Petrogenesis of ultramafic and mafic rock series from the North Pyrenean Castillon massif (Ariège, France): evidence of magma–crust interactions in the Ordovician?

The Castillon massif, in the northern Pyrenees, features a complex of decimetric to metric ultramafic and mafic layers emplaced within metasedimentary series (from the bottom to the top: garnet, sillimanite, and kyanite-bearing gneisses and sillimanite + cordierite-bearing gneisses). Ultramafic and mafic layers and metasediments have been deformed and metamorphosed under granulitic facies conditions during the Hercynian orogenesis. The mineralogical, petrologic, and geochemical characteristics of the studied samples allow us to define two distinct series: 1) a pyroxene-bearing magmatic series (UM-M1) consisting of ultramafic (UM: dunites, harzburgites, and orthopyroxenites) and mafic (M1: norites, gabbro–norites, and gabbros) rocks; and 2) a pyroxene-free and hornblende-bearing series (M2; mela-, meso-, and leucogabbros). The leucogabbros exhibit some characteristics of anorthosites, including the high Al₂O₃ whole-rock content (31 wt%), high An content (An_84–96) in plagioclase, weak rare earth element enrichment, and very positive Eu anomalies. We propose that the rocks of the ultramafic and pyroxene-bearing rock-series (UM-M1 series) are all associated with the same magmatic event and that the M2 series rocks are associated with a distinct separate event. Isotopic data suggest that these formations are Ordovician. The M2 rocks have juvenile Nd isotopic signatures (εNd₍₄₆₀₎ from +4.59 to +8.11), suggesting that they are derived from superheated alumina-rich basaltic or basaltic–andesite melts extracted from a relatively depleted mantle source. In contrast, most of the M1 rocks derived from parental basaltic melts show partially crustal contamination, with only a few clearly juvenile samples (εNd₍₄₆₀₎ from +0.45 to +6.59). We propose a geodynamic evolution for the Castillon massif involving a two-stage genesis and the emplacement of the two series. First, the emplacement of olivine-saturated basaltic melts in a deep metasedimentary crust resulted in the M1 series. The second step involves the emplacement of alumina-rich basaltic or basaltic–andesite melts to produce the M2 hornblende-bearing series (mela-, meso-, and leucogabbros) devoid of pyroxenes. The leucogabbros show strong similarities with common anorthosites although they have not been previously observed in the Variscan Pyrenees.

1 Introduction

Ultramafic and mafic (UMM)-layered magmatic complexes are commonly associated with felsic rocks metamorphosed under granulite facies conditions in the continental lower crust (e.g., Vielzeuf, 1980; Windley et al., 1981; Féménias et al., 2005; Correia et al., 2012). These UMM complexes, which play a key role in crustal growth processes, offer crucial insights into the geological mechanisms responsible for the generation, evolution, and emplacement of magmas (e.g., see references above and Hildreth and Moorbath, 1988; Annen et al., 2005; Polat et al., 2009; Polat et al., 2012). However, the formation mechanism of UMM-layered complexes remains contentious. Layered UMM intrusions are characterized by melt–solid interaction processes (Féménias et al., 2005) and occur at all scales from thick (several km) units (Correia et al., 2012) to very thin microlayers (Zingg, 1996; Philpotts and Dickson, 2002; Féménias et al., 2005). Their variable mineralogical and chemical compositions and textures reflect a complex magmatic history (e.g., fractional crystallization, cumulative processes, and magma mixing) and late- to post-magmatic processes (Féménias et al., 2005). In an orogenic context, UMM complexes are commonly exposed as post-collisional bodies related to late-extensional events (e.g., Vielzeuf and Kornprobst, 1984; Chai et al., 2008; Sun et al., 2007; Wang et al., 2009; Han et al., 2014; Kilzi et al., 2016; Lemirre, 2018; Lemirre et al., 2019). UMM rocks are nonetheless often considered evidence of basaltic components involved in crustal growth (Sisson and Grove, 1993; Grove et al., 1997; Kemp and Hawkesworth, 2004).

In the North Pyrenean massifs, UMM rocks associated with granulitic felsic series occur in massifs, which include, from east to west, the following: Agly (Fonteilles, 1976), Bessède (Albarède and Fourcade, 1969), Saint Barthélémy (Zwart, 1965; de Saint Blanquat, 1989; Lemirre, 2018; Lemirre et al., 2019), Castillon (Roux, 1977; Vielzeuf, 1984; Pin, 1989; Driouch, 1997; Lemirre, 2018; Lemirre et al., 2019), and Ursuya in the Basque country (Boissonnas et al., 1974; Vielzeuf, 1984; Lemirre, 2018). UMM rocks also crop out in small tectonic slices along the North Pyrenean Fault: Treilles (Vielzeuf, 1984; Pin, 1989), Lherz (Monchoux, 1972; Vielzeuf, 1980; Vielzeuf, 1984), and Aunac and Bléchin, southeast of the Castillon massif (Roux, 1977). Pin (1989) suggested that the mafic rocks observed in some North Pyrenean massifs (Saleix and Castillon) are derived from mantle sources. Among the North Pyrenean massifs, the Castillon massif is the best candidate for investigating the origin and evolution of these types of UMM-layered complexes as this area displays the most abundant and diversified granulitic UMM rock series, spanning from dunites to hornblende leucogabbros, the latter showing a corundum + sapphirine association, a very rare association in the European Variscan chain.

In this contribution, we present a new petrological, geochemical, and Sr–Nd isotopic study of the Castillon massif to characterize UMM series, trace their sources, and define their petrogenetic evolution in the context of their Paleozoic evolution.

2 Regional geological context

The Pyrenees, located along the border of France and Spain, constitute a polyorogenic belt that trends WNW–SSE and record Variscan (Carboniferous and Permian) and Alpine (Upper Cretaceous to Oligocene) cycles (Barnolas and Chiron, 1996; Ford et al., 2022; Gamisel-Muzás et al., 2025). The Variscan Pyrenees belong to the external zone of the southern flank of the West European Variscides (Matte, 1991; Barnolas and Chiron, 1996). Formations affected by and associated with the Variscan orogeny crop out in the axial zone and in the North Pyrenean massifs, separated by a major discontinuity, that is, the North Pyrenean Fault (Figure 1A; De Sitter, 1964). The Castillon massif in Ariège, immediately west of the Trois Seigneurs massif (Figure 1A), occurs within the North Pyrenean Albo–Aptian metamorphic series, with the North Pyrenean Fault defining its southern border (Figure 1A). Only a few studies have been devoted to the Castillon massif (Raguin, 1938; De Sitter and Zwart, 1960; Thiebaut, 1964; Monchoux and Roux, 1973; Monchoux and Roux, 1975; Roux, 1977; Vielzeuf, 1984; Pin and Vielzeuf, 1983; De Saint Blanquat, 1989; Barnolas and Chiron, 1996; Driouch, 1997; Kilzi, 2014; Lemirre, 2018).

Figure 1

Figure 1. (A) Simplified geological map of the Pyrenees, Spain/France, showing the location of the North Pyrenean Castillon massif (modified from Denele et al., 2007). (B) Simplified geological map of the North Pyrenean Castillon massif (modified from De Saint Blanquat, 1993).

In our study, we focus on the Couret du Loup area in the southwestern part of the Castillon massif (Figure 1B). In this area (Figure 2A), a series of boudinaged UMM layers are emplaced within a metasedimentary series. The entire formation is affected by the same synmetamorphic foliation and granulitic metamorphism. The foliation attitude (trending 70°N and dipping 30°–40°N) (Figures 2A,B) and the metamorphic evolution, which, at the scale of the massif, decreases from a granulite facies at the bottom of the series to amphibolite facies at the top, allow for defining a polarity (Roux, 1977; Driouch, 1997; Lemirre, 2018). This polarity corresponds in the Couret du Loup area to the following succession (moving from bottom to top): 1) garnet and sillimanite gneisses associated with UMM rocks; 2) garnet and cordierite gneisses associated with mafic rocks; and 3) garnet and sillimanite kinzigites without UMM rocks. The conditions of granulite facies metamorphism, as defined based on the mineral associations from metasediments, correspond to P = 7–8 kb and T = 800 °C–850 °C (Roux, 1977), P = 5–6 kb and T = 700 °C–750 °C (Vielzeuf, 1984), and P = 7 ± 0.5 kb and T 875 °C ± 25 °C (Lemirre, 2018). The UMM occurs as layers sometimes deformed as “boudins” within the granulitic felsic series. The layers trend ENE–WSW and dip approximately 30°N. Several cross sections allow defining a synthetic stratigraphic sequence (Figure 2B; Roux, 1977; Driouch et al., 1989; Driouch, 1997; Kilzi, 2014; Lemirre, 2018). At the bottom of the cross section, the UM series is mainly composed of dunites and harzburgites, sometimes associated with orthopyroxenites and hornblendites. In the middle portion of the section, a pyroxene (Px)-bearing mafic rock series (M1) is found, comprising norites, gabbro–norites, and gabbros; in the upper part of the cross section, hornblende (Hbl) gabbros devoid of Px (M2) are present, with rare corundum + spinel + sapphirine-bearing leucogabbros at the top.

Figure 2

Figure 2. (A) Geological map of the Couret du Loup Area of Castillon massif, showing the distribution of ultramafic and mafic (UMM) rock-type bodies. (B) Schematic synthetic log of the Couret du Loup area of the Castillon massif, showing the overall distribution of UMM rocks and their relationships with metamorphic hosting rocks (leptynites and kinzigites); AZ, axial zone; Crd, cordierite; Hbl, hornblende; s.s., sensu stricto; NPF, North Pyrenean Fault; NPZ, North Pyrenean zone.

3 Materials and methods

Forty-eight ultramafic and mafic samples were selected for whole-rock and mineral chemical analyses. Mineral major element compositions were determined at the Laboratory Geosciences Environnement Toulouse (GET), CNRS-Toulouse University, OMP, Toulouse, France. Analyses relied on a CAMECA SX 50 microprobe with SAMx automation, using wavelength-dispersive spectrometry (WDS). The electron microprobe applied on anhydrous minerals used an accelerating voltage of 15 kV, a beam current of 20 nA, and a focalized beam. For amphiboles, we used the same accelerating voltage but with a beam current of 10 nA and a defocalized beam. Matrix corrections were undertaken using PAP (Pouchou and Pichoir, 1984).

In situ trace element analyses of clinopyroxene and amphibole were carried out using LA-ICP-MS. Analyses were performed at the Geosciences Montpellier Laboratory (Montpellier, France) under analytical conditions similar to those used by Marchesi et al. (2009), using a Thermo Finnigan ELEMENT XR high-resolution (HR) ICP-MS coupled with a Geolas (Microlas) automated platform housing a 193 nm Compex 102 laser from Lambda Physik. The reference materials NIST SRM 610 and NIST SRM 612 were used as external standards to control accuracy. The CaO concentration determined using an electron microprobe served as an internal standard for both clinopyroxene and amphibole. The laser was set at 4–5 J/cm², with a repetition (pulse) rate of 5–10 Hz and a spot size of 50–100 μm. The relative precision and accuracy for laser analysis ranged from 1% to 10% for most elements but were closer to 15% for Nb and Ta. The theoretical detection limits for each element ranged between 10 and 60 ppb, except for Sc and V (100 ppb), Ti (2 ppm), and Ni and Cr (0.7 ppm). Data reduction and processing were performed using Glitter software (Griffin et al., 2008).

For the analysis of major and trace elements, whole-rock UMM samples were crushed and powdered using an agate mill. Whole-rock major and minor element compositions were determined using X-ray fluorescence (XRF) at the Ecoles de Mines, Saint-Étienne, France, following the study by Gruffat (1992). Trace elements, including transition metals (Ni, Cr, and Co), rare-earth elements (REEs), high-field-strength elements (HFSEs), and large-ion lithophile elements (LILEs), were determined through inductively coupled plasma mass spectrometry (ICP-MS), using an Agilent 7500a system at the ALS Laboratory, Sevilla, Spain [except for ultramafic (UM) lithologies, performed on a Thermo Scientific Element XR at the GET laboratory, following the studies by Barrat et al. (1996), Barrat et al. (2007)]. The analysis of Sr and Nd isotopes (38 samples) required dissolving 100 mg of powder from each sample in a mixture of concentrated HF/HNO₃, followed by processing on a hot plate at 95 °C for 24 h. The sample solutions were then dried and dissolved again in 1 mL concentrated HNO₃. After drying, the residue was dissolved in 2N HNO₃ and processed through subsequent Eichrom Ln-Spec 50–100, Thru-Spec 100–150, and Sr-Spec 100–150 dedicated resins (Pin et al., 2014). Isotopic measurements were performed using a Finnigan MAT 261 mass spectrometer at the GET laboratory. Mass fractionation corrections for Sr and Nd isotopic ratios were made on the basis of ⁸⁶Sr/⁸⁸Sr = 8.3752 and ¹⁴⁶Nd/¹⁴⁴Nd = 0.7219. The NBS 987 and La Jolla standard solutions yielded values of ⁸⁶Sr/⁸⁸Sr = 0.710250 ± 35 and ¹⁴³Nd/¹⁴⁴Nd = 0.510850 ± 20. Typical blanks were 20 pg for Nd and 300 pg for Sr.

4 Results

In this study, we present the petrographic, mineralogical, and geochemical characteristics of series of ultramafic rock, Px-bearing mafic rock, and Hbl gabbro rock. Supplementary Table S1 presents the main mineralogical and petrological features, and Supplementary Tables S2–S4 summarize the whole-rock analyses. Trace element compositions of Hbl and clinopyroxene are provided in Supplementary Tables S1–S3.

4.1 Field relationships, petrography, mineralogy, and geochemistry

4.1.1 Ultramafic rocks

The UM series consists of Hbl, olivine (Ol), orthopyroxene (Opx), and spinel (Spl), along with magnetite (Mag) and ilmenite (Ilm) as accessory components (abbreviations according to Whitney and Evans, 2010). Depending on the modal composition (17 point-counting analyses) and using the classification of Streckeisen (1974), several petrographic types can be defined: dunites, harzburgites, Hbl harzburgites, orthopyroxenites, and Hbl orthopyroxenites (Figure 3A). Among the UM rocks, some appear very homogeneous (dunites), whereas others, essentially the harzburgites, consist of alternating centimeter-thick layers of various UM types (see below).

Figure 3

Figure 3. Modal composition (point counting) and classification of ultramafic and mafic (UMM) rocks of the Castillon massif. (A) Ultramafic rocks; (B) mafic rocks: pyroxene-bearing mafic rocks and hornblende gabbros. Hbl, hornblende; Px, pyroxene; Ol, olivine; s.s., sensu stricto.

Dunites (two investigated samples) occur as metric layers at the bottom of the series and alternate with harzburgite (Figure 4A). They are nearly entirely composed of fine-grained crystals (0.3–0.5 mm), defining a polygonal texture (adcumulate, Figures 4B,C). They consist of 90%–95% Ol (Fo_88–90), approximately 3% Opx (Wo_0.2–1.6 En_88.7–92.3 Fs_7.4–10.6), and rare interstitial amphibole (Amp) (magnesio-hornblende 0.8 < Mg# < 0.92). Small grains of Cr–Al Spl with an Mg# of 56–59 and a Cr# of 26–32, along with Mag and Ilm, are included within the olivine (Kilzi, 2014). The analyzed dunite (Supplementary Table S1) displays relatively low SiO₂ (36.61 wt%) and Al₂O₃ (1.28 wt%) contents, a high Mg# (88.4), and high Ni (2445 ppm) and Cr (4450 ppm) contents (Supplementary Table S1). On the chondrite-normalized REE diagram (Figure 5), dunite shows a slightly depleted flat REE pattern [total REE = 0.4 ppm; (La/Yb)_N = 0.94]. The mantle-normalized trace element pattern shows Cs, Th, U, and Nb enrichment, positive Pb and Hf anomalies, and negative Ba and Zr anomalies (Figure 5).

Figure 4

Figure 4. Microphotographs of dunites (polarized light). (A) Macro- and microlayering in ultramafic rock outcrop, dunite alternating with harzburgite. (B,C) Microphotographs of dunite texture (polarized light) samples (B) 09CA10 and (C) 09CA06. Spl, spinel; Ol, olivine; Opx, orthopyroxene; Chr, Chromite.

Figure 5

Figure 5. Chondrite-normalized REE and primitive mantle-normalized trace elements in Castillon dunite and orthopyroxenite.

Harzburgites, fine- to medium-grained (0.5–2 mm) rocks, are the dominant rock types. They comprise alternating centimeter-thick layers differing in their modal compositions (Figure 3A): 1) Hbl-poor harzburgite layers consisting of 55%–90% Ol, 10%–45% Opx, and less than 5% Amp; 2) Hbl harzburgite layers consisting of 70%–85% Ol, 10%–20% Opx, and 15%–20% Amp; and 3) Hbl-rich harzburgite layers consisting of 40%–50% Ol, 20%–30% Opx, and more than 30% Amp (Figure 3A). All harzburgite types contain 1%–5% accessory minerals, Spl grains with an Mg# of 28–76 and a Cr# of 39–81, Ilm with high FeO and TiO₂ contents (44 wt% and 52 wt%, respectively), hematites (Hem), and Fe–Ni–Cu sulfides. Serpentine (Srp) and talc (Tlc) occur as common alteration products in the harzburgite samples. Others display secondary anthophyllite (Ath) fibers, 0.1–1 mm long, with a mean Mg# of 88. Additionally, the harzburgites are layered (Figure 6), including millimetric- to centimetric-thick layers of 1) orthopyroxenite consisting of (up to 90%) prismatic Opx, approx. 4% Ol, approx. 3% Amp, approx. 2% Spl (Al₂O₃: 37.8–58.8 wt%; Cr₂O₃: 8.4–23.6 wt%; FeO(T): 14–23 wt%), and Ilm crystals; and 2) 2–3-cm-thick Hbl layers consisting of large prismatic Hbl (>90% of the modal content). In the harzburgites, the chemical composition of Ol and Opx is homogeneous and close to the dunite mineral composition, that is, Fo_88–90 for Ol and Wo_0.2–1.6 En_88.7–92.3 Fs_7.4–10.6 for Opx. Amphiboles are magnesio-hornblendes [Leake et al., 1997, in Kilzi (2014)], having a composition that varies slightly, with Mg# decreasing from 92 in the Hbl-poor harzburgites to 81 in the Hbl-rich harzburgites.

Figure 6

Figure 6. Photograph and microphotograph of the sample 10CA08. (A) Hand specimen (approx. length 25 cm); (B) hornblendite; (C) hornblende (Hbl)-poor harzburgite; (D) Hbl harzburgite; and (E) Hbl-rich harzburgite. Ol, olivine; Opx, orthopyroxene.

Whole-rock analyses (Supplementary Table S1) of the Hbl-poor harzburgites (three investigated samples) display relatively low SiO₂ (40–41.2 wt%) and Al₂O₃ (average 1.4 wt%) contents and a high Mg# (89–92), along with high Ni (1837–2641 ppm) and Cr (900–2260 ppm) contents. They exhibit (samples 09CA09 and 09CA02, Figure 7) 1) weakly REE-enriched patterns, with ∑REE = 1.6–2.8 ppm; 2) moderately LREE (Light Rare Earth Elements)-enriched patterns of (La/Yb)_N = 1.1–2.7 and (La/Sm)_N = 0.5–1.9; 3) flat to moderately depleted HREE (Heavy Rare Earth Elements) of (Gd/Yb)_N = 1.38–1.78; and 4) a slightly positive or negative Eu anomaly (Eu/Eu* = 0.97–1.15). The mantle-normalized trace element patterns exhibit Cs, Th, U, and Nb enrichments; positive Cs, Pb, and Hf anomalies; and negative Ba, Sr, and Zr anomalies (Figure 7).

Figure 7

Figure 7. Chondrite-normalized mineral rare earth element (REE) and primitive mantle–normalized trace element contents of Castillon harzburgites. Hornblende (Hbl)-poor harzburgites; Hbl harzburgites; and Hbl-rich harzburgites. For all figures, whole-rock values are presented as filled squares and solid lines, and hornblende values are presented as dashed lines.

Hbl harzburgites (Supplementary Table S1; seven investigated samples) display slightly higher Al₂O₃ and CaO contents (up to 1.46 wt% and 1.59 wt%, respectively) and a high Mg# (90–91). They exhibit REE-enriched patterns with total REE = 1.85–3.54 ppm, (La/Yb)_N = 1.89–4.76, and Eu/Eu* = 0.94–1.75 (Figure 7).

Hbl-rich harzburgites (five investigated samples) have higher SiO₂ (up to 44.79 wt%), Al₂O₃, and CaO contents (1.5–5.62 wt% and 1.27–5.63 wt%, respectively) and lower MgO (average 37.3 wt%) contents than dunites and other harzburgites with an Mg# close to 89 (Supplementary Table S1). They also display high Ni (1550–2330 ppm) and Cr (1300–2600 ppm) contents. Hbl-rich harzburgites have the highest REE content (total REE up to 40 ppm) among all harzburgitic rock types. They are moderately LREE-enriched and HREE-depleted, with (La/Yb)_N = 7.71–14.3, (La/Sm)_N = 2.15–3.16, and (Gd/Yb)_N = 2.14–4.47 (Figure 7), and are characterized by slightly negative Eu anomalies (Eu/Eu* = 0.94–1.62). Their mantle-normalized trace element patterns show enriched Cs, Th, La, and Ce; positive Ba, Nb, Pb, and Nd anomalies; and negative Rb, U, Sr, and Zr anomalies (Figure 7).

Orthopyroxenite and Hbl-bearing orthopyroxenites (three investigated samples) occur as centimetric layers alternating with dunites and harzburgites. They are almost entirely composed of fine-grained crystals (0.3–0.5 mm) defining an adcumulate texture. They consist of 85%–95% euhedral Opx associated with 5%–10% Amp, and minor Ilm and Spl. Opx and Hbl have compositions comparable to those found in harzburgites. Orthopyroxenite and Hbl orthopyroxenites have a major element composition close to that of Hbl-rich harzburgites with SiO₂ (44–47.51 wt%), Al₂O₃ (2.62–5.54 wt%), CaO (2.78–5.78 wt%), and MgO (30–33.05 wt%). Their Mg# value is also high, approaching 90. In contrast, orthopyroxenite displays weak REE enrichment (∑REE = 5.7 ppm), moderately depleted to slightly enriched LREE patterns (La/Yb)_N = 1.22 and (La/Sm)_N = 0.7, and no Eu anomaly (Figure 5). Finally, its mantle-normalized trace element pattern is characterized by Cs enrichment, positive Nb and Nd anomalies, and negative Ba, Sr, and Zr anomalies (Figure 5). Hbl orthopyroxenites differ (Supplementary Table S1) by their REE total contents (up to 60 ppm) with (La/Yb)_N = 8.25 to 10.13 and (La/Sm)_N = 2, related to the Hbl modal abundance.

To summarize, the Castillon ultramafic rock types correspond to a layered sequence with the alternation of mainly dunites, orthopyroxenites, and harzburgites with minor hornblendite layers. REE contents in the ultramafic rocks reveal a strong correlation between their Amp modal composition and REE patterns, where an increase in the Amp content is commonly associated with higher total REE concentrations and higher La/Yb ratios (Supplementary Table S1).

4.1.2 Pyroxene-bearing mafic series (M1)

Located above the UM series, the Px-bearing mafic series occurs as layers, 1 dm to 1.5 m thick, sometimes strongly boudinaged within the felsic granulites (Figure 2B). The series is composed of norites [plagioclase (Pl)+Opx + Hbl], gabbro–norites [Pl + Opx + clinopyroxene (Cpx)+Hbl], and gabbros “stricto senso (s.s.)” (Pl + Cpx + Hbl; Figure 8). Layers of diverse types of Px-bearing mafic rocks alternate with layers of Hbl gabbros (M2, see below). Moreover, contacts between the two types of mafic rocks appear to have been parallelized by deformation; however, in detail, they are irregular, sinuous, and often fingered, leading to the inclusion of fragments of norite and gabbro–norite within the Hbl gabbro. Pyroxene-bearing mafic rocks show two scales of igneous layering: 1) layering at the outcrop scale with alternating decametric to metric layers of norites and gabbro–norites, and 2) layering at the hand specimen scale with alternating thin, centimeter-thick layers of gabbro–norite and norite (Figure 8).

Figure 8

Figure 8. Photographs and microphotographs of the pyroxene-bearing mafic rocks of the Castillon massif (M1 series): pyroxene-bearing mafic rocks. (A,B) Igneous layering in norite and gabbro–norite; (C) norite texture (sample 10CA12, polarized light); (D) gabbro–norite texture (sample 10CA0, polarized light). Hbl, hornblende; Ol, olivine; Opx, orthopyroxene; Cpx, clinopyroxene; Bt, biotite; Pl, plagioclase.

Norites (four investigated samples), located at the bottom of the Px-bearing mafic series, occur as boudinaged layers, varying between 10 and 50 cm in thickness within the gneisses. They consist of Pl + Opx + Amp + Opq ± biotite (Bt). The variations in Hbl, Opx, and Pl modal proportions (Hbl: 5%–30%, Opx: 5%–40%, and Pl: 20%–60%) allow distinguishing several types of norite, spanning from Hbl leuconorites to Hbl melanorites (Figure 3B). Norites consist of rounded and subhedral Opx (hypersthene ranging in composition from Wo_0.8–1.2 En_60–67 Fs_32–38 to Wo_1.1–1.9 En_54–56 Fs_43–45), Pl (labradorite Ab_41–47 An_53–59 Or_0.2–0.51), and fine-grained Ilm with high FeO (44 wt%) and TiO₂ (52 wt%) contents. These minerals are cumulus and are associated with subhedral to anhedral intercumulus edenite (6.5 < Si < 6.85 and an Mg# ranging from 68 to 71); together they define a mesocumulate texture (Figure 8C). Norites display moderate variations in SiO₂ (45.5–48.8 wt%), Al₂O₃ (15–16.6 wt%), MgO (8.5–10 wt%), FeO(T) (9.9–12.3 wt%), CaO (6–11 wt%), and Mg# (55–64). They have significant Cr (230–470 ppm) and Ni (131–256 ppm) contents (Supplementary Table S2). On chondrite-normalized REE- and mantle-normalized trace element diagrams (Figure 9), norites are characterized by 1) moderately enriched REE patterns, with ∑REE = 61–80 ppm; 2) moderately enriched LREE patterns, with (La/Yb)_N = 1.8–2.4 and (La/Sm)_N = 1.3–2; 3) moderately depleted HREE patterns, with (Gd/Yb)_N = 1.2–1.6; 4) a weak negative Eu anomaly, with Eu/Eu* = 0.7–1; 5) Cs and Rb enrichments; 6) positive U and Pb anomalies; and 7) a slightly negative Nb anomaly.

Figure 9

Figure 9. Chondrite-normalized mineral rare earth element (REE) and primitive mantle–normalized trace element contents of Castillon massif pyroxene-bearing mafic rocks. REE and trace element values of norites for whole-rock samples (light blue–filled diamonds and solid light blue lines) and hornblende samples (dashed lines). REE and trace element values for whole-rock gabbro–norite (dark blue–filled diamonds and dark blue solid lines), whole-rock gabbro sensu stricto (s.s.) (light blue–filled diamonds and light blue solid lines), clinopyroxene (Cpx) gabbro–norite (light blue dashed lines), and hornblende (Hbl) gabbro–norite (dark blue dashed lines).

Gabbro–norites (six investigated samples) occur as 1–4-m-thick layers located at the top of the noritic series. They are medium- to coarse-grained (1–3 mm) characterized by alternating Hbl-rich and Hbl-poor thin layers (0.5–1 cm thick). Typical samples consist of approx. 35%–40% Opx, approx. 30%–40% Cpx, approx. 20%–25% Pl, approx. 5%–20% Amp, and <3% accessory minerals, mainly Ilm. Opxs are hypersthenes (Wo_0.8–1.4 En_52–56 Fs_42–48; Kilzi, 2014) and are slightly more ferriferous than those of norites. Cpxs are diopsides (Wo_44.5–48 En_37.5–39 Fs_12.5–19), with an Mg# grading from 65 to 73. They occur as anhedral and locally poikilitic crystals, including Pl, Opx, and Ilm grains (Figure 8D). Poikilitic Amps are edenites and magnesio-hastingsites, following the classification of Leake et al. (1997), with an Mg# of 53–62 (Kilzi, 2014). Pls are labradorites (Ab_50–52 An_47–50 Or_0.1–0.9). Gabbro–norites display notable variations in SiO₂ (39–48 wt%), Al₂O₃ (14.5–18.6 wt%), and CaO (9–14 wt%) contents, with a low Mg# (43–60) (Supplementary Table S2). These variations are related to the heterogeneous repartition (microlayering) of some constituent mineral phases, such as Cpx and Amp. They have moderate Cr (23–334 ppm), Ni (45–85 ppm), and Zr (10–2 ppm) contents (Supplementary Table S2). On chondrite-normalized diagrams (Figure 10), gabbro–norites appear moderately REE-rich (∑REE = 46–85 ppm), with (La/Yb)_N = 0.5–2, (La/Sm)_N = 0.5–1.5, and (Gd/Yb)_N = 1.3–2.1. They show minor negative or positive Eu anomalies (Eu/Eu* = 0.9–1.3). Mantle-normalized trace element patterns are characterized by Cs enrichment and negative anomalies in Rb, U, Zr, and Pb (Figure 9). Gabbros s.s. (Le Maitre et al., 2004) occur within the gabbro–norite as centimetric Cpx-rich layers. In these gabbro layers, anhedral Cpxs are diopsides (Wo_47.8–49.7 En_36.7–39.8 Fs_9.5–15.5), with a high CaO content (approx. 24 wt%) and Mg#, ranging from 70 to 81 (Kilzi, 2014). Anhedral poikilitic patches of Amp (up to 1 cm in size) are mainly magnesio-hastingsites, with an Mg# ranging from 52 to 60 and Si varying from 5.4 to 6.6 apfu (Kilzi, 2014).

Figure 10

Figure 10. Photographs and microphotographs of the mafic rocks of the Castillon massif (M2 series): hornblende (Hbl) gabbros. (A) Igneous layering in a Hbl melagabbro and Hbl mesogabbro outcrop; (B) Hbl melagabbro (sample 10CA05, polarized light); (C) Hbl mesogabbro (sample 10CA07, polarized light); (D) Hbl leucogabbro [corundum–spinel–sapphirine (Crn–Spl–Spr)] (sample 09CA05, polarized and analyzed light); (E) Corundum crystals surrounded by spinel (Spl) and sapphirine (Spr) in a Hbl leucogabbro. Pl, plagioclase.

Gabbro layers (two investigated samples) have a low SiO₂ content (approx. 39 wt%) and high CaO (approx. 17 wt%) and TiO₂ (approx. 3.4 wt%) contents. Their Mg# is relatively low (approx. 62–63). They have moderate Cr (250–280 ppm), Ni (144–156 ppm), and Co (70–75 ppm) contents, along with high Zr (253–319 ppm) contents (Supplementary Table S2). They show strong REE enrichment (∑REE = 239–245 ppm) and very strong REE fractionation, with (La/Yb)_N = 10, (La/Sm)_N = 3, and (Gd/Yb)_N = 2.7, but no Eu anomalies (Figure 9). Mantle-normalized trace element patterns are characterized by Cs enrichment; negative Rb, Ba, U, and Sr anomalies; and positive Th, Nb, Pb, and Nd anomalies (Figure 9).

4.1.3 Hornblende gabbros (M2)

Hornblende gabbros (M2 series; 16 samples) occur as layers alternating with Px-bearing UMM series (M1). Locally, Hbl gabbros crosscut the Px-bearing rocks and include them in the form of decimetric fragments. Hbl gabbros are composed of Hbl, An-rich Pl (bytownite–anorthite), and low amounts of opaque minerals (<2%). Their modal composition (Amp/Pl) evolves along the cross section and allows us to define, from the bottom to the top, three Hbl gabbro types (Le Maitre et al., 2004; Figure 3B): Hbl melagabbros (75%–82% Hbl vs. 18%–20% Pl), Hbl mesogabbros (almost 55%–60% Hbl vs. 40%–45% Pl), and Hbl leucogabbros (15%–35% Hbl vs. 65%–85% Pl). Hbl leucogabbros are characterized by a specific mineral association: Pl + Hbl + corundum (Crn)+Spl + sapphirine (Spr).

Hornblende melagabbros (two investigated samples; Figure 10B) are mainly associated with norites and gabbro−norites (Figure 2B), and rarely with ultramafic rocks. In these rocks, Amp consists mainly of magnesio-hastingsites, with Si varying from 6 to 6.5 apfu, FeO(T) ranging from 10.30 to 13.02 wt%, and Mg# ranging from 62 to 72 (Kilzi, 2014). Pls are An-rich bytownites (An_82–88). Locally, Hbl melagabbros include centimetric to decimetric fragments of edenite labrador-bearing norites of the M1 group. Hbl melagabbros display SiO₂ contents ranging from 39 to 47.8 wt% and an Mg# ranging from 55 to 71. They are moderately REE-enriched (∑REE = 22–26 ppm), with (La/Yb)_N = 0.8–1, (La/Sm)_N = 0.85–1, and (Gd/Yb)_N = 1–1.3, have positive Eu anomalies (Eu/Eu* = 1.6–3.5), and display a relatively flat REE pattern (Supplementary Table S3; Figure 11). Their mantle-normalized trace element patterns have positive Pb and Sr anomalies and negative Zr anomalies (Figure 11).

Figure 11

Figure 11. Chondrite-normalized rare earth element (REE) and primitive mantle–normalized trace elements of whole rock and associated hornblende of (Hbl) gabbros from the Castillon massif with i) whole-rock Hbl melagabbros (brown-filled triangles and brown solid lines) and associated hornblende (brown dashed lines); ii) whole-rock Hbl mesogabbros (red-filled triangles and red solid lines) and associated hornblende (red dashed lines); and whole-rock sapphirine-bearing Hbl leucogabbros (yellow-filled triangles and yellow solid lines) and associated hornblende (yellow dashed lines).

Hbl mesogabbros (six investigated samples; Figure 10C) occur in the middle part of the UMM series. In these rocks, Pl is anorthite (An_91–96), and Amp has an intermediate magnesio-hastingsite to pargasite composition, with FeO(T) ranging from 8 to 11.7 wt% and an Mg# ranging from 69 to 78.7 (Kilzi, 2014), being slightly more Mg-enriched than Hbl melagabbros. Hbl mesogabbros have an SiO₂ content of approx. 44 wt%, except for one sample with an SiO₂ content of 48 wt%, and an Mg# ranging from 73 to 77 (Kilzi, 2014). In the chondrite-normalized plots (Supplementary Table S3; Figure 11), Hbl mesogabbros appear less REE-enriched than Hbl melagabbros, with ∑REE = 5–10 ppm, a positive slope of LREE (La/Yb)_N = < 1, and (La/Sm)_N < 1. They show comparable HREE patterns (Gd/Yb)_N = 0.7–0.8) and Eu anomalies (Eu/Eu* = 3.5–3.8). Normalized trace element patterns show the same positive (Pb and Sr) and negative (Zr and P) anomalies as Hbl melagabbros, with the latter exhibiting a slightly greater enrichment of trace elements (Nb to Lu; Figure 11).

Hbl leucogabbros (eight investigated samples) occur as 1–20-m-thick layers at the top of the M2 mafic series (Figure 2). They mostly consist of a Pl–Amp association defining a mosaic texture (Figure 10D). Pls are anorthites (An_93–96), and Amps are pargasites, with Si contents varying from 6.15 to 6.45 apfu, FeO(T) ranging from 5 to 6 wt%, and an Mg# of 86–89 (Kilzi, 2014). Locally, leucogabbros exhibit Crn, Spl, and Spr paragenesis associated with granulitic metamorphism (Monchoux and Roux, 1975; Abraham et al., 1977). On the chondrite-normalized REE diagrams, Hbl leucogabbro displays weak REE enrichment (∑REE = 2–6 ppm), with (La/Yb)_N = 1–12, (La/Sm)_N = 1–6, and (Gd/Yb)N = 1–2.4, along with a strongly positive Eu anomaly (Eu/Eu* = 3–8.2). Mantle-normalized trace element patterns show negative Th, Nb, Ce, Pr, Zr, Sm, and Nd anomalies and positive Sr, Pb, U, and Eu anomalies (Figure 11).

In summary, Hbl gabbros show a remarkable modal and chemical evolution from Hbl-rich melagabbros at the bottom of the series to Pl-rich leucogabbros at the top. This evolution (Supplementary Table S3; Figure 12) is underlined by 1) a decrease in the whole-rock SiO₂ content from 47.80% to 38% and an increase in Al₂O₃ from 16.40% to 29.4%; 2) an increase in whole-rock Mg# grading from 55 to 86.2 and Mg# Hbl grading on average from 62 to 89 (Kilzi, 2014). This evolution is coeval with a decrease in FeO both in Hbl and whole rocks (12%–5% for Hbl and 11%–3% for whole rock) (Supplementary Table S3), whereas MgO remains nearly constant; 3) an increase in Cr from 600 to 1280 ppm and Ni from 59 to 652 ppm (Supplementary Table S3); 4) a progressive depletion in whole-rock REE content, grading on average from 26 to 2 ppm (Supplementary Table S3); and 5) an increased Eu anomaly (Eu/Eu*), grading from 1.6 to 8.2 (Supplementary Table S3). This evolution is singular and remarkable because the less evolved leucogabbros are at the top of the Hbl gabbro pile, whereas the more evolved melagabbros are at the bottom (Figure 2B). It is consistent with an increase in the Pl/Amp ratio from the bottom to the top of the series (Supplementary Table S1).

Figure 12

Figure 12. Major elements versus SiO₂ in ultramafic and mafic (UMM) rock series of the Castillon massif. Al₂O₃; FeO(T); MgO; Cr; CaO; Ni.

4.2 Isotopic signatures

All measured data and isotopic standards are listed in Supplementary Table S4.

4.2.1 M1 ultramafic rocks

Three Hbl orthopyroxenites, one Hbl-poor harzburgite, and one Hbl-rich harzburgite were processed for isotopic measurements. They display relatively low measured Nd values and intermediate Sr isotopic signatures (¹⁴³Nd/¹⁴⁴Nd: 0.512455–0.512690 and ⁸⁷Sr/⁸⁶Sr: 0.704448–0.708936).

4.2.2 M1 pyroxene-bearing mafic rocks

These samples display a wide range of isotopic signatures, both within individual sample types and across the entire sample set. Norites (five samples) are characterized by high ⁸⁷Sr/⁸⁶Sr values (0.70741–0.71196) and variable ¹⁴³Nd/¹⁴⁴Nd values (0.512306–0.512673), gabbro−norites (six samples) show relatively restricted ⁸⁷Sr/⁸⁶Sr values (0.70357–0.70577) and variable ¹⁴³Nd/¹⁴⁴Nd values (0.512617–0.513032), and the two gabbros display very similar isotopic compositions (⁸⁷Sr/⁸⁶Sr: 0.70864 and 0.70865; ¹⁴³Nd/¹⁴⁴Nd: 0.512676 and 0.512689).

4.2.3 M2 Hbl gabbros

The three analyzed isotopic signatures of Hbl melagabbros vary from 0.704624 to 0.706722 for ⁸⁷Sr/8⁶Sr and from 0.512683 to 0.512983 for ¹⁴³Nd/¹⁴⁴Nd. Hbl mesogabbros (nine samples) present ⁸⁷Sr/⁸⁶Sr values of 0.703644–0.705087 and ¹⁴³Nd/¹⁴⁴Nd values of 0.512455–0.513160, whereas Hbl leucogabbros (seven samples) are characterized by consistent ⁸⁷Sr/⁸⁶Sr (0.703655–0.704460) and slightly variable ¹⁴³Nd/¹⁴⁴Nd (0.512632–0.513164) values. It is noteworthy that prior to age correction, M2 samples exhibit the highest non-radiogenic ⁸⁷Sr/⁸⁶Sr and radiogenic ¹⁴³Nd/¹⁴⁴Nd values among the entire sample collection.

4.3 Estimates of ages and initial isotopic compositions of the Castillon UMM rocks

To discuss their geodynamic significance, isotopic data should be corrected for radioactive decay. However, the age of the UMM series of the Castillon massif, along with those of the other North Pyrenean UMM massifs, remains poorly constrained in the absence of unambiguous zircon U−Pb data. In the Castillon massif, zircon U−Pb data are from the metasedimentary series alternating with UMM series, and they provide ages between 300 and 320 Ma (Kilzi et al., 2016; Lemirre, 2018; Lemirre et al., 2019). These ages are clearly related to metamorphism/anatexis and, therefore, cannot be considered emplacement ages. We can reasonably assume that these magmas were formed at an earlier age, certainly pre-Variscan. It seems unlikely that the formation of this type of magma was synchronous with the major metamorphic episode that affected the Castillon massif.

Several methods can be used to estimate the emplacement age of our samples, including whole-rock isochrons, model ages, and the convergence of initial isotopic signals over time (isochrone best fit). It is notable, initially, that no remarkable whole-rock isochrone can be drawn using Rb−Sr or Sm−Nd. For example, initial isotopic signatures remain very scattered when the age correction is forced to 330 Ma, even regarding the Nd isotopes, which are typically less sensitive to metamorphic events. A more visual approach than the classic linear regression (Nicolaysen, 1961; York et al., 2004) was applied to the sample families having the most radiogenic Nd values (samples from the M2 group) in an attempt to find an emplacement age. This method consists of calculating the initial isotopic signature of a group of samples, typically, here, the Hbl meso- and melagabbros on the one hand, and the Spr-bearing Hbl leucogabbros on the other hand, and then increasing the age used for the correction while visualizing the standard deviation between the values corrected for a given age (Supplementary Figure S1). When the standard deviation is minimal, the measurement group converges toward a similar initial value. Therefore, it can be assumed that the corresponding age is the magmatic age. The basic assumption of this type of calculation is that the samples from the same petrological group come from a single parent liquid. This assumption is reasonable in the case of the Castillon massif, given the small volumes of magma involved and its limited geographical extent. Indeed, for the ages ranging from 460 to 520 Ma, it seems that all but three samples from the M2 series exhibit similar initial Nd isotopic signatures. To verify this outcome, we computed the TDM2 model ages (Tilhac et al., 2017), which are considered to be the lowest age at which the parental melts of M2 gabbros may have been extracted from their sources. In this instance, we considered a potential modification of Sm/Nd ratios at the age of 330 Ma, also known as the Variscan metamorphic event. To produce the model age estimates, the ¹⁴⁷Sm/¹⁴⁴Nd ratios are set to zero between 300 Ma and TDM2, in accordance with the assumptions made by Tilhac et al. (2017). Ultimately, with the exception of the three previously mentioned samples, the minimum TDM2 ages of the M2 gabbros range from 337 to 492 Ma.

There are several approaches for correcting isotopic data for emplacement age in old, metamorphosed magmatic series emplaced deep in the crust. One approach focuses on obtaining a snapshot of the magmatic system at the time of a major metamorphic episode to trace possible fluid–magma/rock interactions or anatexis processes (Downes and Leyreloup, 1986; Jesus et al., 2020); alternatively, we can attempt to estimate the initial isotopic signatures during the emplacement of magmatic rocks (Andonaegui et al., 2012; Orejana et al., 2017). Relying on the geological, petrologic, and geochemical considerations discussed above, an age correction to 460 Ma was applied to all our samples, favoring the second hypothesis. We propose that the Variscan metamorphic episode, which affected the Castillon massif, did not fully reset the radioactive geochronometers, allowing us to discuss these isotopic signatures in a geodynamic context of emplacement.

4.3.1 Ultramafic series

After age correction, the ultramafic samples are characterized by low, crustal-contaminated isotopic signatures. The harzburgites fall in the middle of the dataset, having slightly positive εNd₍₄₆₀₎ (+0.45 to +0.76) for variable radiogenic ⁸⁷Sr/⁸⁶Sr₍₄₆₀₎ (0.70396–0.70842). The orthopyroxenites display an intermediate but still contaminated isotopic signature (+1.21 to +3.48; 0.70396–0.70419).

4.3.2 M1 pyroxene-bearing mafic series

The two analyzed gabbros (M1 series) exhibit homogeneous, juvenile εNd₍₄₆₀₎ signatures (+5.70 and +5.90) but highly radiogenic ⁸⁷Sr/⁸⁶Sr₍₄₆₀₎ (0.70749 and 0.70797). The gabbro−norites (M1) display highly variable age-corrected isotopic signatures, with their εNd₍₄₆₀₎ varying from +1.66 to +6.59 and ⁸⁷Sr/⁸⁶Sr₍₄₆₀₎ from 0.70353 to 0.70572. Their initial Nd and Sr isotopic ratios remain roughly anti-correlated, suggesting that both radio-chronometers have been affected by mixing/contamination that may have resulted in this variability (Depaolo, 1981). Norites (M1) are also characterized by highly variable initial isotopic signatures. Their εNd₍₄₆₀₎ values vary from −1.62 to +3.95 and ⁸⁷Sr/⁸⁶Sr₍₄₆₀₎ from 0.70322 to 0.7064. However, in contrast to gabbro−norites, their signatures are positively correlated in the Nd–Sr diagram (Figure 15), which has no clear petrologic or geochemical significance.

To summarize, for the UM and M1 series, we identify facies having clearly heterogeneous isotopic signatures, ranging from partially crustal contaminated to clearly juvenile. Only the ultramafic facies remain systematically contaminated as their low initial Nd and Sr contents make them more sensitive to the crustal contamination process.

4.3.3 Hornblende mesogabbros, melagabbros, and leucogabbros (M2 series)

This series has clearly juvenile, depleted mantle-related, isotopic signatures that are some of the highest ever reported in southern Europe for Ordovician ultramafic/mafic lithologies (Lotout et al., 2017; Montero et al., 2009; Navidad et al., 2010; Navidad et al., 2018; Orejana et al., 2017; Pin and Marini, 1993; Casas et al., 2015; Solá et al., 2008; Andonaegui et al., 2012; Andonaegui et al., 2016). Their εNd₍₄₆₀₎ signatures are +5.13 to +7.11 (although one is −0.24) for Hb meso- and melagabbros and +4.59 to +8.31 (however two with +0.08 and −0.99) for Hb-leucogabbros. For these samples, ⁸⁷Sr/⁸⁶Sr₍₄₆₀₎ varies from 0.70334 to 0.70590 for Hb meso-/melagabbros and 0.70307 to 0.70417, with the exception of the three previously identified samples (10CA04B: 0.70473, 10CA01: 0.70590, and 10CA07C: 0.70376).

In summary, the M2 samples show very juvenile initial isotopic signatures in Nd, without any trace of continental crust contamination. For the discussion, we consider the isotopic signatures of this group as those of magmatism extracted from the underlying mantle. The other isotopic signatures are either partially or completely modified through crustal contamination or Variscan metamorphism.

5 Discussion

5.1 Origin and relationships among the UM, M1, and M2 series

The ultramafic rocks from the Castillon massif are mostly harzburgites associated with minor dunites and orthopyroxenites as decimeter- to meter-thick layers intercalated within the original sedimentary series.

The harzburgites are known as depleted-mantle rocks that are more or less metasomatized but can also occur as cumulate components in layered intrusions. In the latter case, they are mostly related to crystal settling or magma mixing processes in the magma chamber of Ol-saturated mafic magmas (e.g., Raedeke and McCallum, 1984; Cawthorn, 1996; Femenias et al., 2003; Charlier et al., 2015). The ultramafic rocks from the Castillon massif do not display a deformation typical of mantle tectonite textures but rather that of cumulative textures. Moreover, the NiO content of the harzburgite Ol, which is essentially lower than 0.35 wt% (for CaO contents up to 0.03 wt%; Kilzi, 2014), is, therefore, more similar to magmatic than mantle Ol, the latter having a relatively constant NiO content of approximately 0.4 wt% at Fo₉₀ (Sato, 1977; Takahashi E, 1986). For example, the Ol of the Lherz harzburgites have a NiO content greater than 0.41 wt% and a CaO content of less than 0.015 wt% (Le Roux et al., 2007). The field relationships between the different rock types are clear, with ultramafic rocks at the bottom of the pile, overlain by Px-bearing mafic rocks, followed by Hbl gabbros occurring at all levels, and the leucogabbros at the top of the section.

The Castillon UMM series involves cumulates regrouped into two different suites. The first is a Px-bearing suite that includes ultramafic (dunites and harzburgites) and M1 mafic rocks (norite, gabbro–norite, and gabbros s.s.). The second suite comprises Hbl gabbros, devoid of Px, whose modal contents vary in space from Amp-rich melagabbros at the bottom of the pile to plagioclase-rich leucogabbros at the top. The plotting of major element compositions on Harker diagrams against SiO₂ and Mg# (Figures 12, 13, respectively) highlights both the gap between the ultramafic rocks and the mafic rock series in terms of compatible element contents (i.e., MgO, Cr, and Ni) and the location of M2 Hbl gabbros being more aluminous and calcic but less ferriferous than the M1 Px-bearing mafic rocks, in accordance with plagioclase fractionation. It is worth noting that the M2 Hbl gabbros display an Mg# plotting in an intermediate position between ultramafic rocks and M1 Px-bearing mafic rocks (Figure 13).

Figure 13

Figure 13. Major elements versus Mg# (100 MgO/MgO + FeO) in ultramafic and mafic (UMM) rock series of the Castillon massif. Al₂ O₃; TiO₂; CaO; Cr; Ni.

The plots of the average REE and trace element contents of each rock-type on chondrite-normalized and primitive mantle–normalized trace element diagrams (Figure 14) show several patterns. First, in the UM rocks and the M1 Px-bearing mafic rocks, there is a continuous evolution of the REE content, with a progressive and regular REE enrichment. Second, an evolution in harzburgites is observed in the REE pattern as the increase in LREE is associated with the progressive increase in the modal content of Amp. Third, the REE content and patterns in the M2 Hbl gabbros differ greatly from those of M1 mafic rocks by exhibiting low LREE and HREE contents and strong Eu positive anomalies. Moreover, in the primitive mantle–normalized diagram (Figure 15), M2 Hbl gabbros are distinguished from the UM and M1 Px-bearing rocks by having positive U, La, and Sr anomalies and negative Th and Nb anomalies. The evolution of M2 Hbl gabbros from mela- to leucogabbros is controlled by the fractional crystallization of plagioclase and cumulate processes, with Amp concentrating at the bottom of the rock series and plagioclase in the upper part.

Figure 14

Figure 14. Chondrite-normalized rare earth element (REE) diagrams of the mean REE budget of each petrographic group of the ultramafic and mafic (UMM) rock suites of the Castillon massif. Primitive mantle–normalized trace element diagrams of the mean trace element budget of each petrographic group of the UMM suites of the Castillon massif.

Figure 15

Figure 15. εNd₍₄₆₀₎ vs. ⁸⁷Sr/⁸⁶Sr₍₄₆₀₎ for the Castillon ultramafic and mafic (UMM) series compared against Ordovician mafic and metamorphic isotopic data available from southern Europe (Pin and Marini, 1993; Solá et al., 2008; Montero et al., 2009; Casas et al., 2015; Lotout et al., 2017; Navidad et al., 2010; Navidad et al., 2018; Orejana et al., 2017; Andonaegui et al., 2012; Andonaegui et al., 2016). The two mixing lines illustrate theoretical mixing between 1) a mafic component (⁸⁷Sr/⁸⁶Sr = 0.7028; [Sr] = 135 ppm; εNd = +8; [Nd] = 9.88 ppm) and a crustal component (hornblendite, Aguilla massif (Kilzi et al., 2016): ⁸⁷Sr/⁸⁶Sr = 0.70876; [Sr] = 128.5 ppm; εNd = −8.35; [Nd] = 9.1 ppm); 2) an ultramafic component (⁸⁷Sr/⁸⁶Sr = 0.7028; [Sr] = 8 ppm; εNd = +8; [Nd] = 1.22 ppm) and a different crustal component (orthopyroxene hornblendite, Gloriettes massif (Kilzi et al., 2016): ⁸⁷Sr/⁸⁶Sr = 0.7079; [Sr] = 4 ppm; εNd = −10; [Nd] = 8.8 ppm).

To conclude, the petrographic and geochemical study on the Castillon massif allows us to define two magmatic series. The first is a Px-bearing UMM series (M1 series) that evolves by fractional crystallization from dunite to gabbro–norite and gabbros s.s. The second is an Hbl gabbro series (M2 series) that evolves from mela- to leucogabbros, depending on the plagioclase modal content. The leucogabbros have specific characteristics: i) Crn + Spl + Spr paragenesis; ii) a high An content (An_93–96) in plagioclase, reaching 85% of the mode in some rocks; iii) a high Al₂O₃ whole-rock content (31%); and iv) REE and trace element patterns similar to the plagioclase REE pattern. These characteristics are also found in numerous anorthosites (Ashwal and Myers, 1994; Huang et al., 2012; Polat et al., 2017; Karmakar et al., 2017), sakenites (Raith et al., 2008), and amphibolites (Berger et al., 2010). Pyroxene-bearing UMM series similar to the M1 series in the Castillon massif have been found in many areas of the North Pyrenean massifs (Fonteilles, 1976), Bessède (Albarède and Fourcade, 1969), Saint Barthélémy (Zwart, 1965; de Saint Blanquat, 1989; Lemirre, 2018; Lemirre et al., 2019), Ursuya in the Basque country (Boissonnas et al., 1974; Vielzeuf, 1984; Lemirre, 2018), and in the European Variscan chain (Wilson et al., 2004 and references therein). In contrast, only one occurrence of the M2 series (Hbl gabbros M2) has been observed, being described by Berger et al. (2010) in the Variscan French Massif Central.

5.2 Age of the Castillon magmatic event

The M2 Hbl gabbro series exhibits consistent εNd(t) for ages between 460 and 520 Ma and a minimum TDM2 range between 337 and 492 Ma. The consistency between these two independent calculations suggests that these magmas formed between 337 and 492 Ma, at least for the most juvenile samples of Castillon. In the literature, an early–mid Ordovician “Sardic” episode (ca. 470 Ma) of magmatic activity has been recognized, leading to the emplacement of voluminous aluminous granites as protoliths of the orthogneissic Aston (Denèle et al., 2009; Mezger and Gerdes, 2016), Hospitalet (Denèle et al., 2009), Canigou (Cocherie et al., 2005), Roc-de-France (Cocherie et al., 2005; Castiñeiras et al., 2008), and Albera (Liesa et al., 2011; Martinez et al., 2011; Casas et al., 2015; Navidad et al., 2018) massifs and coeval basic series such as the Cortalet metabasites in the Catalanides (Liesa et al., 2011; Navidad et al., 2018) and the Pierrefitte basaltic and andesitic series in the western part of the Pyrenees (Calvet et al., 1988; Cabanis and Pouit, 1988). The Cortalet series has been dated (U–Pb zircon) at 460 ± 3 Ma (Navidad et al., 2018), and the Pierrefitte series is considered Ordovician based on its stratigraphic position (Calvet et al., 1988). Therefore, it can be assumed that all UMM samples from Castillon belong to the same or chronologically similar magmatic event(s). Furthermore, Henry et al. (1998) showed that an errochron of 462 ± 20 Ma agrees with whole-rock and separated minerals on garnet pyroxenites collected at multiple Pyrenees sites (Lherz, Moncaup–Arguneos, Montcaut–Hourat, Freychinede, Pic de Geral, Fontete Rouge, and Caussou). As this age agrees with our proposed emplacement age range for the Castillon UMM, it might be interpreted as the age of a major thermal event that influenced the mantle at that time.

5.3 Amphibole and magmatic evolution

Hornblende is a ubiquitous mineral within the UMM rocks of the Castillon massif. It clearly testifies to a hydrated magmatic melt having crystallized or circulated within the investigated samples. Amphibole controls all or part of the whole-rock trace element signature, at least for the most impoverished samples (ultramafic rocks; Figures 5, 7). In ultramafic rocks, the chondrite-normalized REE and primitive mantle trace element patterns of whole rock and amphiboles (Figures 5, 7) are very similar (see Supplementary Table S1 for Hbl LA-ICP-MS data). For example, the whole-rock and Hbl REE patterns of the Hbl and Hbl-rich harzburgites run very parallel (Figure 7), with more enriched Amp patterns and the whole-rock ∑REE increasing. The primitive mantle-normalized trace element patterns also run parallel to each other (Figures 5, 7). Nonetheless, Hbl-poor harzburgite exhibits some singularities. The REE pattern of its Amp is slightly different from that of the whole rock and also differs from those of Amp from Hbl and Hbl-rich harzburgites. In norites and gabbro–norites, the REE patterns of whole rock and Hbl are nearly parallel (Figure 9), whereas those of Amp are more REE-enriched. For example, the average total REE in norites is 70 ppm compared to 235 ppm for Hbl, whereas in the gabbro–norites, the total REE averaged 63 ppm compared to 200 ppm for Hbl. Trace element patterns are also parallel to each other with the same anomalies. In contrast, the clinopyroxene from the gabbro–norites displays quite different trace element patterns in regard to both the concentration and elemental anomalies for elements such as Th, U, Nb, Pb, and Sr (Figure 9; Supplementary Table S2).

From isotopic data, we can postulate that the melt from which Hbl crystallized in Hbl- and Hbl-rich harzburgites and in M1 Px-bearing mafic rocks probably has a crustal origin, more particularly derived from the lower continental crust. Indeed, the samples in these two rock groups have the lowest Nd isotopic compositions (εNd₍₄₆₀₎ down to −1.62) while retaining moderately radiogenic Sr isotopic signatures (0.703–0.706), which is one of the characteristics of the lower crust (e.g., Rudnick and Fountain, 1995; Kemp and Hawkesworth, 2004). In Hbl gabbros (M2 series), the REE and trace element content and patterns of whole rocks and coexisting Hbl are almost identical (Figure 11; Supplementary Table S3). It is worth noting that the REE whole rock and Amp patterns are comparable to typical plagioclase patterns, with a strong Eu positive anomaly (Eu/Eu* reaching 8) and a strong (La/Sm)_N ratio reaching 6 (Figure 11). Huang et al. (2012), studying the Fiskenæsset metamagmatic UMM complex, have also reported such uncommon Amp features. These authors did not provide a formal interpretation of how Hbl REE composition could shift to plagioclase composition, particularly in terms of Eu content. We propose that this could occur through diffusion of Eu from plagioclase to Hbl, mainly in Hbl leucogabbros, during the granulite metamorphic episode related to the Variscan orogeny (Roux, 1977; Lemirre, 2018).

5.3.1 Nature and evolution of the M2 hornblende gabbro series

Among the UMM series of the Castillon massif, the Hbl gabbro forms a singular series. These rocks are characterized by a high-modal content of An-rich plagioclase (up to 96) and a high Mg# (up to 86), but they lack Px. They display chemical similarities with anorthosites and sakenites; therefore, the formation and evolution of the Hbl gabbro can be analyzed based on these similarities. The origin of anorthosites has been widely discussed (e.g., Ashwall, 2010; Ashwal and Myers, 1994; Ashwal et al., 1998; Latypov et al., 2020; Takagi et al., 2005; Polat et al., 2017). Latypov et al. (2020) considered that the parental magmas of anorthosites are alumina-rich basaltic or basaltic andesitic residual melts derived from deep-seated magma chambers. The transfer of the differentiated melt from the magma chamber to shallower magmatic reservoirs implies a subsequent pressure decrease and a superheating process that leads to the crystallization of An-rich plagioclase by shifting the melt composition to the divariant domain of plagioclase as a liquidus phase before the crystallization of a second phase when the monovariant curve is reached (Latypov et al., 2020).

Given that the Hbl leucogabbros in this study share similarities with anorthosites, we propose that a similar process could be at the origin of the Castillon M2 series, in which Hbl corresponds to the second crystallizing mineral phase. The crystallization of An-rich plagioclase under high-temperature conditions (Latypov et al., 2020) would result from the rapid uplift of the hydrated melt from which the Mg-rich Hbl from the M2 series could crystallize as an intercumulus phase. In the Castillon massif, the anorthositic characteristic of the M2 Hbl gabbros is more pronounced in the leucogabbros located at the top of the series than in the underlying meso- and mela-Hbl gabbros. This could be associated with the increase in the modal proportion of the most An-rich plagioclase toward the top of the series through a density control of the plagioclase segregation during the crystallization toward the top of the pile (Higgins et al., 2015; Namur et al., 2011; Polat et al., 2017). Most rocks from the Hbl gabbro M2 series are characterized by very juvenile Nd isotopic signatures, ruling out any crustal contamination of their mantle-derived parental melt. However, almost all samples from the UM M1 series are affected by this crustal contamination process. The latter display εNd₍₄₆₀₎ values between +5.70 and +6.64, i.e., signatures similar to those from the Hbl-bearing series. Therefore, it is possible that the magmatic events that led to the genesis of both series are associated. This assumption implies that the crustal contamination by lower crust fluids, as indicated by the isotopic Nd–Sr characteristics, has heterogeneously affected the entire investigated Castillon massif, although preferentially its lower sections where UM and M1 rocks dominate (Figure 2B).

5.4 Petrogenetic hypothesis

The North Pyrenean Castillon magmatic complex consists of meter- to decameter-thick UMM rock–bearing layers interlayered within infra-Silurian detrital sedimentary series; the entire complex is affected by Hercynian granulitic metamorphism. Our study defines two magmatic suites, both containing magmatic Hbl. The first suite consists of ultramafic rocks and an M1 Px-bearing mafic suite (UM, norites, gabbro–norites, and gabbros s.s.) that are derived from similar hydrated—note the occurrence of magmatic Hbl—parental basaltic magmas at different levels of differentiation. The second suite, an M2 Px-free Hbl gabbro (Ca-rich plagioclase + Amp) with a modal composition spanning from melagabbros very rich in Hbl, is found in the center of the full (M1 and M2) magmatic pile and is capped by anorthositic-like leucogabbros at the top of the pile. The key features of this pair lead us to propose a hypothesis for the origin and evolution of the UMM series from the Castillon massif. We propose a two-step origin and evolution. The first step corresponds to the history of the cumulate ultramafic and Px-bearing mafic rock M1 series. The second is associated with the Hbl gabbro M2 series.

Step 1 involves the emplacement of Ol-saturated basaltic melts that crystallize ultramafic rocks (dunites, harzburgites, and orthopyroxenites) and Px-bearing mafic rocks (norites, gabbro–norites, and gabbros s.s.). These magmatic rocks derive from the crystallization of basaltic parent magmas, likely having the same affinity but at distinct stages of differentiation in relation to petrogenetic processes, e.g., fractional crystallization, crystal settling, or magma mixing in magma chambers. Simultaneously, during the crystallization processes, these rocks were heterogeneously percolated by a crustal melt, leading to an increase in the Amp mode with the occurrence of some hornblendite layers interbedded with harzburgites and an increased total REE content, higher (La/Yb)_N ratios, positive Rb, Ba, and Th anomalies in some samples, and the subsequent scattering of isotope data. Few samples were preserved from this contamination, as evidenced only by a single norite and a pair of gabbro–norites still displaying a juvenile isotopic composition.

Step 2 is the emplacement of a second melt corresponding to a superheated alumina-rich basaltic or basaltic–andesite magma differentiated in a deeper magma reservoir. This melt is emplaced at the same structural level as the melt from step 1 and crystallizes the Hbl mela-, meso-, and leucogabbros. The geochemical characteristics of these rocks contrast with those of the M1 series. Their geochemical characteristics indicate an anorthositic tendency increasingly pronounced toward the top of the pile (leucogabbros). These Hbl mela-, meso-, and leucogabbros underwent a metamorphic event, leading to the formation of Spl, Spr, and gedrite (Kilzi, 2014), together with only a slight heterogeneous crustal contamination, as most samples preserve the juvenile isotopic signature.

In almost all the investigated rock types, Amp—a ubiquitous mineral—controls the chemical budget and evolution as the Hbl REE patterns are almost identical to those of the corresponding whole rocks. The Hbl gabbros, particularly the Hbl leucogabbros, are an exception because, for these, the REE budget and whole-rock patterns, along with the Amp itself, are controlled by plagioclase. Variscan metamorphism is developed and underlined by the crystallization of anthophyllite in some ultramafic rocks and by a secondary mineral assemblage of Spl + Spr + gedrite in the Hbl leucogabbros (Kilzi, 2014). This metamorphism does not fundamentally alter the overall composition of the rocks, thereby indicating that the petrological evolution of the Castillon massif was clearly the result of magmatic phenomena.

5.5 Geodynamic scenario

The geodynamic context in which these UMM series were emplaced can be developed using data obtained from samples having the most juvenile isotopic signatures, that is, those least likely to have been affected by crustal contamination. We, therefore, focus on the Hbl gabbro M2 series, having εNd₍₄₆₀₎ values between +4 and +8. As discussed previously, the εNd₍₄₆₀₎ values obtained in this study are among the highest published for Ordovician samples from southern Europe. Localities having similar signatures (from similar types of sample) include the Massif Central (Pin and Marini, 1993) and the Órdenes complex in northwestern Spain (Andonaegui et al., 2016). Such isotopic signatures for mafic samples, whose emplacement environment—layers in metasedimentary rocks—resembles that of our samples, are characteristic of either a pre-rifting stage (Pin and Marini, 1993) or continental subduction (Andonaegui et al., 2016). For either hypothesis, this implies the genesis of basaltic liquids from a more or less contaminated depleted mantle and a migration into and emplacement of these liquids in the continental crust. The concomitance of radiogenic isotopic values and the absence of negative Nb anomalies suggest an origin closer to that of an oceanic plateau or early rifting magmatism (Pin and Marini, 1993). The parental magmas of our rock series would thus be generated either within a mantle plume or the lithospheric mantle and then emplaced within the lower metasedimentary continental crust, thus being more or less contaminated during crystallization. The hypothesis of mantle superplume activity in the Ordovician has been proposed by Casas et al. (2024) based on felsic volcanism in the Pyrenees and Mouthoumet (Corbières, France) massifs.

Our two-stage hypothesis explaining the origin of the pair of identified magmatic series is possible under such a scenario and could explain why the Hbl-bearing series are less affected by crustal contamination. Indeed, the emplacement of the first series, the ultramafic and Px-bearing M1 mafic series, would have already been affected and modified by the deep crust, leaving the parental melt of the Hbl M2 series relatively free of contamination.

6 Conclusion

This study on the UMM rock series of the Castillon massif advances our geological understanding of the North Pyrenean massifs. We identified two magmatic suites. The first is a cumulate UM suite comprising dunites, harzburgites, and orthopyroxenites, with minor hornblendites at the bottom of the pile, covered by a Px-bearing mafic series of mainly norite and gabbro–norite, heterogeneously contaminated by crustal melts/fluids. This M1 suite is similar to the previously described mafic magmatic series in the North Pyrenean massifs. The second magmatic suite is a predominantly juvenile differentiated suite of Hbl gabbros, having Hbl leucogabbros with anorthositic characteristic at the top of the pile. These leucogabbros, typified by a Crn + Spl + Spr association, have rarely been observed in the European Variscan chain. However, their isotopic similarities with some mafic rocks from the Massif Central and northwestern Spain suggest that they are probably more widely represented. The Ordovician age attributed to this magmatic evolution assigns UMM Castillon magmatism to the early–mid Ordovician Sardic episode (ca. 460 Ma), whereas the metamorphism and structuring of the massif are placed in the Carboniferous (330 Ma, Variscan). The geochemical similarities of the Hbl gabbros, particularly of the Hbl leucogabbros, with anorthositic complexes are noteworthy. The juvenile isotopic signatures of the Hbl gabbros have been rarely observed in southern Europe for this geological period. These signatures are commonly associated with a mantle plume event affecting the lower continental crust or a passive sedimentary margin. The Castillon massif may be a relic of such geodynamic events during the Ordovician.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material; further inquiries can be directed to the corresponding author.

Author contributions

MK: Writing – original draft, Writing – review and editing. MG: Writing – original draft, Writing – review and editing. YD: Writing – original draft, Writing – review and editing. MB: Writing – original draft, Writing – review and editing. PD: Writing – original draft, Writing – review and editing. LB: Writing – original draft, Writing – review and editing.

Funding

The author(s) declared that financial support was not received for this work and/or its publication.

Acknowledgements

The authors are grateful to Julien Berger for his help and to A.-M. Cousin for her assistance with figures. The authors thank the two reviewers for their helpful comments and corrections and the handling editor for the editorial work. The authors dedicate this work to the memory of Louis Roux, who worked for many years, particularly during his PhD thesis, on the Castillon massif.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The authors MG, LB declared that they were an editorial board member of Frontiers at the time of submission. This had no impact on the peer review process and the final decision.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/feart.2025.1609542/full#supplementary-material

References

Abraham, K., Monchoux, P., Roux, L., and Seifert, F. (1977). Observations sur les exsolutions dans les orthoamphiboles des roches à sapphirine des Pyrénées. Bull. Minéralogie 100, 329–333. doi:10.3406/bulmi.1977.7157

CrossRef Full Text | Google Scholar

Albarède, F., and Fourcade, S. (1969). Sur la présence de gneiss à orthopyroxène dans le socle hercynien du massif nord-Pyrénéen de Bessèdes-de-Sault (Aude). Comptes Rendus l’Académie Sci. Paris 269, 1823–1826.

Google Scholar

Andonaegui, P., Castiñeiras, P., Cuadra, P. G., Arenas, R., Martínez, S. S., Abati, J., et al. (2012). The corredoiras orthogneiss (NW iberian massif): geochemistry and geochronology of the Paleozoic magmatic suite developed in a peri-Gondwanan arc. Lithos 128-131, 84–99. doi:10.1016/j.lithos.2011.11.005

CrossRef Full Text | Google Scholar

Andonaegui, P., Sánchez-Martínez, S., Castiñeiras, P., Abati, J., and Arenas, R. (2016). Reconstructing subduction polarity through the geochemistry of mafic rocks in a Cambrian magmatic arc along the gondwana margin (Órdenes complex, NW iberian massif). Int. J. Earth Sci. 105, 713–725. doi:10.1007/s00531-015-1195-x

CrossRef Full Text | Google Scholar

Annen, C., Blundy, J. D., and Sparks, R. S. J. (2005). The genesis of intermediate and silicic magmas in deep crustal hot zones. J. Petrology 47, 505–539. doi:10.1093/petrology/egi084

CrossRef Full Text | Google Scholar

Ashwal, L. D., and Myers, J. S. (1994). “Chapter 8 Archean anorthosites,” Elsevier Developments in Precambrian geology. Editor K. C. Condie 11, 315–355. doi:10.1016/s0166-2635(08)70226-3

CrossRef Full Text | Google Scholar

Ashwal, L. D., Hamilton, M. A., Morel, V., and Rambeloson, R. A. (1998). Geology, petrology and isotope geochemistry of massif-type anorthosites from southwest Madagascar. Contributions Mineralogy Petrology 133 (4), 391–401. doi:10.1007/s004100050461

CrossRef Full Text | Google Scholar

Ashwall, L. D. (2010). The temporality of anorthosites. Can. Mineralogist 48, 711–728. doi:10.3749/canmin.48.4.711

CrossRef Full Text | Google Scholar

Barnolas, A., and Chiron, J. C. (1996). Synthèse géologique et géophysique des Pyrénées: introduction, Géophysique, Cycle hercynien. Orléans: Bureau de Recherches Géologiques et Minières (BRGM), 729.

Google Scholar

Barrat, J. A., Keller, F., Amossé, J., Taylor, R. N., Nesbitt, R. W., and Hirata, T. (1996). Determination of rare Earth elements in sixteen silicate reference samples by Icp-Ms after Tm addition and ion exchange separation. Geostand. Newsl. 20, 133–139. doi:10.1111/j.1751-908x.1996.tb00177.x

CrossRef Full Text | Google Scholar

Barrat, J. A., Yamaguchi, A., Greenwood, R. C., Bohn, M., Cotten, J., Benoit, M., et al. (2007). Are stannern-trend eucrites ordinary eucrites contaminated by crustal partial melts. Meteorit. and Planet. Sci. 42 (A17)–7. 816.

Google Scholar

Berger, J., Féménias, O., Ohnenstetter, D., Plissart, D., and Mercier, J. C. (2010). Origin and tectonic significance of corundum-kyanite-sapphirine amphibolites from the Variscan French massif central. J. Metamorph. Geol. 28, 341–360. doi:10.1111/j.1525-1314.2010.00866.x

CrossRef Full Text | Google Scholar

Boissonnas, J., Le Pochat, G., Thibault, Cl., and Bernatsky, M. (1974). Carte géologique au 1/50 000. Orléans: Bureau de Recherches Géologique et Minière, Orléans.

Google Scholar

Cabanis, B., and Pouit, G. (1988). “Etude géochimique du volcanisme basique ordovicien de Pierrefitte,” in Contribution 224 Synthèse géologique et géophysique des Pyrénées.

Google Scholar

Calvet, P., Lapierre, H., and Charvet, J. (1988). Diversité du volcanisme Ordovicien dans la région de Pierrefitte (Hautes Pyrénées): rhyolithes calco-alcalines et basalts alcalins. Comptes Rendus l’Académie Sci. Paris 307 (II), 805–812.

Google Scholar

Casas, J. M., Navidad, M., Castiñeiras, P., Liesa, M., Aguilar, C., Carreras, J., et al. (2015). The late Neoproterozoic magmatism in the Ediacaran series of the eastern pyrenees: new ages and isotope geochemistry. Int. J. Earth Sci. 104, 909–925. doi:10.1007/s00531-014-1127-1

CrossRef Full Text | Google Scholar

Casas, J. M., Sánchez-García, T., Díez-Montes, A., Clariana, P., Margalef, A., Valverde-Vaquero, P., et al. (2024). Extent and significance of the Upper Ordovician felsic volcanism in the pyrenees and mouthoumet massifs, 542. London: Geological Society, 433–463.

Google Scholar

Castiñeiras, P., Navidad, M., Liesa, M., Carreras, J., and Casas, J. M. (2008). U–Pb zircon ages (SHRIMP) for cadomian and early Ordovician magmatism in the eastern pyrenees: new insights into the pre-Variscan evolution of the northern gondwana margin. Tectonophysics 461, 228–239. doi:10.1016/j.tecto.2008.04.005

CrossRef Full Text | Google Scholar

R. G. Cawthorn (1996). Layered intrusions (New York: Elsevier), 531.

Google Scholar

Chai, F., Zhang, Z., Mao, J., Dong, L., Zhang, Z., and Wu, H. (2008). Geology, petrology and geochemistry of the baishiquan Ni–Cu-Bearing mafic–ultramafic intrusions in Xinjiang, NW China: implications for tectonics and genesis of ores. Ournal Asian Earth Sci. 32, 218–235. doi:10.1016/j.jseaes.2007.10.014

CrossRef Full Text | Google Scholar

B. Charlier, O. Namur, R. Latypov, and C. Tegner (2015). Layered intrusions (London: Springer), 748.

Google Scholar

Cocherie, A., Baudin, T., Autran, A., Guerrot, C., Fanning, C. M., and Laumonier, B. (2005). U-Pb zircon (ID-TIMS and SHRIMP) evidence for the early Ordovician intrusion of metagranites in the late Proterozoic canaveilles group of the Pyrenees and the montagne noire (France). Bull. Société Géologique Fr. 176, 269–282. doi:10.2113/176.3.269

CrossRef Full Text | Google Scholar

Correia, C. T., Sinigoi, S., Girardi, V. A. V., Mazzucchelli, M., Tassinari, C. C. G., and Giovanardi, T. (2012). The growth of large mafic intrusions: comparing Niquelândia and Ivrea igneous complexes. Lithos 155, 167–182. doi:10.1016/j.lithos.2012.08.024

CrossRef Full Text | Google Scholar

De Saint Blanquat, M. (1989). La faille normale ductile du massif du Saint Barthélémy: âge et signification de l'extension crustale dans la zone nord pyrénéenne. Montpellier: Thèse de doctorat : Géologie, 272p.

Google Scholar

De Saint Blanquat, M. (1993). La faille normale ductile du massif du Saint Barthélémy. Evolution hercynienne des massifs nord-pyrénéens catazonaux considérée du point de vue de leur histoire thermique. Geodin. Acta 6 (1), 59–77. doi:10.1080/09853111.1993.11105239

CrossRef Full Text | Google Scholar

De Sitter, L. U. (1964). Structural geology. 2nd Edition. London: Mc Graw-Hill Book Company, 551.

Google Scholar

De Sitter, L. U., and Zwart, H. J. (1960). Geological map of the central Pyrenees Sheet 4, Valle de Arán, Spain [Document cartographique]/editors; map prepared by W. F. J. Kleinsmiede, H. J. Zwart/Leiden, and J. J. Groen. Leiden, Netherland: Leiden University.

Google Scholar

Denèle, Y., Barbey, P., Deloule, E., Pelleter, E., Olivier, Ph., and Gleizes, G. (2009). Middle Ordovician U-Pb age of the Aston and Hospitalet orthogneissic laccoliths: their role in the Variscan evolution of the Pyrenees. Bull. la Société Géologique Fr. 180, 209–216. doi:10.2113/gssgfbull.180.3.209

CrossRef Full Text | Google Scholar

Denele, Y., Olivier, P., Gleizes, G., and Barbey, P. (2007). The Hospitalet gneiss dome (Pyrenees) revisited: lateral flow during Variscan transpression in the middle crust. Terra Nova 19 (6), 445–453. doi:10.1111/j.1365-3121.2007.00770.x

CrossRef Full Text | Google Scholar

DePaolo, D. J. (1981). Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization. Earth Planet. Sci. Lett. 53, 89–202. doi:10.1016/0012-821x(81)90153-9

CrossRef Full Text | Google Scholar

Downes, H., and Leyreloup, A. (1986). Granulitic xenoliths from the French Massif Central -petrology, Sr and Nd isotope systematics and model age estimates, 24. London, Special Publications: Geological Society, 319–330.

CrossRef Full Text | Google Scholar

Driouch, Y. (1997). “Pétrologie et origine des roches à cordiérite -orthoamphibole de la vallée d'Héas (Hautes Pyrénées). Le magmatisme basique des Pyrénées varisques: pétrologie et évolution. Implications sur l'évolution des Pyrénées varisques. Thèse,” in Université Sidi Mohamed Ben Abdallah. Fès. Maroc. 367p.

Google Scholar

Driouch, Y., Dahmani, A., Debat, P., Seyler, M., and Soula, J. C. (1989). Les formations ultrabasiques et basiques au sein des séries mésocrustales des Pyrénées hercyniennes. Comptes Rendus l'Académie Sci. - Series 2. Paris 308, 1249–1255.

Google Scholar

Femenias, O., Coussaert, N., Bingen, B., Whitehouse, M., Mercier, J-C., and Demaiffe, D. (2003). A Permian underplating event in late- to post-orogenic tectonic setting. Evidence from the mafic–ultramafic layered xenoliths from Beaunit (French Massif Central). Chem. Geol. 199, 293–315. doi:10.1016/S0009-2541(03)00124-4

CrossRef Full Text | Google Scholar

Féménias, O., Coussaert, N., Brassinnes, S., and Demaiffe, D. (2005). Emplacement processes and cooling history of layered cyclic unit II-7 from the Lovozero alkaline massif (Kola Peninsula, Russia). Lithos 83, 371–393. doi:10.1016/j.lithos.2005.03.012

CrossRef Full Text | Google Scholar

Fonteilles, M. (1976). “Essai d’interprétation des compositions chimiques des roches d'origine métamorphique et magmatique du massif hercynien de l'Agly (Pyrénées Orientales),” in Thèse Université de Paris VI. 2 Tomes.

Google Scholar

Ford, M., Masini, E., Vergés, J., Pik, R., Ternois, S., Léger, J., et al. (2022). Evolution of a low convergence collisional orogen: a review of Pyrenean orogenesis. Bull. la Société Géologique Fr. 193, 19. doi:10.1051/bsgf/2022018

CrossRef Full Text | Google Scholar

Gamisel-Muzás, Á., Soto, R., Ayala, C., Mochales, T., Rubio, F. M., Clariana, P., et al. (2025). The geological structures of the Pyrenees and their peripheral basins examined through EMAG2v2 magnetic data. Solid earth. 16, 251–274. doi:10.5194/se-16-251-2025

CrossRef Full Text | Google Scholar

Griffin, W., Powell, W., Pearson, N. J., and O’Reilly, S. Y. (2008). Glitter: data reduction software for laser ablation ICP-MS. Short. Course Ser. 40, 308–311. doi:10.3749/9780921294801.app02

CrossRef Full Text | Google Scholar

Grove, T. L., Donnelly-Nolan, J. M., and Housh, T. (1997). Magmatic processes that generated the rhyolite of glass Mountain, medicine Lake volcano, N. California. Contributions Mineralogy Petrology 127, 205–223. doi:10.1007/s004100050276

CrossRef Full Text | Google Scholar

Gruffat, J. J. (1992). Advances in X-ray analysis, 35. New York: Plenum, 755–756.

Google Scholar

Han, R., Min, K., Ree, J.-H., and Foster, D. A. (2014). Extensional deformation along the southern boundary of the Gyeonggi Massif, South Korea: structural characteristics, age constraints, and tectonic implications. Int. J. Earth Sci. 103, 757–776. doi:10.1007/s00531-013-0985-2

CrossRef Full Text | Google Scholar

Henry, P., Azambre, B., Montigny, R., Rossy, M., and Stevenson, R. K. (1998). Late mantle evolution of the pyrenean sub-continental lithospheric mantle in the light of new 40Ar–39Ar and Sm-Nd ages on pyroxenites and peridotites (Pyrenees, France). Tectonophysics 296, 103–123. doi:10.1016/s0040-1951(98)00139-5

CrossRef Full Text | Google Scholar

Higgins, M. D., Voos, S., and Vander Auwera, J. (2015). Magmatic processes under Quizapu volcano, Chile, identified from geochemical and textural studies. Contributions Mineralogy Petrology 170, 51. doi:10.1007/s00410-015-1209-5

CrossRef Full Text | Google Scholar

Hildreth, W., and Moorbath, S. (1988). Crustal contribution to arc magmatism in the Andes of central Chile. Contributions Mineralogy Petrology 98, 455–489. doi:10.1007/bf00372365

CrossRef Full Text | Google Scholar

Huang, H., Polat, A., Fryer, B. J., Appel, P. W. U., and Windley, B. F. (2012). Geochemistry of the Mesoarchean fiskenæsset complex at majorqap qâva, SW Greenland: evidence for two different magma compositions. Chem. Geol. 314–317, 66–82. doi:10.1016/j.chemgeo.2012.04.020

PubMed Abstract | CrossRef Full Text | Google Scholar

Jesus, A. P., Mateus, A., Benoit, M., Tassinari, C. C. G., and Bento dos Santos, T. (2020). The timing of sulfide segregation in a Variscan synorogenic gabbroic layered intrusion (Beja, Portugal): implications for Ni-Cu-PGE exploration in orogenic settings. Ore Geol. Rev. 126, 103767. doi:10.1016/j.oregeorev.2020.103767

CrossRef Full Text | Google Scholar

Karmakar, S., Mukherjee, S., Sanyal, S., and Sengupta, P. (2017). Origin of peraluminous minerals (corundum, spinel, and sapphirine) in a highly calcic anorthosite from the Sittampundi layered complex, Tamil Nadu, India. Contributions Mineralogy Petrology 172 (8), 1–23. doi:10.1007/s00410-017-1383-8

CrossRef Full Text | Google Scholar

Kemp, A. I. S., and Hawkesworth, C. J. (2004). “Granitic perspectives on the generation and secular evolution of the Continental Crust,” in Treatise on geochemistry 12, the crust. Editor R. Rudnick (Oxford: Elsevier Science), 349–410.

CrossRef Full Text | Google Scholar

Kilzi, M. A. (2014). Etude pétro-chimique des formations mafiques et ultramafiques associées aux formations granulitiques, métamorphiques et anatectiques des Pyrénées hecyniennes (massif de Castillon et massif de Gavarnie). Toulouse: Thèse Université Paul Sabatier.

Google Scholar

Kilzi, M. A., Grégoire, M., Bosse, V., Benoit, M., Driouch, Y., de Saint Blanquat, M., et al. (2016). Geochemistry and zircon U-Pb geochronology of the ultramafic and mafic rocks emplaced within the anatectic series of the variscan Pyrenees: the example of the gavarnie-heas dome (France). Comptes Rendus Geosci. 348, 107–115. doi:10.1016/j.crte.2015.06.014

CrossRef Full Text | Google Scholar

Latypov, R. M., Chistyakova, S., Costin, G., Namur, O., Barnes, S., and Kruger, W. (2020). Dynamics of evolving magma chambers: textural and chemical evolution of cumulates at the arrival of new liquidus phases. Sci. Rep. 10, 3839. doi:10.1038/s41598-020-60778-w

PubMed Abstract | CrossRef Full Text | Google Scholar

Le Maitre, R. W., Streckeisen, A., Zanettin, B., Le Bas, M. J., Bonin, B., and Bateman, P. (2004). Igneous Rocks. A Classification of Igneous Rocks and Glossary of Terms. Le Maître R.E. U.K: Cambrigde University Press, 236.

Google Scholar

Le Roux, V., Bodinier, J. L., Tommasi, A., Alard, O., Dautria, J. M., Vauchez, A., et al. (2007). The lherz spinel lherzolite: refertilized rather than pristine mantle. Earth Planet. Sci. Lett. 259, 599–612. doi:10.1016/j.epsl.2007.05.026

CrossRef Full Text | Google Scholar

Leake, B. E., Woolley, A. R., Arps, C. E. S., Birch, W. D., Gilbert, M. C., Grice, J. D., et al. (1997). Nomenclature of amphiboles: report of the subcommittee on amphiboles of the international mineralogical association commission on new minerals and mineral names. Can. Mineralogist 35, 219–246. doi:10.1180/minmag.1997.061.405.13

CrossRef Full Text | Google Scholar

Lemirre, B. (2018). Origine et développement de la thermicité dans les Pyrénées varisques. Géologie Appliquée. Univ. Paul. Sabati Er - Toulouse III.

Google Scholar

Lemirre, B., Cochelin, B., Duchene, S., De Saint Blanquat, M., and Poujol, M. (2019). Origin and duration of late orogenic magmatism in the foreland of the Variscan belt (lesponne – Chiroulet – neouvielle area, French Pyrenees). Lithos 336-337, 183–201. doi:10.1016/j.lithos.2019.03.037

CrossRef Full Text | Google Scholar

Liesa, M., Carreras, J., Castineiras, P., Casas, J. M., Navidad, M., and Vila, M. (2011). U-Pb zircon age of Ordovician magmatism in the Albera massif (eastern Pyrenees). Geol. Acta 9, 93–101. doi:10.1344/105.000001651

CrossRef Full Text | Google Scholar

Lotout, C., Pitra, P., Poujol, M., and Van Den Driessche, J. (2017). Ordovician magmatism in the Lévézou massif. Int. J. Earth Sci. 106, 501–515. (French Massif Central): tectonic and geodynamic implications. doi:10.1007/s00531-016-1387-z

CrossRef Full Text | Google Scholar

Marchesi, C., Garrido, C. J., Godard, M. F., Belley, F., and Ferré, E. (2009). Migration and accumulation of ultra-depleted subduction-related melts in the Massif du Sud ophiolite (New Caledonia). Chem. Geol. 266, 171–186. doi:10.1016/j.chemgeo.2009.06.004

CrossRef Full Text | Google Scholar

Martínez, F. J., Iriondo, A., Dietsch, C., Aleinikoff, J. N., Peucat, J. J., Cirès, J., et al. (2011). U-Pb SHRIMP-RG zircon ages and Nd signature of lower Paleozoic rifting-related magmatism in the Variscan basement of the Eastern Pyrenees. Lithos 127, 10–23. doi:10.1016/j.lithos.2011.08.004

CrossRef Full Text | Google Scholar

Matte, P. (1991). Accretionary history and crustal evolution of the variscan belt in the Western Europe. Tectonophysics 196, 309–337. doi:10.1016/0040-1951(91)90328-p

CrossRef Full Text | Google Scholar

Mezger, J. E., and Gerdes, A. (2016). Early Variscan (Visean) granites in the core of central 1003 pyrenean gneiss domes: implications from laser ablation U-Pb and Th-Pb studies. Gondwana 1004 Res. 29, 181–198. doi:10.1016/j.gr.2014.11.010

CrossRef Full Text | Google Scholar

Monchoux, P. (1972). Roches à sapphirine au contact des lherzolites pyrénéennes. Contributions 1013 Mineralogy Petrology 37, 47–64. doi:10.1007/bf00377306

CrossRef Full Text | Google Scholar

Monchoux, P., and Roux, L. (1973). Péridotites et amphibolites à saphirine du massif de Castillon 1007 (Pyrénées airiègeoises), 279. 499. C.R. Ac. Sc. Paris.

Google Scholar

Monchoux, P., and Roux, L. (1975). Les amphiboles orthorhombiques riches en aluminium et 1010 magnésium des amphiboles à saphirine des Pyrénées. Bull. de la Société Française de 1011 Minéralogie de Cristallogr. 98 (2- 3), 166–170. doi:10.3406/bulmi.1975.6978

CrossRef Full Text | Google Scholar

Montero, P., Bea, F., Corretgé, L. G., Floor, P., and Whitehouse, M. J. (2009). U-Pb ion microprobe dating and Sr and Nd isotope geology of the Galiñeiro Igneous complex. Lithos 107, 227–238. doi:10.1016/j.lithos.2008.10.009

CrossRef Full Text | Google Scholar

Namur, O., Charlier, B., Pirard, C., Hermann, J., Liégeois, J.-P., and Vander Auwera, J. (2011). 1020 anorthosite formation by plagioclase flotation in ferrobasalt, and implications for the lunar 1021 crust. Geochimica Cosmochimica Acta 75, 4998–5018. doi:10.1016/j.gca.2011.06.013

CrossRef Full Text | Google Scholar

Navidad, M., Castiñeiras, P., Casas, J. M., Liesa, M., Fernández Suárez, J., Barnolas, A., et al. (2010). Geochemical characterization and isotopic age of Caradocian magmatism in the northeastern iberian Peninsula: insights into the late Ordovician evolution 1026 of the northern Gondwana margin. Gondwana Res. 17, 325–337. doi:10.1016/j.gr.2009.11.013

CrossRef Full Text | Google Scholar

Navidad, M., Castiñeiras, P., Casas, J. M., Liesa, M., Belousova, E., Proenza, J., et al. (2018). Ordovician magmatism in the Eastern Pyrenees: implications for the geodynamic 1030 evolution of Northern Gondwana. Lithos 314-315, 479–496. doi:10.1016/j.lithos.2018.06.019

CrossRef Full Text | Google Scholar

Nicolaysen, L. (1961). Graphic interpretation of discordant age measurements on metamorphic 1033 rocks. Ann. N. Y. Acad. Sci. 91 (1), 198–206. doi:10.1111/j.1749-6632.1961.tb35452.x

CrossRef Full Text | Google Scholar

Orejana, D., Villaseca, C., and Merino Martínez, E. (2017). Basic Ordovician magmatism of the 1036 Spanish central system: constraints on the source and geodynamic setting. Lithos 284-285, 608–624. doi:10.1016/j.lithos.2017.05.012

CrossRef Full Text | Google Scholar

Philpotts, A. R., and Dickson, L. D. (2002). Millimeter-scale modal layering and the nature of 1040 the upper solidification zone in thick flood-basalt flows and other sheets of magma. J. 1041 Struct. Geol. 24, 1171–1177. doi:10.1016/s0191-8141(01)00099-2

CrossRef Full Text | Google Scholar

Pin, C. (1989). Essai sur la chronologie et l'évolution géodynamique de la chaîne hercynienne d’Europe. Clermont-Ferrand, France: Université Clermont-Ferrand, 470.

Google Scholar

Pin, C., and Marini, F. (1993). Early Ordovician continental break-up in variscan Europe: Nd-Sr 1050 isotope and trace element evidence from bimodal igneous associations of the Southern 1051 Massif Central, France. Lithos 29, 177–196. doi:10.1016/0024-4937(93)90016-6

CrossRef Full Text | Google Scholar

Pin, C., and Vielzeuf, D. (1983). Granulites and related rocks in Variscan median Europe: a 1044 dualistic interpretation. Tectonophysics 93, 47–74. doi:10.1016/0040-1951(83)90233-0

CrossRef Full Text | Google Scholar

Pin, C., Gannoun, A., and Dupont, A. J. (2014). Rapid, simultaneous separation of Sr, Pb, and Nd by 1054 extraction chromatography prior to isotope ratios determination by TIMS and MC-ICP-MS. J. Anal. Atomic Spectrom. 29, 1858–1870. doi:10.1039/c4ja00169a

CrossRef Full Text | Google Scholar

Polat, A., Kerrich, R., and Windley, B. (2009). Archaean crustal growth processes in southern 1058 West Greenland and the southern Superior Province: geodynamic and magmatic constraints, 318. London: Geological Society, 155–191.

Google Scholar

Polat, A., Fryer, B. J., Samson, I. M., Weisener, C., Appel, P. W. U., Frei, R., et al. (2012). Geochemistry of ultramafic rocks and hornblendite veins in the Fiskenæsset layered 1063 anorthosite complex, SW Greenland: evidence for hydrous upper mantle in the Archean. 1064 Precambrian Res. 214-215, 124–153. doi:10.1016/j.precamres.2011.11.013

CrossRef Full Text | Google Scholar

Polat, A., Frei, R., Longs taffe, F. J., and Woods, R. (2017). Petrogenetic and geodynamic origin 1067 of the Neoarchean Doré Lake complex, Abitibi subprovince, superior province, Canada. Int. J. Earth Sci. 107, 811–843. doi:10.1007/s00531-017-1498-1

CrossRef Full Text | Google Scholar

Pouchou, J., and Pichoir, F. (1984). Analyse d’échantillons stratifiés à la microsonde 1071 électronique. J. de Physique Colloques 45, C2–50.

Google Scholar

Raedeke, L. D., and McCallum, I. S. (1984). Investigations in the stillwater complex: part II. Petrology and petrogenesis of the ultramafic series. J. Petrology 25, 395–420. doi:10.1093/petrology/25.2.395

CrossRef Full Text | Google Scholar

Raguin, E. (1938). Contribution à l’étude des gneiss des Pyrénées. Bull. la Société 1074 Géologique Fr. 8, 11–36.

Google Scholar

Raith, M. M., Rokotodrazafy, R., and Sengupta, P. (2008). Petrology of corundum-spinel-sapphirine-anorthite rocks (sakenites) from the type locality in southern Madagascar. J. Metamorphic Geology 26 (6), 647–667. doi:10.1111/j.1525-1314.2008.00779.x

CrossRef Full Text | Google Scholar

Roux, L. (1977). L'évolution des roches du faciès granulite et le problème des ultramafitites dans le massif de Castillon (Ariège), 530p. Toulouse: Thèse de doctorat.

Google Scholar

Rudnick, R. L., and Fountain, D. M. (1995). Nature and composition of the continental crust: a lower crustal perspective. Rev. Geophys. 33, 267–309. doi:10.1029/95rg01302

CrossRef Full Text | Google Scholar

Sato, H. (1977). Nickel content of basaltic magmas: identification of primary magmas and a measure of the degree of olivine fractionation. Lithos 10, 113–120. doi:10.1016/0024-4937(77)90037-8

CrossRef Full Text | Google Scholar

Sisson, T. W., and Grove, T. L. (1993). Experimental investigations of the role of H2O in calcalkaline differentiation and subduction zone magmatism. Contributions Mineralogy Petrology 113, 143–166. doi:10.1007/bf00283225

CrossRef Full Text | Google Scholar

Solá, A. R., Pereira, M. F., Williams, I. S., Ribeiro, M. L., Neiva, A. M. R., Montero, P., et al. (2008). New insights from U–Pb zircon dating of early Ordovician magmatism on the 1091 northern Gondwana margin: the Urra formation (SW Iberian Massif, Portugal). Tectonophysics 461, 114–129. doi:10.1016/j.tecto.2008.01.011

CrossRef Full Text | Google Scholar

Streckeisen, A. (1974). Classification and nomenclature of plutonic rocks. Geol. Rdsch. 63, 773–786. doi:10.1007/BF01820841

CrossRef Full Text | Google Scholar

Sun, W., Ding, X., Hu, Y.-H., and Li, X.-H. (2007). The golden transformation of the Cretaceous 1095 plate subduction in the west Pacific. Earth Planet. Sci. Lett. 262, 533–542. doi:10.1016/j.epsl.2007.08.021

CrossRef Full Text | Google Scholar

Takagi, D., Sato, H., and Nakagawa, M. (2005). Experimental study of a low-alkali tholeiite at 1–5 kbar: optimal condition for the crystallization of high-An plagioclase in hydrous arc tholeiite. Contributions Mineralogy Petrology 149, 527–540. doi:10.1007/s00410-005-0666-7

CrossRef Full Text | Google Scholar

Takahashi, E. (1986). Melting of a dry peridotite KLB-1 up to 14 GPa: implications on the Origin of peridotitic upper mantle. J. Geophys. Res. Solid Earth 91, 9367–9382. doi:10.1029/jb091ib09p09367

CrossRef Full Text | Google Scholar

Thiebaut, J. (1964). Etude géologique du massif de la forêt de Castillon. Bull. Soc. Hist. Nat. Toulouse 99, 366–389.

Google Scholar

Tilhac, R., Gregoire, M., O’Reilly, S. Y., Griffin, W. L., Henry, H., and Ceuleneer, G. (2017). Sources and timing of pyroxenite formation in the sub-arc mantle: case study of the Cabo Ortegal 1106 complex, Spain. Earth Planet. Sci. Lett. 2017. doi:10.1016/j.epsl.2017.07.017

CrossRef Full Text | Google Scholar

Vielzeuf, D. (1980). Existence d’un fragment de « complexe stratifié » (pyriclasites et anorthosites) associé à la série granulitique du port de Saleix (Pyrénées ariégeoises). Bull. 1110 Minéralogie 103, 287–296. doi:10.3406/bulmi.1980.7406

CrossRef Full Text | Google Scholar

Vielzeuf, D. (1984). “Relation de phases dans le faciès granulite et implications géodynamiques. 1113 L'exemple des granulites des Pyrénées,” in Thèse clermont ferrand 359p.

Google Scholar

Vielzeuf, D., and Kornprobst, J. (1984). Crustal splitting and the emplacement of pyrenean 1116 lherzolites and granulites. Earth Planet. Sci. Lett. 67, 87–96. doi:10.1016/0012-821x(84)90041-4

CrossRef Full Text | Google Scholar

Wang, W., Wei, C., Wang, T., Lou, Y., and Chu, H. (2009). Confirmation of pelitic granulite in 1119 the Altai orogen and its geological significance. Chin. Sci. Bull. 54, 2543–2548. doi:10.1007/s11434-009-0041-6

CrossRef Full Text | Google Scholar

Whitney, D. L., and Evans, B. W. (2010). Abbreviations for names of rock-forming minerals. Am. Mineralogist 95, 185–187. doi:10.2138/am.2010.3371

CrossRef Full Text | Google Scholar

Wilson, M., Neumann, E.-R., Davies, G. R., Timmerman, M. J., Heeremans, M., and Larsen, B. T. (2004). “Permo-carboniferous magmatism and rifting in Europe,” in Geological society, 1123. London: Special Publications, 1–10.

Google Scholar

Windley, B. F., Bishop, F. C., and Smith, J. V. (1981). Metamorphosed layered igneous complexes in Archean granulite-gneiss belts. Annu. Rev. Earth Planet. Sci. 9, 175–196. doi:10.1146/annurev.ea.09.050181.001135

CrossRef Full Text | Google Scholar

York, D., Evensen, N. M., Martinez, M. L., and De Basabe Delgado, J. (2004). Unified equations 1129 for the slope, intercept, and standard errors of the best straight line. Am. J. Phys. 72 (3), 367–375. doi:10.1119/1.1632486

CrossRef Full Text | Google Scholar

Zingg, A. J. (1996). Recrystallization and the origin of layering in the Bushveld complex. Lithos 37, 15–37. doi:10.1016/0024-4937(95)00013-5

CrossRef Full Text | Google Scholar

Zwart, H. J. (1965). Geological map of the Paleozoic of the central Pyrenees, sheet 6, Aston, 1136 France, Andorra, Spain. Leidse Geol. Meded. 33 (1), 191–254.

Google Scholar