Submagmatic to Solid-State Deformation Microstructures Recorded in Cooling Granitoids during Exhumation of Late-Variscan Crust in North-Eastern Sicily

Abstract

Late-Variscan granitoid rocks of trondhjemitic and granitic composition, intruded in migmatitic paragneisses in the north-eastern Peloritani Mountains (southern Italy) at ~310 Ma and ~300 Ma, respectively, exhibit a range of deformation microstructures developed under a shear regime at decreasing temperatures. Non-coaxial deformation is documented by sigmoidal feldspar porphyroclasts, mica fish, and asymmetric boudins affecting tiny andalusite crystals. Late-Variscan shearing during granitoid cooling is constrained by largely represented chessboard patterns in quartz and, especially, submagmatic fractures in plagioclase, indicating deformation at high-temperature conditions (T > 650 °C), in the presence of melt. Submagmatic deformation was extensively superseded by deformation at lower temperatures. Examples of solid state-high temperature deformation-related microstructures (T > 450 °C) include feldspar bulging, quartz grain boundary migration, and subgrain rotation recrystallization. Finally, low temperature subsolidus microstructures (T < 450 °C) consist of quartz bulging, mica kinks, and feldspar twinning and bending. A complete sequence of deformation, operating from submagmatic to low-temperature subsolidus conditions is recorded in both the older and younger granitoids, suggesting a duration of ~20 Ma for shear zone activity during post-collisional exhumation of the Variscan middle crust in southernmost Italy.

1. Introduction

Shear zones (e.g., [1] and references therein) play a fundamental role in shaping the structure and composition of the continental crust, in a number of different ways. Among these, they can control the construction of orogenic edifices [1,2,3,4,5], the extraction, ascent and emplacement of granitoid magmas [1,6,7,8,9,10,11,12,13,14,15], the exhumation of deep crustal rocks [16,17], and the circulation of fluids able to further influence the rheology of the crust [18,19], as well as the mineral assemblage stability [19,20,21,22].

The study of shear zones can therefore provide crucial information on the evolution of the mountain belts. Nevertheless, the role played by the shear zones in controlling the building and exhumation of the orogenic crust may be particularly complicate to assess in poly-orogenic terranes, especially considering that older tectonic structures can be reactivated during subsequent orogenic cycles (e.g., [23,24,25]).

Well constrained Alpine shear zones are long known in the Aspromonte Massif and Peloritani Mountains from the southern Calabria-Peloritani Orogen (CPO; e.g., [1,2,26,27,28,29,30,31,32,33,34,35]). Here, Alpine tectonics locally reworked Mesozoic sedimentary rocks, as well as portions of Variscan and older basements, hampering the reconstruction of the pre-Alpine tectono-metamorphic evolution. In more detail, early- to late-Alpine shear zones were associated first to subduction of basement rocks and related sedimentary covers to deep crustal levels [1,26,27], and then to their exhumation along detachment surfaces, causing the present-day nappe stack structure of the Peloritani Mountains and Aspromonte Massif [1,26,27]. On the other hand, despite late-Variscan shear zones controlled the emplacement of the granitoids from the Serre and Sila batholiths, in central and northern Calabria [15,36,37,38,39,40,41,42,43,44], possible Variscan shear zone occurrence in the southernmost CPO has only been suggested by [45,46]. More recently, [47] highlighted the likely role of late-Variscan shearing in assisting the production of metasomatic trondhjemitic rocks by infiltration metasomatism affecting late-Variscan leucogranites, in the north-eastern Peloritani Mountains.

In this contribution, we describe and discuss a variety of shear-related microstructures developed from submagmatic to low-temperature solid-state conditions, in trondhjemites and leucogranites emplaced during two separate stages of late-Variscan magmatism, at ~310 Ma and ~300 Ma, respectively (e.g., [48,49,50]), to shed light on the deformation mechanisms operating during long-lived shear zone activity associated with post-collisional exhumation.

Although the present-day features of the basement rocks of the Peloritani Mountains are the result of a complex poly-orogenic history involving Cadomian, Variscan, and Alpine tectonic events (e.g., [51] and references therein), we use microstructures, in association with existing geochronological constraints [48,49,50], to provide the first unequivocal evidence for the occurrence of a late-Variscan shearing event in the Peloritani Mountains. Furthermore, the documented complete sequence of deformation, operating from submagmatic to low-temperature subsolidus conditions in both the older and younger granitoids, suggests a duration of ~20 Ma for shear zone activity during post-collisional exhumation of the Variscan middle crust in southernmost Italy. Finally, our results support the location of the Peloritani basement within a peripheral branch of the East Variscan Shear Zone (EVSZ), as previously proposed by [52,53,54,55].

2. Geo-Petrological Background

The Peloritani Mountains, in NE Sicily (Figure 1), make up the southern termination of the Calabria-Peloritani Orogen (CPO), a poly-orogenic basement complex comprising remnants of Variscan and Cadomian mountain belts, incorporated into the Alpine-Apenninic orogenic system ([51]; and references therein). The Peloritani Belt comprises a stack of south-verging Alpine nappes, consisting mostly of Variscan basement rocks with remnants of Meso-Cenozoic covers. The belt is subdivided into two complexes with different tectono-metamorphic histories. In the southern Peloritani, the lower complex consists of sub-greenschist facies Cambrian to Carboniferous volcano-sedimentary sequences with unmetamorphosed Mesozoic sedimentary covers [56,57,58]. The upper complex consists of two tectonic units: (1) the Mandanici Unit, comprising dominant greenschist to lower-amphibolite facies phyllites, with interbedded levels of metabasites and subordinate marbles and calc-schists [26,46,59]; (2) the overlying Aspromonte Unit, exposed also in the adjacent Aspromonte Massif in southern Calabria, which consists of amphibolite-facies paragneisses, augen gneisses and migmatites, with minor marbles and amphibolites, diffusively intruded by late-Variscan granitoid plutons [47,60,61,62,63,64,65,66]. Both granitoids and host rocks are locally affected by an Alpine overprint, at 28–22 Ma [67,68] and P-T (pressure-temperature) peak conditions of ~0.8 GPa at ~600 °C in the southern sector of Peloritani Mountains [26,27,29] that produced pseudotachylytes and cataclastic to mylonitic rocks; furthermore, a weakly metamorphosed Mesozoic-Cenozoic sedimentary cover sandwiched between the Mandanici and Aspromonte units locally occurs ([26]; and references therein).

P-T conditions for the Variscan metamorphism, producing the main field foliation, are in the range of ~0.5 GPa at ~550–680 °C for the Aspromonte Unit in the Peloritani Mountains [69], similar to those obtained for rocks of the same unit in the Aspromonte Massif (0.4–0.5 GPa at 650–675 °C; [70,71]). Relatively high P peak values of 0.8 GPa at ~600 °C ~ and of ~0.9 GPa at ~530 °C were obtained by [46,72], respectively, through thermodynamic modelling of garnet phyllites and schists, from both the Aspromonte and the underlying Mandanici Unit.

In particular, three blasto-deformational events have been detected in upper greenschist-facies metapelites from the north-eastern Peloritani by [46]: D₁, producing an isoclinal folding schistosity (S₁), developed at P ~0.9 GPa and T ~530 °C and only preserved as microfold hinge relics; D₂, a non-coaxial deformational event, associated to metamorphic peak conditions of P ~0.8 GPa and T ~540 °C and generating the field foliation (S₂), that consists of a pervasive crenulation schistosity; D₃, a retrograde non-coaxial shearing event, developed at 0.60–0.30 GPa and 460–420 °C, and producing a schistosity surface well visible only in the porphyroclast pressure shadows, or between fragments of pinch-and-swell garnet porphyroclasts. According to [46], the reconstructed P-T path depicts a typical clockwise P-T evolution, with a prograde stage involving burial at middle-lower crustal levels during collision-related thickening, followed by a retrograde evolution during a post-collisional shearing stage, likely related to late-Variscan exhumation under an extensional regime.

A similar P-T evolution has been outlined by [72] for an amphibolite-facies garnet schist, with a baric peak at P = 0.72–0.82 GPa and T = 565–585 °C (D₁), followed by decompression for which, however, only relatively high pressure of P = 0.46–0.70 GPa at T = 500–600 °C, reflecting the early stages of late-Variscan exhumation (D₂), have been constrained by the authors.

In a more general framework, the whole CPO is considered one of the peri-Gondwanan crustal blocks of the former Galatian superterrane [73] that, with neighbouring areas such as Sardinia, Corsica, West Carpathians and parts of the Alps, belonged to the future south Variscan terranes. These basement areas were located at the northern Paleotethys margin, while terranes such as Saxothuringia, Bohemia and Armorica faced to the north the Rheic–Rhenohercynian Ocean (e.g., [74,75]). The sequential closure of the two oceans led to the final collision and amalgamation of Gondwana, peri-Gondwanan terranes and Laurussia, finally resulting in the formation of Pangea at ca. 300 Ma. Extensive crustal anatexis and granitoid magmatism in the south Variscan terranes, at c. 320–280 Ma, is considered to have been triggered by a change from a compressional to a transtensional/transpressional regime, locally associated with significant crustal thinning and asthenosphere upwelling, producing a complex pattern of strike-slip shear zones, collectively known as East Variscan Shear Zone (EVSZ; e.g., [52,53,55,74]). According to reconstructions by [54,55], the CPO was also possibly part of EVSZ.

The latest stages of the Variscan Orogeny were indeed marked also in the CPO by widespread crustal melting and voluminous granitoid magmatism [15,36,38,47,49,65,76,77]. The late-Variscan magmatism gave rise to the composite Serre and Sila batholiths in central and northern Calabria [15,26,38,48,76,77,78], as well as to isolated anatectic plutons of weakly to strongly peraluminous trondhjemites and strongly peraluminous leucogranodiorite-leucogranites, in southern Calabria and north-eastern Sicily (Aspromonte Unit) [36,47,49].

Unlike the batholiths of the Sila and Serre Massifs, which are 10–13 km thick and are emplaced into high- to low-grade metamorphic rocks, the late-Variscan granitoids of the Aspromonte Unit occur as isolated plutons of a few km², typically emplaced into migmatitic paragneisses [49,61,63,64,79].

Available geochronological constraints indicate that the late-Variscan granitoid magmatism in the southern CPO started with the emplacement of the trondhjemite plutons at ~310 Ma, and was followed by the intrusion of the strongly peraluminous granitoid plutons at ~300 Ma (monazite U-Pb ID-TIMS, [48]; zircon U-Pb SHRIMP, [49,50]). The largest volumes of granitoid magmas were produced afterward more to the north, forming the Serre Batholith, in the 297.3 ± 3.1 Ma to 292.2 ± 2.6 Ma time span (zircon U-Pb SHRIMP, [15] and references therein). Late-Variscan magmatism in the area ended with an extensive subvolcanic activity producing swarms of medium- to high-K calcalkaline dacitic-rhyodacitic to andesitic dykes [44,80,81,82], which intruded both the batholith and the upper crustal host rocks. These products represent the transition from a late-orogenic to a post-orogenic setting, documented by sodic-alkaline and tholeiitic Triassic basalts from northern Calabria [83] and central-western Sicily [84,85], associated with the early breakup of Pangea.

The late-Variscan strongly peraluminous leucogranodiorites and leucogranites from the Aspromonte Unit form plutons up to ~40 km² in size, typically displaying sharp discordant contacts with the upper-amphibolite facies host rocks. The rocks contain variable amounts of Al-silicates, cordierite and garnet, in addition to abundant muscovite and inherited zircon with a wide spread of ages [64], and are mostly considered as S-type granites formed by decompression melting involving muscovite ± biotite incongruent breakdown [49,60,61,79]. In particular, a strong connection with a magma source similar to the paragneisses that represent the dominant rock type in the Aspromonte Unit, has been recently inferred for the Villa S. Giovanni-Capo Rasocolmo leucogranodiorites, based on diagnostic bulk rock geochemical features and zircon inherited patterns [64].

The weakly to strongly peraluminous trondhjemites crop out as small bodies displaying a variety of field relationships with the host high-grade metamorphic rocks; the largest bodies are about 10 km² in size and their contacts range from sharp (discordant or concordant) to gradual. Trondhjemites are sometimes associated in the field with leucogranodiorites–leucogranites. The origin of the trondhjemitic rocks from the north-eastern Peloritani Mountains has been strongly debated. The rocks have been variously interpreted as the products of metasomatic alteration of metasediments [79], isochemical metamorphism of arkoses [86], partial melting of biotite paragneisses [47,60], fluid-assisted metamorphic differentiation of muscovite schists [87], and alkali metasomatism of the strongly peraluminous granites [63].

Two different types of trondhjemites occur in the Peloritani Mountains [47]: (a) magmatic trondhjemites formed at ~310 Ma by water-fluxed melting of metagreywackes at relatively high pressure and low temperature, close to the collision-related baric peak and (b) younger, low-Ca trondhjemitic rocks, formed by alkali metasomatism at expenses of the ~300 Ma leucogranodiorites-leucogranites.

As mentioned earlier, all the granitoid rocks from the north-eastern Peloritani Mountains, were locally affected by shearing. Deformation features are rarely visible in the field, also due to the low biotite amounts (0–7% vol.), associated to a coarse to very coarse feldspar and quartz size. On the other hand, shear-related microstructures are commonly observed in the late-Variscan granitoids, and are particularly well developed in the Forte Cavalli area (FC area, [47]; and references therein).

Post-Variscan shear-zone activity associated with Alpine tectonics is well known in the Peloritani Mountains ([51]; and references therein), even though the development of a discernible Alpine foliation is localized at the tectonic contact between the Mandanici and Aspromonte Units [26], and in a small area northwest of Messina ([88]; and references therein). A more intense Alpine reworking of older basement rocks, producing a pervasive mylonitic foliation and an evident stretching lineation, has been widely documented by several authors in different areas of the Aspromonte Massif [1,2,28,30,33,34,35,70,71].

The entire Alpine deformational history in the Peloritani Mountains has been constrained by [26] by studying the mylonitic rocks developed at the contact between the Aspromonte and Mandanici units.

Such mylonites form a quite continuous horizon of ~100 m in thickness due to repetition by near-isoclinal folding, dislocated by recent faults at the boundary between the uppermost high-grade metamorphic rocks and the underlying medium to low-grade metapelites [26]. Quartz c-axis patterns together with pseudotachylyte geothermometry constrained the shearing temperature from ≤600 °C to 450 °C for the exhumation of the tectonic slices forming the Upper Complex [26,51]. Subsequently, the shear zone developed under shallower conditions, accompanied by greenschist-facies assemblages, up to very low grade to cataclastic brittle regime.

In contrast, no ultimate evidence of Variscan shearing in the Peloritani Mountains has been up to date provided. [45] proposed that late-Variscan shear-zone activity was responsible for the nappe stacking of the Peloritani orogenic edifice, based on an age of 301 ± 2 Ma obtained by ³⁹Ar–⁴⁰Ar dating of white mica from a mylonitic augen gneiss from the tectonic contact between the Aspromonte Unit and the underlying Mandanici Unit. Nevertheless, the age of ~300 Ma is considered to reflect the climax of high-temperature Variscan metamorphism throughout the Peloritani Mountains and the Aspromonte Massif (e.g., [48,89,90]), and not the nappe stacking for which an Alpine tectonic phase at 28–22 Ma has been soundly demonstrated [1,26,67].

Fiannacca et al. [46] obtained for a garnet schist of the Mandanici Unit a clockwise P-T path consistent with late-Variscan crustal thickening up to deep crustal levels (P = 0.9 GPa at ~530 °C) followed by exhumation along an extensional shear zone under greenschist facies conditions. Nevertheless, also in this case, no indisputable evidence about the Variscan age of the shear zone activity, was obtained.

On the other hand, clear evidence of shearing affecting late-Variscan cooling granitoids in the Forte Cavalli area of north-eastern Peloritani (Figure 2) has been recently provided by [47]. The late-Variscan granitoids in this area consist of trondhjemites and granites, which are associated in the field with upper amphibolite-facies rocks mainly comprising metasedimentary migmatites, with minor granitoid gneisses and amphibolites. Shear-related deformation, including also considerable grain size reduction, is considered by [47] to have played a significant role in aiding fluid infiltration and associated production of metasomatic trondhjemites at expense of part of the original granites; according to the above authors, the metasomatic fluids were likely released during crystallization of the large volumes of granitic magmas of the Serre Batholith.

3. Granitoid Rocks in the Study Area

In this section we report the main field and petrographic features of the studied rocks, largely deriving from previous work from part of the same team [47], but integrated here by new field and thin-section images, to provide the reader with an exhaustive background on these rocks, before presenting the deformation microstructures which are the focus of the present work. Studied granitoid rocks crop out in the Forte Cavalli area, in the north-eastern Peloritani Mountains (Figure 2), intruding upper-amphibolite facies rocks composed of dominant biotitic paragneisses and migmatites (metatexites and rare diatexites), with minor amphibolites, micaschists and granitic gneisses. The main foliation dips 35–40° to NW and two main systems of brittle structures occur, with NE-SW and NNW-SSE directions.

Based on petrographic and geochemical features (see [47] for a detailed description), two main types of granitoid rocks occur in the FC area: trondhjemites s.s., with typical CaO contents of 1.5–3.0 wt.%, hereafter called trondhjemites (emplacement age: ~310 Ma; [19,20]) and low-Ca granitoids (emplacement age: ~300 Ma; [48,49,50]). The latter have CaO < 1.50 wt.%, range in composition from granite to low-Ca trondhjemite. Both rock groups are light-colored, with a very similar appearance in the field. In fact, although grain size is typically coarse to very-coarse for the trondhjemites and medium to coarse in the low-Ca granitoids, many exceptions occur.

Trondhjemites make up subconcordant to discordant bodies with thickness ranging from a few metres to ~300 metres (Figure 3a,b). They mainly consist of plagioclase and quartz, with minor biotite (<0–7 vol%), muscovite (0–5 vol%), tartan-twinned K-feldspar (0–5 vol%), and accessory zircon, apatite, ilmenite and monazite. Common secondary mineral phases produced by deuteric alteration are clinozoisite–epidote s.s., sericite, chlorite and opaques. Texture is mainly hypidiomorphic, coarse- to very coarse-grained and inequigranular (dominant grain-size >1.5 cm), with plagioclase and quartz typically larger than the other mineral phases; plagioclase can occur as crystals up to 4 cm long (Figure 4a). K-feldspar occurs mostly as medium-sized interstitial patches and thin rims bordering plagioclase crystals. A small amount of myrmekites and metasomatic turbid plagioclase, grown at the expense of primary K-feldspar, locally occur (Figure 4b).

Low-Ca granitoids include granites and low-Ca trondhjemites, cropping out as discordant bodies up to 100 m thick (Figure 3c,d). Both granitic and trondhjemitic varieties are characterized by a dominant medium to coarse-grained texture (dominant grain-size <7 mm), but inequigranular textures with feldspar megacrysts up to 5 cm long sometimes occur (Figure 4c–f).

The granites include granites s.s. and alkali feldspar granites, with K-feldspar (either microcline and orthoclase) mainly occurring as medium–coarse-sized subhedral crystals or megacrysts, sometimes intergrown with large quartz. Muscovite (3–6 vol%, sometimes up to more than 15% vol) is usually more abundant than biotite. Zircon, apatite, monazite, together with rare garnet and andalusite, occur as accessory phases; deuteric phases consist of chlorite, sericite, clinozoisite and opaques. Small amounts of myrmekites and turbid plagioclase replacing microcline, are often present.

Low-Ca trondhjemites, derived from alkali metasomatism of the granites, comprise transitional, low-Ca s.s. and high-Na low-Ca varieties reflecting progressive replacement of magmatic K-feldspar (either microcline and orthoclase) by secondary turbid plagioclase and myrmekites ([47]) (Figure 4e–h). The metasomatic trondhjemites are dominated by plagioclase and quartz, like the magmatic trondhjemites, but have many similarities with the granites, such as the dominant medium-coarse grain size, similar muscovite contents (2–6 vol%; up to 15 vol%), and the presence of andalusite and garnet among the accessory phases.

All the rock types exhibit shear-related deformation microstructures (Figure 5) that will be described in detail in the following section. It is here only highlighted that tectonic grain-size reduction has been considered by [47] to have played a significant role in facilitating fluid infiltration and metasomatic reactions by producing an increase of both the rock permeability and surface-controlled reactivity. Evidence for a link between shearing and metasomatism is provided by common growth of myrmekites and secondary plagioclase along mm-sized shear zones and over sigmoid feldspar porphyroclasts. In particular, in the metasomatic granitoids, the final texture results from the relative number of shear-related microstructures and late to post-shearing crystallization of pseudomorphic secondary plagioclase.

4. Materials and Methods

During the field survey, 144 specimens of both magmatic and metamorphic host rocks were collected. In particular, in this paper, we dealt with a set of thirty thin sections over the entire sample collection, which have been selected to study in detail the microstructural features in the two granitoid units of different emplacement ages: trondhjemites (15 samples; c. 310 Ma) and low-Ca granitoids (15 samples, c. 300 Ma). We acquired high resolution images of the entire thin sections by using a flatbed scanner (Epson Perfection 3490 Photo) at a resolution of 4800 dpi (both at crossed and parallel polarizers), together with several photomicrographs of deformed domains taken by an Axios Camera mounted on a Zeiss Microscope (model Axioscope 2), both hosted at the Department of Biological, Geological Environmental Sciences, Earth Science Section, University of Catania.

5. Shear-Related Microstructures

Studied rocks display a magmatic texture, with euhedral to subhedral micas and plagioclase, subhedral to anhedral K-feldspar and anhedral quartz, that do not define a magmatic foliation.

Nevertheless, all the granitoids exhibit, at different extents, several types of shear-related deformation microstructures (Figure 6, Figure 7 and Figure 8). Deformation affected all the rocks in the study area (~70 km²) but, in the field, it did not produce thick horizons of intensely sheared rocks. Rather, shear-related mesoscopic structures (e.g., foliation, preferred orientation of minerals) are weakly developed and only discernible at few locations. On the contrary, deformation is more evident at the thin-section scale, and restricted to domains typically forming networks of variously spaced and oriented ‘bands’ (e.g., Figure 6, Figure 7 and Figure 8). Protomylonitic structures occur in the most sheared samples from both rock groups (e.g., Figure 6a,b). Recrystallized fine-grained minerals (Qtz, Pl, Wm) in the strain shadows of plagioclase porphyroclasts have been observed (Figure 6i).

In the following two sub-sections we describe in detail shear-related microstructures developed in the two groups of rocks (trondhjemites, emplaced at ~310 Ma, and low-Ca granitoids, emplaced at ~300 Ma), from high-temperature sub-magmatic conditions to progressively lower temperatures.

5.1. Trondhjemites

The trondhjemites (Figure 6a,c–e and Figure 7) exhibit clear evidences of deformation at sub-magmatic conditions. The most striking evidence is represented by the widespread occurrence of submagmatic fractures (Figure 6c and Figure 7a–d), indicating deformation in the presence of melt. These microstructures are typical of rigid or semi-rigid crystal mushes, where a melt fraction still occurs and late magmatic crystals grow in fractures breaking apart early-crystallized minerals (e.g., [91,92,93]). In the studied rocks, medium- to fine-grained euhedral-subhedral plagioclase and larger interstitial quartz crystallize within veins propagating into large fractured plagioclase (Figure 6c and Figure 7c,d), consistently with melt migration into dilatation sites. Magma flow within these fractures is further testified by impressive examples of micro-scale filter-pressing (Figure 7c), with mechanical accumulation of plagioclase crystals against narrowing portions of the fracture (Figure 7c), or single euhedral crystals which remained caught in such fracture narrowings during flow of the residual magma (Figure 6c). In Figure 6c it is interesting to note the occurrence of a second system of submagmatic fractures, (sub-horizontal in Figure 6c), affecting the same plagioclase crystal; the two systems appear to be correlated since they coexist in the same portion of the crystal, whereas the other part of the crystal is fracture-free. The second system is formed by a set of narrow fractures which reveal their submagmatic nature because they are also confined within the plagioclase.

Deformation at submagmatic conditions is also supported by chessboard pattern extinctions in quartz (Figure 7d,e) [94], resulting from combined activation of basal <a> and prism <c> slip systems, which is mostly referred to occur at T above 650 °C for pressure up to 1.0 GPa [95,96,97,98].

Evidences of solid state-high temperature (HT-Ss) deformation come from grain boundary migration (GBM) recrystallisation [95,96,97,99,100] of quartz crystals (500–700 °C; [98]) with examples of inequigranular grain size and interlobate and sutured crystals boundaries, where the new forming grains tend to obliterate the neighbour ones (Figure 7h). Ribbon-like quartz crystals with undulose extinction, showing elongation parallel to the preferred stretching direction, occur locally (Figure 7g,h). Quartz and plagioclase crystals with irregular and lobate boundaries, as a result of bulging (BLG) recrystallisation, indicate sub-solidus deformation at 450–600 °C ([98]; Figure 7f). Few examples of myrmekitic intergrowths may also indicate deformation at high- to medium- temperatures. Lower temperature sub-grain rotation (SGR) recrystallisation [95,96,97,98,99,100,101] occurs locally as oriented polycrystalline quartz aggregates (Figure 7h) indicating deformation temperatures of 400–500 °C [98]. Rarely, quartz forms a core-and-mantle structure.

Bulging and impingements of plagioclase porphyroclasts with quartz grains (Figure 6f and Figure 7f) suggest deformation temperatures between 500 °C and 700 °C [63,65]. Core and mantle microstructures, typical of HT solid-state deformation, are also quite diffuse (Figure 7i). Very-fine grained quartz, and subordinately plagioclase, recrystallize at the boundaries of large plagioclase porphyroclasts. A temperature of deformation ranging between 400 and 500 °C is mainly indicated by deformed polysynthetic albite twins in plagioclase and, in rare cases, by eye-catching examples of kink-folded crystals (Figure 7m). Sometimes, in the rare K-feldspar crystals deformation twins also occur. Bulging (BLG) recrystallization of quartz crystals [95,98] indicates for these rocks temperature of deformation under 400 °C (Figure 7i). Other microstructures here observed developed probably below 450 °C [98], mainly consisting of undulose extinction of quartz (Figure 7g,i), mica kink banding (Figure 7l), feldspar bending and fracturing (Figure 7d,f,m); subgrains and new grains characterizing recrystallization domains in quartz and feldspars have also been observed (Figure 7i).

Evidences of late cataclastic-flow have been also observed, consisting of mm-thick conjugate sets of cataclastic shear bands, with fragmented quartz and feldspar grains (Figure 7a–d).

5.2. Low-Ca Granitoids

The rocks belonging to the second suite of granitoids, consisting of granites and metasomatic trondhjemites with the same low-Ca amount (dated at ~300 Ma [48,49,50], Figure 8), also show microstructures typically associated to submagmatic conditions, as well as to solid state deformation, from high- to low- temperature conditions.

High-temperature deformation microstructures, such as submagmatic fractures (Figure 6d and Figure 8a–d) and quartz chessboard pattern (Figure 8e,m), are also frequent in these granitoids.

Submagmatic fractures affect large K-feldspar and plagioclase crystals. It is, however, to be highlighted that, also in the case of fractures, apparently cutting the secondary plagioclase, the fractures affected the feldspar when it was still a magmatic K-feldspar. This is illustrated in Figure 4f–h, where the interfaces between the original K-feldspar and the quartz veins acted as a preferential path for the fluids involved in the metasomatic crystallization of the pseudomorphic secondary plagioclase.

Another aspect to highlight is that, in the low-Ca granitoids, quartz is the sole mineral phase filling the fractures. In this case the submagmatic nature of the fractures is only testified by the continuity between the quartz in the vein and the large quartz patches surrounding the feldspar, demonstrating that all the quartz crystallized at the same stage from the residual melt. In the same rocks there are many examples of other types of quartz-filled fractures that are instead clearly post-magmatic since they cross indistinctly all the mineral phases, including quartz (Figure 8f).

Evidences of solid state high-temperature deformation (T > 530 °C) are provided by quartz grain boundary migration (GBM) recrystallisation [95,97], as frequent as in the trondhjemites, and indicating temperature of deformation between 500 °C and 700 °C ([98]; Figure 8g). At deformation temperatures between 450 °C and 600 °C [63] dislocation climb becomes possible in plagioclase and recrystallisation starts to be important, occurring mainly by BLG (Figure 8d,g). Subgrain rotation recrystallisation [95,101] occurs in quartz crystals, where old grains are partially replaced by new forming ones, suggesting temperature of deformation of ~500 °C [98] (Figure 8f). Finally, small dynamic recrystallised quartz grains around older quartz cores or plagioclase porphyroclasts forming core-and-mantle structures, are common in this suite of rocks (Figure 8h–l).

Subsolidus low-temperature conditions are dominantly represented by deformation twins in plagioclase, and in a few cases, in K-feldspar, which according to [98], they typically occur between 400 °C and 500 °C (Figure 8m). Deformation microstructures developed under relatively lower temperature (~400 °C) mostly consist of mica fish and stretched andalusite (Figure 8i), as well as kinking of micas (Figure 8l). Quartz shows lobate and sutured grain-boundaries due to bulging, indicating deformation at temperature of ~300–400 °C (Figure 8m), even if it is less frequent than in trondhjemites. Ribbon-like quartz is also frequent, defining a clear micro-scale anisotropy for these rocks (Figure 6b,h and Figure 8h). The occasional presence of subgrains elongated subparallel with respect to a preferred stretching direction suggests that deformation protracted during cooling. Plagioclase shows intracrystalline plastic deformation evidenced by twin planes bending (Figure 8m). Undulose extinction, as well as deformation bands, occur frequently in quartz grains (Figure 8l), pointing to deformation at temperature under 300 °C [98]. Mica fish (Figure 6l) and kink bands indicate intracrystalline plastic deformation (Figure 8i–l). Later, brittle fracturing occurred at lower temperatures (T < ~250 °C; [98]). Rare examples of tectonic grain size reduction by cataclastic flow are usually confined to narrow shear bands.

6. Discussion

6.1. A Late-Variscan Shear Zone in the Peloritani Mountains

Until now, only Alpine shear zones had been well-constrained in the Aspromonte Unit from Peloritani Mountains and Aspromonte Massif, associated first to early-Alpine subduction and then to late-Alpine exhumation [26,27,29,34]. Alpine mylonitic events in the region took place in a compressional regime and the whole orogenic cycle developed up to upper amphibolite facies, but entirely subsolidus, conditions ([51]; and references therein). In particular, Alpine shearing mostly produced LT deformation microstructures (T ≤ 450 °C), and an evident Alpin foliation developed along the tectonic contacts between the Mandanici and Aspromonte Units ([26]; and references therein) (Figure 1). Peak metamorphic conditions of ~0.8 GPa at ~600 °C [26,29] and the development of an evident Alpine foliation were only attained at the above tectonic contacts and in a small area northwest of Messina ([90] and references therein), many km apart from the study area.

Conversely, non-coaxial deformation microstructures detected in this study in late-Variscan cooling granitoids provide, for the first time, evidence for a late-Variscan shear phase in the Peloritani Mountains.

The occurrence of rare micro-boudins affecting mm-sized andalusite crystals in the studied low-Ca granitoids (Figure 8i) might indicate a general extension-type deformation. This finding is consistent with the occurrence of stretched and pinch-and-swell garnet porphyroclasts in phyllites from the central Peloritani Mountains [46], leading those authors to tentatively suggest a late- to post-orogenic shearing stage related to late-Variscan exhumation.

Finally, the observed gradual transition from submagmatic to low-temperature subsolidus microstructures, developed at progressively decreasing temperatures, points to a continuous change in the deformation mechanisms during cooling of the granitoids (e.g., [19,95,102,103]). Although an Alpine overprint might have been superimposed on the Variscan microstructures, all the above exposed arguments are reasonably in favor of a continuous evolution of late-Variscan shearing. This shearing involved the granitoid bodies at suprasolidus conditions, as documented by the widespread occurrence, in both groups of studied granitoids, of typical submagmatic deformation microstructures such as chessboard extinction in quartz and submagmatic fractures (Figure 7). In this respect, it is to be remarked that chessboard subgrain pattern in quartz have been alternatively considered to reflect deformation at submagmatic (e.g., [92,102,103,104,105,106,107]) or HT subsolidus (e.g., [19,99,108,109,110]) conditions. Nevertheless, independently from their suprasolidus vs. subsolidus categorization of the chessboard pattern, most researchers agree that this microstructure indicates deformation at T > 650 °C, which typically correspond to melt-present conditions in a crystallising felsic granitoid system (e.g., Figure 3 in [91]).

In addition, several examples of submagmatic veins provide a direct evidence of deformation in presence of melt, therefore clearly indicating that the shear zone was active, in the Aspromonte-Peloritani sector of the CPO, during late-stage crystallization of the late-Variscan granitoid magmas.

It is thus possible to trace a spatial connection and a temporal continuity with the activation of the crustal-scale shear zone, evolving from deep-seated to upper crustal conditions, which controlled the emplacement of the granitoid magmas of the 13 km-thick Serre Batholith, in central Calabria, between ~297 and ~292 Ma ([15,36,38,42,77,111,112]; and references therein). Syn-shearing crystallization of late-Variscan granitoids (~304-300 Ma; [48]) has been also ascertained for the Sila Batholith, in northern Calabria, where continuous shear-related deformation from magmatic to low-temperature solid state conditions has been documented by [39,40,43,44].

At a larger continental scale [51], in agreement with [54,55], proposed that the formation of the Sila and Serre Batholith in northern and central Calabria could be framed in the activity of East Variscan Shear Zone (EVSZ) [52,53,54,55].

The EVSZ is a branch of a complex network of coeval strike-slip shear zones, [54,55,113,114,115,116,117], which were involved in the final assembly of Pangea, up to around 300 Ma, in regions of oblique collision of Gondwana, Gondwana-derived microcontinents and Laurussia [118,119,120]. In particular, the EVSZ has been mostly described as a dextral transpressive shear zone running from the external massifs of the Alps to the Corsica-Sardinia-Maures–Tanneron Massifs [53,121,122,123]. Padovano et al. [54,55] have recently proposed that also the Northern Apennines and the CPO were involved in the EVSZ, and that all these regional shear zones were characterized by a transpressional and/or transtensional kinematics, associated with exhumation of deep-middle crust and syntectonic granitoid emplacement.

The late-Variscan shear zones from the Sila and Serre Massif of northern and central Calabria are long known, whereas the proposition from [54,55] that also Peloritani Mountains could be involved in the EVSZ was based on the late-Variscan Ar-Ar muscovite age of 301 ± 2 Ma obtained for a mylonitic augen gneiss by [45]. The latter authors proposed indeed that the age of 301 ± 2 Ma was associated with mylonitisation during thrusting of the Aspromonte Unit over the Mandanici Unit. Nevertheless, the contact between the two units is marked by a 30 m-thick cataclastic to mylonitic shear zone that also involves Triassic to Cretaceous greenschist-facies metasediments from the original cover of the Mandanici Unit ([124]; and references therein), therefore demonstrating the Alpine age of that shear zone. In fact, the age of ~ 300 Ma in the mylonitic augen gneisses cannot be univocally ascribed to a shearing phase, because it corresponds to the age of the thermal peak of the Variscan metamorphism in the CPO, which is recurrently recorded in both sheared and unsheared rocks (e.g., [48,90,125,126]).

On the other hand, shear-related deformation microstructures in the presence of late-Variscan melt found for the first time in this study are a clear evidence of late-Variscan shearing, and might therefore better support the hypothesis that the Peloritani Mountains were also possibly involved in the EVSZ.

In this respect, it is interesting to note that [51] suggested a different location of the CPO massifs with respect to strike-slip domains developed in the last stages of the EVSZ activity, based on the distribution of the late-Variscan magmatic products. In particular, by considering the observed strict connection between activation of late-Variscan shear zones and emplacement of large syn-shear granitoid complexes along the EVSZ, the presence of only small anatectic plutons, rather than composite batholiths, in the Aspromonte-Peloritani has been tentatively considered related to their more peripheral position in the shear zone system.

This suggestion appears now consistent with the style of deformation observed in the granitoids from the Peloritani Mountains which do not form thick horizons of strongly foliated rocks, like the syn-tectonic Serre and Sila granitoids, but are instead characterized by less pervasively sheared rocks.

Another aspect to be considered is the age of the recorded shearing events in NE Sicily, with reference to those recorded in different segments of the EVSZ. ([122] and references therein) remark indeed that the EVSZ is a network of interconnected shear zones that developed progressively according to a tectonic linkage model in the time interval of about 335–300 Ma.

The end of late-Variscan shearing in the CPO can be set between ~292 Ma, which is the age of weakly deformed upper crustal granodiorites from the Serre Massif [15,127] and the age, similar within error, of 281 ± 8 Ma of undeformed felsic dykes intruding the sheared granitoids of the Sila Batholith [44].

On the other hand the oldest dated syn-shear magmatic rocks in the CPO are the studied trondhjemites from the Peloritani Mountains, emplaced at ~310 Ma [49,50], that indicate it as the minimum age for initiation of the shear zone activity. Rb-Sr Bt cooling ages [48,128] together with thermobaric modelling [112] suggest that the granitoids emplaced into the upper crust of Aspromonte-Peloritani and Serre areas took more than ~ 10 Ma to cool at a T of ~300 °C; this would indicate that deformation in the ~300 Ma low-Ca granitoids lasted until ~290 Ma, indicating a total duration of the shear zone event of ~20 Ma, from intrusion of the trondhjemites at ~310 Ma to LT subsolidus deformation in the ~300 Ma granites.

6.2. Supra- to Sub-Solidus Deformation in Granitoids Cooling at Middle to Upper Crustal Depth

Deformation microstructures developed from a submagmatic to low-temperature solid-state regime in trondhjemites and granites from the Peloritani Mountains give the opportunity to investigate the relationships between fabric development and syn-tectonic cooling of granitoids emplaced at middle to upper crustal conditions. Changing temperatures during prograde or retrograde orogenic evolution are largely considered to exert the main role influencing the mechanisms of deformation in crustal rocks (e.g., [98,129]). Nevertheless, other factors can also affect the rheological behaviour of the crust, such as the rock composition and grain size, the limiting pressure, the strain rate, the flow vorticity, and the presence of a fluid or a melt phase [130,131].

In the Peloritani Mountains, two continuous sequences of shear-related deformation at submagmatic to low-temperature subsolidus conditions developed during cooling of granitoids emplaced at different times and different depth. The two granitoid intrusions have a similar composition from a rheological point of view, having the same silica content [47] and being composed by comparable amounts of quartz, feldspars and micas. Furthermore, evidences of deformation in the presence of melt occur in both granitoids. The main difference between them is that the grain size is dominantly coarse to very-coarse in the trondhjemites and medium to coarse in the younger low-Ca granitoids. We do not have information about the strain rate and the flow vorticity and, therefore, the factors that we take into account in addition to temperature, to assess their possible role in controlling the development of specific deformation mechanisms in a natural shear zone, are the grain size and the confining pressure.

In this latter respect, precise pressure condition estimates are not available for the Peloritani Mountains granitoids, also due to the lack of suitable mineral assemblages but, according to [47] and [65], the ~310 Ma trondhjemite magmas intruded the migmatite host rocks after leucosome solidification and migmatite equilibration at ~0.5 GPa (i.e., ~14 km) (e.g., [69,70]). On the other hand, final crystallization of the ~300 Ma granites is likely to have occurred at upper crustal levels, at a pressure of ~0.3 GPa (i.e., ~8 km), as suggested by the occurrence of local granophyric and graphic textures, which are typically associated to crystallization under strong undercooling conditions (e.g., [132,133]). Low-pressure crystallization of the granitic magmas is also supported by the occasional presence of magmatic andalusite in the studied low-Ca granitoid rocks. It is therefore interesting to compare the different types of diagnostic deformation microstructures developed in the trondhjemites, that were intruded at deeper crustal levels (~14 km) during the early evolution of the shear zone and in the granites that were intruded 10 Ma later at shallow depth (~8 km).

Comparison between the two granitoid associations shows similar sequences of deformation microstructures, developed at progressively lower temperatures, that indicate continuous shear zone activity during cooling of the granitoids.

Unlike the syn-tectonic tonalites and granodiorites of the Serre and Sila batholiths, both Pelolritani Mountains trondhjemites and granites do no show a pre-full crystallization [134], or magmatic fabric [108]. This fabric indicates a free rotation of the crystals in the magma, with a melt fraction greater than ~40% (e.g., [108]). The absence of magmatic fabrics in the trondhjemites would indicate that shearing probably started soon after the emplacement of the trondhjemitic magmas. This would imply, in turn, that magma emplacement assisted the nucleation of the shear zone in the study area, (e.g., [10,135]).

Neves et al. [135] (and references therein) have indeed suggested that nucleation of shear zones in crystallizing plutons is a plausible mechanism, due to marked reduction of the effective viscosity in melt-bearing rocks, even at low melt fractions (<10–20%), compared to melt-free rocks, that could therefore favor strain localization in these weak domains.

Activation of early shear zone domains in the trondhjemite plutons according to these considerations is, however, at odds with the lack of magmatic fabrics also in the granites, that were emplaced when the shear zone was already operating in the area for several million years. On the other hand, the recurring spatial association of “old” trondhjemites and “young” granites throughout the Peloritani Mountains and Aspromonte Massif, as visible in Figure 1, suggests that the ascent pathways used by the trondhjemite magmas were then re-used by the granites.

Furthermore, for both types of granitoid plutons, a similar tight connection between emplacement and tectonics, is suggested by the elongation of the plutons along the same trends, which is especially evident in the Aspromonte Massif (e.g., E-W elongated Punta d’Atò pluton; [34]).

In this respect, the absence of a pre-full crystallization fabric in the granites might indicate that the granite magmas were emplaced by reusing the ascent pathways of the trondhjemites during a phase of tectonic quiescence. Shearing in the Peloritani-Aspromonte sector of the CPO then resumed, at a time when those granites were largely crystallized, and the main shear zones of the CPO started to operate in central and northern Calabria, leading to syn-tectonic construction of the Serre and Sila batholiths and to associated 13-km displacement of deep-crustal gneisses, that were finally exhumed to shallow upper crustal levels at ~ 290 Ma (e.g., Mammola paragneisses [15,41,111,136]).

An alternative explanation is that also the peloritanian granitoid magmas were emplaced syntectonically, but the magmatic fabric was obliterated by subsequent deformation, possibly already at near solidus conditions, involving differential grain rotation, grain interference, and grain fracturing ([108]; and references therein).

Despite the apparent lack of magmatic fabrics in the studied trondhjemites and low-Ca granitoids from the Peloritani Mountains, quartz chessboard extinction and, especially, submagmatic fractures, indicate that deformation started anyway in the presence of melt. In particular, submagmatic fractures are considered to reflect deformation at low melt fractions <0.4 and, more typically, within the 0.3–0.1 range (e.g., [91,103,137]).

A fundamental aspect to be considered is that both granitoid units share the same sequence of fabrics developed from submagmatic to low temperature subsolidus conditions (Figure 9). This implies overprinting of low temperature microstructures on earlier submagmatic and HT subsolidus relics, making difficult to establish the relative role of the different deformation mechanisms in the two granitoid units and their relative implications. Nevertheless, some significant inferences can be drawn by estimating the incidence of each recorded deformation microstructure in both units.

By first reviewing the microstructures typical of submagmatic conditions observed in the two granitoid units, common and distinctive features can be observed. Submagmatic fractures in feldspars and chessboard patterns in quartz are quite equally distributed in the two groups of rocks, suggesting that the larger grain size of the trondhjemites, or the relatively greater emplacement depth, did not make them particularly prone to fracture or deform at low melt fractions.

Most of the SS-HT deformation microstructures, such as feldspar bulging, sub-grain rotation dynamic recrystallization, and grain-boundary migration also have a similar distribution in the two granitoid types of rocks.

Similarly, ribbon-like quartz is largely represented in all the granitoid types, as well as quartz grain boundary migration and subgrain rotation. In contrast, core and mantle structures of quartz grains are distinctly more frequent in low-Ca granitoids than in the trondhjemites. This diversity might reflect the significant grain-size difference in the two rock types, due to the higher amount of grain boundaries per volume unit in the finer-grained low-Ca granitoids.

Regarding the SS-LT microstructures, no significant difference has been detected in the two granitoid units in terms of incidence of both quartz bulging, plagioclase deformation twinning and mica kinking.

The similar distribution of deformation microstructures observed in both trondhjemites and low-Ca granitoids supports the proposal that temperature is the main factor controlling the activation of specific deformation mechanisms (e.g., [98,129]), compared to the confining pressure and rock grain size. The only exception might be the different occurrence of core-and-mantle structures in quartz that appears to be higher in the finer-grained low-Ca granitoids.

We observed also that the most frequent microstructure, in both granitoids, is the ribbon quartz, whereas the rarest is the quartz GBM recrystallization. This suggests that the dominant mechanism activated in the deformed cooling granitoids, and/or better preserved during subsequent deformation at lower temperatures, was HT subgrain rotation associated to a shear strain gradient (e.g., [142]). This explanation might account, in turn, for the coupled low incidence of the quartz GBM in both studied granitoids, by considering that both ribbon quartz and grain boundary migration can develop at similar temperatures, but at sensibly different strain conditions. A quite similar conclusion is that ribbon quartz may have reworked previously formed GBM related microstructures.

Finally, submagmatic fractures are also quite rare, but this is an expected result, due to the very transient near-solidus conditions at which these microstructures can develop.

6.3. Implications for Shear-Assisted Exhumation

Microstructural evidences obtained for the granitoids of Forte Cavalli area in the north-eastern Peloritani Mountains, integrated by existing geochronological data, indicate long-lived late-Variscan shear zone activity coeval with granitoid magma emplacement and cooling for a period of ~20 Ma. The earliest granitoid magmas were produced by water-fluxed melting of metagreywackes at pressure likely close to 1.0 GPa ([47]; and references therein), near the collision-related baric peak, giving rise to the trondhjemitic plutons that emplaced at ~310 Ma [49,50]. According to the evidence obtained in this study, shearing started just after the emplacement of the magmatic trondhjemites, as documented by development of submagmatic fractures and chessboard patterns in quartz, and continued during cooling of the magmatic body, as shown by repeated superposition of subsolidus deformation microstructures of progressively lower temperature (Figure 7, Figure 8 and Figure 9).

In the contiguous Pizzo Bottino pluton, the trondhjemites intrude metagreywacke metatexites with trondhjemite leucosomes that are considered to have been also produced by water-fluxed melting at PT conditions of ~1.0 GPa and 700 °C, starting at around 340–330 Ma [65]. The migmatites were then slightly heated during decompression up to Ms-dehydration melting conditions, producing granitic leucosome melt and, finally, totally solidified before the emplacement of the trondhjemite pluton at ~310 Ma and ~0.5 GPa ([65] and references therein). The migmatite path, from early migmatization at ~1.0 GPa to final solidification at ~0.5 GPa, implies a significant decompression with associated post-collisional crustal exhumation of ~14 km.

No evidence for shear zone involvement in this early exhumation stage has been found in the above mentioned migmatites, that appear to have recorded shear-related deformation at high temperature, but only after their complete solidification [65]. On the other hand, submagmatic fractures appear to be absent also in the migmatites from the Forte Cavalli area, but chessboard extinction in quartz is locally present, suggesting possible deformation in the presence of melt. In this respect, it is however to be considered that the rapid transition of the host rocks in the study area from unmigmatized paragneisses to metatexites approaching the granitoid complex, with rare diatexites in the inner part (Figure 2), suggests that migmatization might have been induced by magma emplacement in the upper amphibolite facies paragneisses.

After magma intrusion and extensive crystallization, the trondhjemite plutons were uplifted, together with their metamorphic host rocks, and were then intruded by the granite magmas at ~300 Ma [47,48,49,50] and at lower pressure of ~0.3 GPa. In this case, the presence of shear-related microstructures developed in the trondhjemites from submagmatic to low temperature subsolidus conditions indicate that this second-stage exhumation, from a depth of about 14 km to a depth of about 8 km was assisted by shear zone activity.

Then, also the younger granites were in turn affected by shearing during cooling, starting to record deformation at sub-magmatic conditions, with diffuse formation of submagmatic fractures and chessboard patterns in quartz, followed by subsolidus deformation at progressively lower temperature.

The occurrence of domino-torn boudins in tiny andalusite crystals, in conjunction with stretched and pinch-and-swell garnet porphyroclasts in phyllites from the central Peloritani Mountains [46], suggests that late-Variscan exhumation involving the Peloritanian granites occurred under an extensional regime.

On the whole, considering that migmatites formed at ~1.0 GPa were intruded by granites at ~0.3 GPa, implies that late-Variscan post-collisional evolution involved ~20 km of exhumation of deep continental crust.

Shear-related deformation continued during cooling of the granites down to low-temperature conditions in a ductile and, finally, brittle regime.

7. Conclusions

Investigation of the microstructures exhibited by late-Variscan trondhjemites (~310 Ma) and granites (~300 Ma) intruded in high-grade paragneisses from the Peloritani Mountains (southern Italy) revealed, for the first time, that both granitoids were affected by shear-related deformation from submagmatic (T > 650 °C) to low temperature solid state conditions (T < 300 °C). Non-coaxial deformation is documented by sigmoidal porphyroclasts, mica fish and domino-torn boudins developed in tiny andalusite crystals. The occurrence of chessboard patterns in quartz and, especially, of submagmatic fractures noticeably indicates deformation in the presence of melt and, therefore, reliably constrains the activity of the shear zone during the solidification of the granitoid magmas, at the final stages of the Variscan orogenic cycle. This result could support the proposed location of the Peloritani Variscan basement within a peripheral branch of the East Variscan Shear Zone [54,55].

Submagmatic deformation was extensively superseded by continuous subsolidus deformation at progressively lower temperatures. Examples of solid state-high temperature deformation microstructures (T > 450 °C) include feldspar bulging and quartz grain boundary migration recrystallization, quartz ribbons and core-and-mantle structures. Finally, low temperature subsolidus microstructures consist of quartz subgrain rotation recrystallization, bulging, mica kinks, and feldspar twinning and bending. Each type of deformation microstructure is often placed side by side with others of lower or higher temperature, documenting the progressive change of the deformation mechanisms in the single minerals and mineral aggregates upon decreasing temperature. Comparison of the shear-related microstructures developed in the older trondhjemites, emplaced at higher depth (~0.5 GPa) and younger granites, emplaced at shallower crustal levels (~0.3 GPa) reveals no systematic relationship between deformation mechanisms and emplacement depth, or rock grain size, during different evolutionary stages of the shear zone. The only exception might be the higher incidence of core and mantle structures of quartz grains in the low-Ca granitoids, possibly reflecting the higher amount of grain boundaries per volume unit in finer-grained rocks.

The two continuous sequences of submagmatic to low-temperature shear-related deformation, recorded in both the older and younger granitoids, suggest a duration of ~20 Ma for shear activity during syn-magmatic post-collisional exhumation of the Variscan middle crust in southernmost Italy.