Introduction

The Variscan Orogen resulted from the Paleozoic convergence between Gondwana and Laurussia that culminated with the assembly of Pangea (Matte 2001; Arenas et al. 2016a; Díez Fernández et al. 2016). This orogen, which extends throughout central and western Europe, in parts of north and northwest of Morocco as well as in the Appalachian Mountains of northeastern America, contains the suture of the Rheic Ocean, other minor sutures, and sections of the thinned paleomargin of Gondwana, which hosted a long-lived (Avalonian–Cadomian) active magmatic arc (Dalziel 1997; Fuenlabrada et al. 2010; Pereira et al. 2012a, b; Albert et al. 2015; Andonaegui et al. 2016). Variscan deformation is penetrative and strong along the orogen, and Alpine overprinting affected some areas. Yet, the Bohemian Massif in Germany, the Central and Armorican massifs in France and Iberian Massif in Iberian peninsula (Fig. 1), contain well-preserved low-strain and low-grade Neoproterozoic to Early Paleozoic sedimentary series (Linnemann et al. 2000; Fernández-Suárez et al. 2002; Murphy et al. 2002; Pereira 2015). The correlation between the sedimentary series of these nowadays distant massifs, as well as improving the knowledge of the primary location of the paleobasins along the (North African) Gondwana margin, are being actively investigated through the study of paleofauna provinces and stratigraphic record, as well as provenance studies.

Fig. 1
figure 1

Zonation of the European Variscan Orogen (Díez Fernández and Arenas 2015). Based on Franke (1989), and Martínez Catalán (2011). Location of Figs. 2 and 3 is shown

Full size image

It is widely accepted that the Late Paleozoic Gondwana–Laurussia interplay produced a complex dismembering and juxtaposition of terranes as a result of their progressive collision. From a paleogeographic point of view, adjoining sections of the Gondwana margin dispersed along the Variscan Orogen should have an analogous sedimentary record between Late Ediacaran and Early Cambrian times. However, the common azoic nature of many of these series and the complex deformation associated with Variscan tectonics, make it difficult to recognize the primary configuration of the sedimentary series and to establish correlations between them. For this reason, valuable information can be obtained from whole-rock major and trace element and Nd isotopic geochemistry, combined with U–Pb radiometric ages and Hf systematics of detrital zircon (Fernández-Suárez et al. 2002; Linnemann and Romer 2002; Pereira et al. 2006, 2015; Linnemann et al. 2008; Fuenlabrada et al. 2020).

The Sm–Nd isotopic geochemistry of siliciclastic rocks provides an excellent approach to the average composition, provenance and isotopic sources of the materials which form the upper continental crust (McLennan et al. 1990; McLennan and Hemming 1992). In provenance studies, siliciclastic rocks may provide more significant information than the associated igneous rocks (granitoids), since the juvenile material that may be involved in the generation of many igneous rocks can be mixed with older initial isotopic sources. This methodology has been scarcely used to investigate the provenance of the azoic series involved in the Variscan Orogen, unlike the many published papers focused on U–Pb geochronology of detrital zircon grains (e.g., Fernández-Suárez et al. 2007; Abati et al. 2010). Only combined U–Pb–Hf data on detrital zircon can provide comparable results in this sense (Avigad et al. 2012; Morag et al. 2012).

In the Iberian Massif, whole-rock major and trace element geochemistry and Nd isotopic geochemistry of siliciclastic rocks have been used in recent studies to investigate the tectonic setting and provenance of the sedimentary series deposited during the Ediacaran–Cambrian transition (Ugidos et al. 2003; Pereira et al. 2006; Guijarro et al. 2008; Fernández-Suárez et al. 2014; Fuenlabrada et al. 2016, 2020). However, the possible correlations of these Ediacaran series along the Variscan Orogen, between the Iberian and Bohemian massifs, and based on geochemical grounds that analyse the rock as a whole have not been explored in detail. In this sense, this work will improve the knowledge of the oldest siliciclastic series described in the Iberian Massif, the so-called Serie Negra Group (Carvalhosa 1965), including a new whole-rock geochemical and Sm–Nd isotopic database. These data are compared with the available isotopic (Nd) data from equivalent series described in the Bohemian Massif (Linnemann and Romer 2002), with the aim of testing the affinity of their isotopic sources between the Ediacaran series from both currently distant regions. Such comparison may help to refine reconstructions for the paleobasins once located along the (North African) margin of Gondwana that are now (dismembered and) spread as disconnected individual blocks of Cadomian basement throughout central and western Europe (Stephan et al. 2019a, b).

Geological setting

In the SW Iberian Massif, the lower part of the stratigraphic record of the continental allochthons of the Ossa–Morena Complex (Díez Fernández and Arenas 2015; Arenas et al. 2016b; Fig. 2) is part of a Cadomian basement that encompasses rocks from Neoproterozoic to Early Cambrian age. The Serie Negra Group (Carvalhosa 1965) is composed by a succession of metapelites and metagreywackes, interbedded with metamafic rocks and some layers of calc-silicate rocks and black quartzites. The thickest exposures of the Serie Negra Group occur in three key areas along NW–SE trending regional structures. From north to south these areas are: the Obejo-Valsequillo Domain (Ordóñez-Casado 1998; Bandrés 2001), the Olivenza–Monesterio antiform (Eguiluz and Quesada 1980; Montero et al. 1999) (Fig. 2) and the Aracena-Almadén de la Plata region (Ábalos 1987). Metamorphism in the Serie Negra Group ranges between greenschists facies conditions and pervasive migmatization. Among these key areas, the Olivenza–Monesterio antiform contains the most complete and well-preserved cross section to the Serie Negra Group, which shows a thickness exceeding c. 3000 m (Eguíluz 1988). The core of this antiform contains three stratigraphic formations, which from older to younger are referred to as the Montemolín, Tentudía and Malcocinado formations. Montemolín and Tentudía formations are traditionally considered to be two members of the Serie Negra Group. The c. 1000 m thick Montemolín Formation consists of metapelites, mica schists, quartz-rich schists, metagreywackes, metasandstones, abundant amphibolites (Sánchez-Lorda et al. 2013, 2016) toward the top and interbedded black quartzites. This formation may show extensive migmatisation towards the bottom (Montero et al. 1999). The Tentudía Formation is composed of metasandstones, metagreywackes, slates, phyllites, metacherts, black quartzites, layers of marble and micaschists. The overlying unconformable volcaniclastic Malcocinado Formation (Fricke 1941; Eguíluz et al. 2000) is composed by meta-conglomerates, metasandstones, metapelites, meta-andesites and meta-rhyolites. The Serie Negra Group together with the Malcocinado Formation are unconformably covered by the Early Cambrian Torreárboles Formation (Liñán and Fernández-Carrasco 1984), composed by fluvial to shallow marine deposits with conglomerates, metasandstones and slates (Fig. 2).

Fig. 2
figure 2

a Geological map of the Ossa–Morena Complex including the south part of the Central Iberian Zone. Based on the 1:1.000.000 geological map of Spain and Portugal (IGME 2014; Rojo-Pérez et al. 2019). b Geological map of the Iberian Massif showing the distribution of allochthonous terranes in NW and SW Iberia (Díez Fernández and Arenas 2015). BAO Beja-Acebuches Ophiolite, CMU Cubito-Moura Unit, CU central unit, IOMCO Internal Ossa-Morena Complex Ophiolites, OVD Obejo-Valsequillo Domain, PG–CVD Puente Génave-Castelo de Vide Detachment, PTSZ Porto-Tomar Shear Zone. c Schematic stratigraphic column of the Serie Negra Group. TE and MO stars represent the sampling locations within the Olivenza–Monesterio Antiform

Full size image

A maximum depositional age of c. 565–541 Ma has been obtained for the Tentudía Formation using detrital zircon grains (Schäfer et al. 1993; Casado 1998; Linnemann et al. 2008). The Malcocinado Formation probably straddles the Ediacaran–Cambrian boundary (Pereira 2015). The Torreárboles Formation is Early Cambrian (Liñán and Fernández-Carrasco 1984), with a maximum depositional age of c. 540–532 Ma (Eguíluz 1988; Perejón et al. 2004; Pereira et al. 2011). Large massifs of igneous rocks, mainly of granitic–tonalitic composition, intruded into the Serie Negra Group between c. 600 and 540 Ma (Ordóñez-Casado 1998; Sánchez-García et al. 2003; Linnemann et al. 2008; Alvaro et al. 2014). This Ediacaran igneous suite was generated in relation to the prominent Avalonian–Cadomian peri-Gondwanan magmatic arc (Sánchez-García et al. 2013; Albert et al. 2015; Andonaegui et al. 2016). The chronology of this igneous suite suggests an age of c. 600 Ma for the undated oldest levels of the Montemolín Formation. This chronology is also compatible with the relationship of the Serie Negra Group with the opening of a fore-arc or back-arc basin in the peri-Gondwanan realm and the generation of associated supra-subduction zone type oceanic lithosphere (Arenas et al. 2018).

A similar Cadomian basement is widely exposed in other sections of the Variscan Orogen. In the eastern branch of the orogen, in the northern part of the Bohemian Massif (Fig. 3), the Saxo-Thuringian Zone (SXTZ, Kossmat 1927) contains several Cadomian stratigraphic sequences with scarce Variscan overprint (Linnemann and Buschmann 1995), and probably equivalent to the Serie Negra Group (Pereira et al. 2006; Linnemann et al. 2008). The main outcrops are distributed along the northern margin of the SXTZ (Fig. 3), constituting a total of six siliciclastic series composed essentially by turbiditic metagreywackes, slates, phyllites, schists, metamafic rocks, black metacherts and some layers of quartzites and metaconglomerates toward the upper part of some formations. The oldest maximum depositional age for these formations is considered at c. 629 Ma, according to the U–Pb zircon dating of a granite pebble (Frohnberg Formation; Linneman et al. 2000). Moreover, a minimum depositional age at c. 543 Ma has been estimated using detrital zircon grains extracted from greywacke turbidites and a glaciomarine tillite (Linnemann et al. 2008, 2014, 2018). All these SXTZ formations are intruded by Ediacaran and Cambrian granitic rocks, with ages ranging between c. 540 and 537 Ma (Linnemann et al. 2000, 2010a, 2014, 2018). The Paleozoic sequence overlaying these Neoproterozoic series contains incomplete Cambro-Ordovician series and a more complete stratigraphic succession reaching the Lower Carboniferous (Brause 1968; Linnemann et al. 2010b).

Fig. 3
figure 3

Geological sketch of the Saxo-Thuringian Zone in the Bohemian Massif, including a schematic stratigraphic columns of the Ediacaran–Lower Cambrian series. Stars represent the location of the metasedimentary series. Based on Linnemann and Schauer (1999), Linnemann and Romer (2002) and Linnemann (2007)

Full size image

Sample selection and methodology

Sample selection

Twenty-two samples of siliciclastic rocks belonging to the Serie Negra Group were collected in the Olivenza–Monesterio antiform. Ten samples belong to the Montemolín Formation (MO-1 to MO-11) and were collected in the northern limb of the antiform (Fig. 2). Each sample was extracted from an individual layer of metagreywacke (dm-scale in thickness). The remaining twelve samples belong to the Tentudía Formation (TE-1 to TE-12). They are predominantly fine-grained metagreywackes collected from individual layers and in equivalent proportions in both limbs of the Olivenza–Monesterio antiform (Fig. 2). All rocks sampled are fresh, without evidence of significant alteration. The outcrops visited during sampling indicate that the metasedimentary rocks sampled in the Montemolín Formation seem to accumulate more deformation (strain) than those collected in the Tentudía Formation, although collectively, all of the samples show low-grade metamorphism and slight to moderate strain. Yet, sampling was targeted to the sections with lowest strain.

In thin section, Montemolín samples show abundant micaceous minerals, including muscovite, biotite and chlorite, along with low deformed quartz-feldspathic levels. The samples exhibit heterogeneity in quartz and feldspar grain size, indicative of textural immaturity. This formation presents primary or tectonometamorphic compositional layering. On the other hand, in the Tentudía samples phyllosilicates are even more abundant, as well as opaque minerals and rock fragments. In general, these samples present a smaller grain size than Montemolín Formation and show higher plagioclase content in relation to K-feldspar.

Whole-rock major and trace elements analysis

Crushing and powdering of the rock samples were performed at the facilities of the Complutense University of Madrid. The analysis of major and trace elements were carried out in the Activation Laboratories (ActLabs), Ontario (Canada). Lithium metaborate/tetraborate was used for fusion of the samples, and the elements were measured by Inductively Coupled Plasma Mass Spectrometry (ICP-MS). The general precision was calculated at ~ 0.01%, and for MnO and Ti2O at ~ 0.001%. The elemental composition of the Ediacaran metagreywackes was analysed to constrain their provenance and isotopic sources. The results of the analyses are included in Tables 1 and 2 and in the diagrams shown in Figs. 4, 5 and 7.

Table 1 Whole major and trace element data of Montemolín Formation (Late Ediacaran)
Full size table
Table 2 Whole major and trace element data of Tentudía Formation (Late Ediacaran)
Full size table
Fig. 4
figure 4

Chemical diagrams for Montemolín and Tentudía formations. a Chemical classification diagram based on major elements (Herron 1988). b Maturity and chemical weathering diagram based on major element distribution (after Linnemann and Romer 2002). c Chondrite normalized REE plot (Nakamura 1974); red line corresponds to the PAAS (Post Archean Australian Shale; Taylor and McLennan 1985). d PAAS normalized trace elements diagram

Full size image
Fig. 5
figure 5

Trace element diagrams with tectonic setting discrimination fields (after Bhatia and Crook 1986)

Full size image

Whole-rock Sm–Nd isotopic analysis

Sm–Nd isotope analyses were performed using Isotope Dilution Thermal Ionization Mass Spectrometry (ID-TIMS) at the Geochronology and Isotope Geochemistry Facility of the Universidad Complutense de Madrid. Samples were spiked with a mixed 149Sm-150Nd tracer and then analysed in an IsotopX–Phoenix mass spectrometer (TIMS), following a single collection and a dynamic multicollection mode for Sm and Nd, respectively. To correct procedural and instrumental mass fractionation, the resulting 143Nd/144Nd ratios were corrected for 142Ce and 144Sm interferences and normalized to 146Nd/144Nd = 07219 value (O'Nions et al. 1979). To have control over the drifts from La Jolla reference value (Lugmair et al. 1983), this standard was analysed along with the samples, yielding an average value of 143Nd/144Nd = 0.511851 for six replicates, with an internal precision of ± 0.000007 (2σ). The analytical error for the 147Sm/144Nd and 143Nd/144Nd ratios was estimated lower than 0.1% and 0.006%, respectively, and Nd procedural blanks were always under 0.1 ng. The results are included in Table 3 and plotted on the diagram in Figs. 6, 7 and 8.

Table 3 Whole rock Nd isotope data of Montemolín and Tentudía Fms. (Late Ediacaran)
Full size table
Fig. 6
figure 6

εNd vs. age diagram showing TDM values for the Late Ediacaran siliciclastic rocks of the Tentudía and Montemolín formations. The range of Nd model ages of selected regions are shown for comparison (Linnemann and Romer 2002; Fuenlabrada et al. 2016; Rojo-Pérez et al. 2019)

Full size image
Fig. 7
figure 7

εNd(T) vs. CaO diagram for Late Ediacaran series of the Ossa-Morena Complex (SW Iberian Massif) and the Saxo-Thuringian Zone (Bohemian Massif), showing the influence of young mafic or felsic contribution, or less evolved components (after Linnemann and Romer 2002). The blue band shows the range where the majority of samples plot

Full size image
Fig. 8
figure 8

εNd vs. age diagrams showing the TDM values for the Tentudía and Montemolín Late Ediacaran series, compared with the Late Ediacaran series from the Saxo-Thuringian Zone. The ranges of Nd model ages of other coeval SW Iberian series (Fuenlabrada et al. 2016; Rojo-Pérez et al. 2019) are also plotted. Saxo-Thuringian data are extracted from Linnemann and Romer (2002)

Full size image

Whole-rock and isotope geochemistry results

The geochemistry of major elements in siliciclastic rocks is used to describe and constrain their nature and degree of alteration, as well as to characterize their provenance and tectonic setting during sedimentation (Nesbitt and Young 1982; Bhatia 1983; Taylor and McLennan 1985; Bhatia and Crook 1986). The use of immobile trace elements such as REE, Th, Co; Cr, Nb, Ti, Sc, Y, Zr, Hf, and especially their ratios, are excellent proxies to infer the likely tectonic setting and discriminate between felsic/mafic contributions from the source area, since they remain virtually undisturbed through weathering, transport, sedimentation and metamorphic processes (Nesbitt et al. 1980; Taylor and McLennan 1985; Wronkiewicz and Condie 1987; Feng and Kerrich 1990). The distribution of REE in detrital rocks reflects the average composition of the source area and, consequently, the REE-patterns are a means to discriminate between a crustal or mantle influence in the source. In this sense, certain crustal models consider Nd model age of siliciclastic rocks as an approximation to the average resulting of the extraction of the different terrigenous components from the melt source (McCulloch and Wasserburg 1978 and others), so Nd model ages may act as tracers of the crustal/mantle evolution of the detritus reaching a particular sedimentary basin (McLennan et al. 1990; McLennan and Hemming 1992).

Composition, classification and tectonic setting

The Montemolín and Tentudía formations define homogeneous series considering their major element composition (Tables 1 and 2). The SiO2 average content for both formations (65.10 and 66.23 wt%, respectively) are slightly lower than the average described by Condie (1993) for the Upper Continental Crust (UCC) (66.79 wt%), while the average value for Fe2O3 (5.69 and 5.10 wt%, respectively) and Al2O3 (15.75 and 15.41 wt%, respectively) are slightly higher than typical UCC values (4.51 and 14.99 wt%, respectively). In both Montemolín and Tentudía formations, the mobile elements K2O (avg. 2.78 and 2.59 wt%, respectively) and Na2O (avg. 3.23 and 3.21 wt%) have an average content similar to the UCC range (3.1 and 3.35 wt%, respectively); only two samples (TE-03 and TE-10) show enriched and depleted contents, respectively, for these two elements compared to the average values of the group (Tables 1 and 2). The most significant difference between both Montemolín and Tentudía formations, in comparison with the UCC, appears in lower values of CaO content (1.93, 0.73 and 3.36 wt%, respectively). The higher amount of CaO in the Montemolín Formation than in the Tentudía Formation is likely linked with the occurrence within the Montemolín Formation of abundant mafic material (Montemolín amphibolites; Eguíluz et al. 1990). The average compositional ranges of TiO2, MnO, MgO and P2O5 are 0.75 and 0.73 wt%; 0.07 and 0.06 wt%; 2.38 and 2.23 wt%; 0.18 and 0.20 wt%, for Montemolín and Tentudía rock samples, respectively.

Whole-rock anhydrous chemical compositions of the Montemolín and Tentudía samples have been plotted in the classification diagram published by Herron (1988; Fig. 4a). The rocks sampled in both formations appear at the outcrop as little deformed common greywackes. However, they exhibit a relative compositional heterogeneity when plotted in this classification diagram. A part of the Tentudía and the whole Montemolín samples appear represented into the shale field. This depiction, in the case of Montemolín Formation, seems motivated for their relatively low SiO2 contents. While for Tentudía Formation, such plotting can be explained by its low K2O values, which may reflect a significant post-depositional weathering characteristic of this element. It is appropriate to use other indexes and immobile ratios for accurate measuring of the maturity and weathering of these rocks.

The average values of the SiO2/Al2O3, K2O/Na2O, Al2O3/Na2O, Al2O3/TiO2 ratios for the Montemolín Formation (4.14, 0.91, 4.99, and 21.18, respectively; Tables 1 and 2) show no significant deviation from those in the Tentudía Formation (4.34, 0.67, 4.73, and 21.27, respectively), falling within the range estimated by Condie (1993) for the UCC (4.46, 0.93, 4.47 and 24.18, respectively), which suggests an immature nature. The very low values of the K2O/Na2O and Al2O3/Na2O ratios, with respect to the PAAS (Post Archean Australian Shale, Taylor and McLennan 1985), confirm this immaturity as well as a certain predominance of plagioclase over K-feldspar (except for the TE-03 and TE-10 samples) and white mica in the sedimentary protoliths. The limited effect of weathering processes and the immature character of the sedimentary protoliths from both formations are also supported by a negative correlation between the SiO2 and the rest of major elements, which is reflected in the homogeneous and restricted major element variation of all the samples (Fig. 4b). The CIA (Chemical Index of Alteration; Nesbitt and Young 1982) and PIA (Plagioclase Index of Alteration; Fedo et al. 1995) values for the Tentudía Formation (avg. 63.71 and 76.30, respectively) are slightly higher than those of the Montemolín Formation (avg. 60.24 and 71.51, respectively), and lower than those of the PAAS (avg. 73.70 and 79.02, respectively). These values are consistent with a limited degree of post-depositional alteration and a scarce influence of the sedimentary transport over plagioclase and K-feldspar.

The REE chondrite-normalized fractionation patterns (Nakamura 1974) in Montemolín and Tentudía formations exhibit an enrichment of the LREE (avg. LaN/SmN: 3.90 and 3.79, respectively; Tables 1 and 2), and relatively flat patterns for the HREE, with GdN/YbN values close to unity (avg. 1.64 and 1.51, respectively). The metagreywacke samples from the Tentudía Formation display REE patterns (avg. La/YbN 9.51) closer to those of the PAAS (9.06), than samples of the Montemolín Formation (avg. La/YbN 10.68) (Fig. 4c). All the samples show slightly negative Eu anomalies, whose average values for the Montemolín and Tentudía formations are 0.83 and 0.82, respectively (Eu/Eu* calculated according to Taylor and McLennan 1985).

Trace elements such as La, Zr, Th, Nb, Sc, Y, Ti, and Co, as well as their ratios, are useful tectonic setting and provenance discriminators for (meta)sedimentary rocks (Bhatia and Crook 1986), since some of these elements are directly associated with felsic (Th y La) or mafic (Sc and Cr) sources. Th/Nb values for Montemolín and Tentudía formations (avg. 0.85 and 0.95, respectively; Tables 1 and 2) are slightly higher to those of PAAS (0.77), whereas the Th/Sc ratio obtained for the Montemolín Formation (0.62) is lower than Tentudía Formation and PAAS (0.76 and 0.91, respectively). A relative enrichment in Sc in the Montemolín Formation is in agreement with its CaO values. The average values in Montemolín and Tentudía formations for the La/Th (4.29, and 3.48, respectively) and Zr/Sc (15.15 and 18.29, respectively) ratios are higher than the estimated for the UCC and the PAAS (2.7 and 13.13, respectively). These values point to a mixed source, dominated by an intermediate-felsic component (Cullers 2002). This interpretation is also suggested by the low values observed in both formations in the Ti/Zr ratio (23.21 and 20.71, respectively); and low contents of Cr (avg. 88 and 85, respectively), and Ni (avg. 42 and 39, respectively).

A set of major and trace elements analyses (following Thompson 1982) was normalized to PAAS and plotted into a multivariate diagram for both Ediacaran metasedimentary series (Fig. 4d). Both formations display very similar patterns. They show depletion in the LILE elements, with larger variability in the metagreywackes from the Tentudía Formation, especially in K and Sr, probably due to some post-depositional alteration. Instead, the metagreywackes from Montemolín Formation show a significant enrichment in Sr, linked to the high content in CaO described above. In both formations, LILE elements exhibit a slightly positive slope, while the HSFE element (Zr, Hf, Sm, HREE, and Sc) display flat patterns close to unity (Fig. 4d). A Ti negative anomaly is present in both formations, as expected for a dominant felsic provenance. The patterns of Montemolín and Tentudía formations are consistent with the features described for sediments deposited in an active margin setting (Winchester and Max 1989).

The Montemolín and Tentudía formations lack of large deviations of their geochemical composition from references such as PAAS and UCC. Low Na, Ca, K and Sr contents may indicate weak post-depositional weathering/recycling alteration, but with no significant consequences for using immobile trace elements on provenance and tectonic setting discrimination diagrams. Ternary diagrams (Bhatia and Crook 1986; La–Th–Sc, Th–Co–Zr/10 and Th–Sc–Zr/10) (Fig. 5) provide a valuable tool to distinguish between four depositional settings for sedimentary series. All samples plot into the same field (B), suggesting a similar geodynamic environment for their sedimentation during Ediacaran times (Fig. 5). This tectonic setting would be characterized by the presence of a magmatic arc, built over a thinned continental crust. The La/Sc ratios for both Ediacaran formations (2.65 and 2.58 for Montemolín and Tentudía Formations, respectively) close to the PAAS value (2.38; Taylor and McLennan 1985), and Ti/Zr ratios all above 20, strengthen the aforementioned geodynamic environment interpretation, since the values are close to the range for the continental island arc setting defined by Bhatia and Crook (1986).

Sm–Nd isotope geochemistry

To constrain the paleo-location of the basins where the Neoproterozoic sedimentary series were deposited, additional information to the major-trace element geochemical features is provided by the Sm–Nd system. These two elements are strongly immobile, thus reflecting an average of the original isotope composition from the mantle/crust source rocks. In this context, Nd model ages (TDM) in siliciclastic rocks may be regarded as a proxy for the average age of extraction of their constituents from a depleted mantle, and therefore, the study of the Nd isotope composition of sedimentary sequences provides an excellent tool to understand crustal evolution (McCulloch and Wasserburg 1978; Allegre and Rousseau 1984; McLennan et al. 1990), as well as to constrain the mantle/crust provenance of terrigenous sediments.

Sm–Nd isotope data from the metagreywackes of the Tentudía and Montemolín formations are given in Table 3 and plotted in Fig. 6. Age assignment for the calculation of εNd(T) is based on the stratigraphic and structural features of the selected samples (see “Geological setting”). The reference depositional age considered for the Montemolín Formation (older part of the Ediacaran Serie Negra Group) is 600 Ma, and 565 Ma for the Tentudía Formation (upper part). The metagreywackes from Montemolín Formation show homogeneous 147Sm/144Nd ratios, which vary between 0.1104 and 0.1191, being very close to the value defined for the UCC (0.12), while the metagreywackes from Tentudía Formation display larger variability in 147Sm/144Nd ratios, ranging from 0.1049 to 0.1296. Both groups of ratios remain within the typical range established by Zhao et al (1992) for clastic sediments (0.1–0.13), and far from the upper limit suitable for Nd model ages calculation (0.165; Stern 2002). fSm/Nd values represent the deviation of the 147Sm/144Nd ratios from the present CHUR values (Table 3). This factor calculated for the samples of the Montemolín Formation yields a very restricted value range (− 0.39 to − 0.44), reflecting a considerable isotopic homogenization. On the other hand, the metagreywackes from Tentudía Formation display a slightly wider range of values, varying from − 0.34 to − 0.47, which may suggest a lower influence of recycling processes.

All of the metagreywacke samples from both formations display very negative ɛNd(0) values, ranging between − 13.6 and − 17.1 for the Montemolín Formation; and between − 11.7 and − 18.1 for the Tentudía Formation. The εNd(t) follows a similar trend in both formations: metagreywackes from Montemolín Formation show an εNd(600) values ranging between − 7.6 and − 10.5, while in the metagreywackes from Tentudía Formation the εNd(565) ranges from − 6.9 to − 11.5. For the entire Serie Negra Group, these negative values together with the low ranges of the fSm/Nd values are consistent with a provenance from old and recycled sources mixed with juvenile components (McLennan and Hemming 1992). The fact that both formations have such similar εNd(t) values suggests that their sources probably remained constant throughout the sedimentation period that encompasses both formations.

The calculated TDM model ages for both formations of the Serie Negra Group plot within a restricted range of Paleoproterozoic ages (Fig. 6). The age vs. εNd diagram shows a clear overlap for both formations, with TDM values varying from 1747 to 1892 Ma for the Montemolín Formation, and from 1686 to 1918 Ma in the Tentudía Formation. This figure also contains a synthesis of Nd model ages from the Serie Negra Group exposed in the Obejo-Valsequillo Domain (Rojo-Pérez et al. 2019), and from the Lower Alcudian section of the Central Iberian Zone (CIZ; Fuenlabrada et al. 2016), together with other Nd model ages from Central Europe (Linnemann and Romer 2002), are represented in Fig. for discussing paleogeographic affinities: old isotopic signatures from the West African Craton and the Amazonian Craton, conjoined with more juvenile ones such as those from West Avalonia in Atlantic Canada and Carolina.

Comparison with Central European correlatives: Saxo-Thuringian Zone series.

The affinity between Variscan sectors, which according to geotectonic and paleontological criteria share a common pre-Paleozoic evolution, has been studied and debated before (Murphy et al. 2002; von Raumer and Stampfli 2008; von Raumer et al. 2015; Ballèvre et al. 2014; Martínez Catalán et al. 2020). Nevertheless, geochemistry and especially isotopic approaches have not been extensively considered to correlate now distant, but potentially equivalent, metasedimentary sequences that made the Cadomian basement involved in the Variscan Orogen (Liew and Hofmann 1988; Pereira et al. 2006). This kind of comparison can help to understand Cadomian regions with low quality or lack of exposures and/or subjected to intense deformation and metamorphism. The new geochemical Sm–Nd isotopic data obtained herein for the Cadomian basement of the Ossa–Morena Complex, can be compared with equivalent coeval stratigraphic series described in the Cadomian basement of the Saxo-Thuringian Zone. Different Nd isotopic data from the Ediacaran series of the Bohemian Massif have been published (Linnemann and Romer 2002; Linnemann et al. 2004). From these data, only samples that can be considered lithologically comparable to those from SW Iberia have been used to perform a Nd isotopic comparison. These samples are mainly Late Ediacaran metagreywackes and metapelites and consist of a group composed by 23 samples. These siliciclastic rocks belong to four different units of the Bohemian Cadomian basement, which include the following successions: Frohnberg, Leipzig, Clanzschwitz and Rothstein formations; Lausitz and Weesenstein groups. Lithological details of the Late Ediacaran series from Iberian and Bohemian massifs, and the number of rocks sampled in the different locations are presented in Figs. 2, 3 and 8.

The diagram CaO vs. εNd(T) (Fig. 7) shows the influence of felsic or mafic (more juvenile) contributions in the siliciclastic rocks, as well as the contribution of old crustal areas. CaO contents in the Tentudía Formation show uniform distribution, similar to the distribution followed by the Saxo-Thuringian rocks, in both cases suggesting mixed felsic–mafic sources. However, the samples from the Montemolín Formation show a different trend directed toward less evolved source areas. The higher CaO values (avg. 1.93 wt%) are consistent with the presence of a significant volume of metamafic igneous rocks interbedded in this sequence. In Fig. 7, the Ediacaran samples show a wide variation in CaO with little changes in εNd(t) values. Most samples fall into a restricted range of εNd(t) values, ranging between − 7 and − 11 (Figs. 7 and 8). Three samples show less negative values of εNd(t), probably due to the addition of juvenile material derived from an active magmatic arc. The homogeneous distribution, with little or no variation in εNd(t), can be interpreted as evidence for a common isotopic source area, which should have remained available for a long time in the (North African) margin of Gondwana. This uniformity and availability are also reflected in the εNd(0) values of the Bohemian series compared. Linnemann and Romer (2002) obtained highly negative εNd(0) values for the SXTZ series, ranging between − 16.6 and − 12.8, except for the three samples mentioned above with larger juvenile input.

The εNd vs. time diagram presented in Fig. 8 shows a comparison between the isotopic composition of the SXTZ (North Bohemia) samples and the siliciclastic rocks of the Serie Negra Group. The Nd model ages obtained for the Bohemian samples define a relatively narrow age group ranging between 1570 and 1976 Ma. This age range is somewhat greater than that defined by the Montemolín and Tentudía formations of the Serie Negra Group. However, the two groups of samples are similar in relation to their isotopic sources, which very likely indicate a common source area for the North Bohemian and SW Iberian massifs during Ediacaran times. Their source area must have remained available at least during the depositional time of all the Ediacaran series. As a result, these data seem to indicate a common palaeographic context in the (North African) margin of Gondwana for the two regions. Ediacaran proto- North Bohemia and proto- SW Iberia were probably part of the same (or closely connected) sedimentary basin(s), where turbiditic siliciclastic series that shared common source areas were deposited (Fig. 9).

Fig. 9
figure 9

Paleogeographic reconstruction at c. 570 and 540 Ma, showing the paleo-position inferred for the deposition of the Late Ediacaran series located in the Ossa-Morena Complex (OMC), Armorican Massif (AM), Saxo-Thuringian Zone (SXTZ) and Central Iberian Zone (CIZ) along the (North African) margin of Gondwana. The arrows indicate the location of the source areas and the direction of sediment transport into the peri-Gondwanan basin. Gondwana paleogeography based on von Raumer and Stampfli (2008), Meert and Bruce (2008) and Díez Fernández et al. (2010)

Full size image

Discussion

In the southwest part of the Iberian Massif, the Ossa–Morena Complex represents the most external section of the Gondwana margin (Díez Fernández and Arenas 2015). This section and its prolongation in the allochthonous complexes of NW Iberia, was affected by an intense Variscan deformation (Arenas et al. 2016a; Díez Fernández et al. 2016). The Variscan orogenic evolution resulted in an intricate amalgamation of terranes with different origin and tectonothermal evolution, which make it difficult to correlate units along the orogen. The (North African) margin of Gondwana has been interpreted as a long-lived magmatic arc, probably active in the period ranging between c. 750 and 500 Ma (Linnemann et al. 2004; Rodriguez-Alonso et al. 2004; Pereira et al. 2006, 2012b; Andonaegui et al. 2016). The activity of this magmatic arc can be tracked by both the related igneous rocks and through the sedimentary record generated during its progressive dismantling.

Associated with the (North African) margin of Gondwana, the deposition of the Ediacaran Serie Negra Group has been traditionally considered as related to the opening of a peri-Gondwanan back-arc (Bandrés 2001). However, recent interpretations have also considered the possible deposition in a fore-arc basin opened during a short event of roll-back of the peri-Gondwanan trench that occurred at c. 600 Ma. The roll-back event favoured the generation of new oceanic lithosphere of supra-subduction zone type (Arenas et al. 2018; Díez Fernández et al. 2019). This new buoyant oceanic lithosphere was promptly obducted onto the magmatic arc as one of the ophiolites preserved in the Ossa–Morena Complex (Calzadilla Ophiolite). The major and trace element geochemistry, as well as the Nd isotopic data provided in this contribution for the Montemolín and Tentudía formations, are compatible with both (back- or fore-arc) tectonic settings affecting the thinned margin of Gondwana.

The constant highly negative εNd(t) values obtained for the Montemolín and Tentudía formations, constrained to a narrow interval ranging between − 6.9 and − 11.5, are paired to a relatively narrow interval of Paleoproterozoic Nd model ages ranging between 1686 and 1918 Ma. The limited variability of their Nd T(DM) model ages seems to indicate that they shared a common-source scenario for a long time. The detrital input derived from this source area remained essentially unchanged at least during the depositional time of the Montemolín and Tentudía formations, which ranges between c. 600 and 541 Ma. Altogether, these Nd model ages replicate the outcome obtained for the Serie Negra Group in the northern part of the Ossa–Morena Complex, in the Obejo-Valsequillo Domain (Rojo-Pérez et al. 2019).

The Nd model ages obtained for the Serie Negra Group representing the Cadomian basement in the SW Iberian Massif are almost identical to those obtained for coeval sedimentary series described in the Saxo-Thuringian Zone of the Bohemian Massif (Linnemann and Romer 2002). Their very similar TDM values seem to indicate a common crustal provenance for the Ediacaran siliciclastic rocks that can be found along the Variscan Orogen, at least between SW Iberia and North Bohemia. A common provenance from old continental sources, previously located in the Gondwana mainland, is suggested by the highly negative values and the scarce variability obtained in the εNd range shown by these series. Nd model ages ranging between 2 and 1.6 Ga can only be obtained considering Paleoproterozoic and/or Archean sources, variably mixed with more juvenile material coming from the most external margin and/or the magmatic arc. An Ediacaran sedimentary basin located in the Gondwanan margin in front of the West African Craton, would probably account for the isotopic features recorded in the Ediacaran siliciclastic series from SW Iberia and North Bohemia (Fig. 9). This is in agreement with previous correlations suggested between both regions currently separated c. 2000 km (Fernández-Suarez et al. 2002; Linnemann et al. 2008; Pereira et al. 2015). The isotopic data considered in this contribution provide evidence about the correlation and peri-Gondwanan equivalence between the Cadomian basement of SW Iberia and North Bohemia. These data also indicate that both domains were located in a similar paleo-position in the African margin of Gondwana, although they could be separated from one another by a distance that must be shorter than the current one. Superimposed Variscan tectonics would have enlarged their primary distance by moving along strike the various tectonic blocks that make the Variscan Orogen. Consequently, the Ediacaran siliciclastic series located in SW Iberia and North Bohemia were probably deposited in a shared (long) sedimentary basin located in the Gondwana margin in front of the West African Craton, and whose length measured along the margin is yet to be constrained.