Strike-slip shear zones of the Variscan orogen are used to derive the evolution of paleostrain and discuss the kinematics of the waning stages of the Gondwana-Laurussia collision during the amalgamation of Pangea. In the Iberian Massif, the recognition of three late Carboniferous deformation events related to strike-slip tectonics (D3, D4, D5) in the Trancoso-Pinhel region (Portugal) reveals that late orogenic transcurrent deformation was episodic and occurred in a short period of time (<15 m.y.). Early stages of strike-slip deformation included dextral and sinistral shear zones and orogen-parallel upright folds (D3; ca. 311 Ma). These structures followed the development of extensional shear zones (D2) during the tectonothermal reequilibration of the orogen. D3 structures were deflected and folded by the sinistral D4 Juzbado-Penalva do Castelo shear zone, dated as ca. 309–305 Ma by SHRIMP (sensitive high-resolution ion microprobe) U-Pb zircon dating of synkinematic granitoids. D3 and D4 structures were folded under east-west compression (D5) influenced by the strike-slip movement of the dextral Porto-Tomar shear zone. Variscan movement along the Porto-Tomar shear zone started ca. 304 Ma (onset of the Buçaco basin and syn-D5 granites), but ceased before ca. 295 Ma (age of the final closure of the Ibero-Armorican arc and crosscutting granites). The contrasting geometry, kinematics, and timing of these strike-slip shear zones are explained by deformation partitioning upon a rheologically inhomogeneous crust with structural and tectonothermal anisotropies generated during previous deformation. The convergence vector between Gondwana and Laurussia during D3–D5 remained the same, and was equivalent to the vector that explains the previous tectonic record (D2) in central and northwestern Iberia.


Structures with contrasting geometry and kinematics can be (1) the result of a single phase of deformation operating differently from one place to the other due to, e.g., local variations of finite strain, strain vorticity, thermal conditions, rheological rock properties, and presence of fluids; (2) independent from each other and derive from different phases of deformation; or (3) the result of progressive deformation associated with gradual changes of pressure-temperature conditions, e.g., emplacement and cooling of a syntectonic granitoid (e.g., Carreras et al., 2004). Distinguishing between these three scenarios is challenging when the structures under consideration share geometrical properties, and/or when the time period for their development is shorter than the time resolution of absolute dating methods. Establishing the geometry and timing for structures formed during orogeny is key to reconstructing the associated convergence and/or divergence and evolution. Inferences about evolving paleostrain can help to build tectonic models for orogens where significant parts of the paleogeographic references are modified, such as in the case of Paleozoic orogens that are now dispersed over different tectonic plates. Therefore, identifying the full sequence of structures in a given region is a fundamental step toward reconstructions of large-scale tectonic settings.

Strike-slip shear zones are common in orogenic systems, and are considered one of the tools to determine relative plate movements (e.g., Shelley and Bossière, 2002). Block extrusion or escape tectonics may produce conjugated faults (e.g., Tapponnier and Molnar, 1979; Tapponnier et al., 1982), so using a suitable general picture, including timing and fault trace, is advised for producing tectonic models based on the study of strike-slip shear zones. Here we provide a case study of the complexity regarding the structural evolution, kinematic interpretation at the scale of the orogen, and dating of strike-slip shear zones formed in a collisional orogen. In the Variscan orogen, some major geotectonic domains are bounded by strike-slip shear zones (Fig. 1). Defining their kinematics and timing beyond the limits of analytical methods is thus necessary to better understand the processes involved in the building of long-lived orogenic systems such as the Variscan. The strike-slip shear zones analyzed here formed at an advanced stage in the late Paleozoic collision of Gondwana and Laurussia, and therefore provide a closer view into the kinematics and finite strain during the final amalgamation of Pangea.

In the Iberian Massif of the Variscan orogenic system, strike-slip shear zones and upright folds are considered the result of deformation partitioning in a transpressional regime. However, the timing for transpression has been assigned to contrasting periods in the Variscan orogeny (e.g., Iglesias Ponce de Leon and Choukroune, 1980; Dias and Ribeiro, 1998; Dias et al., 2010, 2013; Díez Fernández and Martínez Catalán, 2012; Martínez Catalán, 2012). The debate is centered on orogen-parallel upright folds accompanying the strike-slip shear zones. Were they developed during the early stages of collision (e.g., Dias et al., 2010, 2013), or at an advanced stage (e.g., Martínez Catalán, 2012)? These opposite views are not mutually exclusive, because detailed structural studies have consistently demonstrated the existence and interference of earlier and later upright folds (Díez Balda, 1986; Díez Balda et al., 1990a, 1990b, 1995; Díez Balda and Vegas, 1992; Díez Fernández et al., 2013; Díez Fernández and Pereira, 2016).

Even if a late orogenic nature is admitted for the strike-slip shear zones of the Iberian Massif, an increasing amount of geochronological data suggests that their formation took place over a short time interval, from ca. 315 to 300 Ma (Regêncio Macedo, 1988; Rodríguez et al., 2003; Valle Aguado et al., 2005; Gutiérrez-Alonso et al., 2015; Díez Fernández and Pereira, 2016). Consequently, models considering all of the strike-slip shear zones as coeval have emerged (e.g., Gutiérrez-Alonso et al., 2015).

Isotopic dating of tectonic fabrics is a tool to determine the evolution of orogenic systems. However, the principle of superposition, when applied to the recognition of the sequence of phases of deformation using field observations, is not rivaled by any radiometric dating method in terms of relative timing. Such observations are particularly useful when the sequence of geological processes considered is fast enough not to be fully tracked by isotopic dating methods. Here we present field structural data from an area of the Iberian Massif selected because of its potential to examine the relationships between the late strike-slip shear zones of the Variscan orogen in the Trancoso-Pinhel region. Our data reveal the existence of three different pulses of strike-slip deformation in the Central Iberian Zone of the Iberian Massif between ca. 311 Ma and ca. 300 Ma. Structural data are reinforced by SHRIMP (sensitive high-resolution ion microprobe) U-Pb zircon dating of a synkinematic granitoid. We use our structural and geochronological data to demonstrate the diachronous and stepped character of the late Carboniferous strike-slip shear zones of the Iberian Massif, and constrain the kinematic framework of the waning stages of the Variscan orogeny.


The Iberian Massif is located in the southern branch of the Variscan orogen (Fig. 1), which is conventionally attributed to the progressive collision between Gondwana and Laurussia in the late Paleozoic (Matte, 2001; Faure et al., 2009; Martínez Catalán et al., 2009; Arenas et al., 2014; Díez Fernández et al., 2016). The Central Iberian Zone represents a section of the margin of Gondwana underlying a set of far-traveled allochthonous terranes that spread across the Iberian Massif (Fig. 1; Díez Fernández and Arenas, 2015). The allochthonous terranes are tectonic slices of continental and oceanic crust that were originally located outboard of and along the margin of Gondwana before collision (e.g., Arenas et al., 2016), during which they were transported onto inner sections of the margin (e.g., Martínez Catalán et al., 2009; Díez Fernández et al., 2016). Late Carboniferous strike-slip shear zones (Fig. 2) cut across folds and crustal-scale thrusts formed during the early stages of collision and the emplacement of allochthonous units, and are spatially and temporally related to extensional shear zones and elongated granitic massifs. Accompanying the strike-slip shear zones, late upright folds affect the thrust nappes, the extensional shear zones, and the granitoids (Fig. 3).

The Trancoso-Pinhel region (Fig. 2) is located in the southern branch of the Ibero-Armorican arc and near the broad axial zone of the Central Iberian arc (Fig. 1). The section consists of Ediacaran to Ordovician metasedimentary rocks that alternate with rare bodies of orthogneiss, and are intruded by abundant and variably deformed ca. 331–304 Ma granitoids (Díez Fernández and Pereira, 2016, and references therein) (Figs. 4A, 4B). Variscan metamorphism ranges between greenschist and granulite facies conditions (Regêncio Macedo, 1988; Valle Aguado et al., 1993; Pereira, 2014) and the grade of regional metamorphism increases downsection. The regional structure is defined by an upright synform, cored by Ordovician quartzites and black phyllites (Marofa synform), that continues for tens of kilometers to the southeast and northwest (Fig. 2; Tamames-Marofa-Sátão synform). Strike-slip shear zones, such as the Huebra and the Juzbado-Penalva do Castelo shear zones, are important local structures (Figs. 2, 4A, and B4B).


The Trancoso-Pinhel region is affected by five main phases of deformation (D1–D5), and by a set of late faults (Díez Fernández and Pereira, 2016). Upright folds characterized by an axial planar foliation (S1) represent D1. D2 records development of the Pinhel shear zone, a low-dipping, ductile, extensional shear zone with a penetrative medium- to high-grade foliation (S2) that overprints D1 structures. D2 was responsible for telescoping of regional metamorphic isograds, exhumation of migmatized rocks, and the progressive deformation of synkinematic granitoids. D2 metamorphism created two contrasting metamorphic domains, a low-grade domain (LGD) characterized by greenschist facies metamorphic rocks that occupy the hanging wall of the Pinhel shear zone, and a high-grade domain (HGD) featuring amphibolite to granulite facies metamorphic rocks that occupy the footwall (Figs. 3 and 4; Díez Fernández and Pereira, 2016).

The third phase of deformation (D3) produced widespread, shallowly plunging upright folds along with a penetrative, low-grade axial planar foliation (S3). The plunge of D3 folds varies from the east and west, producing the pinch and swell map pattern that characterizes the region. To the south of the Marofa synform, the Almeida-Malpartida granitic massif is an intrusion that is discordant with respect to its previously deformed and metamorphosed host rocks (Fig. 4B). The N50°–60°E trending boundary of this nonfoliated biotite-rich granitic massif cuts across the LGD-HGD boundary (i.e., Pinhel shear zone) and the axial traces of D3 folds.

D3 is also characterized by strike-slip shear zones that, together with D4 and D5, are described in the following.

Huebra and Tamames Shear Zones (D3)

The Huebra shear zone occurs along the northern boundary of the Mêda-Escalhão massif (Figs. 2 and 4). This shear zone shows subvertical foliation (S3; Fig. 4A) and subhorizontal elongation lineation (L3e; Fig. 5A). The strike of foliation (Fig. 4E) and trend of lineation (Fig. 5D) are parallel to the trace of the shear zone. Foliation planes show slight variations in dip direction (to the north and to the south) and in inclination values. The plunge of lineations also varies, and may be gently to the east, horizontal, or gently to the west.

Strain in the Huebra shear zone is heterogeneous. Deformation concentrates along the northern boundary of the Mêda-Escalhão massif, and diminishes progressively to the north and to the south. The deformation gradient is steeper to the south, where the Mêda-Escalhão massif is bound by a narrow band of mylonites with subvertical foliation (Fig. 6A; for color versions, see the GSA Data Repository Item1), which grades into a nonpenetrative, subvertical planar fabric defined by the preferred orientation of feldspar phenocrysts. To the north of the shear zone, the metasedimentary rocks show more spaced crenulation cleavage and increasingly open upright folds (Fig. 6B) compared to those that characterize the high-strain core (Fig. 6C). Fold axes (F3) in the domains with higher strain coincides with L3e (Fig. 6D). Upright folds of the high-strain domain, the transitional domain, and the domain located away from the core of the shear zone have similar trends (see following for details) and exhibit similar variations in plunge, but they do not interfere with each other, suggesting that they also developed during D3.

In the high-strain domain, S3 is accompanied by the shallowly plunging L3e, and displays sinistral asymmetric fabrics (e.g., sigma objects, C’, C’-S and S-C structures). Discrete strike-slip shear zones (decimeter scale) are dispersed to the north of the core of the Huebra shear zone (not represented in the map [Figs. 4A and 5A]). The lateral continuity of these shear zones could not be established due to limited exposure. D3 fold axial traces located to the north of the Huebra shear zone show a fan-like pattern converging to the southeast in the core of the Huebra shear zone, thus providing regional-scale criteria for sinistral kinematics (Fig. 5C).

Regional mapping of the N80°–90°E trending Huebra shear zone suggests that it may merge with the dextral, N130°–140°E trending Malpica-Lamego shear zone. Both shear zones are parallel to the axial planes of major D3 folds (Fig. 2).

The Tamames shear zone (Díez Fernández and Pereira, 2016) was described by Díez Balda et al. (1990a), who identified a strike-slip shear zone trending N130°–140°E along the northeastern limb of the upright fold that was named the Tamames syncline. Orthogonal shortening associated with the strike-slip shear zone produces shallowly plunging upright folds (attributed to the third phase of deformation) that interfere with D1 folds and produce minor domes and basins. A Z-shaped deflection of D1 folds in that area suggests a right-lateral movement for the Tamames shear zone (Díez Balda et al., 1990a).

The relative timing between the Huebra and Tamames shear zones could not be established directly. They are separated by another strike-slip shear zone (Fig. 2; Juzbado-Penalva do Castelo shear zone; see following), which impedes the use of in situ field criteria to address this question. However, because both shear zones formed together with D3 folds, this particular set of strike-slip shear zones can be considered coeval.

Juzbado-Penalva do Castelo Shear Zone (D4)

The Juzbado-Penalva do Castelo shear zone (JPSZ) is a strike-slip system that is parallel to the northern limb of the Tamames-Marofa-Sátão D3 synform in the Trancoso-Pinhel region (Fig. 4B). Mapping to the east of these two major structures indicates that the JPSZ cuts and deflects this D3 fold in a sinistral sense, affecting its trace for more than 100 km (Fig. 2). JPSZ represents the fourth phase of deformation in the study area (D4) and separates the D3 Huebra shear zone in the north from the D3 Tamames shear zone in the south (Fig. 2).

To the north and south of the JPSZ, D3 structures recover their regional N130°–140°E trend. This is evident at a large scale (analysis of D3 fold axial traces; Fig. 2) and can be also observed in the foliation map of the Trancoso-Pinhel region (Fig. 4B). The orientations of the structures in the study area are controlled by the N80°–90°E to N60°–70°E trend of the JPSZ. There are two contrasting structural imprints associated with the JPSZ (D4); an early ductile deformation (D4.1), and a later brittle one (D4.2). We relate these two types of structures to the deformation induced by the JPSZ because they affect S3, they only occur adjacent to this shear zone, and their shear planes are systematically parallel to the trace of the JPSZ.

Early Deformation Related to the JPSZ (D4.1)

One type of D4 ductile structure is a subvertical and spaced crenulation cleavage (S4) that is irregularly distributed in the tectonic block located to the south of the JPSZ (Fig. 6E). This cleavage overprints S0–S3. The local shortening related to S4 is responsible for some of the folded pattern exhibited by S3 along the core of the Marofa synform. The interference between this D3 fold and larger folds associated with D4 crenulation produces a hook-type fold pattern at map scale (Fig. 4D). The apparent asymmetry of macro-D4 folds is due to the south-dipping geometry of S3 in this area. D4 folds and fabrics diminish in intensity to the south of the JPSZ as S3 becomes steeper and has a more oblique trend relative to the trace of the Marofa synform.

Another style of D4 ductile deformation is represented by a subvertical mylonitic to ultramylonitic band (S4) that occurs in the quartzites defining the northern limb of the Marofa synform (Fig. 6F); the long subvertical fault here may approximate the location of a mylonitic shear band (Fig. 4B). Although these two structures are not strictly the same, they are intimately related at a regional scale. The width and internal structure of this shear band in the study area are unknown because exposures are limited, of poor quality (Fig. 7A), and reworked by the fault (see following).

D4 shearing is progressively less intense and heterogeneous with increasing distance from the mylonitic shear band located along the northern limb of the D3 Marofa synform, which probably represents the core of the JPSZ. Other conjugate D4 shear bands broaden around the metasedimentary (migmatitic) host of the São Pedro–Vieiro granitic massif (Fig. 7B), located in the northern limb of the Marofa synform. This granitic massif is also deformed by D4 shear bands (Fig. 7C), so the extent of D4 ductile deformation to the north of the Marofa synform is ∼2 km.

In the metasedimentary rocks, S4 is a low-grade foliation defined by the reorientation of pre-S4 mineral grains into parallelism to the cleavage or shear planes, and by a newly formed paragenesis consisting of fine-grained quartz + white mica + chlorite + opaque minerals. S4 in the shear bands displays sinistral kinematic criteria (Figs. 7B, 7C) compatible with the deflection of major structures at a regional scale. S4 follows the local trend of the JPSZ, and is vertical or steeply dipping to the south (Fig. 8A). S4 contains a shallowly plunging, N80°–90°E to N60°–70°E elongation lineation defined by quartz and mineral aggregates, with plunge varying from the west to the east (L4e; Figs. 7D, 7E, and B8B). D4 crenulation (F4) shows the same overall orientation and plunge variation as L4e.

São Pedro–Vieiro Massif: Petrography and Structure

The São Pedro–Vieiro massif is an elongated body of muscovite granite; the long axis in map view is parallel to the JPSZ (Fig. 4B). The granitoid body consists of quartz, K-feldspar, plagioclase, muscovite, and minor tourmaline, apatite, chlorite, zircon, opaques, and some rare biotite. Most of this massif is affected by sinistral D4 shear bands (Fig. 7C), although there are sections of low strain and even undeformed rock where primary igneous textures are preserved.

S4 in this granitoid body is formed in relation to sinistral C-S structures at solid-state conditions (Fig. 7F). C and S planes are defined by aligned quartz (ribbons) and muscovite. Grain size is significantly smaller for the dynamically recrystallized grains that, together with reoriented original igneous grains, define the main mylonitic foliation. Reoriented grains, particularly muscovite, have larger size, intracrystalline deformation (undulose extinction), and micro–kink folds. K-feldspar shows textures related to granular flow, such as the development and migration of subgrains as well as partial transformation into plagioclase. K-feldspar may also show sigma shape and usually displays twins and symplectite. Plagioclase is fractured and displays arched twins. C shear planes may be defined by chlorite, suggesting low-grade metamorphic conditions for the development of S4.

With the exception of the D4 shear bands, field observations indicate that this granitic massif is not affected by previous phases of deformation, and they have not been detected by petrographic analysis in thin section. This observation is significant in the study area, where D2 and/or D3 produce a penetrative but variably intense planar fabric in the synorogenic granitoids (Díez Fernández and Pereira, 2016). These observations and the elongate nature of this granitic massif parallel to the trace of the JPSZ indicate that its intrusion can be considered synkinematic relative to D4.

Late Deformation Related to the JPSZ (D4.2)

D4 brittle deformation is represented by a narrow but poorly exposed fault zone that cuts and removes portions of the main D4 mylonitic band. The fault zone is at least 20 m wide, dips to the north, and consists of a complex set of gauges and breccias (Fig. 9A). The fault trace is parallel to the ductile JPSZ and to the mylonitic shear band that defines the core of the JPSZ; therefore, this fault is considered part of the same system. However, deformation conditions were colder and the fabrics indicate that this is not a strike-slip fault.

A preliminary analysis of kinematic criteria in the fault zone did not provide clear indicators, in part because the fault zone cuts across a wide variety of rocks (phyllites, gneisses, migmatites, and granites) and tectonic fabrics, with different pre-fault kinematic histories (D1–D4). Some dip-slip lineations and striations suggest a dominant dip-slip movement, although in some sections shallowly plunging lineations (plunging 20°–50° to the west) may correspond to previous strike-slip deformation. Whether this is a normal or a reverse fault can be solved using a regional approach. The lack of a wide LGD to the north of this fault, and the juxtaposition of the HGD and the Marofa synform cored by low-grade metasedimentary rocks can be explained by an upthrown movement of the northern block in the hanging wall of this fault, implying that it is a north-dipping high-angle thrust (Fig. 4D; Pereira et al., 2014). Coeval lateral movements or later brittle deformation and reworking cannot be ruled out.

Another type of late deformation is represented by a crenulation cleavage that has a low to moderate dip and subhorizontal crenulation lineation. This fabric produces open mesofolds that interfere with D3 folds (Fig. 9B). The only crosscutting relationship observed indicates that this is a post-D3 event, but no criteria are available to assess its timing relative to the fabrics associated with the JPSZ or the later events (described in the following). The distribution of this type of deformation is random and local.

Steeply Plunging Folds (D5)

The shear bands related to D4 (Fig. 9C) and previous structures are variably affected by north-south subvertical folds (D5; Figs. 7D and 9D). These folds occur together with a nonpenetrative subvertical crenulation cleavage (S5; Figs. 8C and 9E) and a crenulation lineation (F5; Figs. 7D and 9D) that plunges between 30° and 80° to the south and southeast, largely controlled by the local orientation of S0–S4 (Fig. 8D). The amplitude of mesoscale D5 folds may range between 2 and 3 m (Fig. 7D). These folds occur together with centimeter-scale folds (Fig. 9D), which may be accompanied by a close-spaced crenulation (Fig. 9E).

D5 folds occur throughout the Trancoso-Pinhel region, and are particularly concentrated in places where the lithological units and the local structural elements define very open map-scale vertical folds. For example, the trace of the JPSZ has a wavy pattern in the Trancoso-Pinhel region. The trace varies from N70°–80°E to N100°–110°E and then to N60°–70°E (Fig. 4B). D5 folds are more abundant around areas showing these kind of north-south inflections and therefore we consider them as major D5 folds and responsible for much of the undulate structural trend of the region. Moreover, the length of the limbs that define these inflections is different at a regional scale. The two main inflections depicted by the Marofa synform and the JPSZ define an open, Z-shaped vertical fold. Such asymmetry is compatible with a north-south dextral shearing.

Late Faults

The latest phase of deformation is a set of subvertical faults that cuts all the previous structures and rocks, including the latest Variscan granitoids (Figs. 4A, 4B). Even the largest faults show limited offsets of the regional tectonometamorphic structure. The fault set is divided into two principle families of approximately north-northeast–south-southwest and northwest-southeast trends. The group trending north-northeast–south-southwest dominates in the northern part of the study area.


The Almeida-Malpartida and São Pedro–Vieiro massifs were selected for a U-Pb geochronological study. These two massifs are good candidates for obtaining an absolute age reference for the strike-slip shear zones of the region. Dating of the nonfoliated Almeida-Malpartida biotite-rich granite constrains the age of the third deformation phase. This massif is discordant across the Pinhel shear zone and the D3 axial traces recognized in the host metamorphic rocks (Fig. 4B). Dating of the foliated and syn-D4 São Pedro–Vieiro muscovite-rich granite provides an age for the movement along the JPSZ, which, as discussed here, was developed after D3 and before D5.

Analytical Technique

Two samples of granite (TP-4, São Pedro–Vieiro; TP-13, Almeida-Malpartida; see location in Fig. 5) were crushed, disk-milled, sieved (300 μm), concentrated on a Wilfley table, and separated via heavy liquids (bromoform and methylene iodide) at Universidade de Évora (Portugal) and Universidad Complutense de Madrid (Spain). Care was taken to select representative grains from fractions using no magnetic discrimination. Zircon grains were hand-picked under binocular microscope; the most transparent were selected to calculate protolith crystallization ages. At the IBERSIMS Lab (SHRIMP Ion-Microprobe Laboratory of the University of Granada), zircon grains of each sample plus several grains of standards were cast on a 3.5-cm-diameter epoxy mount, polished, and documented using optical (reflected and transmitted light) and scanning electron microscopy (secondary electrons and cathodoluminescence). After extensive cleaning and drying, mounts were coated with ultrapure gold (8–10 nm thick) and inserted into a SHRIMP IIe/mc for analysis. Each selected spot was rastered with the primary beam for 120 s prior to the analysis, and then 6 scans were analyzed, following the isotope peak sequence 196Zr2O, 204Pb, 204.1background, 206Pb, 207Pb, 208Pb, 238U, 248ThO, 254UO. Every mass in every scan is measured sequentially 10 times with the following total counting times per scan: 2 s for mass 196; 5 s for masses 238, 248, and 254; 15 s for masses 204, 206, and 208; and 20 s for mass 207. The primary beam, composed of 16O16O+, is set to an intensity of ∼5 nA, with a 120 μm Kohler aperture, which generates 17 × 20 µm elliptical spots on the target. The secondary beam exit slit is fixed at 80 μm, achieving a resolution of ∼5000 at 1% peak height.

All calibration procedures are performed on the standards included on the same mount. Mass calibration is done on the REG zircon (ca. 2.5 Ga; very high U, Th, and common Pb content). Every analytical session started measuring the SL13 zircon, which is used as a concentration standard (238 ppm U). The TEMORA-2 zircon (416.8 ± 1.1 Ma, Black et al., 2003), used as an isotope ratios standard, was then measured every four unknowns. Data reduction and age calculations were done with the SHRIMPTOOLS software (www.ugr.es/fbea). Crystallization ages and concordia plots were obtained using Isoplot (Ludwig, 2008).


Results of U-Pb SHRIMP analyses are listed in Table DR1 and presented in Figure 10. The uncertainties for individual analyses in the text, Table DR1, and concordia diagrams are given at the 1σ level, whereas the uncertainties on weighted mean 206Pb/238U ages in the text and Figure 10 are given at the 2σ level (uncertainties have 95% confidence limits).

Cathodoluminescence images of the majority of zircons show complex internal structures. Composite grains have cores (xenocrysts) with oscillatory and banded zoning, accompanied by dark or light rim overgrowths of variable widths (20–70 µm). Most of the analyses were performed on dark overgrowths. Few simple grains show oscillatory zoning.

Foliated Granite

We obtained 18 U-Pb SHRIMP analyses for sample TP-4. Ages were calculated assuming 204Pb correction. The older two zircon grains have Cryogenian (693 Ma) and Tonian (961 Ma) ages, suggesting inheritance. The remaining 16 spots in single grains and in rim overgrowths of composite grains spread along the concordia curve from ca. 352 to 287 Ma, and give a weighted mean 208Pb/238Th age of 302.2 ± 8.3 Ma (mean square of weighted deviates, MSWD = 2.4; Fig. 10A). Some of the spread observed could be due to the presence of antecrysts and/or caused by the combination of inheritance and Pb loss. In this group of grains, 7 grains in the age range of ca. 332–296 Ma gave a weighted mean 208Pb/238Th age of 309 ± 11 Ma (MSWD = 0.6; Fig. 7A). A concordia age of 307.8 ± 3.1 Ma (MSWD = 1.8; Fig. 10A) is considered the best estimate for the crystallization age of the foliated granite.

Nonfoliated Granite

We performed 25 U-Pb SHRIMP analyses on sample TP-13. Ages were calculated using uncorrected ages; 6 of the zircon grains yield 206Pb/238U ages of ca. 569–491 Ma that probably represent xenocrysts. A weighted mean 206Pb/238U age of 296.0 ± 7.5 Ma with a very poor fit (MSWD = 49; Fig. 10B) is obtained with the remaining 20 spots, quite scattered in the concordia curve from ca. 336 to 281 Ma. Most of the scattering corresponds to analyses performed on zircon overgrowths that probably formed during hydrothermal activity related to late magmatic processes. A group of 6 analyses in the age range of ca. 313–300 Ma yields a weighted mean 206Pb/238U age of 304.0 ± 3.0 Ma (MSWD = 0.57; Fig. 10B). That age coincides with the age obtained from the upper intercept of an apparent common Pb discordia line on concordia at 303.0 ± 2.1 Ma (MSWD = 0.89; Fig. 10B), which is taken as the probable crystallization age of the post-D3 granite.


The analysis of orientation, kinematics, and timing of strike-slip shear zones is presented here as a tool for constraining finite paleostrain and inferring relative plate movements in orogeny. A distinction between structures sharing geometrical features and kinematics along with a correlation of structures with contrasting geometry and kinematics (upright folds and strike-slip shear zones) have been made based on field criteria (crosscutting relationships), isotopic dating, and regional analysis.

Relative Timing of Strike-Slip Shear Zones

The Huebra and Tamames shear zones coincide with the formation of D3 folds. The Huebra shear zone has regional trend and kinematics similar to those of the JPSZ. Consequently, both strike-slip shear zones could have been formed together. However, our study shows that the structures associated with D3 (at least the upright folds) are deflected and locally folded by the JPSZ (D4). In central Iberia, such relative timing between strike-slip structures was previously described in the eastern section of the JPSZ (e.g., Villar Alonso et al., 1992), where even the interference between north-south–trending D3 folds and later east-west–trending D4 folds associated with the JPSZ can be observed (Fig. 9F; Díez Fernández et al., 2013). We cannot exclude the possibility that the Huebra shear zone remained active or was reactivated during subsequent sinistral deformation associated with the JPSZ, because their shear planes almost coincide at a regional scale. In addition, minor sinistral strike-slip shear zones trending N70°–120°E have been recognized to the east and southeast of the Trancoso-Pinhel region, and are considered as post-D2 and postdating development of the Tamames synform (Díez Balda, 1986). The Tamames synform has been interpreted as a D3 structure (Díez Fernández and Pereira, 2016), so we propose that the aforementioned sinistral shear zones represent yet another example of D4 strike-slip structures.

The JPSZ can be mapped northwest into the so-called Douro-Beira shear zone (Fig. 2). This alignment defines a single D4 strike-slip shear zone that is deflected clockwise, from a northeast-southwest trend in the eastern part to a northwest-southeast trend in the western part.

The fifth phase of deformation (D5) in the Trancoso-Pinhel region produced a series of north-south–trending folds accompanied by subvertical crenulation cleavage and steeply plunging fold axes. Equivalent folds are irregularly distributed throughout basement areas nearby and, in some cases, are accompanied by minor north-south– to north-northeast–south-southwest–trending dextral strike-slip shear zones (e.g., Jiménez Ontiveros and Hernández Enrile, 1983; Gil Toja et al., 1985; Díez Balda et al., 1990b; Díez Montes and Gallastegui, 1992; Villar Alonso et al., 1992; Valle Aguado et al., 2000; Mellado et al., 2006). D5 folds affect the trace of JPSZ, and formed by east-west shortening. Previous studies considered these folds to be responsible for accommodating lateral movements during the last pulses of the JPSZ (e.g., Valle Aguado et al., 2000). That interpretation is inconsistent with the fact that D5 folds actually affect the strike-slip shear zone with which they are supposed to be related (Figs. 7D and 9C). Therefore, we provide an alternative interpretation for these folds.

Kinematic models of partitioned transpression postulate that the simple shear and pure shear components of strain are decoupled into zones of deformation (Dewey et al., 1998). This way the simple shear component is preferentially accommodated by strike-slip faults, whereas the pure shear component produces shortening within the intervening tectonic blocks (Fig. 11A; Teyssier et al., 1995; Dewey et al., 1998). Rotation of intervening blocks in response to the displacement along their bounding strike-slip faults is expected (Mount and Suppe, 1987). The Porto-Tomar shear zone is a north-south–trending strike-slip shear zone associated with lateral dextral movement and east-west subhorizontal shortening, the latter manifested as reverse faults and local upright folds (Ribeiro, 1974; Ribeiro et al., 1980). This shear zone drags and cuts D3 and D4 structures (Fig. 2). At a large scale, the Porto-Tomar shear zone is considered responsible for the clockwise deflection of the Juzbado-Penalva-Douro-Beira shear zone (Martínez Catalán, 2011b), and deflection of other strike-slip shear zones of southwest Iberia, such as the Coimbra-Córdoba shear zone (Fig. 1; Pereira et al., 2010).

According to the relative timing, geometry, and kinematics of D5 structures in the study area (and in central Iberia in general), we propose that D5 is the far-field response to a broad east-west shortening related to the Porto-Tomar shear zone. In our view, the Porto-Tomar shear zone is a partitioned transpressional shear zone, in which the high-strain zone accommodated north-south dextral movements (simple shear), while the reverse faults and D5 folds account for coeval east-west orthogonal shortening (pure shear) distributed over the tectonic blocks bounded by this shear zone (Fig. 11A). The clockwise deflection of former D3 and D4 structures along the Porto-Tomar shear zone resembles the model of high-drag distributed shear proposed by Mount and Suppe (1987), so dextral strike-slip deformation was probably coupled into a broad shear zone for a time. An equivalent rotation of tectonic blocks can be observed for D3 and D4 strike-slip shear zones. D3 folds rotate toward the shear planes of the D3 Huebra shear zone (Fig. 5C), whereas those folds and related fabrics are also deflected by the D4 JPSZ (Figs. 2 and 4). Rotation associated with D3 and D4 is restricted to a relatively narrow section along the shear zones, thus making it possible to observe the primary orientation of structures away from the shear zones.

Considering the discussion here, at least three pulses of strike-slip movement along steeply dipping shear zones can be identified: (1) the Huebra and Tamames shear zones along with other shear bands of lesser regional extent (D3; Fig. 11B), (2) the JPSZ and its continuation into the Douro-Beira shear zone (D4; Fig. 11C), and (3) the Porto-Tomar shear zone (D5; Fig. 11D). Each of these shear zones is accompanied by its own set of upright folds and minor shear bands.

In shear zone networks the faults are not all simultaneously active. Shearing can consist of a pulsating process associated with strong partitioning (e.g., Carreras, 2001; Carreras et al., 2010). Progressive shearing gives rise to interference patterns that do not imply polyphase tectonics (e.g., intramylonitic folding). Thus, differences in radiometric ages or deduced relative timing based on structural analysis do not exclude the possibility of partitioned deformation in space and time. Detailed structural analysis carried out in the study region shows the existence of progressive deformation histories where deformation phases overlap. Spatial strain partitioning is suggested by the development of individual strike-slip shear zones. However, deformation was also partitioned through time. Radiometric dating indicates differences in age that cover an interval of time of ∼15 m.y. (see following), and crosscutting relationships between strike-slip shear zones give the impression of a sequence of individual phases of deformation in the region, i.e., D3–D5. Consequently, the strike-slip shear zones of the Iberian Massif are not coeval, and likely formed during progressive strike-slip deformation.

Absolute Timing of Strike-Slip Shear Zones

Any model to explain the postextensional (post-D2) evolution of the Central Iberian Zone must take into account that ensuing late Carboniferous strike-slip deformation was stepped in space and time (as inferred from the sequence of post-D2 events presented in this study; D3, D4, D5). This strike-slip deformation probably occurred in a short period of time (ca. 317–304 Ma; Regêncio Macedo, 1988; Rodríguez et al., 2003; Valle Aguado et al., 2005; Gutiérrez-Alonso et al., 2015; Díez Fernández and Pereira, 2016).

Timing of D3 deformation has been constrained in the Trancoso-Pinhel region to between 317 ± 9 Ma (age of the youngest syn-D2–D3 granitoid) and 311 ± 6 Ma (age of a syn-D3 granitoid) by means of structural data and U-Pb isotopic dating of synorogenic granitoids (Díez Fernández and Pereira, 2016). The age of D3 broadens a little if the age obtained for strike-slip deformation along the D3 Malpica-Lamego shear zone (ca. 310–307 Ma; Rodríguez et al., 2003; Gutiérrez-Alonso et al., 2015) and the age of its synkinematic granitoids (ca. 320–314 Ma; Simões, 2000; Rodríguez et al., 2007) are considered. All these ages are in agreement with the first absolute age estimations made for the tectonic fabrics associated with late D3 upright folds of the Iberian Massif, dated as 314 ± 6 Ma (Rb-Sr and K-Ar methods; Capdevila and Vialette, 1970; corrected by Ries, 1979).

The U-Pb age of ca. 303 Ma obtained for the postkinematic Almeida-Malpartida granite places an upper limit on the timing of D3 folding. The age of 307.8 ± 3.1 Ma obtained for the São Pedro–Vieiro granite is consistent with the previously published K-Ar age of 305 ± 8 Ma (muscovite; Regêncio Macedo, 1988) for the same granitic body, and helps to constrain the timing of the fourth deformation phase related to the movement in the JPSZ. These geochronological data are in agreement with an Ar/Ar date of 309 ± 2 Ma obtained from mica that defines the S4 foliation in the eastern section of the JPSZ (Gutiérrez-Alonso et al., 2015). The Aguiar da Beira granite, dated as 304 ± 8 Ma (U-Pb in zircon; Costa, 2011), cuts the JPSZ and all the deformed granitoids in the Trancoso-Pinhel region (Fig. 4B). A similar age of 304 ± 2 Ma was estimated for the crystallization of the Villavieja de Yeltes granite (Fig. 2) (Gutiérrez-Alonso et al., 2011), the emplacement of which was controlled by the development of an extensional structure related to a north-south shear zone formed after the JPSZ (Mellado et al., 2006). The Cota-Viseu granodioritic-monzogranitic massif (Fig. 2), which was dated as 305 ± 4 Ma (Valle Aguado et al., 2005), also postdates the JPSZ. Assuming some minor overlap in the isotopic ages, D4 can be restricted to a range ca. 311–304 Ma. However, given the consistency and concordance of the age data, this interval could be further restricted to ca. 309–305 Ma.

The Porto-Tomar shear zone has been considered a transcurrent fault that might have been active during a significant period of the Variscan orogeny (Ribeiro et al., 1990; Dias and Ribeiro, 1993; Shelley and Bossière, 2000; Chaminé et al., 2003). However, the fault trace and structural imprint associated with this strike-slip system are late orogenic (Ribeiro et al., 1980, 2007, 2016; Pereira and Silva, 2001; Martínez Catalán et al., 2007; Pereira et al., 2010), and the sedimentary control exerted by this fault over adjacent basins started in the Late Pennsylvanian (e.g., Gama Pereira et al., 2008; Machado et al., 2011).

The Buçaco basin (Fig. 2) is a pull-apart basin, the development and filling of which are associated with the Porto-Tomar shear zone (Domingos et al., 1983; Wagner, 2004; Flores et al., 2010). The basin is early Gzhelian (palynology data; Machado, 2010), i.e., ca. 304 Ma (Cohen et al., 2013). Because the entire basin is folded and affected by normal faults and thrusts formed during movement of the strike-slip system, the Porto-Tomar shear zone must have been active after ca. 304 Ma. According to the relative timing of phases of deformation and the correlation established in this work between D5 and the Porto-Tomar shear zone, the age of the latter should be younger than ca. 309–305 Ma (proposed age for D4). This age estimate is in agreement with the onset of the Buçaco basin ca. 304 Ma, with the Ar/Ar age obtained from syntectonic mica grown in foliated granite of the Porto-Tomar shear zone (307 ± 5 Ma; Gutiérrez-Alonso et al., 2015), and with the overprinting of the strike-slip deformation on ca. 308 Ma granites (Pereira et al., 2010) from the southern branch of the Porto-Tomar shear zone. Moreover, the development of north-south D5 strike-slip shear zones locally controls the emplacement of granitic magmas (e.g., Villavieja de Yeltes granite; Mellado et al., 2006). The crystallization age of this granite (Gutiérrez-Alonso et al., 2011) provides further evidence for the ca. 304 Ma age for the onset of the Porto-Tomar shear zone.

Some (e.g., Martínez Catalán, 2011a) consider that the Porto-Tomar shear zone extends to the crustal-scale transcurrent shear zones of the Armorican Massif, thus depicting the Ibero-Armorican arc (e.g., Ribeiro et al., 2007). The age of this orocline has been constrained by means of paleomagnetic and U-Pb data to ca. 310–295 Ma (Weil et al., 2010; Gutiérrez-Alonso et al., 2011). The postkinematic Lavadores granite, located at the northern branch of the Porto-Tomar shear zone (Fig. 2), yielded a Permian age of 294 ± 3 Ma (U-Pb zircon; Martins et al., 2014), so the age of the Porto-Tomar shear zone must be older.

Kinematic Analysis

The sequence of individual phases of deformation that emerge from our analysis has been used to derive the evolution of paleostrain and consider the kinematics of the waning stages of the Gondwana-Laurussia collision. The relative timing between strike-slip structures observed in the Trancoso-Pinhel region favors kinematic models where structures formed progressively (e.g., Martínez Catalán, 2011a) and questions those that assume they all were formed at the same time (e.g., Gutiérrez-Alonso et al., 2015).

The strike-slip shear zones studied in this paper are among the largest of the central part of the Iberian Massif, and their noncoeval nature implies that these shear zones cannot be integrated directly in an extrusion model featuring conjugated faults. The existence of conjugated faults, i.e., major faults accompanied by minor subsidiary faults with oblique trace and opposite kinematics, is possible for some shear zones of the Iberian Massif (e.g., Iglesias Ponce de Leon and Choukroune, 1980; Díez Fernández and Martínez Catalán, 2012). Such a model was used as part of the working hypothesis in previous studies that attempted integration of all the strike-slip shear zones in a single tectonic process (e.g., Shelley and Bossière, 2002; Gutiérrez-Alonso et al., 2015). Those approaches need to be revised for the case of major shear zones such as Porto-Tomar, Coimbra-Córdoba, Juzbado-Penalva do Castelo, and Malpica-Lamego, among others, because these shear zones are not coeval.

The D3 Tamames and Huebra shear zones represent the onset of intracontinental deformation after a period of orogenic extensional collapse (D2; Díez Fernández and Pereira, 2016). The orthogonal components of ongoing convergence were distributed across the orogen and produced D3 upright folds. In the northwest and central Iberian section of the orogen, dextral strike-slip structures at this stage are more dominant than sinistral ones. The Tamames shear zone, and more important, the Malpica-Lamego shear zone (Figs. 1 and 2), which are several hundreds of kilometers long (Coke and Ribeiro, 2000; Llana-Fúnez and Marcos, 2001), are related to the development of D3 upright folds (Díez Fernández and Martínez Catalán, 2012; Pamplona et al., 2015) and in some instances were exploited by the intrusion of deep-sourced magmas (e.g., Gallastegui, 1993), thus indicating their crustal-scale nature. Sinistral shear zones located nearby are interpreted to represent subsidiary shear zones equivalent to bookself-type structures (Iglesias Ponce de Leon and Choukroune, 1980; Shelley and Bossière, 2000). Accordingly, and given that the trace of these shear zones is parallel to the structural trend of the orogen, convergence likely included a right-lateral component, which was concentrated in strike-slip shear zones such as Malpica-Lamego and Tamames, among others.

The coexistence of late Carboniferous strike-slip shear zones and widespread upright folds during D3–D5 suggests that they formed during partitioned transpression (Fig. 11A). The northwest-southeast orientations of D3 folds provide a constraint for the maximum shortening direction at this stage, i.e., northeast-southwest (in present-day coordinates) assuming limited rotation of the folds during transpression. If the relative movement between intervening landmasses included a dextral lateral component, and convergence was partitioned into strike-slip shear zones and folds (e.g., Sanderson and Marchini, 1984), the orientation of the convergence vector for D3 would approach a north-northeast–south-southwest direction (Fig. 11B). Similar conclusions can be reached for D4 and D5 (Figs. 11C, 11D), so convergence did not change its orientation significantly from D3 through D5, but produced a sequence of structures with different orientations that have contrasting kinematics. Deformation partitioning associated with convergence could have been focused along local rheological anisotropies, thus resulting in strike-slip shear zones with different orientations. The intrusion of granitic batholiths, the existence of large-scale shear zones (either extensional or strike slip), and the contrasting rheological behavior between thermally different zones are the most likely factors controlling the orientation and location of D3–D5 shear zones. The thermal reequilibration of the orogen was ongoing long after D2, as indicated by the presence of syn-D3, and even syn-D4 and syn-D5 granitoids. Deformation registered during D3–D5 in the study area took place under low-grade greenschist facies conditions (chlorite zone). The presence of hotter and less viscous material underneath, capable of sourcing synkinematic granitoids, implies a rheologically inhomogeneous crust. The trace of late strike-slip shear zones could be controlled by such heterogeneities. As an example of orogenic reworking and conditioned deformation in the Trancoso-Pinhel region, the D3 Huebra shear zone and D4 JPSZ are parallel to a D2 extensional shear zone occurring at both sides of a granitoid massif located in the core of a large-scale antiform (i.e., Pinhel shear zone; Díez Fernández and Pereira, 2016). Therefore, the oblique trace of sinistral shear zones in the Iberian Massif relative to major dextral shear zones is not necessarily imposed by the orientation of the stress field, as it could be expected for a conjugated fault system. We suggest a control on the trace of major strike-slip shear zones exerted by previous structures, which would have conferred a conjugated appearance to some of the late Variscan structures.

In the Late Devonian, convergence affecting the margin of Gondwana included a dextral component (oblique continental subduction; Díez Fernández et al., 2012a). Lateral translations continued to be significant into the early Carboniferous, when the oblique convergence was coeval with (1) the exhumation of high-pressure rocks, (2) the emplacement of allochthonous nappes (Díez Fernández and Martínez Catalán, 2012), (3) the generation of a subduction-related tectonic mélange (Arenas et al., 2009), and (4) the subsequent extensional collapse of the orogenic hinterland (Díez Fernández et al., 2012b). The development of late strike-slip shear zones under dextral transpression during the late Carboniferous indicates the persistent character of right-lateral displacements between Gondwana and Laurussia throughout the late Paleozoic.

Strike-Slip Shear Zones and the Iberian Oroclines

The Ibero-Armorican arc is a late orogenic or postorogenic orocline that bends the entire Variscan orogen of Iberia (Weil et al., 2013), thus conferring a dominant northwest-southeast trend to the major structures that occupy its southern branch (Fig. 1). In that branch, the Central Iberian arc represents another orocline that seems to be coupled to the Cantabrian orocline (Aerden, 2004), the core of which is located to the north and convex in the opposite direction, i.e., representing a roughly equivalent orogenic curvature that encompasses all of Iberia. Although the mechanics, kinematics, and even the existence and geometry of these plate-scale vertical folds are still discussed (e.g., Martínez Catalán, 2011a; Shaw et al., 2012; Weil et al., 2013; Martínez Catalán et al., 2014; Pastor-Galán et al., 2015; Dias et al., 2016), the development of strike-slip shear zones is considered to be intimately related to the evolution of the Iberian oroclines (e.g., Martínez Catalán, 2011a; Weil et al., 2013; Gutiérrez-Alonso et al., 2015).

Whereas the Ibero-Armorican arc affects many of the strike-slip shear zones (note their curved pattern in Fig. 1), none of them conform to the arched geometry of the Central Iberian arc. The JPSZ and Porto-Tomar shear zone, for example, cut across the axial zone of the Central Iberian arc without being affected by it. Accordingly, (1) the Ibero-Armorican arc is the latest orocline formed in this part of the orogen (Martínez Catalán, 2011a); (2) the Ibero-Armorican arc was nucleated after ca. 309–304 Ma, which is the putative age of the onset of the Porto-Tomar shear zone (this age is also supported by abundant paleomagnetic and structural data; Weil et al., 2010, 2013); and (3) the Central Iberian arc was nucleated before the onset of strike-slip shear zones in the Central Iberian Zone (i.e., before ca. 317–311 Ma; Martínez Catalán, 2012).

The Ibero-Armorican arc is thought to be formed under north-northeast–south-southwest compression (present-day coordinates; Weil et al., 2013), a trend that fits the vector of convergence proposed for D3–D5 in the study case. However, the Trancoso-Pinhel region is located along one of the branches of the Ibero-Armorican arc and, consequently, the original orientation of the vector of convergence for D3–D5 must have been rotated to some extent during orocline development. In any case, the original vector of convergence for D3–D5 should have a more east-west orientation relative to the north-northeast–south-southwest compression suggested for the Ibero-Armorican arc.


The late Carboniferous strike-slip shear zones of the Iberian Massif are not coeval, and therefore can be used to derive and discuss the evolution of the strain and kinematics during the waning stages of the Gondwana-Laurussia collision. Structural analysis performed in the Trancoso-Pinhel region of the Central Iberian Zone reveals that during the late Carboniferous strike-slip deformation was stepped, and created different sets of dextral and sinistral subvertical shear zones in fewer than ∼15 m.y. under partitioned transpression. The strike-slip shear zones analyzed are among the largest in the central part of the Iberian Massif. Therefore, major dextral and sinistral shear zones cannot be integrated in an extrusion model featured by (coeval) conjugated faults. Even though the trace and kinematics of the analyzed strike-slip shear zones are remarkably different, no major changes in the orientation of continental convergence are required to explain the sequence of late Carboniferous structures observed. Such differences, however, can be explained by a rheologically heterogeneous lithosphere by the end of the Variscan orogeny. Mechanical anisotropies were abundant after the gravitational and thermal reequilibration of the orogen, which produced widespread magmatism and extensional shear zones that constituted weak zones for absorbing subsequent lateral movements upon superimposed compression. The widespread nature of late upright folds across the Iberian Massif indicates a tectonic origin for such compression, likely derived from further convergence between Gondwana and Laurussia. A kinematic analysis of the late strike-slip shear zones suggests that they all may have resulted from dextral oblique convergence. This type of convergence also explains the previous tectonic evolution recorded in central and northern Iberia related to folds, thrusts, and extensional faults, and points out the long-lasting nature of lateral movements during the amalgamation of Pangea.


We thank Stephen Johnston and Brendan Murphy for insightful revisions. Financial support was provided by Fundação para a Ciência e Tecnologia (Portugal) through the research project GOLD (Granites, Orogenesis, Long-term strain/stress, and Deposition of ore metals) (PTDC/GEO-GEO/2446/2012: COMPETE: FCOMP-01–0124-FEDER-029192). Díez Fernández appreciates financial support from Fundação para a Ciência e Tecnologia (Portugal) through its postdoctoral program (SFRH/BPD/85209/2012). This is IBERSIMS (SHRIMP Ion-Microprobe Laboratory of the University of Granada) publication 41, and a contribution to IGCP (International Geoscience Programme) Project 648 (Supercontinent Cycle and Global Geodynamics).

1GSA Data Repository Item 2017250, containing field photographs and Table DR1: U-Pb geochronology data of samples TP-4 and TP-13, is available at http://www​.geosociety.org/datarepository/2017, or on request from editing@geosociety.org.