg in chloritoid and Si (a.p.f.u.) in white mica. The intersection is highlighted by a black ellipsoid. The x-axis represents temperature in degrees Celsius, and the y-axis represents pressure in gigapascals. Panel C: A phase diagram overlays garnet compositional isopleths for the garnet mantle (Bulk II; XPrp=0.09, XSps=0.09, XAlm=0.75, XGrs=0.06) with XMg in chloritoid and Si (a.p.f.u.) in white mica. The x-axis represents temperature in degrees Celsius, and the y-axis represents pressure in gigapascals. Panel D: A phase diagram overlays garnet compositional isopleths for the garnet rim (Bulk III; XPrp=0.10, XSps=0.05, XAlm=0.72, XGrs=0.13) and Si4+ (a.p.f.u.) in white mica. The x-axis represents temperature in degrees Celsius, and the y-axis represents pressure in gigapascals.

Mantle crystallization corresponds to P–T conditions of c. 2.7–3.1 GPa and 540–570°C (Figure 11c), based on the measured garnet composition (X_Prp = 0.09, X_Sps = 0.09, X_Alm = 0.75, X_Grs = 0.06) and Bulk II (Table S4, Figure S8a in Supplementary Material 1). The modelled Si content of white mica in this field (c. 3.45–3.50 a.p.f.u.) matches the measured values. The X_Mg of chloritoid rims and inclusions in the outer core of garnet (0.17–0.19) is also consistent with the inferred P–T range in the Car–Grt–Cld–Wm–Omp–Ky–Lws–Rt + Coe field. No coesite relics were identified, likely due to the abundance of fractures in garnet crystals limiting inclusion preservation. Moreover, most quartz inclusions in the garnet mantle are not barometrically isolated, leaving open the possibility that some originally represented coesite. Calculated modal abundances (∼6 vol% garnet and chloritoid, ∼1.5 vol% lawsonite) are consistent with petrographic observations (Figure S9 in Supplementary Material 1). Kyanite is predicted at ∼10 vol% but is not observed in thin section, either due to replacement during decompression or a slight overestimation of bulk Al₂O₃; a T–X(Al₂O₃) pseudosection at 2.7 GPa confirms that a variation of ∼2 wt% is sufficient to suppress kyanite stability (Figure S10; Supplementary Material 1). Omphacite and carpholite (∼7 and ∼10 vol%, respectively) were likely replaced during decompression, as discussed for the previous sample. The peak temperature estimated from the equilibrium assemblage diagrams is, within uncertainties, consistent with the RSCM temperature of 503°C ± 23°C.

The third diagram (Bulk III; Table S4; Figure S8b in Supplementary Material 1) constrains garnet rim crystallization to c. 1.2–1.6 GPa and 550–560°C (Figure 11d), based on the rim composition (X_Prp = 0.10, X_Sps = 0.05, X_Alm = 0.72, X_Grs = 0.13). At these conditions kyanite is no longer stable and paragonite is instead predicted, consistent with petrographic observations. Unlike 23PIN43, amphibole is not predicted for DD20, in agreement with petrographic observations (Section 5.b.).

9.d. P–T results for metapelite FONT2

THERMOCALC average P–T calculations were performed for sample FONT2 using two different sets of mineral compositions (Table S3 in Supplementary Material 2). These calculations indicate peak metamorphic conditions of 507°C ± 14°C and 2.9 ± 0.3 GPa (σ fit = 1.08) and 501°C ± 13°C and 2.7°C ± 0.3 GPa (σ fit = 0.90), respectively (Figure S11 in Supplementary Material 1). The temperatures estimated with THERMOCALC are consistent with those obtained using the RSCM method (504°C ± 24°C). The calculated pressures are consistent with the occurrence of coesite relics in this sample and with the peak P–T conditions inferred from the PERPLE_X thermodynamic modelling of the other two samples. The mineral assemblages observed are comparable to those predicted under similar P–T conditions by equilibrium assemblage diagrams calculated for an average metapelitic bulk composition (Smye et al. Reference Smye, Greenwood and Holland2010; Manzotti et al. Reference Manzotti, Schiavi, Ballèvre and Nosenzo2025 b).