Zircon and allanite U-Pb ID-TIMS ages of vaugnerites from the Calzadilla pluton, Salamanca (Spain): dating mantle-derived magmatism and post-magmatic subsolidus overprint

Vaugnerite was first described by Fournet (1861) in the XIX^th century and named after the Vaugneray region, close to Lyon (France). These rocks are dioritic, with a striking (vaugneritic) texture consisting of amphibole and/ or decussate, very large biotite crystals protruding in a matrix of plagioclase, amphibole, biotite, alkali feldspar, ± quartz, and ± clinopyroxene as essential constituents, with accessory apatite, zircon, allanite, and titanite. This texture gives the rock a characteristic apparent isotropic fabric.

Typically, vaugnerites form small stocks or occur as enclaves in calc-alkaline and peraluminous leucogranites that are aligned with shoshonitic series (García de los Ríos, 1981; Sabatier, 1991; López-Moro and López-Plaza, 2001). Vaugnerites are not very important in terms of volume, but occur scattered throughout the entire Variscan Orogenic Belt of Iberia (e.g.Gil Ibarguchi, 1982; Sabatier, 1991; López-Moro, 2000; Castro et al., 2003; Gallastegui, 2005; Scarrow et al., 2009; Vegas et al., 2011; Molina et al., 2012; von Raumer et al., 2014).

The particular mineralogy of vaugnerites reflects the high K, LILE, and LREE contents of these rocks at relatively low silica content (intermediate rocks). This particular geochemical signature of vaugnerites and other K-Mg-enriched rocks is obtained from the partial melting of a metasomatically veined mantle, whereas the MgO, Cr, and Mg# as typical mantle signatures reflect that the mantle material and the high contents of K, LILE, and LREE are derived from the metasomes (e.g.Foley, 1992; Prelević et al., 2010; 2015). The metasomes represent a material that was originally derived from the subducted slab and its sedimentary cover, and reacted with the mantle above the subducting plate (e.g.Prelević et al., 2012; 2015; Soder et al., 2016). The metasomes melted when the thermal structure of the lithospheric mantle changed (e.g.Prelević et al., 2010; 2015) and, therefore, the formation of vaugnerite melts may be related to particular tectonothermal events (von Raumer et al., 2014) as, for instance, post-orogenic extension (e.g.Janousek and Holub, 2007; Prelević and Foley, 2007).

In the Iberian Massif, vaugnerites are present in Finisterre (Gil Ibarguchi, 1982), the Bayo-Vigo zone (Gallastegui, 2005), Sanabria region (Chacón Muñoz, 2005; Vegas et al., 2011), and the Tormes Dome (García de los Ríos 1981; López-Moro, 2000). Among the vaugnerites of the Central Iberian Zone only those from Bayo-Vigo have been dated (319.6 ± 0.7Ma; Rodríguez et al., 2007). Although the Tormes Dome is the area in the Iberian Massif with the most important occurrences of vaugnerite, i.e. there are 6 plutons located in the Sayago region (SW Zamora) and one pluton in the Campo Charro region (Central Northern Salamanca) (Fig. 1); the age of these rocks is only constrained by contact relations. In order to understand the Variscan plutonism in this area, it is important to understand the temporal and genetic relation between the vaugnerites and the surrounding plutonic rocks and, ultimately, the role of the mantle as heat source for crustal melting and as the vaugnerite source. Therefore, in this study, we present U-Pb ID-TIMS zircon and allanite ages to constrain the temporal relationship between vaugnerites and surrounding granitoids.

The studied area is part of the Iberian Massif, a large area with Palaeozoic and Ediacaran rocks in western Iberia that have been variably overprinted during the Variscan Orogeny (Fig. 1). The Tormes Dome (TD) represents the southernmost prolongation of the “plutono-metamorphic belts” of the NW Iberian Massif (Fig. 1). The present-day geometry resulted from the superposition of late Variscan structures that controlled the location of zones of extensive partial melting and multiple granite intrusion. The TD composite batholith includes: i) syn-kinematic two-mica granites, including equigranular and porphyritic granites (Fig. 1; López-Plaza and López-Moro, 2004), which are abundant throughout the NW Iberian Massif and known as ‘‘Older two-mica Granites’’ or ‘‘Peraluminous Suite’’ (Capdevila et al., 1973; Castro et al., 2002), and ii) pre/syn-kinematic biotite granitoids and related rocks, with three different associations: IIa) a dioritic-tonalitic association of high-K calc-alkaline affinity; IIb) a granodioritic association, with granodiorites and monzogranites (“Older Granodiorites”), often including spatially, but not genetically, associated vaugnerites and hybrid rocks; and IIIc) a monzonitic association with monzogabbros, monzonites, quartz-monzonites, and scarce monzogranites.

FIGURE 1
Simplified geological map of the Tormes Dome (López-Plaza and López-Moro, 2004), showing the major lithological units, location of vaugnerite plutons and available geochronological data of the Variscan rocks discussed in the text. Augen gneisses ages are after Talavera et al. (2013) and granitoid ages after Valverde Vaquero et al. (2007), López-Moro et al. (2017), and Ferreira et al. (2000).

The metamorphic rocks in the TD consist of, from bottom to top:

• Migmatized felsic gneisses and migmatized Ediacaran-Lower Cambrian sandstones and pelites (the Lower Unit in Escuder Viruete et al., 1994). The felsic gneisses comprise garnet-bearing, fine-grained gneisses and Early Ordovician augen gneisses (López-Moro, 2000; Bea et al., 2006). The migmatized Ediacaran-Lower Cambrian rocks have been partially metamorphosed into garnet-cordierite-biotite-sillimanite paragneisses (Gil Ibarguchi and Martínez, 1982).

• Lower Cambrian slates and schists with a low-grade Variscan metamorphic overprint, also known as the Upper Unit in Escuder Viruete et al. (1994). These materials represent the highest structural level of the dome. These lithologies are overlain by Ordovician-Silurian platform sediments.

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. D₁ deformation is closely related to Barrovian-type metamorphism at upper amphibolite facies conditions (M1; 700-740ºC and 800-900MPa, Escuder Viruete et al., 2000) that peaked at ca. 332Ma (Valverde-Vaquero et al., 2007). Early Variscan compressional structures were variably overprinted during the D2 extensional event, which resulted in a nearly isothermal decompression to a pressure of 300MPa, giving rise to a low-pressure/high-temperature paragenesis (M2), penetrative subhorizontal tectonic foliation, extensive migmatization and anatexis (Escuder Viruete et al., 1994; 2000). The migmatites yielded ages at 325-320Ma (Valverde-Vaquero et al., 2007). The third deformational phase (D3) was the consequence of NNESSW compression (Iglesias and Choukroune, 1980; López Plaza, 1982), which resulted in i) subvertical to moderately NE-vergent synforms and antiforms, and ii) strike-slip shear zones, being the sinistral Juzbado-Penalva Shear Zone (Fig. 1) the most prominent of them. Voluminous syn-D3 granitoids yield ages at 316–310Ma (Valverde-Vaquero et al., 2007; López-Moro et al., 2012), whereas sinistral movements along the Juzbado-Penalva Shear Zone and the Villalcampo Shear Zone have been dated at 309Ma and 306 ± 3Ma, respectively (using the ⁴⁰Ar/³⁹Ar method, Gutiérrez-Alonso et al., 2015).

The vaugneritic Calzadilla pluton

The vaugneritic Calzadilla pluton is a small NE-SW elongated pluton of a length of 2.5km and a width of around 600m (López-Plaza et al., 2007; Fig. 2). The pluton is surrounded by two-mica leucogranites that locally crosscut the external contact of the pluton, although the contacts between these rocks generally are concordant. Based on mineral association, texture, and grain size distribution, different facies of vaugnerites s.s. and associated intermediate rocks have been distinguished (Fig. 3) (see González Sánchez, 2016): a) fine-grained intermediate rocks without vaugnerite texture; b) mesocratic biotite - amphibole vaugnerite; c) leucocratic biotite - amphibole vaugnerite, and d) leucocratic biotite ± amphibole vaugnerite. Fine-grained facies surrounds vaugnerites s.s. to the NE, SW and SE and locally also to the NW (Fig. 2). Vaugnerites s.s. display zoning at pluton scale, consisting of different compositional bands close to the NE-SW direction. The zoning ranges from the least evolved mesocratic facies to the East-Southeast, to the most evolved facies to the West, largely reflecting the depletion in amphibole (Fig. 2). The leucocratic biotite ± amphibole vaugnerite is the most abundant facies, and the mesocratic rocks are the least abundant facies and tend to occur in meter-scale boulders (Fig. 4). The peripheral finegrained facies or border facies displays a marked strain anisotropy parallel to the exo-contact. Both foliation and exocontact show a centripetal pattern, with different dip directions in the South and the North of the pluton, reflecting a lopolith-like body (Fig. 2) In contrast, the geometry of the vaugnerite s.s. facies displays an internal pattern, which appears to be, to a certain extent, independent of the border facies, showing an upwards magmatic differentiation, as it can be inferred by the progressive decrease in amphibole (Fig. 2). Both types of facies, despite being presumably coeval, manifest two different prevailing effects: the early magmatic pulse of the border facies absorbed the deformational strain at the floor of the pluton, whereas the subsequent internal vaugneritic facies accommodated the magmatic evolution inside the pluton. Consistently, this pattern seems to account for a two-stage construction of the lopolith-like body with a steep feeder conduit located at the SE part of the pluton (Fig. 2).

FIGURE 2
Geological map of the Calzadilla Pluton (above) and interpretative cross section (below, modified from González Sánchez, 2016). The location of the sample used for dating is marked by a star (UTM coordinates, X: 747.913; Y: 4550.010; Zone 29, Datum: ETRS89).

FIGURE 3
A-C) Hand specimens of vaugnerites s.s. and associated rocks. A) Leucocratic biotite ± amphibole vaugnerite; B) leucocratic biotite - amphibole vaugnerite; C) mesocratic biotite - amphibole vaugnerite; D) fine-grained intermediate facies without vaugnerite texture (border facies).

FIGURE 4
Outcrop photograph showing the characteristic boulders of leucocratic biotite - amphibole vaugnerite at the sampling site. The picture was taken in an E-W direction and looking eastwards.

Vaugnerites from the Calzadilla pluton (in Ledesma pluton) and other vaugnerites from the Tormes Dome consist of amphibole, biotite, plagioclase, quartz and interstitial K-feldspar as major components, whereas apatite is the most characteristic accessory mineral (López-Moro, 2000). Less common accessory minerals are zircon, allanite, monazite, titanite, epidote, magnetite-hematite, chalcopyrite and pyrrhotite. The most distinctive feature of the vaugnerites is their decussated texture, defined by large crystals of biotite and amphibole (Fig. 3A; B; C).

Plagioclase is the most abundant mineral, ranging between 31 and 41% in volume. It forms up to 4mm large crystals, corresponding texturally to the “complex plagioclase phenocrysts” (A-Type) of Sabatier (1991). There is a large difference in the anorthite content between the partially resorbed cores (An_67-55) and the non-resorbed part of these phenocrysts. The composition of the latter plagioclase shows a relatively small variation (An_35-25), being similar to the one of small (<600μm) plagioclase phenocrysts and abundant plagioclase crystals in the groundmass, which suggests simultaneous crystallization. Rims and matrix plagioclase correspond to the B-Type of Sabatier (1991).

The amphibole occurs as isolated crystals of 100400 microns, sometimes hosting small crystals of clinopyroxene, or in millimeter polycrystalline clots consisting of amphibole, or amphibole, sulphides and biotite (López-Moro and López-Plaza, 2004a). Amphibole modal content ranges between 0.1% (in leucocratic biotite ± amphibole vaugnerites) and 19.7% (in mesocratic biotite – amphibole vaugnerites, González Sánchez, 2016). Compositionally, vaugnerite s.s. amphiboles are magnesiohornblende and, occasionally, actinolite (López-Moro and López-Plaza, 2004a).

There are three types of biotite, which is rather abundant as phenocryst and in the matrix (15-29 vol.%): a) large poikilitic, up to 5mm crystals, hosting apatite, plagioclase, clinopyroxene, amphibole and, locally, allanite; b) small tabular crystals (100-200μm), associated with amphibolic clots; and c) very small rounded crystals (≈10μm), hosted by plagioclase (López-Moro, 2000).

K-feldspar occurs either as an interstitial phase or as a component of up to 4mm-large ocelli. Its abundance varies between 5.7 and 17.1 vol.%, from dark to light rocks, respectively. K-feldspar is dominantly orthoclase (rarely adularia), as indicated by its monoclinic symmetry and 2Vx angle values of between 52° and 74° (López-Moro and López-Plaza, 1998).

Clinopyroxene typically forms euhedral or subhedral 400-800μm-sized crystals. It is relatively uncommon (0.682 vol.%) and generally absent in the most differentiated samples. Present as diopside (En37-38Fs13.9-15.6Wo47-48.7; López-Moro, 2000), it is similar to the clinopyroxene from the vaugnerites of the French Central Massif (Michon, 1987; Sabatier, 1991).

Allanite is a common accessory mineral in the various vaugnerite types. Its modal content is always <1%. It commonly exhibits crystal zoning, as well as a conspicuous alteration that is also recorded in biotite crystals with a marked chloritization (Fig. 5). Apatite is always present as fluorapatite. In melanocratic vaugnerites, it forms small crystals, whereas in leucocratic vaugnerite it occurs as larger crystals, typically rounded. Titanite is irregularly distributed and may locally reach high amounts (>1% in vol.), particularly in association with mafic minerals, such as amphibole and biotite. A local enrichment may reflect a secondary origin of titanite resulting from the redistribution of Ti from mafic minerals. Ilmenite is the major Ti-phase in amphibole-free quartzmonzonite, where it is abundant and occurs as inclusions in biotite and, less commonly, in the groundmass. Zircon usually occurs in euhedral grains (Fig. 5), ranging between 250 to 100 microns and, in most cases, is hosted in large biotite crystals. Nevertheless, the largest crystals are presented together with framework silicates.

FIGURE 5
Microphotographs of dated minerals. A) Allanite with alteration in plane light and B) crossed polars; C) mineral concentrate with an euhedral zircon grain and abundant apatite crystals (planeline view). Mineral abbreviations are after Whitney and Evans (2010).

Zircon and allanite concentrates were obtained from a leucocratic biotite - amphibole vaugnerite (Figs. 2 and 4) by using the Wilfley table, Frantz isodynamic magnet separator, and heavy liquids (bromoform and methylene iodide). Inclusion-free, perfectly clear zircon crystals were selected under the binocular microscope. Before dissolution, zircon was rinsed in warm 7N HNO₃ to remove U attached to the mineral surface, and cleaned in warm H₂O and acetone. Zircon and a ²⁰⁵Pb-²³⁵U mixed tracer were weighted into small teflon vials. The spike solution was dried before the addition of concentrated HF. The vials were put in a Parr autoclave and the zircon was dissolved at 210°C for four days. After drying the samples, zircon was redissolved in 6N HCl in the autoclave at 210°C overnight. After the samples had been dried again and taken up in 3N HCl, Pb and U were separated using the ion-exchange chromatography procedure described by Krogh (1973).

Allanite and a ²⁰⁵Pb-²³⁵U mixed tracer were weighted in screw-top beakers and dissolved using concentrated HF overnight on a hot plate at 140-160°C. Lead and U were separated using HCl-HBr ion-exchange chromatography and U was cleaned using the HNO₃-HCl ion-exchange chromatography procedure described in Baumgartner et al. (2006).

Pb and U were loaded together on single Re-filaments using a silica-gel emitter and H₃PO₄ (Gerstenberger and Haase, 1997), and measured at 1200°-1260°C and 13501400°C, respectively, on a Finnigan MAT262 multicollector mass-spectrometer using Faraday collectors and ion counting. Geochronological data plotting were performed using the Isoplot program of Ludwig (2000).

Six multi-grain zircon samples were analyzed. Care was taken to select only clear crystals, free of inclusions and fractures. Zircons from these fractions show a broad range of U contents from 213 to 1,236ppm (Table 1). Five fractions are concordant and one is slightly discordant (Fig. 6). Four fractions overlap within the error and define a weighted ²⁰⁶Pb/²³⁸U age of 318.1 ± 1.4Ma (2σ; Fig. 6). The remaining two samples are shifted to slightly lower ²⁰⁶Pb/²³⁸U ages and were not used for age calculation. The slight offset of the two samples to a younger age is interpreted as due to a minor Pb-loss. As the dated zircon does not show any indication of inheritance, the age of 318.1 ± 1.4Ma is interpreted as the age of the vaugnerite emplacement.

TABLE 1
U-Pb analytical data of zircons from Calzadilla vaugnerite pluton (leucocratic biotite - amphibole vaugnerite)

^a Lead isotope ratios corrected for fractionation (0.1% /a.m.u.).^b Lead corrected for fractionation, blank, tracer contribution, and initial lead with ²⁰⁶Pb/²⁰⁴Pb = 17.6±0.3, ²⁰⁷Pb/²⁰⁴Pb = 15.55±0.05, ²⁰⁸Pb/²⁰⁴Pb = 38.1±0.3. During the measurement period; total blanks were less than 15pg for lead and less th an 1 pg for uranium. Errors on the isotop ic ratios (values in brackets) at 2-sigma level (%).^c Apparent ages were calculated using the constants recommended by IUGS (Steiger and Jäger 1977).

FIGURE 6
Concordia diagram for zircons from the Calzadilla vaugnerite pluton (amphibole-bearing biotite vaugnerite facies). Fractions represented as dashed ellipses were not used for age calculation. The inset shows the ²⁰⁶Pb/²³⁸U ages of the various fractions. Grey fractions were used to calculate weighted average ²⁰⁶Pb/²³⁸U age. MSWD = mean square of weighted deviates.

The analyzed allanite samples have markedly lower U content (130 to 186ppm) and markedly higher Pb content (113 to 198ppm) than zircon (compare Table 2 with Table 1). Most of the Pb in allanite is common Pb. Therefore, the measured ²⁰⁶Pb/²⁰⁴Pb and ²⁰⁷Pb/²⁰⁴Pb ratios are not very radiogenic (²⁰⁶Pb/²⁰⁴Pb ranges from 23.07 to 44.12; Table 2), and the calculated ²⁰⁶Pb/²³⁸U and ²⁰⁷Pb/²³⁵U ratios have large uncertainties, in part due to uncertainties in the common Pb composition, and also due to the limited formation of radiogenic Pb. Therefore, the allanite samples are not shown in a concordia diagram. Instead, they are shown in the ²⁰⁶Pb/²⁰⁴Pb vs.²³⁸U/²⁰⁴Pb diagram (Fig. 7). The allanite samples scatter about a reference line of 275Ma, i.e. they all fall below the reference line of 318Ma, which corresponds to the U-Pb zircon age and represents the best age estimate for the emplacement of the vaugnerite. In the ²⁰⁶Pb/²⁰⁴Pb – ²⁰⁷Pb/²⁰⁴Pb diagram (not shown), the allanite samples fall on a straight line with some excess scatter. Because of their high Th/U ratio (Table 2), the various allanite samples may show excess of ²⁰⁶Pb. Excess of Pb originates from ²³⁰Th, which is part of the ²³⁸U decay series. Minerals that strongly favor Th over U incorporate significantly more ²³⁰Th than they would if ²³⁰Th and ²³⁸U were in activity equilibrium. The surplus of ²³⁰Th eventually results in an excess of ²⁰⁶Pb (Schärer, 1984). Correction of excess of Pb, however, is too insignificant to remove the scatter in the ²⁰⁶Pb/²⁰⁴Pb vs.²³⁸U/²⁰⁴Pb diagram, and would reduce the age rather than increase it. As textural relations demonstrate that allanite is a magmatic phase, the U-Pb systematics of allanite implies that it has been disturbed by a younger event, which may have resulted in partial or complete resetting of the U-Pb system of allanite.

TABLE 2
U-Pb analytical data of allanite from Calzadilla vaugnerite pluton (leucocratic biotite - amphibole vaugnerite)

^a Concentrations determined by isotope dilution using a mixed ²⁰⁵Pb-²³⁵U tracer.^b Lead isotope ratios corrected for isotopic fractionation (0.1% / a.m.u.), blank and tracer contribution. Analytical uncertainties at 2-sigma level are better than 0.1% for Pb isotope ratios, and 1% for ²³⁸U/²⁰⁴Pb.^c Calculated for an age of 275Ma using ²⁰⁸Pb rad/²⁰⁶Pb rad (estimated using measured ²⁰⁶Pb/²⁰⁴Pb and ²⁰⁸Pb/²⁰⁴Pb, corrected for common lead with ²⁰⁶Pb/²⁰⁴Pb = 17.6 and ²⁰⁸Pb/ ²⁰⁴Pb = 38.1).

FIGURE 7
²⁰⁶Pb/²⁰⁴Pb - ²³⁸U/²⁰⁴Pb diagram showing allanite samples scattering around a reference line of 275Ma. Line of 318Ma (zircon age) is shown as a reference.

U-Pb zircon ID-TIMS ages of vaugnerites from the Iberian Massif overlap with zircon and monazite U-Pb of granitoids and anatectic rocks in the same area, which implies that mantle melting (vaugnerites) and crustal melting (granites and anatectic rocks) are coeval. The advective heat input from the mantle, i.e. the heat carried by the vaugnerites, ultrapotassic and rocks of shoshoniticaffinity, is insufficient to account for the large-scale crustal melting. Instead, the pervasive anatexis of the area is related to the extensional regime that prevailed at 320-318Ma. In such a scenario, the quasi-isothermal decompression triggers anhydrous partial melting by mica-dehydration (crustal melts) in the crust and preferential melting of metasomatic domains in the upwelling lithospheric mantle (vaugnerites). Thus, even though vaugnerites, granites and anatectic melts are derived from different reservoirs and may be petrogenetically entirely unrelated, they are coeval as the same process (large-scale crustal extension) controlled the production of the melts.

The U-Pb age of allanite is distinctly younger than the emplacement of the vaugnerite, implying that the U-Pb system of allanite may have been disturbed by subsolidus overprint during a later event at a lower temperature and pressure. The disturbance of the U-Pb system of allanite may have occurred at ca. 275Ma (Permian), i.e. a period characterized by an important hydrothermal activity in the Central Iberian Zone, with the formation of tungsten, gold deposits and episyenites.

This work has been funded under the DGCYT project “Geochemical, tectonic and experimental approach of the crustal recycling processes and the mantle-crust interaction: Genesis and emplacement of the granitoids from the Tormes Dome” (CGL2004-06808-C04-04). F.J. López-Moro thanks Dr. F. Lucassen and Dr. J. Glodny for their assistance at the GeoForschungsZentrum Potsdam.