1 Introduction

During Late Carboniferous to Early Permian times, the Variscan orogenesis faced end and subsequent tectonic collapse of the orogen. The kind of geodynamic model that actually led to the break-up of Pangea and the formation of intracontinental Permian basins accompanied by magmatic activity in the present-day Europe area is still largely debated (e.g. Ménard & Molnar, 1988; Echtler & Malavieille, 1990; Burg et al., 1994; Rottura et al., 1998; Cagnard et al., 2004; Casini & Oggiano, 2008; Le Pichon et al., 2019, 2021; Fréville et al., 2022). Several mechanisms have been proposed to be responsible for the Early Permian phase of crustal extension and lithospheric thinning. Many interpretations cluster around three hypotheses: (i) a tectonic regime characterized by dextral transtension (e.g. Arthaud & Matte, 1977; Cassinis & Perotti, 1994) and wrenching associated to the transition from Pangea A to Pangea B configuration (Muttoni et al., 2003; Schaltegger & Brack, 2007; Pohl et al., 2018), (ii) an orogenic collapse stage (Dewey, 1988; Ménard & Molnar, 1988; Malavieille et al., 1990; Malavieille, 1993) at the end of the Variscan orogenesis, accompanied by almost pure N-S directed extension (in the present-day coordinates), (iii) lithospheric extension and thinning at the initial stage of continental rifting (Lardeaux & Spalla, 1991; Siletto et al., 1993). Several fault-controlled intra-continental basins of Late Carboniferous to Late Permian age are preserved to the north of the present-day Alps belt (e.g. von Raumer, 1998; McCann et al., 2006), in the Southalpine Domain (e.g. Cadel et al., 1996; Cassinis et al., 2012; Berra et al., 2016; Zanchi et al., 2019; Locchi et al., 2022), but also within the axial belt of the Eastern and Western Alps (e.g. Capuzzo & Wetzel, 2004; Veselá et al., 2008).

The formation of intra-continental basins occurred together with a diffuse magmatic activity, recorded as emplacement of gabbro to gabbro-diorite bodies close to the lower crust–mantle transition (Sinigoi et al., 1991; Zingg et al., 1990; Tribuzio et al., 1999; Müntener et al., 2000; Rebay & Spalla, 2001), emplacement of intermediate to acidic intrusive complexes at mid to shallow crustal levels (e.g. Rottura et al., 1998; Steck et al., 2013; Bergomi et al., 2017; Boscaini et al., 2020) and volcanic activity at the surface, within or in the proximity of the intra-continental basins (Cadel et al., 1996; Cortesogno et al., 1998; Schaltegger & Brack, 2007; Cassinis et al., 2012; Willcock et al., 2013).

Geochemical and geochronological studies of the Late Carboniferous—Early Permian magmatic activity in the Alps have been mainly focused on the gabbroic bodies within the Austroalpine and Penninic domains north of the Periadriatic Fault (e.g. Rebay & Spalla, 2001; Spalla et al., 2014; Manzotti et al., 2015; Petri et al., 2017; Bergomi et al., 2017; Ballèvre et al., 2020 with references), where original field relationships between magmatic-volcanic rocks and hosting metamorphic/sedimentary rocks have been severely overprinted by Alpine deformation and metamorphism.

In the Southalpine Domain deformation and metamorphism related to the Alpine collision were much less severe and allow to a valuable reconstruction of the architecture and stratigraphic evolution of the Permian basins and volcanic units, together with the intrusive relationships and the thermal aureole associated to magmatic bodies (e.g. Wyhlidal et al., 2012). Most of available data however cluster in the western (i.e. Ivrea-Verbano Zone; Sinigoi et al., 1991; Schaltegger & Brack, 2007) and eastern sectors (Marocchi et al., 2008; Bargossi et al., 2010; Boscaini et al., 2020) of the Southalpine Domain (Fig. 1), with the central Southern Alps (CSA) that appear to be under-investigated. Geochronological and geochemical data are limited to the volcanics occurring in the Laghi Gemelli Group (Brack & Schaltegger, 1999; Schaltegger & Brack, 2007; Gretter et al., 2013; Berra et al., 2015), on the intrusive bodies of the Val Trompia area (Locchi, 2023) and to the Val Biandino Intrusive Suite (Pohl et al., 2018). Even if geochronological data obtained from high-retentivity geochronometers (e.g. U–Pb on zircon) are available for the Val Biandino Intrusive Suite, no complete geochemical datasets considering all major and trace elements commonly used to decipher the petrogenesis and evolution magmatic complexes (De Capitani & Liborio, 1988; De Capitani et al., 1994) are available.

Fig. 1
figure 1

Tectonic scheme of the Alps with highlighted all the occurrences of Permian volcanic and intrusive rocks. Small number in black refer to references reported in Table SUPP1 (age compilation of volcanic and intrusive rocks in the Southalpine Domain) and below. Only ages obtained from high-retentivity isotopic system have been preferred where available for each magmatic complex. CSA: central Southern Alps; ESA: eastern Southern Alps; WSA: western Southern Alps. [1] Klötzli et al., 2014; [2] Sinigoi et al., 2011; [3] Peressini et al., 2007; [4] Quick et al., 2003; [5] Garuti et al., 2001; [6] Vavra et al., 1999; [7] Pin, 1986; [8] Quick et al., 2009; [9] Schaltegger & Brack, 2007; [10] Pinarelli et al, 1988; [11] Pohl et al., 2018; [12] Berra et al., 2015; [13] Gretter et al., 2013; [14] Brack & Schaltegger, 1999; [15] De Capitani et al., 1994; [16] Borsi et al., 1966; [17] Macera et al., 1994; [18] Barth et al., 1994; [19] Morelli et al., 2012; [20] Rottura et al., 1998; [21] Marocchi et al., 2008; [22] Del Moro & Visonà, 1982; [23] Boscaini et al., 2020; [24] Avanzini et al., 2010; [25] Bargossi et al., 2010; [26] Piccin et al., 2009; [27] Visonà et al., 2007; [28] Klötzli et al., 2003

In this work we provide, based on new geochemical and geochronological data, insights on the magmatic evolution of the Val Biandino Intrusive Suite. We show how, despite its small size, the study of the Val Biandino Intrusive Suite could help in the understanding of the Early Permian magmatism in the Southalpine Domain. The fact that intermediate and felsic plutons that were emplaced at different crustal levels crop out in a relative small area allow to investigate their genetic relationships, shading some lights on the mechanisms of magma genesis and evolution of the magmatic systems during the Early Permian extensional phase in southern Europe.

2 Geological setting

The Val Biandino Intrusive Suite (Pohl et al., 2018) crops out in the central sector of the Southalpine Domain (central Southern Alps, CSA in Fig. 1), defined as the area comprised between Lake Como and the Giudicarie Fault System. The CSA are a thick-skinned fold-and-thrust belt that extends from the Periadriatic Fault to the north, active at least since the Oligocene (Schmid et al., 1989; Zanchetta et al., 2023), and the Po plain to the south. In the CSA, sedimentary and magmatic rocks of Permian age are chiefly exposed at the core of three regional anticlines, called the Orobic Anticlines (De Sitter & De Sitter-Koomans, 1949; Schönborn, 1992). These anticlines occur just to the south of the Orobic-Porcile-Gallinera thrust system, a regional scale structure along which the Variscan basement was thrust over the Upper Pennsylvanian to Lower Triassic volcano-sedimentary cover (Laubscher, 1985; Schönborn, 1992; Blom & Passchier, 1997; Carminati et al., 1997; Zanchetta et al., 2011; D’Adda & Zanchetta, 2015; Zanchetta et al., 2015; Mittempergher et al., 2021). To the south of the Orobic Anticlines and the Adamello batholith (Fig. 1) another regional anticline occurs, the Monte Alto anticline (Schönborn, 1992). This regional-scale fold is responsible for the exposure of the Permian rocks below the Triassic and younger successions of the CSA (Fig. 1).

Despite the locally intense Alpine deformation that affected the Upper Pennsylvanian to Lower Triassic rocks of these anticlines, two distinct basins have been reconstructed: the Orobic Basin, corresponding to the three Orobic Anticlines, and the Collio Basin (or Val Trompia basin), corresponding to the Monte Alto anticline (Cassinis et al., 2007, Cassinis et al., 2012). Another small basin has been identified to the NW of the Collio one: the Boario Basin (Fig. 1; Cassinis et al., 2012). These basins developed from the Early Permian as fault-controlled basins (Zanchi et al., 2019; Locchi et al., 2022) in an extensional continental setting. Large amounts of volcanic products and siliciclastic sediments with thickness in the order of a few kilometres were accumulated (De Sitter & De Sitter-Koomans, 1949; Casati & Gnaccolini, 1967; Cassinis et al., 1988; Cadel et al., 1996; Berra & Carminati, 2010; Berra et al., 2016). Significant differences exist among the basins in terms of architecture, stratigraphic evolution and volcanic vs. siliciclastic prevalence (e.g. Cassinis et al., 2012).

The Val Biandino Intrusive Suite occurs at the western end of the Orobic Basin, that extends in a E-W direction for about 45 km, with a present width of less than 10 km (measured N-S). The oldest sediments deposited above the exposed Variscan basement, which forms the sole of the Orobic Basin, are fluvial conglomerates (Conglomerato Basale, late Pennsylvanian?) with sandy matrix and clasts consisting of polycrystalline quartz and rare metamorphic clasts (Casati & Gnaccolini, 1967; Cadel et al., 1996). Above this unit that irregularly covers the metamorphic basement, the Permian succession starts with the Cabianca Volcanite (the Lower Collio Fm. Auctorum) consisting of pyroclastic flows, tuffs and minor terrigenous deposits and continues with mainly siliciclastic sediments with rare tuffs of the Pizzo del Diavolo Fm. (Upper Collio Auctorum) (Berra & Felletti, 2010; Gaetani et al., 2012; Berra et al., 2016). The Pizzo del Diavolo Fm., the Cabianca Volcanite and the Conglomerato Basale are grouped together in the Laghi Gemelli Group (Berra et al., 2016).

In the Valsassina area the total thickness of the Permian succession is about 1200 m (Gaetani et al., 1986; Sciunnach, 2003). The end of the Permian cycle is marked by a gap during the Middle Permian, followed by the deposition of red fluvial conglomerates of the Lopingian Verrucano Lombardo, that rest with an angular unconformity above the Pizzo del Diavolo Fm. (Gaetani et al., 1986; Sciunnach, 2001a). The thickness of the Verrucano Lombardo displays significant lateral variations, testifying for the occurrence of a complex paleo-topography, likely controlled by active tectonics possibly due to a partial inversion of the Cisuralian basins (Sciunnach, 2003), at the end of the pre-Verrucano Permian volcano-sedimentary cycle. In the western part of the Orobic Anticline the Laghi Gemelli Group is overlying a metamorphic basement made of quartz-rich two-mica gneiss (referred in the literature as the Morbegno Gneiss, Cornelius, 1916) with minor micaschists, quartzites and amphibolites (Froitzheim et al., 2008).

2.1 The Val Biandino Intrusive Suite

The basement of the Valsassina area, at the western end of the central Southern Alps, was deformed and metamorphosed during the Variscan orogeny (e.g. Siletto et al., 1993). During the Early Permian, several magmatic bodies ranging from granitic to dioritic in composition (Pasquarè, 1967; De Capitani, 1982), often accompanied by several leucogranitic and aplitic dikes, intruded the hosting basement rocks. The Val Biandino Intrusive Suite has been subdivided in two distinct intrusive suites: the “Val Biandino pluton” and the “Valle di San Biagio Granite” (Porro, 1897; Crommelin, 1932; De Sitter & De Sitter-Koomans, 1949; Pasquarè, 1967). The Val Biandino pluton, here referred as the VBQD (“Val Biandino Quartz-Diorite” of Pohl et al., 2018) consists of four major bodies (Fig. 2). In addition to these major bodies, small stocks and lenses also occur (Fig. 2). The pink-colored porphyric “Valle di San Biagio Granite” (VSBG of Pohl et al., 2018), located to the NW of the Biandino pluton, is exposed in four erosional windows (Fig. 2) beneath the Southalpine sedimentary cover, here made by the continental deposits of the Verrucano Lombardo and the marly sandstones of the Servino (Gaetani et al., 1986).

Fig. 2
figure 2

Simplified geological map of the Valsassina northern sector. The Val Biandino Intrusive Suite is divided into two units: the “Val Biandino Quartz-Diorite” (VBQD) and the “Valle di San Biagio Granite” (VSBG). The VBQD crops out in a tectonic window bounded by the Grassi Shear Zone (GSZ; Froitzheim et al., 2008) and the Sassi Rossi Normal Fault (SRNF; Sciunnach, 2001b) intruding the late Variscan metamorphic basement of the central Southern Alps. The VSBG crops out in small erosional windows to the NW of the SRNF, below the Lopingian Permian Verrucano Lombardo. Analysed samples are reported in the map. A, B, C, and D are the four major stocks of the VBQD

The magmatic rocks of the VBQD, and their hosting paragneiss of the Southalpine basement, crop out in an area that extends in a E-W direction for about 15 km with a N-S maximum width of 7–8 km (Fig. 2). The contact between the basement with the overlying sedimentary-volcanic cover is tectonic and has been interpreted to represents a syn-magmatic detachment fault, the Grassi Detachment (Froitzheim et al., 2008; Pohl et al., 2018), that was responsible for the exhumation of the crystalline basement as a metamorphic core complex. The contact between the VBQD and the VSBG is tectonic and occurs along a NW-dipping high-angle tourmalinite-bearing normal fault (Fig. 2), the Sassi Rossi Fault (Sciunnach, 2001b; Pohl et al., 2018; Locchi 2023). This fault is crosscut atop by the erosional unconformity on which the Verrucano Lombardo conglomerates and sandstones deposited, directly above the Lower Permian rocks exposed at the surface during the Late Permian.

The main intrusive body of the VBQD crops out in the medium tract of the Biandino valley (A in Fig. 2, following the nomenclature of De Capitani & Liborio, 1988) and is made mainly of quartz-diorite accompanied by minor granite lenses and dikes. The second major body (B in Fig. 2) consists of tonalite and gabbro-diorite with granite lenses and dikes concentrated at its rims. Quartz-diorite and gabbro-diorite prevail in the Zucco del Corvo stock (C in Fig. 2), where the occurrence of gabbro have also been reported (Pasquarè, 1967). The westermost body crops out N of Cortabbio (D in Fig. 2) and consists almost exclusively of granodiorite, with some occurrence of fine-grained leucogranitic dikes at its southern rim.

Geochemical data (De Capitani, 1982; De Capitani & Liborio, 1988; De Capitani et al., 1994) indicate that rocks of the VBQD have a typical high-K calc-alkaline affinity and can be divided into two groups, based on the existence of a gap in the silica content. The first group represents the main part of the intrusive bodies, having a SiO2 content lower than 64 wt%, whereas rocks of the other group, with SiO2 higher than 68 wt%, occur as minor lenses and dikes of granitic composition at the rims of larger bodies.

The VSBG granite occurs NW of the VBQD, separated from this by the Sassi Rossi Normal Fault (Sciunnach et al., 2001b; SRNF in Fig. 2). The granite usually displays a pink color in the field with a porphyric to granophyric texture (Porro, 1897; De Sitter & De Sitter-Koomans, 1949).

Available K–Ar (De Capitani et al., 1988) and Rb–Sr (Thöni et al., 1992) whole rock and single mineral ages loosely constrain an intrusion age of 312 ± 48 Ma for the quartz-diorites and tonalities, i.e. the main intrusive bodies. More recent U–Pb LA-ICP-MS zircon data indicate a crystallization age of 289.1 ± 4.5 Ma for the VBQD and 286.8 ± 4.9 Ma for the VSBG (Pohl et al., 2018).

3 Methods

3.1 Whole rock geochemical analysis

The whole rock contents of major, minor and trace elements were determined at the ACME analytical labs, Vancouver (Canada). At least 500 g of fresh un-weathered material for each sample was crushed and pulverized. After quartering an amount of 20 g was used for analysis.

Total abundances of the major oxides were obtained by ICP-ES (Inductively Coupled Plasma Emission Spectroscopy), whereas REE (Rare Earth Element), refractory elements and precious and base metals by ICP-MS (Inductively Coupled Plasma Mass Spectroscopy). Samples were prepared for the ICP-MS analysis by a LiBO2 fusion and dilute nitric digestion for major oxides, REE and refractory elements, whereas precious and base metals were digested in aqua regia. Analytical errors are within 2% for major elements and in the 5–10% range for trace elements, as determined by repeated analysis of the same sample (1 among 5).

3.2 U-Th-Pb dating of zircons

Two samples, a porphyric granite from the VSBG (VBZ43), and one quartz-diorite from the VBQD (VBZ3) were selected for U–Pb zircon dating. Samples were crushed, sieved and separated in two granulometric fractions (125–250 and 250–500 μm). The granulometric fractions were selected following thin section analyses that highlighted that most of the zircon grains population fall within those ranges. Zircon grains were concentrated by Wilfley table, heavy liquids and Frantz magnetic separator. Individual grains were hand-picked, ultrasonically cleaned in de-ionized water, and cast on a 3.5 cm diameter epoxy mount. The mount was then polished and documented at the optical microscope (reflected and transmitted light) and the scanning electron microscope (back-scattered and cathodoluminescence). After cleaning and drying, the mount was gold-coated. Zircon grains were analyzed by a sensitive high-resolution ion probe (SHRIMP) at the IBERSIMS laboratory, Granada University (Spain).

Each analytical spot was rastered with the primary ion beam (16O16O+) for 120 s before analysis, the following isotopes were then measured: 196Zr2O, 204Pb, 204.1background, 206Pb, 207Pb, 208Pb, 238U, 248ThO, 254UO. The spot diameter on mount was ca. 18 μm. The SL13 reference zircon (U: 210 ppm, Black et al., 2004) was used as standard for U concentrations. U/Pb ratios were calibrated using the TEMORA-II reference zircon (417 ± 1 Ma, Black et al., 2003). Common lead (204Pb) was corrected from the measured 204Pb/206Pb using the model of terrestrial evolution of Pb of Cumming and Richards (1975).

4 Field aspects and petrography of the Val Biandino Intrusive Suite

The intrusive rocks of the VBQD and VSBG, together with their hosting Variscan metamorphic basement, crop out in a tectonic window (Froitzheim et al., 2008) within the Permian volcanic and sedimentary succession (Fig. 2). The basement-cover contact is tectonic along the Grassi Shear Zone, or stratigraphic in the hangingwall of the same structure (Fig. 2). To the north and west of Cima d’Agrella (Fig. 2) the Verrucano Lombardo directly overlies in non-conformity the metamorphic basement and the Early Permian intrusive rocks of the VSBG, these ones showing evidence of exposure (paleosoils) at the surface before the deposition of the Verrucano Lombardo (Sciunnach, 2001a). Elsewhere the contact between the basement and the cover rocks is almost invariably marked by a ductile shear zone (the Grassi Shear Zone, GSZ, in Fig. 2), locally overprinted by cataclastic deformation, that has been interpreted as a detachment fault (Froitzheim et al., 2008; Pohl et al., 2018).

Both the VBQD and the VSBG intruded the Southalpine basement which reached amphibolite facies metamorphic conditions during the Variscan orogenic cycle (Milano et al., 1988; Diella et al., 1992). The basement mainly consists of two-mica paragneiss (Morbegno Gneiss; Cornelius, 1916) with minor micaschists, quartzites and amphibolites. The equilibrium mineralogical association at peak conditions consists of Qz + Pl + Bt + Chl + Ms + Gar + Ky + St (Froitzheim et al., 2008; mineral abbreviation here and in all the text after Whitney & Evans, 2010). Retrogression at lower greenschist facies conditions was completed during the Variscan orogenic cycle (Spalla et al., 1999; Zanoni and Spalla, 2018).

Along the Grassi Shear Zone (GSZ, Fig. 2) the paragneiss displays a mylonitic fabric with a well-developed stretching lineation. Biotite is replaced by chlorite that, together with muscovite, marks the mylonitic foliation. The stretching lineation, W of Rif. Grassi (Fig. 2), is highlighted by the occurrence of quartz rods, up to 20–25 cm in length (Fig. 3a). Finite strain associated with the GSZ rapidly decreases away from the contact, as highlighted by the occurrence of small-scale isoclinal folds that have been completely transposed within the shear zone. In the GSZ segment close to the Rif. Grassi, the mylonites are overprinted by brittle deformation associated with the development of cataclastic bands displaying a S-verging reverse motion, that affect both the basement and the overlaying Monte Cabianca Volcanite Small scale, asymmetric SE-verging folds interpreted as coeval with this brittle deformation phase, also occur. The vergence of these folds and the fact that they display brittle slip along their long limbs with kinematic indicators pointing to a S-vergent hanging wall motion (Fig. 3b), are consistent with the kinematics of the Late Cretaceous-Paleocene Orobic Thrust (Fig. 2; Zanchetta et al., 2011; Zanchetta et al., 2015).

Fig. 3
figure 3

a Mylonites with quartz rods related to the Grassi Shear Zone (GSZ, Fig. 2) in the hosting basement of the VBQD; b Fault planes with a top-to-SE reverse motion developed in the paragneiss of the Variscan basement close to Rif. Grassi (Fig. 2); c Brittle planes overprint the quartz fabric related to subgrain rotation recrystallization in the mylonites of the Grasi Shear Zone

4.1 The Val Biandino Quartz-Diorite (VBQD)

The VBQD consists of several stocks and minor lenses of mafic to felsic magmatic rocks that crop out in area comprised between Valtorta and Cortabbio (Fig. 2). Quartz-diorite, diorite and gabbro-diorite form the main bodies of the VBQD (A, B, C and D in Fig. 2; following De Capitani et al., 1988), with granodiorite chiefly occurring within bodies A and D.

Field surveys revealed that felsic types such as granite, two-mica leucogranite and aplitic dikes are confined at the rims of the main bodies, postdating their emplacement (Fig. 4a).

Fig. 4
figure 4

Field aspect of the VBQD intrusive rocks. a Leucocratic granitic dike intruding the Variscan basement just S of Rif. Grassi (Fig. 2); b The quartz-diorite of stock A (Fig. 2) as observable in the field; c Quartz-diorite of stock D (Fig. 2) intruding the basement paragneiss with apophyses of the main body that crosscut the regional foliation and sills of a few mm in thickness that expand along the foliation planes; d Basement xenoliths in the VBQD granodiorite (sample VBZ32, Fig. 2) that is in turn included in a granite dike

Diorite and quartz-diorite are the most abundant magmatic types. They are usually medium-grained to fine-grained with an isotropic texture (Fig. 4b). A poorly-developed magmatic foliation has been observed at the rims of bodies A and D, marked by the shape preferred orientation of biotite crystals. The magmatic phase assemblage consists of Pl + Bt + Amp + Qz + Kfs + Ap + opaque minerals (Fig. 5 and Table 1). The occurrence of orthopyroxene in these rocks is reported by De Capitani et al. (1988), but was not observed in the samples from this study, even if cummingtonite has been recognized in several samples, and may represents the alteration product after hydration of orthopyroxene.

Fig. 5
figure 5

Quartz-diorites of the VBQD a at parallel polars. b two generation of plagioclase with large crystals (Pl 2) having a saussuritized core and Ab-rich second generation forming smaller interstitial crystals with polysynthetic twin. c Cordierite crystals within the granites of the VBQD; d same as (c) at crossed polars

Table 1 Samples location and petrography

Plagioclase is almost invariably zoned, mainly occurring as the interstitial phase between biotite and amphibole. Biotite occurs as highly pleochroic crystals with a reddish brown colour. Amphibole occurs in prismatic crystals reaching a few millimetres in length and are often partially replaced by biotite (Fig. 5a).

Gabbro-diorite form small stocks to the south of body B (Fig. 2) and at the southern margin of stock C. The magmatic assemblage is similar to that described for diorite and quartz-diorite except for the lack of biotite and K-feldspar. The occurrence of gabbro has been reported by Pasquarè (1967) within stock C (Fig. 2), but, based on our observations, only gabbro-diorite and diorite crop out there.

Granite and granodiorite are less abundant than mafic-intermediate rocks, with granodiorite that mainly occurs in the northern area of body A (Fig. 2) and, to a minor extent, as small lenses within body D, despite the fact that this magmatic body is historically known as the “Cortabbio Granodiorite” no clear crosscutting relationships have been observed between the granodiorite and mafic types, but the occurrence of dioritic and tonalitic enclaves within a granodiorite body SW of Rif. Grassi (Fig. 2) suggests that more differentiated rocks are younger or at least coeval with respect to diorite and quartz-diorite. The magmatic phase assemblage of studied granodiorite samples is made of Qz + Kfs + Bt + Pl ± Amp ± Ms + Ap + Zrn + opaque minerals. Granodiorite of the VBQD show the same mineralogical and fabric features across the whole complex. They are medium grained with an isotropic texture. An oriented fabric is seldom observed within sills at the rims of body D, intruded parallel to the main foliation in metamorphic host rocks. In these cases, the shape preferred orientation of biotite marks a poorly defined magmatic foliation parallel to sills rims.

The occurrence of two plagioclase generations is a common feature. The first one shows large crystals (up to 10 mm) typically with saussurite growing on Ca-rich cores and sericite replacing Ca-poor rims. The second generation, usually almost completely preserved, is made up of smaller interstitial plagioclase frequently displaying an albite twinning (Fig. 5b). Chlorite and sericite are present as alteration products on biotite and K-feldspar, respectively. Myrmekitic rims frequently occur along boundaries between K-feldspar and plagioclase or K-feldspar and quartz. Amphibole in the granodioritic types are partially substituted by biotite.

The VBQD comprises also two-mica leucogranite and aplitic dikes. Two-mica granitic lenses and dikes are mainly concentrated along the rims of bodies A and B. They occur as small lenses and dikes up to several meters in thickness (Fig. 4a). Granite are usually fine-grained and equigranular, with a typical magmatic assemblage made of Qz + Kfs + Pl + Ms + Bt ± Crd ± Ap ± Zrn ± Aln ± Ttn + opaque minerals. Epidote, chlorite and calcite occur as secondary phases. The presence of cordierite (Fig. 5c) suggests a peraluminous character for the granite, whose significance will be discussed in detail in Sect. 7.2.3. If the crosscutting relationships between dioritic and granodioritic bodies are unclear, the intrusion of the two-mica leucogranite clearly postdates the main mafic and intermediate magmatic bodies. This is clear at the NE margin of stock A, where crosscutting relationships between intermediate rocks and late-stage granitic and leucogranitic dikes can be observed. Here dioritic to tonalitic enclaves with more or less equidimensional shape and angular edges, are hosted in a granitic mass, containing also basement xenoliths made of partially digested paragneiss, showing evident signs of partial melting (Fig. 4d). Other inclusions are made of polycrystalline quartz nodules and minor granodiorite with abundant biotite.

4.2 The Val di San Biagio Granite (VSBG)

The VSBG crops out in the central part of Valsassina, north of the VBQD (Fig. 2). The VSBG consists of a coarse-grained pink-coloured granite (Fig. 6a), with a typical porphyric texture. The mineralogy is Qz + Kfs + Pl + Bt ± Ms + Zrn + Ap + opaque minerals. Degassing cavities, filled with quartz, calcite and zeolites commonly occur close to the contact with the hosting basement rocks. Mafic and dioritic enclaves, together with basement xenoliths, are quite common in the southernmost body, close to the Sassi Rossi Normal Fault (SRNF in Figs. 2, 6b), but only rarely observed elsewhere. The Sassi Rossi Normal Fault is characterized by a several decimetres thick cataclasites, mainly developed at the expense of rocks in the footwall of the fault plane. Locally, the cataclasites are impregnated by dark brown to black tourmalinite (Fig. 7), a feature already reported from other Permian faults close to magmatic rocks (e.g. Zhang et al., 1994; De Capitani et al., 1999; Zanchi et al., 2019; Locchi et al., 2022, Zanchetta et al., 2022). The SRNF terms abruptly both toward NE and S (Fig. 2) where the Verrucano Lombardo rests in non-conformity directly on the fault plane. These field relationships suggest that the SRNF was active in a time interval comprised between the intrusion of the VSBG and the deposition of the Lopingian Verrucano Lombardo (Sciunnach, 2001b).

Fig. 6
figure 6

a The typical porphyric texture of the VSBG granite as it appears in outcrops; b) The VSBG intruding Variscan basement NW of the SRNF (Fig. 2); c Myrmekites developed at the contacts between K-feldspar and plagioclase crystals; d detail of figure c

Fig. 7
figure 7

a Foliated cataclasites along the Sassi Rossi Normal Fault (SRNF, Fig. 2); b Tourmalinite vein of ca. 2 cm thickness injected along the SRNF fault plane.

4.3 Contact metamorphism associated to the Val Biandino Quartz-Diorite

Thermal metamorphism associated with the emplacement of the Val Biandino intrusives affected the hosting paragneiss up to 150–200 m away from the magmatic bodies. The first observable effect is the crystallization of a new generation of muscovite and biotite, growing statically on the Variscan foliation. Biotite along the regional foliation appears strongly decoloured, whereas the second generation occurs both as fresh single flakes and as pluri-millimetric aggregates (Fig. 8a), made by tabular lamellae of red-brownish Ti–rich biotite. A second generation of muscovite has been also observed in some samples.

Fig. 8
figure 8

a Static overgrowth of biotite crystals in the thermal aureole of the VBQD on pre-existing biotite in the host paragneisses, close to the contact with stock A (Fig. 2). b Pinite + quartz aggregates after cordierite in the basement paragneisses (stock B, Fig. 2). c Statically growth of corundum (Crn) and sillimanite (Sil) within the contact aureole. d Figure (c) at crossed polars. The size of the corundum crystals allow to rule out a contamination during thin sections preparation and polishing

Small sillimanite crystals with a prismatic habit (Fig. 8c and d) have been recognized in samples collected less than 80 m from the contact between the paragneiss and the diorite of body B (Fig. 2). The occurrence of andalusite and cordierite in biotite gneiss within the contact aureole was reported by Pasquarè (1967) in the host rocks of body C (Fig. 2), but neither minerals have never been observed in our samples. The recognition of cordierite could be hampered by the fact that it is easily replaced by pinite (chlorite + muscovite, Fig. 8b), especially if late-stage hydrothermal fluids circulated during the final stage of pluton emplacement, as suggested by the observed occurrence of tourmaline-bearing quartz veins close to the intrusive contact of bodies A and D. Newly formed, thin K-feldspar rims around plagioclases occur in paragneiss as far as ca. 100 m from the intrusive contact.

The foliated fabric of hosting paragneiss is completely obliterated within 10–15 m from the intrusive contact, with paragneiss that are replaced by massive dark-gray coloured hornfels, sometimes with bluish shade. Small (up to 150 μm of diameter) subeuhedral corundum crystals (Fig. 8c and d) have been recognized in one sample. The occurrence of corundum formed in response to the thermal metamorphism in the contact aureole of a Southalpine Permian intrusion has been also reported by Whylidal et al. (2012) for the contact aureole of the Brixen granodiorite. The occurrence of corundum suggests temperature above 680 °C reached in the close proximity of the margins of the intrusive bodies (Montel et al., 1986).

5 Geochemistry

A total of 11 samples where crushed and processed for bulk rock geochemical analysis. The analyses were performed on samples collected in different parts of the intrusive complex and selected to be representative of the lithological variety, from gabbro-diorite to granite (Table 2).

Table 2 Whole rock major (oxide %) and trace element (ppm) abundances in the analysed samples of the Val Biandino Intrusive Complex

The SiO2 content ranges from 52 wt% to 75 wt% and it is negatively correlated with the FeO and MgO contents (Fig. 9) that range 7–10 wt% and 5–10 wt%, respectively. Alkalis wt% (Na2O + K2O) varies from ca. 3 wt% in low-SiO2 rocks, up to 7 wt% in granites of the VBQD (Fig. 9, Table 2) and more than 8 wt% in the VSBG granites (Table 2). K2O is commonly higher than Na2O, with the only exceptions of granodiorites of lenses A-D and the diorite of lens D (Table 2).

Fig. 9
figure 9

a SiO2 vs. K2O wt% diagrams displaying the high-K calc-alkaline affinity of the VBQD and VSBG rocks; b alkali-silica classification diagram that shows the positive correlation between silica and alkali and also the gaps existing in the distribution of the SiO2 content in the rocks of the VBQD

The TAS classification diagram and the K2O-SiO2 variation diagram indicate a high-K calc-alkaline affinity (Fig. 9a and b) for the rocks of the Val Biandino Intrusive Suite, as already pointed out by De Capitani & Liborio (1988) and Thöni et al. (1992). Our data and data available in the literature clearly show that the SiO2 content of the VBQD and VSBG display some gaps, without a continuous trend from mafic to acidic types. Clusters occur around 52 wt%, 60 wt%, 67 wt% and 75 wt% (Fig. 9a and b). The implications of this pattern in terms of magma genesis will be discussed later. The Mg number is negatively correlated with the SiO2 abundance, with some samples displaying very high values (sample VBZ-16, diorite, Table 2).

The VBQD and VSBG calc-alkaline affinity is also evident from trace elements, with LILE enrichment (100 to 1000 times PM, Primitive Mantle, values, Fig. 9c, d, e) with respect to HFSE, and LREE enrichment with respect to HREE (Fig. 9f, g, h). The PM-normalized negative anomalies of Nb, Ta and Ti are coherent with a volcanic arc magmatism (Gill, 1981), with values between 1 and 10 for Ti, 5 -50 for Nb and 9–50 for Ta (Table 2). However, the negative spikes of Nb, Ta and Ti are not coupled with positive spikes of Pb, Sr and Rb, as commonly recognized in volcanic arc magmatism (Gill, 1981). Rb and Th are enriched in the PM-normalized diagram (Fig. 9c, d, e) with respect to Ba, as already observed in calc-alkaline magmatism developed in extensional settings (Broutin et al., 1994).

LREE are significantly fractionated, mainly in leucogranites (Fig. 9h), with La/Smn up to ca. 5 (Table 2). HREE fractionation is instead weak, with some samples that show a poorly outlined “spoon-like” pattern (Fig. 9f, g, h).

6 Geochronology

The intrusion age of the several bodies composing the Val Biandino Intrusive Suite has been dated by several authors from the’90 until recently. More recent results point to an intrusion age of 289.1 ± 4.5 Ma and 286.8 ± 4.9 Ma for the VBQD and the VSBG, respectively (U–Pb on zircon, Pohl et al., 2018). As the uncertainties associated with the available ages are quite large, we tried to obtain more intrusion ages for both the VBQD and the VSBG in area not sampled by previous works. Two samples, one granite from the VBQD (VBZ3) and one porphyric granite from the VSBG (VBZ43) have been selected.

The zircon grains of the VBQD sample (VBZ3) typically range in dimension from 50 up to 200 μm, with the largest ones displaying an elongated prismatic habit, whereas smaller crystals have usually sub-equant prismatic habit. Elongated crystals are richer in inclusions, mainly apatite and rutile, that occur less frequently in small zircons. The width-to-length ratios of larger crystals is 1:3 to 1:4, whereas smaller ones are typically 1:1.5 to 1:2.5. Cathodoluminescence imaging (Fig. 10) revealed an oscillatory zoning of magmatic origin (Corfu et al., 2003) for most of the crystals, with a few of them showing inherited cores and signs of resorption phenomena during the growth. The U content ranges from 208 to 1890 ppm, with the higher values recorded in the inner parts of the elongated crystals. The Th contents ranges from 88 to 990 ppm and it is generally positively correlated with the U content. The Th/U ratios are below 1 and above 0.1 for most of the analytical spots (all data are reported in Table 3). Single 206Pb/238U dates between 271.3 and 302.6 Ma (Table 3) have been measured in zircon grains of the VBZ3 granite, with 14/16 analytical spots that provided dates < 290 Ma. The calculated concordia date, disregarding dates with more that 5% of discordance, is 285.2 ± 1.9 Ma (Fig. 10a).

Fig. 10
figure 10

SHRIMP U–Pb zircon ages of two granites from the VBQD (a) and VSBG (b) units. c Analytical spots on inherited cores within zircon grains. All spot ages are 206Pb/238U ages. Analytical results are reported in Table 3

Table 3 SHRIMP U-Th-Pb isotopic data of zircons from the leucogranites of the Val Biandino Intrusive Suite

Zircon crystals from the VBZ43 sample appear colourless or with a pale pink shade, mainly inclusion-free, with the exception of rare micrometric apatites. Most zircon grains display a subequant prismatic habit with a width-to-length ratio of 1:2–1:2.5 ranging in dimension from 150 to 200 μm. Rare crystals with an elongated prismatic habit reach 300–350 μm, with a width-to-length ratio of about 1:5. In these crystal inclusions are more frequent, mainly consisting of apatite and rutile. Cathodoluminescence imaging revealed a typical oscillatory zoning of magmatic origin (Corfu et al., 2003) for most of the zircon, grains with more complex zoning displayed by a few subequant grains larger than 200 μm (Fig. 10). The U contents of measured spots range from 197 to 2028 ppm (Table 3), with higher values related to the outer rim of elongated crystals. Sevently percent of the measured spots have U-contents below 600 ppm. Th ranges 97–540 ppm, and it is positively correlated with U, except for the spots with U > 1500 ppm. The Th/U ratios are always below 1 and above 0.1. Zircons from the VSBG porphyric granite (sample VBZ43) gave single 206Pb/238U dates between 273.1 and 290.2 Ma (Table 3), with 16/20 analytical spot yielding dates younger than 285 Ma. The calculated concordia date is 283.2 ± 1.9 Ma (Fig. 10b).

7 Discussions

7.1 The Val Biandino Intrusive Suite and the Early Permian magmatism in the Southalpine Domain

In the Val Biandino Intrusive Suite, the main rock varieties are represented by metaluminous to peraluminous high-K calc-alkaline quartz-diorite, granodiorite and granite. Metaluminous high-K calc-alkaline gabbro-dioritic stocks also occur in the eastern area of the complex (Fig. 2). S-type cordierite leucogranite and peraluminous high-K calc-alkaline muscovite leucogranitegranite form small stocks and dikes intruding the mafic to intermediate rocks. North of the VBQD, the emplacement of the 283.2 ± 1.9 Ma (Fig. 10b) VSBG granite possibly marks the end of the intrusive magmatic activity in the area as the occurrence of other magmatic bodies has never been reported. This is hypothesis is also supported by the occurrence of Middle Permian inversion tectonics and the almost complete lacking of sediments and volcanic products of the same age in the central Southern Alps (Gaetani et al., 1986) The onset of subaerial volcanic activity in the Orobic Basin (central Southern Alps, Fig. 1) is constrained at ca. 280 Ma (Gretter et al., 2013; Berra et al., 2015) and likely ended at 270 ± 2 Ma (ignimbritic flows at the top of the Monte Cabianca Volcanite, U–Pb zircon age, Berra et al., 2015). In the study area the “Ponteranica conglomerates”, bearing clast of volcanic rocks, of the Pizzo del Diavolo Formation and the Monte Cabianca Volcanite rest below the Verrucano Lombardo (Fig. 2), but their field relationships with the VSBG are not observable.

The Permian magmatism occurs extensively in the western Southern Alps (WSA in Fig. 1). Mafic (gabbro) to intermediate (norite and diorite) rocks mainly occur in the Ivrea Verbano Zone (Lower and Upper Mafic Complex; e.g., Rivalenti 1975; Sinigoi et al., 1991; Peressini et al., 2007; Sinigoi et al., 2011), whereas eastwards granitic and hypabyssal granitic bodies (“Graniti dei Laghi” and Valganna granophyric granite) intrude the basement of the Serie dei Laghi (e.g., Bakos et al., 1990; Pinarelli & Boriani, 2007) (Fig. 1).

Comparing the mafic to intermediate rocks of the Ivrea Verbano Mafic Complex, at any given SiO2 content they are less enriched in K2O (Fig. 11) than the Val Biandino ones, though they have similar REE patterns (Fig. 12). Minor anatectic granitic bodies, showing similar geochemical features to the leucogranite of the Val Biandino Intrusive Suite, also occur (Fig. 11a, b). U–Pb zircon analyses have provided emplacement ages of ca. 285 Ma (Sinigoi et al., 2011) and 293 ± 6 Ma (Vavra et al., 1999) Ma for amphibole gabbros of the Lower Mafic Complex, of 287 ± 3 Ma for an interlayered norite (Quick et al., 2003; Peressini et al., 2007), of 286–289 Ma (Peressini et al., 2007) for a gabbro of the Upper Mafic Complex, and of 288 ± 3 (Quick et al., 2003) and 285 ± 7 Ma (Pin, 1986) for biotite-rich monzogabbros historically referred to as “diorites” in the literature (Rivalenti, 1975) (Fig. 13). To the east of the Ivrea Vebano Zone, large calc-alkaline granitic subvolcanic bodies, represented by the Graniti dei Laghi (e.g., Pinarelli & Boriani, 2007) and the Valganna granophyric granite (e.g., Bakos et al., 1990), are enriched in K2O, LREE and HREE contents and show flat REE patterns with a more pronounced Eu anomaly than the Val Biandino leucogranite (Fig. 12). The “Graniti dei Laghi” emplaced at 282 ± 1.5 Ma (Montorfano, U–Pb zircon ages, Schaltegger & Brack, 2007), 272–279 Ma (Baveno, Rb–Sr WR-Bt, Pinarelli et al., 2002), 289 ± 3.4 Ma (Roccapietra, U–Pb zircon age, Quick et al., 2009), 273–280 (Valle Mosso, U–Pb zircon ages, Quick et al., 2009) and 281–298 Ma (Valle Mosso, U–Pb zircon age, Klötzli et al., 2014). An emplacement age of 281.3 ± 0.5 Ma (U–Pb zircon age, Schaltegger & Brack, 2007) is instead reported for the Valganna granophyric granite (Fig. 13).

Fig. 11
figure 11

Whole rock data of the major intrusive bodies of Early Permian age in the western (WSA, a and d), central (CSA, b and e) and eastern southern Alps (ESA, c and f). All intrusive bodies are subalkaline with a calc-alkaline affinity. Rocks from the CSA and ESA, independently from their SiO2 content, are always high in K2O, whereas in the WSA more mafic terms are strictly calc-alkaline or even low-K calc-alkaline

Fig. 12
figure 12

Chondrite normalized (McDonough and Sun, 1995) REE pattern of Early Permian intrusive rocks from the western (WSA), central (CSA) and eastern Southern Alps (ESA). Intrusive rocks in the central Southern Alps are basically represented by the Val Biandino Intrusive Complex and the Val Navazze Complex: their REE patterns (in colour) are compared in the diagrams with both the WSA and ESA intrusives having the same SiO2 content

Fig. 13
figure 13

Compilation of available radiometric ages related to the Early Permian magmatic activity in the Southalpine domain. Age of volcanic rocks have been also included. IV-MC: Ivrea-Verbano Mafic Complex; SdL: “Serie dei Laghi” granodiorites and granites; Vg: Valganna granophyre; OB: Orobic Basin; CB: Collio Basin; TB: Tregiovo Basin; Intr.Comp.: intrusive complex in the ESA (Monte Croce, Ivigna and Bressanone); AVD: Athesian Volcanic District. Bars with dashed outline refers to low-retentivity isotopic systems (mainly Rb–Sr whole rock or mica ages); bars with continuous outline are U–Pb zircon ages. [1] Klötzli et al., 2014; [2] Sinigoi et al., 2011; [3] Peressini et al., 2007; [4] Quick et al., 2003; [5] Garuti et al., 2001; [6] Vavra et al., 1999; [7] Quick et al., 2009; [8] Schaltegger & Brack, 2007; [9] Pinarelli et al., 1988; [10] Berra et al., 2015; [11] Gretter et al., 2013; [12] Schaltegger & Brack, 1999; [13] De Capitani et al., 1994; [14] Borsi et al., 1966; [15] Macera et al., 1994; [16] Barth et al., 1993; [17] Morelli et al., 2012; [18] Marocchi et al., 2008; [19] Rottura et al., 1998; [20] Del Moro & Visonà, 1982; [21] Avanzini et al., 2010; [22] Bargossi et al., 2010; [23] Klötzli et al., 2003

A swarm of mafic to acidic dikes and small hornblende-diorite stocks (Appinites, Pinarelli & Boriani, 2007) emplaced between 312–278 Ma (U–Pb zircon age, Klötzli et al., 2014) near the contact of the Serie dei Laghi and the Ivrea Verbano Zone (Fig. 13). The bimodal volcanic activity in the western Southern Alps persisted from 290 to 282 Ma (U–Pb zircon age, Quick et al., 2009) in the Sesia Valley and from 298 to 288 Ma (U–Pb zircon ages, Schaltegger & Brack, 2007) in the Valganna district.

In the central Southern Alps, beside the Val Biandino intrusive Suite, small stocks of quartz-diorite, granodiorite to granite occur (De Capitani et al., 1994) (Val Navazze, Fig. 1). In this area, granodiorite and granite are enriched in K2O, as well as in LREE and HREE contents, with a more pronounced negative Eu anomalies relative to our samples (Val Navazze, Fig. 12g).

In the Collio basin, east of the Orobic basin (Fig. 1) the volcanic activity started at 283 ± 2 Ma (U–Pb zircon age, Schaltegger & Brack, 2007) with ignimbrites and ended at about 280.5 ± 4 Ma (U–Pb zircon age, Schaltegger & Brack, 2007) with the emplacement of the Auccia volcanite. This latter age is in agreement with those determined for a dacitic flow and a rhyodacitic ignimbrite at the top of the successions in Val Daone (277.9 ± 2 Ma, U–Pb zircon age, Gretter et al., 2013) and in Val Rendena (279.2 ± 2 Ma, U–Pb zircon age, Gretter et al., 2013), respectively (Figs. 113).

A large intermediate to acidic Permian magmatic activity took place also in the eastern Southern Alps (ESA in Fig. 1). The Permian intrusions are represented by the plutons of Cima d’Asta, Bressanone (Brixen), Ivigna (Ifinger), Monte Croce (Kreuzberg) and Monte Sabion (Fig. 1). These plutons mainly consist of high-K calc-alkaline granodiorites, monzogranites and S-type leucogranites, with minor tonalities (only at Cima d’Asta). Their emplacement ages have been constrained at 275.5 ± 3 Ma (U-Th allanite age, Barth et al., 1993) for a Cima d’Asta granodioritic sample; at 286.6 ± 2.9 and 293 ± 3 Ma (U–Pb zircon age, Bargossi et al. 2010; Morelli et al., 2012, respectively) for the Ivigna tonalities; at 285.4 and 284.3 ± 1.6 Ma (U–Pb zircon age, Marocchi et al., 2008) for the Monte Croce granodioritic and monzogranitic rocks, respectively. The emplacement age of the Mt. Sabion pluton (Figs. 1, 13) is constrained at 275 ± 9 Ma (Rb–Sr whole rock, Borsi et al., 1966). The Bressanone (Brixen) granodiorite and granite were emplaced at ca. 281 Ma (Rb–Sr whole rock, Del Moro & Visonà, 1982), with associated gabbroic bodies (the Lives, Luson and Chiusa gabbros) having the same age (281.8 ± 0.004, Boscaini et al., 2020). U–Pb zircon age of 277.9 ± 3.4 Ma have been obtained for the Canezza quartz-diorite-tonalite stock (Avanzini et al., 2010) (Fig. 13). Gabbros and diorites, showing medium- to high-K calc-alkaline affinity (Fig. 11), are reported only for the Bressanone (Brixen) and Cima d’Asta plutons. These rocks are locally associated with S-type cordierite granites. Del Moro & Visonà (1982) determined a Rb–Sr whole rock age of 286 ± 5 Ma for a cordierite granite within the Brixen pluton. At any given SiO2 content, all these rocks show chondritic-normalized REE patterns similar to those of the Val Biandino Intrusive Suite (Fig. 12).

The volcanism, represented by the Athesian Volcanic district, started during the Early Permian at 290.7 ± 3 Ma (U–Pb zircon age, Visonà et al., 2007) and ended at 274.1 ± 1.6 Ma (U–Pb zircon age, Marocchi et al., 2008).

7.2 Petrogenesis

The Val Biandino Intrusive Suite, with its mafic components showing a magnesian-potassic affinity and its felsic ones displaying both peraluminous and S-type character, is similar to numerous high-K calc-alkaline mafic–felsic igneous associations worldwide (e.g., Bonin, 2004; Barbarin et al., 2005). Experimental arguments, field evidence, petrographic and geochemical data for these coexisting mafic and felsic suites suggest that their genesis may involve a complex interplay of several petrogenetic processes, including contemporaneous melting of distinct sources and fractional crystallization. To unravel this complex petrogenetic scenario, we attempt here to constrain the source end-members of this mafic–felsic igneous suite.

7.2.1 Gabbro-diorite

In the VBQD, gabbro-dioritic types represent the less evolved rocks. These rocks have geochemical features similar to modern magmatic arc rocks worldwide (e.g., Stern, 2002). Selective enrichment of LILE, U, Pb and LREE, and depletion in HREE and HFSE (e.g., Nb, Ta and Ti) (Table 2), thus resulting in high values of La/Nb (ca. 3), Ba/Nb (43–53), Ba/La (> 15) and Zr/Nb ratios (10–22) and low Ce/Pb (3.6–8.9) (Table 2). The gabbro-diorite has an initial 87Sr/86Sr value of 0.70853 (Fig. 14, data from Thöni et al., 1992; De Capitani et al., 1994; all 87Sr/86Sr values reported hereafter for the Val Biandino Intrusive Suite have been recalculated at 290 Ma) higher than those of E-MORB and Hawaiian ocean island basalts (e.g., Saunders et al., 1980). Such geochemical features can be ascribed either to partial melting of a metasomatized mantle source enriched in LILE and LREE by slab-derived hydrous fluids or melts or to crustal contamination of a MORB-like magma during its emplacement.

Fig. 14
figure 14

87Sr/86Sri isotopic composition of the VBQD (Thöni et al., 1992) compared with the rocks (metasediments and amphibolites) of the Orobic Variscan basement (Boriani et al., 2003) plotted against SiO2 wt% (a) and 1000/Sr (b), respectively. The 87Sr/86Sr ratio has been recalculated at 290 Ma

Any mantle-derived magmas passing through a thick crustal section should interact with it, resulting in some degrees of contamination/assimilation (Hawkesworth et al., 1984; Mahoney, 1988; Carlson, 1991; Hergt et al., 1991). Actually, a simple process of binary mixing cannot explain the compositional features of the gabbro-diorite suite. The lack of a granite end-member associated with the gabbro-diorite makes the mixing hypothesis unfeasible for this suite. The gabbro-diorite have higher Sr abundance (224–244 ppm) and distinct Sr isotopic composition ((87Sr/86Sr) = 0.70853) with respect to those of the local upper crust host rocks (Sr = 84–180 ppm; (87Sr/86Sr) = 0.712–0.718) as represented by the meta-sediments of the Orobic basement (Fig. 14, Sr isotopic ratios recalculated at 290 Ma). This argues against such basement as the potential source of the enrichments observed in the gabbro-diorite suite.

If crustal assimilation for the gabbro-diorites can be excluded, source contamination must be considered as a main factor in their petrogenesis. In the Nb/Ba vs. Nb/Zr diagram (Fig. 15a; e.g., Hooper & Hawkesworth, 1993), gabbro-diorites plot close to the field for subcontinental lithospheric mantle and far away from oceanic island basalt and E-MORB sources. Zr/Ba (0.15–0.4) and Nb/La (< 0.5) values also support a subcontinental lithospheric mantle as source region for the gabbro-dioritic melt. For the transport of trace elements in the mantle wedge, the prevailing model is that different types of metasomatic agents (e.g., fluid, silicic or carbonatite melt) could yield distinctive trace elements and isotopic imprints in a metasomatized mantle (e.g., Menzies et al., 1987). For instance, slab-derived fluids usually carry significant amounts of Si and LILE in the mantle wedge at shallow depths (e.g., Plank and Langmuir, 1993; Stein et al., 1997), whereas slab melts may be the metasomatic agents at greater depths (Kilian and Stern, 2002; Rapp et al., 1999) and produce more sodic metasomatic products. Sun & Stern (2001) used Ba/La, Pb/Ce, Cs/Rb and U/Th ratios to identify fluid enriched sources. These ratios are > 1, which means that magmas derived from fluid-metasomatized sources are enriched in Ba, Pb, Cs and U relative to La, Ce, Rb and Th, respectively. Woodhead et al. (1998) suggested that if the mantle is enriched by melts from subducted sediments, it would be reflected in the Th/Yb ratio in mafic magmas. Consequently, these variables can be used as reliable indicators of potential sediment or fluid contributions from the subducted slab to magma source regions (Woodhead, et al., 2001). Based on mass balance considerations, Hawkesworth et al. (1995) suggested that the elevated Ba/La and La/Nb in mafic rocks from the Basin and Range province (USA) cannot realistically be attributed to crustal contamination of MORB-type magmas, and that such rocks are connected to metasomatized source regions in the subcontinental lithospheric mantle. In our case, high Ba content and Th/Yb, La/Nb, Ba/La, (Hf/Sm)PM and Cs/Rb and low (Ta/La)PM ratios point to a mantle source region enriched by both fluids and sediments from a subducting slab (Fig. 15b–d). The high Ba (ca. 400 ppm) content and Ba/Ce (ca. 10) ratios, along with the poorly fractionated REE patterns [(La/Yb)N = 5–6)] suggest a shallow H2O-metasomatized mantle (e.g. Hartmann & Wedepohl, 1993) containing hydrous phases like amphibole and phlogopite (as suggested by the high Ba contents; O’Brien et al., 1995) as the source for the gabbro-dioritic melt. H2O-dominated slab-derived fluids could explain mantle enrichment in mainly LILE (and lesser LREE; Fig. 9f–h), whereas high 87Sr/86Sr (0.70853) (Fig. 14) may indicate a significant contribution of continental terrigenous components in the source (Thirlwall et al., 1996), as also recorded in the Neogene calc-alkaline and K-rich volcanics occurring in the Betic Cordillera (87Sr/86Sr = 0.70874–0.72073; Zeck et al., 1998; Benito et al., 1999; Turner et al., 1999), in mantle xenoliths from the M. Vulture (S-Italy; e.g., Downes et al., 2002) and in the Finero peridotite (NW-Italy; e.g., Obermiller, 1994).

Fig. 15
figure 15

a The Nb content of the mafic rock types of the VBQD with respect to Zr and Ba (Nb/Ba vs. Nb/Zr diagram of Hooper & Hawkesworth, 1993). Data from this work and literature (De Capitani & Liborio, 1988) show how a sub-continental lithospheric mantle is favourable as a source for the VBQD magmatic rocks. b Th/Yb – Ba/La, c Th/Yb – Cs/Rb and d (Ta/La)PM – (Hf/Sm)PM diagrams (Hartmann & Wedepohl, 1993) that show the sediments and fluids possible contribution to the source of the Biandino gabbrodiorites. e A simple fractionation mechanism of magma derived from a fluid- and/or melt-metasomatized subcontinental lithospheric mantle is supported by the incompatible trace element ratios (Soesoo, 2000) and by the fractional crystallization tendency depicted in the La/Sm vs La diagram for gabbrodiorite of the VBQD. f Relationships between the composition of anatectic melt with respect of the source composition in terms of the Ca/Fe + Mg and Al/Fe + Mg molar ratios (Altherr et al., 2000). The composition of quartz-diorite and granodiorite of the VBQD suggest a contribution derived from partial melting of metapelites-metapsammites (micaschists and paragneisses) and metamorphic mafic rocks (amphibolites) for quartz-diorite

In summary, the available data are not compatible with a magma mixing at the level of emplacement as mechanism that yielded the chemical variations in the gabbro-dioritic suite. A simple fractionation mechanism of magma derived from a fluid- and/or melt- metasomatized subcontinental lithospheric mantle is supported by the incompatible trace element ratios (Soesoo, 2000) and by the fractional crystallization tendency depicted in the La/Sm vs La diagram (Fig. 15e). The absence of positive or negative correlation of La content with LREE fractionation (La/Sm) suggests a simple crystal fractionation process with highly incompatible elements (La) not enriched with respect to less incompatible ones (Sm). Other processes, like partial melting or different degree of crustal contamination, should have resulted in different patterns (Aldanmaz et al., 2000).

7.2.2 Quartz-diorite and granodiorite

The quartz-diorite and granodiorite have different trace element concentrations compared to the gabbro-diorites (Fig. 9), precluding a single partial melting/fractionation/assimilation model for these rocks. If this had happened close similarities in trace element abundances and patter should be expected at least for the gabbro-diorite and the quartz-diorite. In addition, the difference in Sr isotope compositions rules out a direct derivation of the quartz-diorites from the gabbro-diorites (Fig. 14) as they differ of ca. 0.02 in 87Sr/86Sr, recalculated at 290 Ma (Fig. 14).

The low concentrations of Ni and Cr (Table 2) argue against a pure mantle derivation of the quartz-diorite, even though similar low abundances have been reported in the least evolved members of some arc-type and calc-alkaline series (e.g., Perfit et al., 1980). Some trace element characteristics are typical for arc-like magmatism: enrichment of LILE (Ba, Rb, K), relative depletion of HFSE (e.g. Nb) and, consequently, high LILE/HFSE ratios (Fig. 9). Transfer of volatiles might cause enrichment in LILE from dehydrating subducted oceanic crust to the mantle wedge (e.g., Hawkesworth, 1979; Hawkesworth & Vollmer, 1979; Thirlwall & Graham 1984). In the absence of a subduction zone setting, enrichment of LILE may be caused by preferential tapping of enriched mantle sources (e.g., Saunders et al., 1980) or may be caused by AFC (assimilation-fractional crystallization) or MASH (melting–assimilation–storage–homogenization) processes (e.g., Hildreth & Moorbath 1988). The less evolved quartz-diorite samples have high 87Sr/86Sr values (0.7104; Thöni et al., 1992; Fig. 14). While large-scale crustal contamination of magma derived from upper mantle sources could explain the Sr isotopic composition, such contamination would likely be accompanied by strong fractional crystallization, which is not observed (Fig. 9), as granodiorite display almost the same trace element abundances of more mafic terms. These features likely exclude a significant involvement of mantle-derived melts in the genesis of the quartz-dioritic—granodioritic suite.

Compositional differences of granitic liquids produced by partial melting of various sources under variable melting conditions can be distinguished in terms of molar CaO/(MgO + FeO*) vs. molar Al2O3/(MgO + FeO*) (Fig. 15f; Altherr et al., 2000). The dominant mafic composition of the quartz-diorites precludes partial melting of metasedimentary or felsic meta-igneous rocks, with partial melting of mafic to intermediate sources that seems necessary in order to explain the major and trace element contents. Experimental evidence (Beard & Lofgren, 1989, 1991; Rapp et al., 1991; Rushmer, 1991; Wolf & Wyllie, 1994; Rapp & Watson, 1995; Sisson et al., 2005) suggests that dehydrating basaltic material within the mid- to lower crust can produce significant volumes of mafic to intermediate partial melts, particularly in regions characterized by high heat flow. These studies have shown that melting conditions (water-saturated vs. dehydration melting, water fugacity etc.) strongly influence the composition of the melt and of the residuum during partial melting of rocks of basaltic compositions at mid- to lower crustal pressures (e.g., Beard and Lofgren 1989, 1991; Wolf and Wyllie 1989; Sisson et al., 2005). Depending on bulk composition, fluid-absent melting of amphibolite yields 10 – 60% of melt at temperatures of 900–1000 °C, whereas fluid-present melting yields similar amounts of melt at lower temperatures between 850 and 900 °C (e.g., Beard & Lofgren 1991; Rushmer 1991; Wolf & Wyllie 1994; Sisson et al., 2005). However, liquids produced experimentally by fluid-absent melting are quartz-dioritic to tonalitic in composition and coexist with a residuum dominated by plagioclase, pyroxene and Fe-Ti oxides (Beard & Lofgren 1991; Rushmer 1991; Wolf & Wyllie 1994), whereas melts generated in the presence of an H2O-rich fluid phase are “granitic” in composition (rich in SiO2 and Al2O3 and low in Fe2O3 and MgO) and coexist with a residuum of amphibole, clinopyroxene, Fe-Ti oxides and minor plagioclase (Beard & Lofgren 1991).

The most primitive quartz-diorite (SiO2 = 58 wt%) of the Val Biandino Intrusive Suite has chemical features recalling melts derived from a mafic source: moderately high Al2O3 (17.5 wt%), MgO (3.7 wt%), FeO* (6.3 wt%), K2O (2.2 wt%) and Na2O (2.94 wt%) contents and Na2O/K2O (> 1.3) ratio, high LREE (118 ppm) and moderately low HREE (15 ppm) concentrations, and no Eu anomaly (Eu/Eu* ca. 1) (Fig. 9; Table 2). Moreover, it does not show significant HREE depletion (i.e. YbN < 10; Wareham et al., 1997) predicted for melts that equilibrated with residual garnet, suggesting that it formed alternatively under lower pressures. Also, its REE pattern is akin to models involving partial melting of an amphibolitic source (Arth & Barker, 1976; Arth et al., 1978; Sengupta et al., 1983). The patterns show in Fig. 9 with the decreasing of ΣREE and increasing of positive anomaly are very similar to ones expected if hornblende fractionation was operating to produce the melt.

Since melts derived from fluid-absent experiments on partial melting of a mafic source (e.g. Rapp et al., 1991) are generally too sodic in composition to justify the formation of medium- to high-K granitoids, Sisson et al. (2005) suggested that metaluminous to weakly peraluminous K2O-enriched granitoids may derive from fluid-absent partial melting of medium- to high-K Hbl-gabbros, basalts and basaltic andesites (Fig. 16). Although experimental melts obtained at 925–1000 °C and ca. 7 kbar (Sisson et al., 2005) have produced melts having chemical composition comparable to those of the quartz-diorites and granodiorites in terms of Al, Ca and K contents (Fig. 16), some chemical signatures argue against this interpretation. Partial melting of these source rocks produces melts that have higher Fe and Na and lower Mg and Ti concentrations than the quartz-dioritic–granodioritic suite (Fig. 16). Moreover, partial melting of Hbl-gabbro, basalt and basaltic andesites produces melts that are only slightly peraluminous, while some of the samples of the quartz-diorite–granodioritic suite of the Val Biandino Intrusive Suite display higher ASI (Fig. 16a). In addition, the 87Sr/86Sr isotopic composition (0.7102–0.7105; Fig. 14) of the quartz-dioritic–granodioritic suite has a strong crustal signature excluding their derivation from a typical basaltic source.

Fig. 16
figure 16

Composition of the VBQD high-K calc-alkaline rocks compared to the compositions of melts experimentally generated by partial melting of high-K Hbl-gabbros, basalts and basaltic andesites (Sisson et al., 2005)

The K/Na ratios (0.5–0.7) that characterize the quartz-dioritic–granodioritic suite indicate that biotite could be a reactant phase in the melting reaction, thus suggesting that the source underwent concomitant biotite and amphibole incongruent melting. Fluid-absent melting of these two phases produces both peritectic clinopyroxene and orthopyroxene, with the two pyroxenes that, at low pressures, may also be intergrown (Skjerlie & Johnston, 1996). The positive linear correlation for Ti and Ca plotted against Fe + Mg (mol) (Fig. 16e, f) together with the negative correlation of ASI (Fig. 16a) suggests the entrainment in the source of a peritectic assemblage dominated by a Ca–Fe–Mg–Ti bearing phase or assemblage of phases. The entrainment of both peritectic clinopyroxene and ilmenite phases (e.g., Vielzeuf & Montel, 1994) in the magma can generate the chemical variability exhibited by the quartz-dioritic–granodioritic suite. The entrainment of clinopyroxene can be responsible for the positive correlation of Ca and Mg and the negative one of ASI with Fe + Mg, whereas the entrainment of ilmenite can explain the increase in Ti (Fig. 16e). The entrainment of peritectic ilmenite is also supported by the positive correlation that exists between TiO2 and V for the rocks of the quartz-dioritic–granodioritic suite (Table 2). In the quartz-diorite mineral phases that may represent remnants of a peritectic assemblage could be the aggregates of amphiboles that in many calc-alkaline granitic rocks have been described (e.g. Castro and Stephens, 1992; Sial et al., 1998) as fragments of “restites” modified in response to the increase of water activity in the melt during crystallization (Stephens, 2001).

The formation of a peraluminous melt in equilibrium with a clinopyroxene-dominated peritectic assemblage provides constraints on the nature of the source and argues for concurrent biotite and amphibole fluid-absent melting in the genesis of the quartz-dioritic–granodioritic suite. Few experiments have been performed on source rocks containing multiple hydrous phases (Rutter & Wyllie, 1988; Skjerlie & Johnston, 1992, 1996; Skjerlie & Patiño Douce, 1995). Granitic melt and clinopyroxene-dominated peritectic assemblage are produced by partial melting of: (i) immature volcanogenic sediments of andesitic-dacitic composition (e.g., Skjerlie & Johnston, 1996) or potassic arc andesites (Roberts and Clemens, 1993); (ii) interlayered garnet-bearing amphibolites and sillimanite-bearing metapelite (Skjerlie & Patiño Douce, 1995). These sources are able to produce 20–60 vol.% of melt at temperatures of ca. 900 °C while, according to Thompson (2001), metavolcanics containing low-Ca amphiboles and biotite can undergo substantial fluid-absent melting at lower temperatures (T ca. 825 °C). It is important to emphasize that rocks similar to those used as starting material in the experiments of Skjerlie & Patiño Douce (1995) occur in the Paleozoic Orobic basement (Boriani et al., 2012), which is made up of metapelites and metapsammites (“Orobic metasediments” in Fig. 14) interlayered with minor quartzites and lenses of amphibolites (“Orobic amphibolites” in Fig. 14). These latter mainly consist of plagioclase, hornblende and biotite, and their protholiths is interpreted as a basaltic meta-tuffite (Boriani et al., 2003). Finally, the chemical differences between the quartz-diorite and granodiorite may be the effect of temperature increase in a biotite and hornblende-bearing source. However, the origin of granodiorite from partial melting of a metasedimentary source unrelated to the source of the quartz-diorite can not be ruled out.

7.2.3 Leucogranite

Peraluminous granites and leucogranite occurs in the studied area as small stocks and dikes in the VBQD and form almost the entire body of the VSBG. Granites in the VBQD always contain muscovite together with cordierite in most of them. As determined by U–Pb zircon dating (Fig. 10), granites in the two complexes are almost coeval considering uncertainties.

Aluminous silicate minerals, such as cordierite, and muscovite are characteristic minerals of S-type granitoids (e.g., Chappell & White, 2001) suggesting that the granites of the Val Biandino Intrusive Suite most likely have a metasedimentary source (Miller, 1985), also supported by their enrichment of Ba and Rb relative to Sr, high K2O (> 4.0 wt%, Fig. 9) and moderately low CaO (< 2.0 wt%, Table 2) contents, high initial 87Sr/86Sr (0.7145–0.7161; Thöni et al., 1992; Fig. 14) and strongly peraluminous character (ASI > 1.2; Table 2). Cordierite is a low-pressure mineral that appears above the granite solidus in many peraluminous felsic magmas (Clarke, 1995), whose occurrence is typical of low-pressure, high-temperature terranes that underwent extensive crustal melting either in the late thermal stages of continental collision or in the high thermal regime during crustal thinning.

Felsic peraluminous melts, such as the cordierite granite, likely form in response of partial melting of metapelites and metapsammites under fluid-absent (Tompson, 1982; Miller, 1985; Breton & Thompson, 1988; Vielzeuf & Holloway, 1988; Vielzeuf et al., 1990; Vielzeuf & Montel, 1994; Patiño Douce & Harris, 1998) or even dry conditions (e.g., White & Chappell, 1977; Miller, 1985; Clemens & Wall, 1981; Turpin et al., 1990; Holtz & Johannes, 1991). The critical parameters controlling formation of cordierite in magmas are relatively low pressure, and high (Mg + Fe2+), Mg/Fe2+, ASI, Al2O3 activity and fO2 (Clarke, 1995). Two important fluid-absent melting reactions relevant for the formation of S-type granites at high temperature (800–900 °C) are (Breton & Thompson 1988; Vielzeuf & Holloway 1988; Vielzeuf & Montel 1994):

  1. (i)

    Bt + Sil + Qz + Pg = melt  + Crd + Grt ± Kfs;

  2. (ii)

    Bt + Qz + Pg = me lt + Opx ± Grt ± Kfs

At high temperatures (> 800 °C), considerably high degrees of melting (20–40%) are normally associated with these biotite breakdown reactions (Clemens & Vielzeuf 1987; Stevens et al., 1997). Since muscovite is expected to have completely disappeared at lower temperatures during anatexis of metasedimentary rocks, the concentrations of Sr, Ba and Rb are related to the amounts of feldspar and biotite remaining in the residue. The composition of the restite depends, among other factors, on the presence of H2O-rich fluids during melting: the more H2O-rich fluid is available, the larger is the proportion of melt that can be formed, and the smaller the proportion of feldspar that remains in the restite (e.g., Harris & Inger, 1992.; Harris et al., 1995; Inger & Harris, 1993). The cordierite granites have Sr = 82–229 ppm, Ba = 405–574 ppm and Rb = 161–237 ppm (Table 2), resulting in low Rb/Sr (0.7–3.0), Rb/Ba (0.3–0.6) and Sr/Ba (0.2–0.3) ratios that recall a (meta)psammitic source rock (Fig. 17c). This distinction probably reflects the higher Rb/Sr and Rb/Ba ratios of clay-rich sources as compared to clay-poor ones. For instance, the average shale has higher Rb/Sr and Rb/Ba than does the average greywacke (Condie, 1993). The wide ranges of Rb and Sr concentrations in our cordierite granites suggest variable proportions of feldspar and biotite remaining in the source during anatexis. Sylvester (1998) suggested that CaO/Na2O ratio is mainly controlled by the abundance of plagioclase relative to clay within the sedimentary sources. Strongly peraluminous granitic melts produced from plagioclase-rich and clay-poor sources will tend to have higher CaO/Na2O (> 0.3) ratios than melts derived from sources that are plagioclase-poor and clay-rich (Skjerlie & Johnston, 1996). The CaO/Na2O ratios of the cordierite granites are 0.3 – 0.6 (Fig. 17a), in further agreement with the conclusion obtained from the Rb/Sr and Rb/Ba ratios. Plots of Ba versus Pb concentrations in S-type granitic rocks have been used to distinguish between those magmas produced by relatively low-temperature partial melting involving muscovite breakdown and those formed by higher-temperature reactions involving biotite breakdown (Finger & Schiller, 2012). A Pb–Ba diagram (Fig. 17b) for the cordierite granite shows that, as expected, all samples plot in the high-temperature field, also in agreement with the low Al2O3/TiO2 ratios (< 100) (Sylvester, 1998). The cordierite bearing granites of the Val Biandino Intrusive Suite fall within the same Pb–Ba field of Variscan S-type granites, interpreted to had formed by high-T crustal anatexis (Finger & Schiller, 2012).

Fig. 17
figure 17

Composition of the cordierite granites of the VBQD and of granites of the VSBG compared with the compositions of possible rock sources. Most of the rock samples display LILE abundance and ratios (Rb/Ba, Rb/Sr, Ba). Also in terms of major elements (K, and and Ca, d), the VBQD and VSBG granites suggest the contribution of partial melting of metapelites and metapsammites during their genesis. Data of the Variscan S-type granites are from Finger & Schiller (2012)

Moreover, the displayed horizontal data spread, can be interpreted as a source-controlled pattern (Finger & Schiller, 2012). The preservation of source-controlled chemical variations implies that the magmatic system was not or only poorly homogenized during its magmatic evolution.

The high-T and H2O-undersaturated S-type magma that of the cordierite granites could be derived through fluid-absent partial melting (mainly biotite breakdown) of a metapsammite. A possible metasedimentary source could be represented by low- to medium- grade paragneisses and micaschists of the Orobic basement, which are commonly characterized by 55–68 wt% SiO2, 15–24 wt% Al2O3, 2–4 wt% MgO, 3–5 wt% K2O and 2–4 wt% Na2O (Boriani et al., 2003). Moreover, their initial Sr isotope ratios recalculated at 290 Ma (0.71155 – 0.7155, Fig. 14; Boriani et al., 2003) are similar to those of the cordierite granite (0.7145–0.7161; Fig. 14; Thöni et al., 1992). This interpretation is also supported by the Rb, Ba and Sr contents. Because Sr and Ba are compatible in plagioclase, whereas Rb is incompatible (e.g. Harris & Inger, 1992), cordierite granites should have higher Rb/Sr (0.7–3) and Rb/Ba (0.3–0.6) than their metapsammitic source (0.3–1.2. and 0.1–0.3, respectively; Boriani et al., 2003). These geochemical features, and the large number of schists enclaves found in the cordierite granites (Fig. 6b) are consistent with an important contribution from a relatively mature metasedimentary source similar to the low- to medium-grade paragneisses and micaschists of the Orobic basement.

Some samples display a positive Eu anomaly that can be explained by fluid-present incongruent melting of muscovite from a metaspammitic source (Fig. 17), in which the depletion of feldspar in the restite would lead to positive Eu anomaly and Sr enrichment in the melt under disequilibrium conditions (Harris & Inger, 1992).

The VSBG granite has major element composition similar to those defined by some experimental melts on metasedimentary sources (Thompson, 1982; Miller, 1985; Breton & Thompson, 1988; Vielzeuf & Holloway, 1988; Vielzeuf et al., 1990; Vielzeuf & Montel, 1994; Patiño Douce & Harris, 1998) (Fig. 17d). The low Rb/Sr, Rb/Ba and high CaO/Na2O ratios suggest for the VSBG granite a melt derived from a metapsammitic source (Fig. 17a), as for the VBQD cordierite granite. However, some differences can be observed. The VSBG granite show lower MgO, Al2O3, Na2O, and Sr and higher K2O, FeO*, HREE, and Y values than the cordierite granite (Fig. 9). The higher HREE and Y (without modifying LREE, Fig. 9) contents could indicate the entrainment of peritectic garnet, as also supported by lower Al2O3 and Mg# values with respect to the cordierite granite. Therefore, a different magma source could be envisaged for the VSBG granite. We can speculate a low- Mg# metaspammitic source characterized by a more Ca-rich plagioclase and in which less plagioclase undergone melting.

7.3 The Val Biandino Intrusive Suite in the frame of the post-Variscan geodynamic scenario

The Early Permian phase of crustal extension and lithospheric thinning was accompanied by the emplacement of magmatic bodies at different crustal levels and superficial volcanism that lasted ca. 20 Ma, from 295 to 275 Ma (Fig. 13), whit a climax around 280–285 Ma (Schaltegger & Brack, 2007). The extension of the early Permian magmatic belt in Southern Europe could be estimated of ca. 500 km, from Sardinia (Traversa et al., 2003; Renna et al., 2006, 2007; Tribuzio et al., 2009; Boscaini et al., 2020) to the Eastern Alps (Del Moro & Visonà, 1982; Rottura et al., 1998; Marocchi et al., 2008 Boscaini et al., 2020). This belt extended in the southern foreland of the former Variscan orogen, with most of the early Permian magmatic rocks that are now day exposed in the Southalpine, Austroalpine and Penninic domains of the Alps. Despite the short time interval of magmatism duration, significant differences in the geochemical signature occur, mainly expressed in the most basic terms. Gabbros, gabbronorites, troctolites and norites with tholeiitic affinity occur in the Ivrea-Verbano mafic–ultramafic complex (Voshage et al., 1990; Sinigoi et al., 1994; Sinigoi et al., 2016) and in the Central Alps (Braccia-Fedoz gabbro and Sondalo gabbro, Hermann et al., 2001; Tribuzio et al., 1999; Petri et al., 2017), whereas other mafic complexes like whose of the Corsica-Sardinian batholith (Renna et al., 2006, 2007; Tribuzio et al., 2009; Boscaini et al., 2020), the Mt. Collon (Monjoie et al., 2005, 2007) and the minor mafic bodies associated to the Brixen granodiorite (Boscaini et al., 2020) display a subduction-related geochemical signature.

The almost contemporaneous emplacement of these geochemically different mafic magmatic complexes could be interpreted as a response of partial melting processes that affected at different depths an heterogeneous mantle source, composed by depleted or slightly-enriched regions (e.g. Tribuzio et al., 1999) and enriched ones. The subduction-related metasomatism of the enriched mantle source could be related to the effect of the Variscan subduction. However, the occurrence of subduction-related magmatism could not be taken as an evidence of the presence of a slab beneath the early Permian magmatic belt, as partial melting of the metasomatized mantle could have been triggered in response of astenospheric upwelling (Bonin et al., 1998) in the frame of the tectonic extensional regime, several Myrs or decades of Mrys after subduction has ceased. The calk-alkaline geochemical signature of the most mafic terms of the Val Biandino Intrusive Suite could thus be explained by partial melting of an already subduction-modified mantle source, likely during the Variscan orogeny, with produced melts that were also later partially contaminated by a crustal component.

In the intermediate and differentiated rock types of the Val Biandino Intrusive Suite, the crustal component becomes predominant. The occurrence of inherited cores within zircon grains of the leucogranites testifies for this process. 206Pb/238U ages obtained from zircon inherited cores (Fig. 10 and Table 3) are within the 540–520 Ma and 360–300 Ma time intervals, similar to ages obtained from detrital zircon grains from the Orobic basement (Siegesmund et al., 2023).

This supports the occurrence of crustal anatexis promoted by high heat advection due to asthenosphere upwelling and crystallization of mafic bodies in the lower crust or at the mantle/crust boundary.

8 Conclusions

The Val Biandino Intrusive Suite consists of two distinct units: the Val Biandino Quartz-Diorite (VBQD) and the Valle di San Biagio Granite (VSBG). The VBQD is a heterogeneous magmatic unit, whereas the VBSG consists of a single magmatic body of porphyric granite. Textural features of the VSBG granite suggest shallower depth of intrusion with respect to the VBQD. The two complexes were later juxtaposed by the NW dipping Sassi Rossi Normal Fault (SRNF, Fig. 2) of early to middle Permian age.

SHRIMP U–Pb zircon intrusion ages obtained for the most differentiated terms of the VBQD (285.2 ± 1.9 Ma) and for the VSBG (283.2 ± 1.9 Ma) point to slightly younger ages than previously envisaged. Available data suggest instead that the intermediate terms of the VBQD intruded earlier (289.1 ± 4.5 Ma; Pohl et al., 2018), as also testified by the crosscutting relationships observed in the field.

Despite its relatively small size, the Val Biandino Intrusive Suite displays a marked heterogeneity in terms of rock varieties that intruded in a short time interval (ca. 5 Myrs). Geochemical data suggest that gabbro-diorite, quartz-diorite, granodiorite and leucogranite are not co-magmatic. All the rock types of the Val Biandino Intrusive Suite were likely generated with the significant contribution from melting and/or assimilation of the heterogeneous continental crust forming the pre-Permian Orobic metamorphic basement, with the most mafic terms that preserve evidence of partial melting of a subduction-modified mantle source.

The petrogenetic scenario in which the Val Biandino Intrusive Suite formed could be framed in the early Permian crustal extension than led to heat advection from the asthenosphere. Upper mantle partial melting resulted in the underplating of mafic complex at the lower crust–mantle boundary (e.g. Ivrea-Verbano Zone, Sesia–Lanzo, Valmalenco area) that in turn promoted anatexis also in the lower and middle crust, as the crustal signature occurring in several intermediate to acidic intrusive bodies of early Permian age in several domains of the Alps demonstrates.