Introduction

The late-orogenic extension of the hinterland of collisional belts leads to the development of large gneiss-cored domes that typically undergo long-lived magmatic activity including intense migmatization and concomitant generation of decompressional crustal melts (i.e., Sawyer 1994, 1999; Brown 1994, 2001; Vanderhaeghe and Teyssier 2001; Vanderhaeghe 2004). Several realms of the European Variscan Belt (Fig. 1) exhibit the superposition of granitoid suites that record the complete magmatic history of the orogen, including late-orogenic extension-related peraluminous anatectic domains (Martínez and Rolet 1988; Rossi and Cocherie 1991; Timmerman 2008; Faure et al. 2009a, b; Lardeaux et al. 2014; Tabaud et al. 2015). In NW Iberia, there are several gneiss-cored domes: Lugo (Arenas and Martínez Catalán 2003; Reche et al. 1998; Alcock et al. 2009), Sanabria (Díez Montes 2007), Tormes (Martínez 1977; López-Plaza and Gonzalo 1993; Martínez et al. 1988; López-Plaza et al. 1999; Escuder Viruete et al. 1997, 2000) (Fig. 2), Guadarrama (Barbero and Villaseca 2000), Somosierra (Rubio Pascual et al. 2013), Toledo (Barbero 1995; Barbero et al. 1995; Bea et al. 2006a; Castiñeiras et al. 2008), Padrón (Díez Fernández et al. 2012), Peares, and Celanova. All these domes include a structurally lower unit (core) made up of meta-igneous and meta-sedimentary migmatites that constitute the wall rocks to syn-extensional crustal leucogranitoids, whose source is debated (Bea et al. 2003; Fernández-Suárez et al. 2011), and an upper unit, limited by km scale extensional shear zones, with peraluminous syn-kinematic granitoids that also intrude the lower grade metamorphic rocks. For comprehensive reviews on the origin and evolution of these domes, see Martínez-Catalán et al. (2014) and Alcock et al. (2015).

Fig. 1
figure 1

Permian reconstruction of the West European Variscan Belt (according to Ballèvre et al. 2014, modified by Fernández-Lozano et al. 2016). Red star indicates the location of the study area. CZ Cantabrian Zone, WALZ West Asturian-Leonese Zone, CIZ Central Iberian Zone, OMZ Ossa-Morena Zone, SPZ South Portuguese Zone

Full size image
Fig. 2
figure 2

Simplified geological map of the Tormes Dome (López-Plaza and López-Moro 2004, 2008) showing the most important lithological types and indicating the location of the studied samples (red stars) and other available Variscan geochronological data in the region

Full size image

The cost and time efficient U–Pb Laser ablation ICP-MS technique allows for a statistically robust characterization of the zircon age spectrum of granitoid rocks, determining the age of the magmatic event and characterizing the main sources (zircon-bearing rocks) involved in crustal melting. In high-grade domains, with a prolonged history of recurrent migmatization, magmatic zircon may not only be related to the emplacement of the granite, but also the age spectrum of zircon from such rocks may contain several generations of magmatic zircon and inherited zircon from the source and/or assimilated wall rocks. This study shows that crustal recycling/cannibalization may often happen at a fast pace in orogenic scenarios with only short lapses of quiescence; in our case, it seems plausible that a “crustal layer” of ca. 340 Ma granitoids/migmatites was recycled, partially or totally, only 15–20 My after its emplacement.

The present study focuses on the crystallization history of crustal granitoids in the Tormes Dome and the source rocks involved in their generation. The results obtained herein constrain the emplacement age of the studied syn-kinematic granitoids to ca. 320 My ago and suggest that the main sources of magma correspond to different rock types which can be identified in their surroundings (Ediacaran and Ordovician), together with other sources (Devonian and Early Carboniferous) whose nature and origin must be further tested.

Geological setting

The Iberian Massif represents a large area of Palaeozoic and Ediacaran rocks in western Iberia (Fig. 1) and is part of the Variscan orogen of Western Europe. The Iberian Massif includes Proterozoic Gondwanan rocks with north African affinity, which form the basement for Ediacaran-Early Cambrian subduction‐related and Lower Palaeozoic passive margin sequences (e.g., Rodríguez-Alonso et al. 2004; Murphy et al. 2008; Fernández-Suárez et al. 2014; Rubio-Ordóñez et al. 2015). In the late Devonian and Carboniferous, collision of the passive margin of Gondwana (acting as the lower plate) with Laurussia (upper plate) resulted in the Variscan orogen and ultimately the amalgamation of Pangea (e.g., Matte 1986, 2001; Murphy et al. 2009; Kroner and Romer 2013; Kroner et al. 2016). A sequence of magmatic events recorded in rocks of the Variscan orogen of Iberia can be summarized as follows: (1) A subduction‐related Cadomian (ca. 600 Ma) magmatic event dominated by I‐type granitoids and volcanic rocks (Fernández-Suárez et al. 1998; Rubio-Ordóñez et al. 2015). This event is scarcely represented in NW Iberia; (2) a voluminous extension‐related late Cambrian to early Ordovician magmatic event (ca. 490–470 Ma) generally interpreted to be linked to the opening of the Rheic Ocean (Murphy et al. 2006; Bea et al. 2006b; Montero et al. 2007; Díez Montes et al. 2010; Valverde-Vaquero and Dunning 2000; Talavera et al. 2013); (3) carboniferous syn‐orogenic (Variscan) magmatism that began at ca. 350–340 Ma in the hinterland of the orogen (Gallastegui 2005) and ended at ca. 320–315 Ma (e.g., Valle Aguado et al. 2005; Dias et al. 1998; Fernández-Suárez et al. 2000); (4) post‐orogenic magmatism that peaked at ca. 310–295 Ma when voluminous granitoids and some mafic rocks with their extrusive equivalents were emplaced and erupted in both the internal and external zones of the orogen (Fernández-Suárez et al. 2000, 2011; Bea et al. 2006a; Orejana et al. 2009; Gutiérrez-Alonso et al. 2011). This post-orogenic magmatism is observed throughout the entire Variscan belt (e.g., Kroner and Romer 2013).

Geology of the Tormes Dome

Gneiss-cored domes within the Central Iberian Zone (Fig. 1), one of the paleogeographic domains of the Iberian Massif, include the Tormes Dome (TD) (Fig. 2) and the Sanabria and Somosierra domes. The Central Iberian Zone is characterized by abundant magmatic rocks, some of which crop out in the core of the structural gneiss-cored domes, forming characteristic high-grade, commonly migmatized, plutono-metamorphic complexes (Martínez et al. 1988). The present-day geometry of the TD dome resulted from the superposition of (1) extensional Variscan structures that controlled the location of zones of extensive partial melting and multiple granite intrusion and (2) post-Variscan strike-slip shear zones (Fig. 2). The Variscan magmatic rocks of the TD include (1) calc-alkaline (and K-rich) granitoids and associated monzonitic rocks of I-type affinity (López-Moro and López-Plaza 2004), which traditionally have been referred to as “Older Granodiorites” or “Granodiorite Suite” (Capdevila et al. 1973; Castro et al. 2002), and (2) S-type peraluminous leucogranites (López-Plaza et al. 2008; López-Moro et al. 2012), which are abundant throughout the NW Iberian Massif and are known as “Older two-mica Granites” or “Peraluminous Suite” (Capdevila et al. 1973; Castro et al. 2002). Calc-alkaline and K-rich granitoids are made up of fine- to medium-grained biotite ± muscovite granites, porphyritic biotite granites, and porphyritic biotite ± muscovite granites (Fig. 2). S-type peraluminous leucogranites consist of porphyritic two-mica ± sillimanite granites and equigranular granites. The latter are composed of two-mica granites (coarse-grained, medium-grained, and fine-grained granites), tourmaline-bearing granites (mainly coarse-grained granites), garnet-bearing granites (medium- to very coarse-grained), and cordierite granites (medium- to very coarse-grained).

The metamorphic and igneous rocks in the TD have been divided into a Lower Unit and an Upper Unit (Escuder Viruete et al. 1994) separated by an initially shallow dipping extensional ductile detachment with top to the SE displacement. The Lower Unit includes migmatized felsic gneisses and migmatized Neoproterozoic-Lower Cambrian sandstones and pelites, partially metamorphosed into garnet-cordierite-biotite-sillimanite paragneisses (Gil Ibarguchi and Martínez 1982). The felsic gneisses comprise garnet-bearing, fine-grained gneisses, and Early Ordovician augen gneisses (Bea et al. 2006b; Talavera et al. 2013). The Upper Unit represents the highest structural level of the dome and consists of a monotonous sequence of Lower Cambrian slates and schists in lower grade metamorphic conditions.

Variscan D1 deformation produced NW–SE trending asymmetric NE-vergent folds in the upper structural levels, and NE-vergent large-scale recumbent folds and thrust sheets in the lower structural levels. These early Variscan compressional structures were variably overprinted during the D2 extensional event (Escuder Viruete et al. 1994) with large-scale, sub-horizontal shear zones delimiting rocks with different metamorphic grade (upper and lower units, Escuder Viruete 1998). Subsequently, Variscan D3 structures produced upright, open to tight folds of centimeter to kilometer wavelength responsible for the present-day structure of the dome (Iglesias and Choukroune 1980; López-Plaza 1982) (Fig. 2). Coeval with or post-dating the D3 deformation event, western Iberia was affected by large-scale ductile strike-slip shear zones dated at ca. 308 Ma (40Ar/39Ar on muscovite, Gutiérrez-Alonso et al. 2015). The most prominent of these shear zones in the TD is the sinistral Juzbado-Penalva Shear Zone and its conjugated shear zones (the sinistral Pelazas Shear Zone and the dextral NW striking Fermoselle Shear Zone) (Fig. 2), which are responsible for the sigmoidal foliation pattern found throughout the whole area. These large-scale shear zones are coeval with the post-Variscan counterclockwise rotation of the TD during the generation of the Cantabrian Orocline (or Ibero Armorican Arc, Fig. 1) as revealed by recent paleomagnetic studies in this region and its surroundings (Fernández-Lozano et al. 2016; Pastor-Galán et al. 2016).

Previous geochronology

The timeframe of the geological evolution of the area has been largely inferred from structural relationships, some of which are bracketed by units whose age is poorly constrained. The compressional D1 deformation that led to Barrovian-type upper amphibolite facies metamorphism (M1), in the lower unit, is interpreted to have peaked at ca. 332–337 Ma in the Tormes Dome (U–Pb ID-TIMS in monazite, Valverde-Vaquero et al. 2007) and ca. 337 Ma in the Somosierra Dome (SD) (U–Pb ID-TIMS in monazite, Escuder Viruete et al. 1998), although recent 40Ar/39Ar ages in the Somosierra Dome attribute an age of 354–347 Ma to this phase (Rubio Pascual et al. 2013). Major D2 extension, which resulted in nearly isothermal decompression from 800–900 to 300 MPa in the TD (Escuder Viruete et al. 2000) and from 1000–1200 to 400 MPa in the SD (Rubio Pascual et al. 2013), was recorded in low-pressure/high-temperature paragenesis (M2) and resulted in extensive migmatization and anatexis together with D1 isograd telescoping. These migmatites yielded ages at 325–320 Ma (U–Pb ID-TIMS in monazite, Valverde-Vaquero et al. 2007) and the extensional fabric in the SD gave coeval 40Ar/39Ar ages between 323 and 314 Ma (Rubio Pascual et al. 2013). These ages are slightly younger than the 334 Ma Rb–Sr whole-rock isochron age of Beetsma (1995; recalculated with the 87Rb decay constant recommended by IUGS; Villa et al. 2015) for coarse-grained two mica granites from the Barruecopardo area (Fig. 2) and the emplacement age of the biotite ± muscovite porphyritic granites in the neighborhood (320 ± 5, U–Pb ID-TIMS on monazite, Ferreira et al. 2000; 319 Ma U–Pb ID-TIMS on zircon, Gomes et al. 2014; 318–322 Ma U–Pb–Th in monazites, Gloaguen 2006; Gloaguen et al. 2006; 322–317 Ma U–Pb ID-TIMS on zircon, Costa et al. 2014; 330–321 Ma U–Pb SHRIMP, Díez Fernández and Pereira 2017). Migmatization temperatures may have persisted locally at lower-middle crustal levels as a result of the high thermal gradient inherited from D2, as voluminous syn-D3 granitoids yielded slightly younger ages (311 Ma U/Pb monazite, López-Moro et al. 2012; 315–313 Ma U–Pb ID-TIMS, Gomes et al. 2014; 316–310 Ma U–Pb ID-TIMS, Valverde-Vaquero et al. 2007; 316 ± 9 and 311 ± 7 Ma U–Pb SHRIMP, Díez Fernández and Pereira 2017). These ages are also in agreement with the ages obtained in other gneiss-cored domes along the Ibero-Armorican Arc (Augier et al. 2015). Late strike-slip movements along the Juzbado-Penalva and related shear zones have been dated at 308 Ma (Gutiérrez-Alonso et al. 2015 and references therein).

Field description and petrography

About 65% of the plutonic rocks from the TD consist of equigranular two-mica granites, including coarse-grained types (average mineral diameter 4 mm, alkali feldspar crystals being larger than the rest), medium-grained (ø = 1.55 mm), and fine-grained (ø = 1.03 mm) granite types. They occur as sheet-like intrusions on the outcrop to the pluton scale (López-Plaza and López-Moro 2004). Five representative samples (Figs. 2, 3) from four syn-kinematic granitoids of the TD were collected for U–Pb LA-ICP MS dating to constrain their intrusion age and their possible protoliths. The analyzed samples are from plutons of the migmatized TD, except for sample BARR that is from the epizonal Barruecopardo pluton.

Fig. 3
figure 3

Detailed maps depicting the lithologies and structures in the surroundings of each of the samples studied in this work

Full size image

All studied granitoids are peraluminous, with an A/CNK index between 1.13 and 1.28. They are enriched in incompatible and mobile elements, especially Rb and K. REEs are fractionated and have high (La/Lu)n ratios (between 28 and 75). They show negative anomalies for Ba, Sr, Eu, and especially Nb and Ti. Whole-rock oxygen isotope analyses yielded O\(\delta^{ 1 8} \cdot\) values between 10.3 and 12.9‰. These geochemical features are all characteristic for S-type granitoids with crustal affinity (e.g., López-Plaza et al. 2008; López-Moro et al. 2012).

Cordierite-bearing granites from the Pelilla complex (Sample O-45)

The Pelilla complex (Fig. 3c) is a WNW-ESE trending elongated body mainly composed of K-feldspar-rich cordierite granites. A sinistral NE-trending shear zone transecting the pluton controls both the pluton shape and the weak internal foliation. The cordierite granites contain (average of six samples): quartz (21.6 vol%), K-feldspar (37.8% vol%), plagioclase (21.8 vol%), muscovite (8.6 vol%), biotite (4.6 vol%), cordierite (4.5 vol%, may reach locally up to 30 vol%), and accessory sillimanite, andalusite, apatite, zircon, monazite, and opaque minerals. Their composition is monzogranitic, although some of the coarse-grained facies are classified as quartzsyenites. The large K-feldspar megacrysts up to 3 cm are equant and define an apparent isotropic fabric and panidiomorphic texture. Rapakivi texture is not uncommon and consists of mantled feldspars with a subhedral core of K-feldspar rimmed by oligoclase. Anorthite composition of zoned plagioclase crystals ranges from An25 to albite, although oscillatory zoning may also be present in the core. Sillimanite-rich enclaves are surrounded by cordierite that together with the occurrence of andalusite indicate relatively low-pressure conditions during granite emplacement.

Porphyritic two-mica ± sillimanite granites from the Pelazas pluton (Sample 99-49)

Porphyritic two-mica ± sillimanite granite and medium-grained two-mica granite constitute the Pelazas pluton (Fig. 3a, López-Plaza and López-Moro 2008). This pluton was affected by the Pelazas Shear Zone (López-Plaza 1982), a N70ºE trending sinistral shear zone (Fig. 2). Porphyritic two-mica ± sillimanite granites contain abundant pegmatites as well as irregular and elongated K-feldspar-rich pockets at the top of the pluton (dated roof facies) or in sheet-like structures (layered facies). Layered facies may show modal graded layers in the northernmost part of the Pelazas Pluton, where their rhythmic nature corresponds to the B-type of Naslund and McBirney (1996). The layered granite of the Pelazas pluton typically contains swarms of biotite schlieren with gradual transitions from the modal graded layered zone to the schlieren zone. Schlieren are composed of thin, wavy streaks of biotite, sillimanite, and quartz, as well as secondary muscovite. They may be equivalent to the “wispy or streaky layering” of Barrière (1981). The layers of the modal graded layered zone, ranging between 10 cm to more than 1 m in thickness, typically are outward-dipping. The bottom layer commonly shows a thin, almost continuous, biotite-rich lamina. Centimeter-scale biotite- and sillimanite-rich xenoliths are found mainly towards the bottom. Modal grading is not regular, with a general increase of biotite, plagioclase, sillimanite, muscovite, quartz, and accessory minerals and a decrease of K-feldspar towards the bottom. The composition of plagioclase rims changes slightly from the top to the bottom of these layered structures, with values of An13 at the top down to An04 at the bottom (López-Plaza and López-Moro 2008). Two types of muscovite occur: subhedral muscovite commonly associated with biotite and late muscovite, which may be simplectitic in association with sillimanite or palm-like.

Coarse-grained two-mica granites and fine-grained biotite ± muscovite granites of the Ledesma pluton [Samples 04-17(coarse) and 04-18 (fine-grained)]

Coarse-grained two-mica granite (CTMG) and a fine-grained biotite ± muscovite granite (FBG) are the two main granite types of the Ledesma pluton (Fig. 3d), an asymmetric drop-shaped pluton with two contrasting fabrics: a sub-vertical fabric in the external contacts and a widespread flat-lying fabric at the center of the pluton. The sub-vertical fabric is related to strike-slip deformation along the Juzbado-Penalva Shear Zone (Figs. 2, 3d) (307–308 Ma according to Valle Aguado et al. (2005) and Gutiérrez-Alonso et al. 2015). The flat-lying fabric seems to have a tectonic origin as foliation passes through enclaves and internal contacts of the plutons (López-Plaza and López-Moro 2008). The concentric foliation pattern may also suggest a certain effect of radial expansion (Fig. 3d). The FBG is slightly porphyritic, shows xenoliths of migmatites and augen gneiss, as well as xenocrysts of K-feldspar, the latter being more common towards the top of the body. Microgranular mafic enclaves of tonalitic and granodioritic composition are common in this granite type. The modal composition (average of four samples) is as follows: quartz (31 vol%), alkali feldspar (24 vol%), plagioclase (28 vol%), biotite (10 vol%), and muscovite (5 vol%) as essential minerals, and apatite, zircon, and monazite as accessories. The CTMG has an equigranular texture, although it may be locally slightly porphyritic or show schlieren. The modal composition (average of four samples) is as follows: quartz (35% vol%), alkali feldspar (32 vol%), plagioclase (19 vol%), biotite (6 vol%), and muscovite (7 vol%) as essential constituents, and apatite, zircon, monazite, sillimanite, and cordierite. Andalusite is relatively common, occurring in pegmatitic pockets as well as around sillimanite-rich enclaves within the granite. The biotite composition allows both granite types to be discriminated, with more aluminic biotite in the CTMG relative to those of FBG, which leads to their plotting in different fields (aluminopotassic and calc-alkaline, respectively) in the diagram of the Nachit et al. (1985). Although contrasting structural relations have been found (López-Plaza 1980; López-Plaza and López-Moro 2008) both granites are probably coeval.

Medium-grained muscovite ± biotite granite of the Barruecopardo tungsten mining district (Sample BARR)

This granite hosts one of the most important tungsten mineralizations in Spain. The intrusion is a nested pluton that displays discordant relationships with other granites to the north and with Ediacaran-Cambrian age metasediments in the andalusite zone to the south (Fig. 3d). The granite exhibits a weak foliation (NW–SE) that is recorded by scarce biotite. The pluton consists of a volumetrically dominant equigranular medium-grained muscovite ± biotite granite (MMG) and occasionally a finer-grained facies that has been interpreted as a border facies (Díez Montes et al. 1991). Both facies are compositionally and texturally similar. A swarm of centimeter-scale quartz-veins with scheelite ore crops out within the main facies. These dykes developed a greisenization or feldspatization in their salbands, and an important hydrothermal activity was dated using K–Ar in muscovite, yielding a Permian age (287 ± 5 Ma, Antona 1991). The main facies (MMG) exhibits a conspicuous “fly wing” texture with big crystals of biotite of 4–5 mm in size that stands out in a quartz-feldspathic mesostasis of 2–3 mm in size, where alkali feldspar phenocrysts up to 15 mm in size can also be found. Enclaves are rare and small, but decametric country-rock enclaves have been occasionally found in the open-pit of Coto Merladet mine. Petrographically, the main facies is a medium inequigranular granite consisting of quartz (35 wt%), alkali feldspar (17.5 wt%) with scarce perthites, plagioclase (25 wt%; An10-01), muscovite (15.5 wt%), and biotite (6 wt%) as their main constituents and minor amounts of apatite, zircon, monazite, rutile, and sillimanite, the latter either as needles or prismatic crystals, but always enclosed in muscovite. Quartz with undulose extinction and straight borders, plagioclase with curved twin planes and undulose extinction and micas showing curved cleavage planes and incipient mica fish structures are petrographic features that point to a deformation in ductile conditions according to Passchier and Trouw (2005). It should be noted that the main facies suffered a pervasive overprint of subsolidus recrystallization, with a feldespatization of plagioclase and an overall muscovitization that mainly affected the biotite.

Analytical procedures

Zircon and monazites were separated at the GeoForschungsZentrum Helmholtz-Zentrum Potsdam (Germany) and in the University of Salamanca (Spain). Approximately 3 kg of sample was crushed in a jaw crusher and sieved for the fraction 63–400 µm. Concentrates were obtained by using Wilfley table, Frantz isodynamic magnet separator, and heavy liquids (bromoform and diiodomethane). Inclusion-free, perfectly clear monazite crystals and zircon grains were selected from these concentrates by hand-picking under a binocular microscope.

Monazite fractions with a 205Pb–235U mixed tracer were dissolved overnight in concentrated H2SO4 at 220 °C on the hot plate. Pb and U were separated using a HBr–HCl and HCl–HNO3 ion-exchange chromatography procedure, respectively, and were loaded together with silica emitter on single Re-filaments, and were measured at 1200–1260 and 1350–1400 °C, respectively, on a Finnigan MAT262 multicollector mass-spectrometer using Faraday collectors and ion counting at Deutsches GeoForschungsZentrum. The U–Pb analytical data obtained for monazite are shown in Table 1 (Supplementary material).

Zircon crystals of all grain sizes and morphological types were selected, mounted in resin blocks and polished to half their thickness at the Museum für Mineralogie und Geologie (Senckenberg Naturhistorische Sammlungen Dresden). All grains were documented by back-scattered electron (BSE) and cathodoluminiscence (CL) images using a SEM coupled to a HONOLD CL-detector operating with a spotsize of 550 nm at 20 kV to study their internal structure and to select the best areas for laser ablation (Fig. 4). Zircon was analyzed for U, Th, and Pb isotopes by LA-ICP-MS, using a Thermo-Scientific Element 2 XR sector field ICP-MS coupled to a New Wave UP-193 Excimer Laser System. A teardrop-shaped, low volume laser cell by Ben Jähne (Dresden, Germany) was used to enable sequential sampling of heterogeneous grains (e.g., growth zones) during time resolved data acquisition. Each analysis consisted of approximately 15 s background acquisition followed by 35 s data acquisition, using a laser spot-size of 25 µm. Sixty zircon crystals were analyzed from each sample. A common-Pb correction based on the interference- and background-corrected 204Pb signal and a model Pb composition (Stacey and Kramers 1975) was carried out if necessary. Analyses with discordance higher than 10%, (calculated as the difference between the 206Pb/238U and the 207Pb/206Pb ages) were rejected. Raw data were corrected for background signal, common Pb, laser-induced elemental fractionation, instrumental mass discrimination, and time-dependent elemental fractionation of Pb/Th and Pb/U using an Excel® spreadsheet program developed by Axel Gerdes (Institute of Geosciences, Johann Wolfgang Goethe-University Frankfurt, Frankfurt am Main, Germany). Reported uncertainties were propagated by quadratic addition of the external reproducibility obtained from the standard zircon GJ-1 (~0.6 and 0.5–1% for the 207Pb/206Pb and 206Pb/238U, respectively) during individual analytical sessions and the within-run precision of each analyses. Concordia diagrams (2σ error ellipses) and concordia ages (95% confidence level) were produced using Isoplot/Ex 3.75 (Ludwig 2012). The 206Pb/238U ages were taken for interpretation and building probability density plots when appropriate. For further details on analytical protocol and data processing, see Frei and Gerdes (2009). The analytical results are shown in Tables 2–6 (supplementary material).

Fig. 4
figure 4

CL images of representative zircon grains dated in this study with ablation points and corresponding ages. aXX stands for the analysis number in each of the samples (Tables 2–6 in supplementary material)

Full size image

Results

Pelilla granite (sample O-45)

Analyzed zircon crystals are mostly prismatic, showing oscillatory zoning and occasionally inherited cores (Fig. 4). Sixty zircon grains were analyzed, of which 45 yield concordant ages (Table 2, supplementary material). A coherent (concordant and overlapping) group of the 30 younger analyses gives a concordia age of 318 ± 1 Ma, which is interpreted to represent the crystallization/intrusion age of this granite (Fig. 5). Fifteen zircon grains yield older (concordant) 206Pb/238U ages at ca. 350, 500, 550–600, 650, and 1600 Ma (Fig. 5). These older grains are inherited from the protoliths (see discussion below).

Fig. 5
figure 5

Concordia diagrams for a magmatic and b inherited/xenocrystic zircon crystals from sample O-45

Full size image

Pelazas pluton (sample 99-49)

Most of the analyzed zircon grains are rounded with low luminescence and cores (Fig. 4). Only 32 of the analyzed zircon crystals provide concordant ages (Table 3, supplementary material) of which the two youngest yield a concordia age of 325 ± 5 Ma (Fig. 6) that is interpreted to date the crystallization/intrusion age of this pluton. Both zircon crystals are prismatic and show overgrowths (Fig. 4). The other 30 concordant zircon grains are inherited: eleven of them have 206Pb/238U ages within the range of the Cambro-Ordovician magmatic event (“Ollo de Sapo”, see above) in Iberia; fifteen analyses yield ages corresponding to the Cadomian cycle (540–840 Ma, Cryogenian-Ediacaran); one is ca. 917 Ma (Tonian); and three have Paleoproterozoic and Archean cores (ca. 2200–2700 Ma).

Fig. 6
figure 6

Concordia diagrams for a magmatic and b, c inherited/xenocrystic zircon from sample 99-49

Full size image

Ledesma pluton (sample 04-17, coarse-grained facies)

Zircon crystals from this sample feature varied shapes, from prismatic to rounded, most of them display core-rim structures with oscillatory overgrowths and some crystals have non-luminescent rims (Fig. 4). Forty-nine zircon grains yield concordant ages (Table 4, supplementary material). A coherent group of the 16 younger analyses defines a concordia age of 318 ± 2 Ma (Fig. 7), which is interpreted to be the crystallization/intrusion age of this pluton. The inherited 206Pb/238U ages cluster around 340 Ma (5 zircon ages), ca. 500 Ma (19 zircon ages attributed to the “Ollo de Sapo” event), 600 Ma (5 zircon grains from the Cadomian cycle), two of Tonian age and an older Archean core.

Fig. 7
figure 7

Sample 04-17 has two types of magmatic zircon yielding different ages. Concordia diagram (a) shows magmatic zircon ages interpreted to date the crystallization age. Concordia diagram (b) shows magmatic zircon inherited from a magmatic event at ca. 340 Ma. Inherited/xenocrystic zircon data are shown in c. The bold (gray) ellipse in b represents the age of a single concordant monazite fraction dated by TIMS

Full size image

A single monazite fraction of this granite gave a concordant U–Pb age at 343.2 ± 2.7 Ma (Fig. 7; Table 1, supplementary material), which corresponds to the age of the youngest inherited zircon population.

Ledesma pluton (sample 04-18, fine-grained facies)

This sample was collected ca. 100 m SE of sample 04-17, at a fine-grained NS elongated body within the Ledesma pluton. Forty-six out of the 60 analyses give concordant ages (Table 5, Supplementary material). The age of crystallization is tightly constrained by a concordia age of 318 ± 2.5 Ma (identical to that of sample 04-17) obtained from a coherent group of analyses from 8 prismatic zircon crystals (Fig. 8) displaying magmatic oscillatory growth patterns or non-luminescent areas (Fig. 4). Zircon crystals with older concordant ages are inherited. Four zircon grains have 206Pb/238U ages ranging from 338 to 346 Ma yielding a pooled concordia age of 343 ± 2.5 Ma (Fig. 8). An older group of five zircon grains yields Devonian 206Pb/238U ages ranging from 400 to 436 Ma. The youngest four zircon crystals of this group yield a pooled concordia age of 405 ± 3.5 Ma. Uppermost Cambrian zircon crystals (10) with 206Pb/238U ages ranging from 489 to 500 Ma are attributed to the “Ollo de Sapo” event. Eighteen zircon grains of Cadomian age were also identified within the concordant inherited population (206Pb/238U ages of 552–781 Ma) and, finally, a Paleoproterozoic age (ca. 2000 Ma) was found in an inherited core.

Fig. 8
figure 8

Concordia diagrams for a magmatic zircon interpreted to date the crystallization age, b magmatic zircon from a ca. 343 Ma old magmatic event, and c older inherited/xenocrystic zircon crystals obtained from sample 04-18. The bold ellipse in a represents the age of a concordant monazite fraction dated by TIMS

Full size image

A single monazite fraction of this granite gave a concordant U–Pb age at 318 ± 1.4 Ma (Fig. 8; Table 1, Supplementary material), consistent with the crystallization age interpreted from zircon analyses.

Barruecopardo leucogranite (sample BARR)

The sample from the Barruecopardo leucogranite pluton (Pellitero et al. 1976; Arribas 1979; Antona et al. 1992; Sanderson et al. 2008) was collected in a part of the intrusion that hosts a tungsten mineralization that forms a steeply east-southeast dipping sheeted vein system. The veins are made up mainly of quartz, arsenopyrite, and scheelite. Forty-two out of the 60 analyses yielded concordant ages (Table 6, Supplementary material). A coherent group of eleven analyses yields a pooled concordia age of 324 ± 2 Ma (Fig. 9) that is considered to be representative of the crystallization/emplacement age of the Barruecopardo pluton. The majority of the inherited zircon grains (28) have ages within a narrow 206Pb/238U age span from 334 to 348 Ma and yield a concordia age of 344 ± 1 Ma (Fig. 9). Only three older inherited zircon grains were analyzed in this sample: one of them of Devonian age (206Pb/238U age of 390 Ma) and two of them of Ediacaran age (206Pb/238U age of 543 and 563 Ma) (Fig. 9).

Fig. 9
figure 9

Concordia diagrams for a magmatic zircon interpreted to date the crystallization age and b magmatic zircon related to a ca. 343 Ma old event obtained from sample BARR

Full size image

Discussion

The studied samples from the Tormes Dome have yielded intrusion/crystallization ages that cluster around ca. 320 Ma, which provide tight constraints on the age of the extensional activity of this sector of the Variscan belt in Iberia. The intrusion ages are in agreement with previous ages obtained in the Tormes Dome and other gneiss-cored domes in the surrounding areas (see above). Despite the presence of a younger voluminous magmatic event at ca. 310–295 Ma that is represented by nearby large plutons (Gutiérrez-Alonso et al. 2011 and references therein), there is no evidence for growth of new zircon during a thermal overprint or isotopic disturbance of the U–Pb system in the coherent group of 66 zircon ages (from all samples) considered to be of magmatic origin (i.e., crystallized from the anatectic melts that formed the studied plutons).

The variable proportion of magmatic and inherited zircon crystals in the various granite samples largely reflects whether inherited zircon grains derived from the protoliths were dissolved or preserved in the granitic melt. Dissolution of inherited zircon depends on Zr-saturation of the melt, which, in turn, depends on melt composition and melting temperature (e.g., Boehnke et al. 2013). A significant portion of inherited zircon may derive from xenolithic material and restitic schlieren incorporated into the melt. Xenolithic and restitic material may also contain monazite or xenotime, i.e., allow for the apparent inheritance of these minerals. Sample 0-45, which has the apparent highest portion of magmatic zircon (>60%), is derived from a feldspar-enriched cumulate in a cordierite-bearing granite (López-Plaza and López-Moro 2008), whereas sample 99-49, which has rare magmatic zircon crystals (apparently <5%), shows ubiquitous schlieren, layering, varied internal flow structures and centimeter-scale pelitic xenoliths indicative for entrainment of older material (López-Plaza and López-Moro 2008).

The 145 inherited xenocrystic zircon grains found in the studied anatectic granitoids provide data on the possible sources of the melts and the results also shed light on the crustal components of western Iberia. The ages of the inherited zircon grains define five distinct groups that are found in different proportions in the five studied samples (Fig. 10). From younger to older, these age groups are as follows:

Fig. 10
figure 10

Diagram depicting the relative proportions of magmatic and inherited/xenocrystic zircon grouped according to age. For details and explanation of the chosen intervals, see text

Full size image
  1. 1.

    Carboniferous zircon crystals (ca. 345 Ma). This group includes 40 zircon crystals with ages ranging from 334 ± 8 to 357 ± 8 Ma. Zircon crystals of this group are present in samples O-45 (Pelilla), 04-17, 04-18 (Ledesma) and are particularly abundant in sample BARR (Fig. 10). This age is not only restricted to inherited zircon but also has been found in monazite (344 Ma) from granite sample 04-17.

  2. 2.

    Devonian–Silurian zircon crystals. This group is represented by only six zircon grains with scattered ages between ca. 390 and 432 Ma. Zircon grains of this group were found in samples O-45, 04-18 and BARR (Fig. 10).

  3. 3.

    Upper Cambrian–Ordovician zircon crystals. Forty-three zircon grains were identified and their age ranges from ca. 450 to 511 Ma. This age group is present in all samples except in sample BARR.

  4. 4.

    Ediacaran–Cryogenian zircon grains (ca. 540–840 Ma). There are 44 zircon grains falling within this age range with a peak at ca. 565 Ma (Figs. 10, 11). This age group is present in all samples except in sample BARR.

    Fig. 11
    figure 11

    Probability density plot of all concordant zircon ages showing major peaks at ca. 320 Ma, ca. 340 Ma, and ca. 480 Ma

    Full size image
  5. 5.

    Older zircon grains (900–2700 Ma). Eleven zircon crystals yielded ages older than 900 Ma. Different proportions of these older zircon crystals appear in all samples except in sample BARR.

The zircon xenocrysts and inherited cores represent a first-order proxy for the investigated anatectic granites. It should be noted, however, that some protoliths already have inherited zircon populations that may contribute to the age spectrum of the anatectic rocks studied in this work. Based on Rb–Sr, Sm–Nd, and oxygen isotope data, partial melting modeling, and experimental and thermodynamic approaches, the main sources for the anatectic granites of the Tormes Dome have been interpreted to be the Ediacaran to Early Cambrian “Schist-Greywacke sedimentary Complex” and/or the uppermost Cambrian to Middle Ordovician “Ollo de Sapo” magmatic rocks (Holtz and Barbey 1991; Castro et al. 2000; Bea et al. 2003; López-Plaza et al. 2008; García-Arias and Corretgé 2010; López-Moro et al. 2012). Our new data on zircon xenocrysts/cores demonstrate that these Ediacaran to Ordovician rocks represent potential source rocks for the anatectic granites, except for sample BARR that seems not to contain “Ollo de Sapo,” Ediacaran–Cryogenian or older zircon crystals. Furthermore, the abundance of 340-350 Ma old inherited zircon indicates that metamorphic rocks, migmatites, and granites of this age group also significantly contributed to the anatectic granites.

Significance of the ca. 400 Ma old zircon ages

Although there is evidence of ca. 400 Ma magmatic rocks described mostly in the easternmost part of the Variscan belt abundant zircon ages around 400 Ma in NW Iberia are only found in the so called Upper Units and the ophiolite-like rocks present in the allochthonous complexes that are interpreted to represent the closure of the Rheic ocean that caused the Variscan orogeny (e.g., Fernández-Suárez et al. 2007; Sánchez Martínez et al. 2011; Kroner and Romer 2013; Arenas et al. 2014; Mateus et al. 2016). Such a source for this group of zircon ages is not permissible as the rocks that contain them were structurally above the rocks that form the Tormes Dome. In addition, rocks from the allochthonous complexes were not likely involved in the production of anatectic melts during the subsequent extension as they were located in the hanging wall of the main detachments (Rubio Pascual et al. 2013 and references therein). There are, however, some scarce volcanic rocks dated at ca. 400 Ma in northwestern and central Iberia (Loeschke 1983; Hall et al. 1997; Gutiérrez-Alonso et al. 2008) whose origin is interpreted as related to the extension of the northern Gondwana passive margin (Pin et al. 2006) caused by the ridge subduction in the northern flank of the Rheic ocean (Gutiérrez-Alonso et al. 2008). A propagating failed rift departing from the newly formed Paleotethys ocean has also been proposed as a cause for this magmatic event (Armendáriz et al. 2008). This event could have been responsible for the emplacement of (minor?) magmatic bodies that may have been later involved in anatectic melting, thus making a small though statistically significant contribution to the inherited population of the studied granitoids.

The presence of the small but significant population of inherited zircon with ages around 400 Ma is an example of how inherited zircon populations in igneous rocks can point to the presence of events poorly (or not at all) constrained by geological studies of exposed rocks. These events may be minor in many cases but they may be even of global significance as could be the case in our study if the ca. 400 Ma event is indeed the far field effect of the subduction of the Rheic Ocean ridge under Laurussia. It is perhaps pertinent to mention that the largest mercury deposit in the world (Almadén in Central Spain) could be related to this event (Hall et al. 1997; Hernández et al. 1999) as could be the intrusion of diabase dyke swarms in the CIZ (López-Moro et al. 2007).

Significance of the ca. 340 Ma old zircon ages

The abundant (exceptionally so in sample BARR) ca. 340 zircon population lacks a putative source in the neighboring areas, but the large proportion of zircon xenocrysts/cores of this age found in the TD granites begs the question of whether there was a major and hitherto unrecognized magmatic/migmatitic event of such age in the studied area. Magmatic rocks of such age have only been found in northern and central Iberia in the Cangas do Morrazo Vaugnerites (U–Pb in zircon, Gallastegui 2005); the Beariz, Chantada and Carballino granites (U–Pb in monazite cores, Gloaguen 2006; Gloaguen et al. 2006) and volcanic rocks in the southern CIZ (Sierra de San Pedro, Bascones Alvira et al. 1982; Soldevila Bartolí 1992). Igneous rocks of that age have also been reported in Variscan realms that can be correlated with the Central Iberian Zone such as the Cantabrian Zone (Merino-Tomé et al. in press); the Pyrenees (Mezger and Gerdes 2016); the French Massif Central (Galán et al. 1997; Ledru et al. 2001); Corsica (Rossi and Cocherie 1991; Li et al. 2014); see also compilation in von Raumer et al. (2014); and southwestern Iberia (Dallmeyer et al. 1995; Montero et al. 2000; Ordóñez Casado 1998, Ordóñez Casado et al. 2008; Romeo et al. 2006; Jesus et al. 2007; Pin et al. 2008; Pereira et al. 2009, 2015; Braid et al. 2012; Dupuis et al. 2014; Gladney et al. 2014; Cambeses et al. 2015).

In the light of the large number of zircon grains of ca. 340 Ma found in the studied rocks (Figs. 10, 11), and given the widespread presence of magmatic rocks of ca. 340 Ma in the Variscan realm (although there is no consensus on their genesis at orogenic belt scale), it is reasonable to postulate that rocks of such age may indeed be part of the “basement” of the CIZ in general and of the Tormes Dome in particular. Despite the fact that a small proportion of zircon grains of this age are found in the syn-orogenic sediments of the CIZ (Martínez Catalán et al. 2016) and of the Cantabrian Zone, which is the orogenic foreland of the CIZ (Pastor-Galán et al. 2013), magmatic rocks of this age have not yet been described in the study area and/or may have not been yet exhumed and eroded to the surface.

The large variation in the proportions of inherited zircon ages in the studied samples (Fig. 10) raises the issue of whether these variations reflect different volumetric involvements of the crustal sources for each of the studied plutons. As shown in Fig. 11, there are five main zircon populations represented in the studied samples, which are interpreted as representative of the zircon content of the main source rocks. Figure 10 highlights the large variability in the proportions of the different populations from one sample to another which, taken at face value, could be attributed to the presence of a very heterogeneous crustal basement as source for the granitoid melts. In addition to the magmatic zircon ages, the most abundant population corresponds to the ca. 470–490 Ma ones followed by the Cadomian (Ediacaran-Early Cambrian) ones, whose likely sources were the “Ollo de Sapo” magmatic rocks and the Ediacaran “Schist-Greywacke Complex,” respectively. The ca. 400 Ma zircon crystals are only present (in low proportions) in three of the studied samples. Despite their apparent scarcity, we consider them to be potentially representative of an unknown source in the basement as discussed above. The ca. 340 Ma population is absent in one of the samples (99-49), scarce (<10%) in three of them (O-45, 04-17 and 04-18) and dominant in the BARR sample (ca. 70%), which suggests that this source is present, albeit very irregularly distributed in the basement. Such hypothesis is compatible with the presence of plutons of unknown size with intrusion ages around 340 Ma in the CIZ basement and is also in agreement with the lack of extensive Cambro-Ordovician igneous rocks within the upper unit of the Tormes Dome.

Concluding remarks

In addition to the zircon ages obtained and fully described and discussed previously, in a more general context, this paper makes a contribution to some relevant aspects of the significance of granitoids in the study of crustal growth and evolution. The following issues are in our opinion of interest:

  1. 1.

    This paper contributes yet another example to the “paradox” of why crustal granitoids are essentially very similar to one another “on the surface” but display such a wide range of mineralogical, geochemical and isotopic (both stable and radiogenic) features and thus make classification schemes (such as I-S) a rather weak tool (Chappell and Stephens 1988; Castro et al. 1991, 1999; Roberts and Clemens 1993; Aguilar and Villaseca 2010). To some extent, the data presented herein illustrate that much of the multifariousness of granites (Pitcher 1993) is directly inherited from the complexity of their known or unknown sources.

  2. 2.

    This study shows that crustal recycling/cannibalization may often happen at a fast pace in orogenic scenarios with only short lapses of quiescence; in our case study, it seems plausible that a “crustal layer” of ca. 340 Ma granitoids/migmatites was recycled, partially or totally, only 15–20 My after its emplacement.

  3. 3.

    An open question: The most volumetrically significant granitoid suite in the Variscan belt of North and Central Iberia was produced between ca. 305 and 290 Ma, after the main orogenic events (collision and extensional collapse) and its genesis has been linked to lithospheric delamination triggered by oroclinal bending of the mountain belt (Gutiérrez-Alonso et al. 2011). In an imaginary alternative scenario in which oroclinal bending did not occur, would the ca. 320 Myr old crustal leucogranitoids (like those studied herein) have been the last manifestation of Variscan magmatism in Iberia?