Internal structure of the Late Triassic Central Patagonian batholith at Gastre, southern Argentina: Implications for pluton emplacement and the Gastre fault system

The Central Patagonian batholith (CPB; Fig. 1A; Rapela et al., 1991, 1992; Zaffarana et al., 2014) is a suite of Late Triassic calcalkaline plutons that represents a key element in paleogeographic models of pre-Gondwana breakup, as it was emplaced during the transition between the Gondwanide (late Paleozoic) and the Andean (Jurassic to present) tectonic cycles in South America. The emplacement of the CPB has been inferred as syntectonic with the activity of the Gastre fault system, a controversial NW-SE subvertical structure that has been conceived as a major dextral fault cross-cutting Patagonia (Fig. 1A; Rapela et al., 1991, 1992; Rapela and Pankhurst, 1992; Marshall, 1994; König and Jokat, 2006; Martin, 2007; Riley et al., 2016). The Gastre fault system was widely invoked to restore Patagonia to achieve a tectono-stratigraphic correlation between the Paleozoic successions from the Malvinas-Falkland islands and the Cape fold belt of South Africa (Marshall, 1994; Storey et al., 1999; Macdonald et al., 2003; Hole et al., 2016). This correlation also requires ∼100° clockwise rotation of the Malvinas-Falkland islands, which in turn has been independently supported by paleomagnetic results from NE-SW–trending Jurassic dikes cropping out in the islands (Taylor and Shaw, 1989). According to this model, almost all the dextral motion through the Gastre fault system must be younger than 190–180 Ma, which is the age of the paleomagnetically studied dikes (Mussett and Taylor, 1994; Stone et al., 2008, 2009; Hole et al., 2016). Figure 1A shows the inferred trace of the Gastre fault system passing along the outcrops of the Late Triassic CPB. Recent paleomagnetic data from the overlying Lonco Trapial Formation (Fig. 1B) indicate that no clockwise tectonic rotations took place in the Jurassic through the Gastre district (Zaffarana and Somoza, 2012).

Furthermore, local mesoscale geological observations in the type locality of the Gastre fault system argue against the idea that the CPB has been affected by continental-scale Jurassic dextral shearing (von Gosen and Loske, 2004; Zaffarana et al., 2010, 2014). However, a comprehensive structural characterization of the CPB has remained a pending task. In this contribution, we present the results of a study of the internal structure of the plutons that compose this batholith. In many sectors of the batholith, the determination of foliation and lineation is difficult because of the preponderance of magmatic fabrics and because of the common presence of equigranular and in some cases fine-grained plutonic rocks. To solve this, we performed anisotropy of magnetic susceptibility (AMS) measurements in systematically distributed sites (as much as possible) over an area of ∼400 km² around the Gastre locality. The integration of the AMS results with our more local meso- and microscale structural observations results in a suitable picture of the internal structure of the plutonic rocks (Fig. 1B), which has implications to magmatic processes.

The CPB is composed of Late Triassic (time scale of Cohen et al., 2016) calcalkaline plutons (Rapela et al., 1991, 1992; Rapela and Pankhurst, 1992). These granitoids have initial Sm/Nd and ⁸⁷Sr/⁸⁶Sr isotopic ratios intermediate between the primitive, mantle-derived isotopic signatures of the Andean arc-related batholiths and that of the more evolved, widely distributed Permian granitoids of northern Patagonia (Rapela et al., 2005; Pankhurst et al., 2006). These intermediate ratios reflecting mantle and crustal contributions led previous authors to consider the plutons of the CPB as “modified I-type” (Pankhurst, 1990; Rapela et al., 1991, 1992; Rapela and Pankhurst, 1992).

Exposure of the CPB is excellent, with an overall vertical relief of ∼800 m. Access is difficult toward the north, where the granites are covered by the Lower Jurassic volcanic rocks of the Lonco Trapial Formation (Fig. 1B). The CPB is primarily comprised of diorites to granites of the older Gastre superunit, which are in turn intruded by the granites of the Lipetrén superunit (Rapela et al., 1991, 1992; Zaffarana et al., 2014).

The oldest unit of the Gastre superunit comprises equigranular hornblende-biotite granodiorites (Zaffarana et al., 2014). These granodiorites pass gradually to porphyritic biotite-hornblende monzogranites (Zaffarana et al., 2014). Another unit in the Gastre superunit is composed of equigranular biotitic monzogranites, which form stocks of light pink, medium- to coarse-grained equigranular monzogranites that intrude both the equigranular hornblende-biotite granodiorites and the porphyritic biotite-hornblende monzogranites (Zaffarana et al., 2014). The Gastre superunit is intruded by at least three stocks of dioritic to quartz-monzodioritic composition and by dioritic to quartz-dioritic dikes. The stocks and the dikes comprise the hornblende quartz diorites (Zaffarana et al., 2014). The Lipetrén superunit is predominantly composed of more evolved granitic rocks, mostly of biotitic monzo- and syenogranites that intrude all of the units of the Gastre superunit (Zaffarana et al., 2014). Big tabular bodies of the biotitic granites are particularly well exposed to the east of Gastre village at the Cordón de la Cruz (Fig. 1; Zaffarana et al., 2014). Volcanic to subvolcanic equivalents of these granites are difficult to find. Possible volcanic equivalents are the felsites that crop out in several places of the northern studied sector, in all cases close to the contact with the overlying Lonco Trapial Formation (Lower Jurassic) and which were mentioned by Rapela et al. (1991). They are pink to brown aplitic rocks mostly composed of small crystals of quartz and feldspar. These rocks could signal the highest structural level exposed in the batholithic system. The only clearly extrusive rock that could be associated with the magmatic system of the CPB corresponds to a meter-scale outcrop (42°01.169′S, 69°23.198′W) of a gray ignimbrite with eutaxitic texture and a rhyolitic to dacitic composition.

The Gastre superunit yields a Rb-Sr isochron age of 222 ± 3 Ma (data from Rapela et al. [1992] reinterpreted by Zaffarana et al. [2014]) and a U-Pb zircon age of 221 ± 2 Ma that was mentioned (without providing analytical data) by Rapela et al. (2005). The Gastre superunit also yields an ⁴⁰Ar/³⁹Ar stepwise biotite cooling age of 213 ± 5 Ma (Zaffarana et al., 2014). With respect to the Lipetrén superunit, a recently reported U-Pb SHRIMP (sensitive high-resolution ion microprobe) age of 215 ± 1 Ma for its monzogranites (Lagorio et al., 2015) suggests that the previous ⁴⁰Ar/³⁹Ar stepwise biotite cooling age of 206 ± 4 Ma and the Rb/Sr isochron age of 206.4 ± 5.3 Ma discussed by Zaffarana et al. (2014) represent cooling ages associated with moderate cooling rates.

The Lipetrén superunit also comprises the Horqueta Granodiorite (Rapela et al., 1991, 1992), a set of biotitic granodiorites whose age had remained poorly constrained (Zaffarana et al., 2014) until the acquisition of a new U-Pb zircon SHRIMP crystallization age of 213 ± 2 Ma (Lagorio et al., 2015). This age overlaps with the ⁴⁰Ar/³⁹Ar stepwise biotite cooling age of 211 ± 2 Ma reported by Zaffarana et al. (2014). The crystallization ages of the Gastre and Lipetrén superunits suggest that a time span of ∼7 m.y. elapsed between their emplacements.

Outcrops of host rock of the CPB are scarce because, with the exception of the metamorphic Calcatapul Formation (Proserpio, 1978), possible host rocks would be hard to identify in the field because they belong to granites of the previous late Paleozoic intrusive cycle (the Permian granitoids of the Mamil Choique Formation; Cerredo and López de Luchi, 1998; López de Luchi and Cerredo, 2008). The Paleozoic Calcatapul Formation crops out exclusively in the southern side of the Sierra de Calcatapul (near Puesto Yancamil and Puesto Uribe; Fig. 1). It consists of a steeply dipping, dynamically metamorphosed succession of metavolcanic rocks with thin intercalations of phyllites and metaconglomerate lenses (Proserpio, 1978; von Gosen and Loske, 2004; Zaffarana et al., 2010). The Calcatapul Formation shows a subvertical, NW-SE penetrative S₁ foliation developed under greenschist facies conditions that is almost parallel to the bedding planes (von Gosen and Loske, 2004; Zaffarana et al., 2010). The lineation is subvertical at Puesto Yancamil and subhorizontal a few kilometers northwards (von Gosen and Loske, 2004; Zaffarana et al., 2010; P53 and P156 in Table S1 in the Supplemental Materials¹), suggesting that the S₁ planes either formed under partitioned transpression or were produced by separate deformation events. The Permian biotitic porphyritic Yancamil Granite (von Gosen and Loske, 2004) intrudes the Calcatapul Formation at Puesto Yancamil. The Yancamil Granite and its host bear a tectonic foliation and lineation with the same orientation (von Gosen and Loske, 2004; Zaffarana et al., 2010; the granite corresponds to site P148 in Table S1 [see ^{footnote 1}]). At Puesto Yancamil, the NW-SE steep foliation and the steeply dipping lineation of Calcatapul Formation were interpreted as the result of downfaulting deformation or of oblique (sinistral) displacements (Franzese and Martino, 1998; von Gosen and Loske, 2004; Zaffarana et al., 2010). Additionally, some insights about the host rock near the town of Gastre are given by small (decameter- to hectometer-scale) stoped blocks (Yoshinobu et al., 2009) of metamorphic rocks enclosed within the CPB (Zaffarana et al., 2012, 2014).

Microstructural observations were performed on a collection of 110 thin sections. Samples from outcrops having visible foliation were oriented according to planar and, whenever possible, linear structures observed in the field. Fabrics of the CPB range from pre–full crystallization state to superimposed low-temperature solid-state deformation. For practical purposes, the outcrops selected for AMS studies were classified as exhibiting magmatic or solid-state fabrics (“M” or “S”, respectively, in Tables S1–S3 [see ^{footnote 1}]) on the basis of observable megascopic structures in the field.

Sites involved in the magmatic, or M, field classification (example in Figs. 2A–2D) comprise rocks whose microstructures are purely magmatic, submagmatic (with low melt percent; Paterson et al., 1989; Bouchez et al., 1992; Cruden et al., 1999), and/or with a mild overprinting of solid-state deformation, the latter formed especially in high-temperature conditions near the solidus temperature (Blumenfeld et al., 1986; Gapais and Barbarin, 1986; Mainprice et al., 1986; Paterson et al., 1989; Figs. 2A–2D). The high-temperature solid-state fabrics are in all cases coaxial with the magmatic fabrics.

In contrast, the rocks from sites classified as exhibiting S fabrics show solid-state deformation visible at hand-sample scale (Figs. 2E–2H). In this case, microstructures involve grain-size reduction and the development of moderate to strong mineral foliation and, in some cases, mineral lineation (Fig. 2F). Microstructures include subgrains in quartz and incipient bulging of and microfracturing in feldspar. These processes which result in S fabrics occur at relatively low- to medium-temperature conditions (300–500 °C) and/or at fast strain rates (Hirth and Tullis, 1992; Tullis et al., 2000). However, when deformation starts at high-temperature conditions and almost continuously follows up to low-temperature conditions (as is a common case in deformed granites), the resulting fabrics are classified in the field as S as well (Figs. 2G–2H). Mylonitic fabrics are a special type of S fabrics.

In the CPB as a whole, magmatic fabrics (M fabrics) predominate over the solid-state fabrics (S fabrics; Fig. 2I). The orientation of the magmatic and tectonic foliations and lineations measured in the field are detailed in Tables S1–S3 (see ^{footnote 1}). The magmatic foliations are usually well defined and mostly subvertical, whereas the magmatic lineations are rarely observable. The dioritic dikes belonging to the hornblende quartz diorites are generally subvertical, with variable strike, and dispersed throughout the Gastre superunit. We have sampled 11 dikes, all of them subvertical, with predominantly NE-SW or NNE-SSW strike (six dikes), but NW-SE (three dikes) and N-S strikes (two dikes) also occur (Table S2 [see ^{footnote 1}]). Six dikes have magmatic fabric and five have low-temperature solid-state deformation; the solid-state deformational fabrics all have steep foliation planes (70° or higher; Tables S2–S3 [see ^{footnote 1}]).

Ductile shear zones were observed in several localities (Fig. 2I). The high- and low-temperature deformation present at Puesto Yancamil and at Puesto Jaramillo was described by von Gosen and Loske (2004) and by Zaffarana et al. (2010), and the observations in Puesto Yancamil were summarized previously. In this work, we contribute the description of the deformation and kinematics of seven additional mylonitic zones. The first one corresponds to the Calcatapul Formation at Puesto Uribe (Fig. 1; P139 in Table S1 [see ^{footnote 1}]). There, metaconglomerate layers have a NW-SE subvertical foliation and a lineation plunging 50° toward the NW with an oblique (sinistral) sense of shear. The other shear zones are developed in granites of the CPB, where three of them are assigned to the Gastre superunit and the other two to the Lipetrén superunit (Table S3 [see ^{footnote 1}]; Fig. 3). The thickest (∼100 m 215 thick) mylonitic zone has oblique (sinistral) kinematics and is located near Puesto Uribe (“Uribe 216 mylonite” at P138 in Table S3 [see ^{footnote 1}]). It has a suspected Gastre superunit protolith. There, the tectonic foliation strikes NNW-SSE and the lineation plunges steeply (60°–70°) toward the NNW (Fig. 3A). K-feldspar and granitic porphyroclasts of different sizes (from millimeters up to 50 cm) are surrounded by strain shadows defining δ-type or, more commonly, σ-type microstructures, which suggest uplift of the southwest block with respect to the northeast block with a sinistral shear component (Figs. 3A–3D). About 500 m to the northwest of this shear zone, the hosting metapyroclastic rocks of Calcatapul Formation also appear mylonitized (site P139, Table S1 [see ^{footnote 1}]), with the foliation and lineation parallel to that observed in the above-mentioned ∼100-m-thick mylonite. The metapyroclastic rocks contain angular porphyroclasts of quartz, microcline, and zonal plagioclase immersed in a matrix composed of quartz, sericite, and chlorites with evidence of dynamic recrystallization. Porphyroclasts in the metapyroclastic rocks have straight grain boundaries, and nearly all are surrounded by strain shadows and by quarter mats (Hanmer and Passchier, 1991).

The other shear zones that traverse the granites of the CPB are much thinner and of lesser importance. The mylonites near Puesto Rechene have a NW-SE–trending subvertical mylonitic foliation and form a ∼15–20-m-wide zone with a shallowly plunging lineation defined by elongated quartz and feldspar grains (Figs. 3E–3H; “Rechene mylonite” at P143 in Table S3 [see ^{footnote 1}]). Sinistral sense of shear is suggested by the development of S-C′ planes. The ductile shear zone near La Horqueta has a mylonitic foliation in a ∼10-m-wide zone, with NE-SW strike and a moderate dip (40°–50°) to the SE (Figs. 3I–3L; “Horqueta mylonite” at P145 in Table S3 [see ^{footnote 1}]). An oblique (dextral) sense of shear was detected parallel to the moderately SW-plunging mineral lineation. Foliation is regular and defined by lenses of quartz and feldspars with dynamic recrystallization surrounded by bands of biotite and opaque minerals. The rocks are strongly annealed, as quartz and feldspars originally formed by dynamic recrystallization appear now as a mosaic of equigranular grains free of deformation with straight grain boundaries meeting at triple junctions.

The equigranular hornblende-biotite granodiorites of the Gastre superunit are mylonitized at P144 (Figs. 3M–3P; “S-C mylonite” in Table S3 [see ^{footnote 1}]) with an E-W subvertical foliation and a subhorizontal mineral lineation defined by elongated quartz and biotite crystals. The shear zones are thin and parallel, between 50 and 20 cm wide. The S-C structure defines a dextral sense of shear.

The Lipetrén superunit is mylonitized in two mylonitic zones located near the east of the Gastre village. The northern one (site P59 in Table S3 [see ^{footnote 1}]; Figs. 4A–4D) trends NNW-SSE and dips moderately to the west (∼50°–70°), its width is 10 m, and the mylonitic foliation is defined by the alternation of lighter bands composed of quartz and micas with darker bands composed of opaque minerals and micas. The lineation plunges ∼30° to the NW. S-C′ microstructures suggest oblique (sinistral) uplift of the western block with respect to the eastern one. The mylonitic foliation is cut by low-angle brittle fractures filled with quartz, feldspar, and epidote which developed in a younger deformational event (Fig. 4D). The southern shear zone developed in the Lipetrén superunit (site P20 in Table S3 [see ^{footnote 1}]; Figs. 4E–4H) is ∼20 m thick. It consists of fine-grained pink mylonitic rocks with a regular ∼N-S, subvertical foliation. A moderately to steeply plunging lineation is defined by elongated quartz and feldspars. von Gosen and Loske (2004) described uplift of the western block with a sinistral component. The two ductile shear zones that crop out east of the Gastre village show triple junctions among quartz and feldspars, suggesting that they were overprinted by late to post-kinematic annealing.

Anisotropy of magnetic susceptibility (AMS) is a standard technique for petrofabric studies in granitoids (e.g., Archanjo et al., 1995, 2002; Saint Blanquat and Tikoff, 1997; Ferré and Améglio, 2000; McNulty et al., 2000; Neves et al., 2003; Žák et al., 2005; D’Eramo et al., 2006; Stevenson, 2009; Olivier et al., 2016). In particular, AMS provides a determination of the magmatic lineations, which are generally hard to measure in the field or in the laboratory. The three principal axes of the AMS ellipsoid (K_max ≥ K_int ≥ K_min) define a magnetic lineation parallel to K_max and a magnetic foliation as the plane containing K_max and K_int, with K_min being the pole to foliation. The common relationship between the AMS ellipsoid and petrofabric shows the magnetic lineation parallel to the structural lineation (stretching direction) and the magnetic foliation parallel to the structural foliation (flattening plane) (Borradaile and Jackson, 2010).

The low-field AMS survey in the CPB comprises a total of 149 sites (stations; 1219 oriented cores; the number of cores of each station is the “N” in Tables S1–S3 [see ^{footnote 1}]), of which 16 sites were included in previous studies (Zaffarana et al., 2010, 2012; the sites included in previous studies are shown with one or two asterisks in Tables S1–S3 [see ^{footnote 1}]). AMS measurements were performed using a MFK1-B Kappabridge susceptibilimeter. AMS ellipsoids were calculated using matrix averaging routines (Jelinek, 1978) with the programs Anisoft 4.2 (Chadima and Jelinek, 2009). The Gastre superunit was studied in 91 sites (72 bearing magmatic fabric and 19 bearing solid-state deformation fabric), and the Lipetrén superunit was studied in 52 sites (43 with magmatic fabric and nine with solid-state deformation fabric). The host rock was studied in six sites.

Determining the minerals that dominate the magnetic signal is important when analyzing AMS data. To do this, we performed hysteresis curves and isothermal remanent magnetization-backfield analyses in order to gain information on coercive force H_c, saturation magnetization M_s, remanent magnetization M_r, and remanent coercive force H_cr, which allow estimating the domain state (i.e., size) of titanomagnetites. Temperature variation of bulk magnetic susceptibility (thermomagnetic curves) was also applied to further investigate the magnetic properties of rock minerals (e.g., Hrouda, 2003, 2010). Additionally, representative specimens were studied via performing thermomagnetic curves and inspection of polished sections in order to gain further insights into oxide mineralogy.

The degree of magnetization of a material in response to an applied magnetic field is reflected by a dimensionless proportionality constant known as magnetic susceptibility (K). The bulk or mean magnetic susceptibility of a rock reflects the sum of the individual contributions from all of its forming minerals (K_m = [K_max + K_int + K_min] / 3), so magnetic susceptibility has compositional significance. Bouchez (2000) proposed that when bulk magnetic susceptibility is higher than 4 × 10⁻⁴ SI, the magnetic susceptibility is mainly controlled by ferromagnetic minerals (mainly magnetite), and the rock can be classified as “ferromagnetic”. In contrast, when bulk magnetic susceptibility is lower than this value, it is assumed to be largely controlled by paramagnetic minerals (e.g., mafic silicates), and the rock can be classified as “paramagnetic”. In the Lipetrén superunit, there are sites whose mean bulk susceptibility is on the order of 10⁻⁵ SI, therefore they are classified as weakly paramagnetic. The classification of each AMS site is given in Tables S1–S3 (see ^{footnote 1}). In Figures S1–S4 (see ^{footnote 1}), there are also detailed descriptions of the magnetic mineralogy of each unit.

The distribution of K_m shows that sites with K_m < 1 × 10⁻³ SI predominate (Fig. 5A), because even though there is compositional variation between granites and diorites, the acidic terms (granites and granodiorites) predominate. Mean magnetic susceptibility in the Gastre superunit is 19.45 × 10⁻³ SI, so it is comprised by ferromagnetic granites (Bouchez, 2000). Titanomagnetite is in all cases the main ferromagnetic mineral in the Gastre superunit, appearing as isolated grains included in other minerals or forming part of biotite + amphibole clusters. In general, higher values of K_m correlate with higher contents of mafic minerals in the Gastre superunit (the hornblende quartz diorites tend to show the greatest K_m values; see the correlation in K_m in Fig. 5A and the composition patterns within the CPB shown in Fig. 1).

In the Lipetrén superunit, opaque minerals are <1% in volume and are generally present as isolated euhedral to subhedral grains that occur as inclusions in other minerals. A second generation of anhedral opaque minerals is present along biotite cleavage planes and is interpreted to be associated with the alteration of biotite. Average mean magnetic susceptibility for the monzo- and syenogranites of this superunit is 2.95 × 10⁻³ SI. According to their susceptibility value, 30 sites of the biotitic granites of the Lipetrén superunit have been classified as ferromagnetic, eight sites as paramagnetic, and seven sites as weakly paramagnetic (these biotitic granites appear as sites with K_m <1 × 10⁻³ SI in Fig. 5A). Hematite is also present in some cases within the opaque phases of the biotitic granites of the Lipetrén superunit as deuteric alteration of biotite and titanomagnetite (see Supplemental Figure S4 [see ^{footnote 1}]). Magnetic susceptibility of the Horqueta Granodiorite is rather homogeneous, around 37 × 10⁻³ SI, and dominated by multidomain (MD) magnetite (these sites are included in the group with <1 × 10⁻³ SI in Fig. 5A).

The Calcatapul Formation is largely dominated by MD magnetite (Zaffarana et al., 2010). Mean susceptibility values from the Yancamil Granite (1.8 × 10⁻³ SI; Table S1 [see ^{footnote 1}]) are compatible with the ferromagnetic field, however hysteresis loops and the thermomagnetic curves from this unit show that its magnetic susceptibility is actually controlled by the superimposed effects of both magnetite and paramagnetic minerals.

The degree of anisotropy of the AMS ellipsoid can be described by the corrected anisotropy degree P′ (Jelinek 1981), which is a measure of the degree to which the AMS ellipsoid deviates from a sphere. The shape of the AMS ellipsoid can vary from oblate (pancake shaped) to prolate (pencil shaped). The T parameter, where T = (ln F − ln L) / (ln F + ln L) (Jelinek 1981) is one way of describing the ellipsoid shape, where L is the magnetic lineation (K_max / K_min) and F is the magnetic foliation (K_int / K_min). The magnetic ellipsoid is oblate for 0 ≤ T ≤ 1 and prolate for −1 ≤ T ≤ 0.

In general, in both in the Gastre and Lipetrén superunits of the CPB there seems to be a weak correlation between K_m and P′, suggesting that there could be some compositional (magnetite content) control on rock anisotropy (Fig. 6A). This is also reflected in the maps, as the sites with higher K_m generally have higher anisotropy degree (Figs. 5A–5C). However, beyond this weak compositional control in the anisotropy, the magmatic fabrics of both superunits show a direct relationship between anisotropy degree P′ and T, as more anisotropic fabrics are more strongly oblate (higher T; Figs. 5B–5C). In the maps, the highest anisotropy degree values (P´ > 6%) are found in the region between Gastre and Puesto Jaramillo (Fig. 5B). This is also the area where the magmatic (magnetic) foliations of the CPB define a well-ordered NW-SE–striking, steeply dipping pattern (see the next section). In addition, this area tends to concentrate the most strongly oblate magmatic fabrics (T > 0.5; Fig. 5C; compare with the fabric types of Fig. 2I). The anisotropy degree is lower (<6%) toward the north of Puesto Rechene and also in the area between Gastre and La Horqueta (Fig. 5B), where more disordered magmatic fabrics are seen (see the next section). Prolate ellipsoids (with T > 0.5) are very scarce and their location is random (Fig. 5C).

With respect to overprinting of solid-state deformation on the magmatic fabrics, the behavior is slightly different in each superunit. In the Gastre superunit, granites that underwent low-temperature solid-state deformation have lower mean susceptibility values than rocks having exclusively magmatic petrofabric. This could be explained by magnetite alteration to hematite by fluid circulation during low-temperature deformation. An exception to this behavior is the dioritic dikes of the hornblende quartz diorites, which show a slight increase of P′ associated with higher values of K_m. This case could be explained, instead, by magnetite generation due to biotite alteration during low-temperature deformation. In the Lipetrén superunit, P′ values >10% are nearly everywhere seen in rocks with overprinting of magmatic fabrics by low-temperature solid-state deformation (Figs. 6A–6B), suggesting that during solid-state deformation there is magnetite generation in these rocks. With respect to the shape of the AMS ellipsoids, in the sites where the magmatic fabric is overprinted by low-temperature solid-state deformation, oblate ellipsoids (T > 0.5) predominate (Fig. 6B).

In all cases where measurement of magmatic or tectonic foliations was possible in the field, we noted that they are parallel to the corresponding magnetic foliation. Parallelism between the visible lineation and the magnetic lineation in areas of solid-state deformation was also observed, also including the two localities of the Calcatapul Formation. Coaxiality between magmatic-tectonic and magnetic fabrics is independent of the mineral that dominates the bulk susceptibility of the rock. These observations strongly suggest that the magnetic fabric of the rocks appropriately represents their petrofabric and is a suitable tool to map the internal structure of the plutons.

Overall, the magnetic fabrics show roughly coherent patterns throughout the studied area. The more conspicuous feature in the internal structure of the studied granitoids is the presence of NW-SE–trending, dominantly steeply dipping magmatic (magnetic) foliations in the rocks of the Gastre superunit, especially in the Loma Alta area (between Gastre and Puesto Jaramillo; Fig. 7; Table S2 [see ^{footnote 1}]). These NW-SE subvertical magmatic foliations are parallel to slightly oblique with respect to the contacts among units (Fig. 7). The four AMS sites belonging to enclave swarms hosted by the porphyritic biotite-hornblende monzogranites of the Gastre superunit also comprise subvertical NW-SE magmatic foliations in this area (sites marked with an ellipse in Fig. 7 and individualized with stars in Fig. 8). Enclave swarms can be considered as the trace of feeder conduits (Tobisch et al., 1997; Collins et al., 2006) in the CPB. Local deviations of the general subvertical NW-SE trend exist where the contacts of the units of the Gastre superunit display irregular geometry and also toward the north of Puesto Rechene (northern Loma Alta area; see Fig. 7), where the granites of the Gastre superunit also have N-S to NNE-SSW steeply to moderately dipping magmatic (magnetic) foliations (Fig. 8).

Similarly, the magmatic fabrics of the Lipetrén superunit are predominantly parallel to the magmatic fabrics of the Gastre superunit in the Loma Alta area (Fig. 7). However, the magmatic foliations of the Lipetrén superunit deviate from the main NW-SE subvertical trend defined by the magmatic fabrics of the Gastre superunit in the Puesto Jaramillo area (see circled area in Fig. 7). Furthermore, the morphology of the stocks of the Lipetrén superunit in the Loma Alta area is slightly more equidimensional than that of the plutons of the Gastre superunit, which have a more elongated shape. In the area between Gastre and La. Horqueta (Cordón de la Cruz; Fig. 7), the magmatic (magnetic) foliations of the Lipetrén superunit display a disordered pattern, with variable strike and intermediate dip (Fig. 7). The more random magmatic (magnetic) fabrics of the Lipetrén superunit are reflected in the K_min orientation distribution of Figure 7.

The Gastre superunit has predominant magnetic (magmatic) lineations of alternating steep or shallow plunge, although intermediate plunges are also observed (Fig. 8). In the Puesto Jaramillo area, NW-SE shallow magmatic (magnetic) lineations are observed in the Gastre superunit (Fig. 8). In the Lipetrén superunit, the magmatic lineations can be shallowly, intermediately, or steeply plunging (Fig. 8). At La Horqueta, the magmatic (magnetic) lineations of the Lipetrén superunit are predominantly shallowly plunging (Fig. 8).

Solid-state magnetic foliations of the Gastre superunit are nearly everywhere roughly parallel to the fabrics developed in the magmatic stage (Fig. 7). Note that, unlike the magmatic fabrics, the solid-state fabrics in the Lipetrén superunit are parallel to the dominant NW-SE subvertical trend displayed by the Gastre superunit both in magmatic and in solid-state conditions (see circled area near Puesto Jaramillo in Fig. 7). The solid-state lineations in the Gastre and Lipetrén superunits can show either high, intermediate, or shallow plunge, although steeply plunging lineations predominate in both superunits (Fig. 8; Table S3 [see ^{footnote 1}]).

Equilibrium crystallization pressures for the Gastre superunit determined from amphibole compositions by Zaffarana et al. (2014) are reevaluated here in light of the numerous problems that have been noted with the Al content in hornblende as a geobarometer (i.e., Erdmann et al., 2014). One important problem is that the total Al content in amphibole results not only from the pressure-dependent tschermakitic substitution but also from the temperature-dependent edenitic substitution (Erdmann et al., 2014). To infer more real emplacement pressures, we followed the approach of Ribeiro et al. (2016), which consists of comparing the composition of the amphiboles of the Gastre superunit (reported in Zaffarana et al., 2014) with the compositions of experimental amphiboles crystallized under different pressure conditions (e.g., Samaniego et al., 2010). Crystallization and melting experiments show that the Al^T (total aluminum content), Mg≠ (magnesium number; Mg#=100 × mol.Mg/(mol.MgO+mol.FeO)), and Na + K contents of amphiboles vary with pressure and temperature, and that their compositions plot within distinct pressure-dependent compositional fields (Scaillet and Evans, 1999; Prouteau and Scaillet, 2003; Alonso-Perez et al., 2009). The comparison of the amphiboles of the Gastre superunit with those compositional fields demonstrates a good fit for the field of 2.2 kbars (Supplemental Figure S5 [see ^{footnote 1}]), and therefore the emplacement depth of the Gastre superunit is consistent with shallow crustal levels, mostly between 6 and 7 km.

Likewise, the presence of miarolitic cavities and microgranophyric textures in the Lipetrén superunit (Zaffarana et al., 2014) supports the conclusion that it was emplaced in a shallower subvolcanic environment, under pressures of probably <1.5–1.2 kbar (Candela, 1997). Overall, isotopic, geobarometric, and field data document that the Gastre area underwent 3–4 km of exhumation during a time span of ∼7 m.y. in the Late Triassic. Consequently, the CPB comprises adjoining plutons crystallized at slightly different crustal depths (3–4 km of vertical distance). These different crustal depths may have influenced the emplacement mechanisms of the two magmatic suites.

Most of the rocks of the CPB have magmatic fabrics with triaxial to oblate ellipsoids. The most outstanding characteristic when analyzing the magmatic foliations of the Gastre superunit is the presence of a structural corridor of predominantly subvertical NW-SE orientation (Fig. 7). The magmatic foliations can be slightly oblique with respect to the NW-SE trend of the contacts among the units of the Gastre superunit. The plunge of magnetic lineation in the magmatic fabrics of the Gastre superunit varies from steep to shallow (Fig. 8). The magmatic lineation is interpreted as the direction of stretching. In the areas where the magnetic ellipsoids of the Gastre superunit are triaxial, and therefore their magnetic lineation is well defined (Fig. 5C), we find steeply, intermediately, and shallowly plunging lineations interspersed (Fig. 8).

AMS data combined with gravimetric surveys commonly lead to discovery of a granite root or feeder zones (Vigneresse, 1995; Vigneresse and Bouchez, 1997). However, enclave swarms, which are places where mafic and felsic magmas interact, leading to highly heterogeneous rocks, have also been interpreted as possible batholith root zones (Tobisch et al., 1997; Moyen et al., 2003; Collins et al., 2006). The tabular enclave swarms hosted by the porphyritic biotite-hornblende monzogranites of the Gastre superunit could represent root zones within the CPB, and they also bear the predominantly subvertical magmatic NW-SE fabrics (Fig. 7). There, the AMS ellipsoids are strongly oblate, and therefore their intermediately to shallowly plunging magnetic lineation is not so well defined compared to triaxial or prolate sites (Fig. 5C).

The magmatic foliations of the stocks of the Lipetrén superunit are predominantly parallel to those of the Gastre superunit (Fig. 7), but near Puesto Jaramillo the magmatic foliations of the Lipetrén superunit are strongly discordant to the general trend of the batholith (see circled area in Fig. 7). In addition, in the area between Gastre and La Horqueta, the magmatic foliations of the Lipetrén superunit display variable strike and are of intermediate dip in a disordered pattern (Fig. 7). The magnetic lineations (in predominantly triaxial magnetic ellipsoids) are of variable plunge (Figs. 6C, 8). The shallow foliations of the Lipetrén superunit in the area around La Horqueta may correspond to structurally higher emplacement levels, where magma sheets flowed parallel to the roof of the pluton (Fig. 7). In addition, west of Puesto Uribe, von Gosen and Loske (2004) observed that undeformed porphyries of Lipetrén biotitic granites intrude foliated granodiorites ascribed to the Gastre superunit, suggesting that the deformation that controlled the emplacement of the Gastre superunit would not have lasted throughout the emplacement of the Lipetrén superunit.

Several observations throughout the CPB would suggest that it was emplaced by the amalgamation of dike-like magma bodies. For instance, the biotitic granites of the Lipetrén superunit in the area of La Horqueta intrude the equigranular hornblende-biotite granodiorites of the Gastre superunit, forming tabular bodies (Zaffarana et al., 2014, their figure 7A). In addition, the magmatic dikes of the CPB that intrude the NW-SE–foliated Calcatapul Formation at Estancia Yancamil (von Gosen and Loske, 2004) constitute another piece of evidence for dike emplacement. Our paleomagnetic information (work in progress) indicates that outcrops of the CPB having remanence of normal polarity are interspersed with those carrying reversed-polarity magnetization, therefore suggesting that the CPB is formed by multiple plutonic units which were injected episodically, cooling through the blocking temperatures.

Within the CPB, the granites with low-temperature solid-state deformation of the Gastre and Lipetrén superunits predominantly bear NW-SE subvertical foliation planes (Fig. 7), suggesting that the deformation that controlled the magmatic fabrics of the Gastre superunit continued to operate during cooling of both intrusive suites. The lineations of the tectonically deformed granites are either subhorizontal (as in the area of Puesto Jaramillo; Fig. 8) or subvertical (toward the west of Gastre village; Fig. 8). The magnetic ellipsoids of the low-temperature deformed rocks are mostly oblate, but the direction of the tectonic lineation was in all cases observed in the field as well. The presence of magmatic and solid-state vertical, intermediate, and horizontal lineations, together with the predominance of subvertical oblate ellipsoids (Figs. 6C, 7, 8), would suggest that the emplacement of the CPB would have taken place within a transpressive strike-slip zone with partition of the deformation (Fossen et al., 1994; Tikoff and Greene, 1997; Fossen and Tikoff, 1998; Saint Blanquat et al., 1998). Transpressive environments are rather common in magmatic arcs in conditions of slightly oblique convergence (Saint Blanquat et al., 1998). In this context, the less-controlled general morphology and magmatic fabrics of the Lipetrén superunit can be ascribed to its shallower emplacement level compared to the Gastre superunit, which would make it less affected by transpression-induced forces (Saint Blanquat et al., 1998). The exhumation of the CPB was coeval with the development of metamorphic accretionary complexes in the Patagonian Andes (Thomson and Hervé, 2002; Hervé et al., 2003, 2008), as well as with the Peninsula orogeny in West Antarctica (Miller, 1983; Vaughan and Livermore, 2005) and with the Rangitata I orogeny in New Zealand (Bradshaw et al., 1981; Vaughan and Livermore, 2005). This supports the concept that the uplift and denudation of the CPB was associated with an accretionary episode of regional extent along the proto-Pacific margin of Gondwana.

The shallower emplacement level of the Lipetrén superunit (∼3 km or less) would allow roof uplift as one of the possible space-making processes that could have helped the intrusion of stocks. The Gastre superunit has more rectilinear fabric, as is normal in deeper zones within batholiths emplaced in transcurrent environments (Moyen et al., 2003).

Structural observations in the mylonite-hosting Calcatapul Formation at Puesto Yancamil have led several workers to propose a predominantly sinistral shear component (Franzese and Martino, 1998; von Gosen and Loske, 2004; Zaffarana et al., 2010). Moreover, Zaffarana et al. (2010) proposed that transpressive deformation controlled the syntectonic emplacement of the late Paleozoic Yancamil Granite. If the Uribe mylonite (Fig. 3A), which is the thickest mylonite zone within the domain of the CPB, effectively represents a mylonitized granite of the Gastre superunit, then this would be strong evidence of the syntectonic emplacement of the Gastre superunit within a sinistral transpressive regime. This deformation would have lasted from late Paleozoic to Early Triassic times. The Rechene mylonite has sinistral sense of shear, thus its kinematics is concordant with that of the Uribe mylonite. The Horqueta mylonite and the S-C mylonite at site P144 (Figs. 3I–3M) display kinematic indicators of dextral motion, but they have a different trend than the rest of the batholith. Possible dextral solid-state strike-slip motion has been also noted by von Gosen and Loske (2004) in the area of Jaramillo (Fig. 1). These authors also reported a conjugate set of subvertical, dextral and sinistral shear planes affecting Paleozoic rocks and CPB granites at Estancia Yancamil, interpreting this younger deformation as a response to post-CPB NE-SW compression. Finally, the mylonites found at the east of Gastre have a sinistral shear sense component (Figs. 4A–3H), but their tectonic foliation is not parallel with the magmatic fabrics in the same area (Fig. 7). Curiously, the N-S orientation of these mylonitic bands is parallel to the trend of the valley that connects Gastre with La Horqueta (Fig. 7), suggesting that they could have been produced by a younger structure, probably Andean, which could be responsible for the uplift of the Cordón de la Cruz.

Having analyzed the magmatic and tectonic fabrics of the CPB and their possible significance, we interpret that the CPB emplaced syntectonically within a sinistral transpressive system (Fig. 9). At stage 1, the intrusion of the earliest facies of the Gastre superunit took place in a dominantly NNW-SSE– to NW-SE–elongated crustal deformation zone (Fig. 9A). In this sense, the NNW-SSE subvertical magmatic foliations and the steeply dipping lineations of the porphyritic biotite-hornblende monzogranites of the Gastre superunit located north of Puesto Rechene area (Figs. 7–8) could represent early-formed magmatic fabrics. At stage 2, the intrusion zone enlarged during the injection of the main facies of the Gastre superunit: the equigranular biotitic monzogranites, the equigranular hornblende-biotite granodiorites, and the porphyritic biotite-hornblende monzogranites hosting the enclave swarms (Fig. 9B). In this tectonic framework, the subvertical NE-SW orientation is an extensional direction, and it coincides with the predominant orientation of the dioritic to quartz-dioritic dikes of the Gastre superunit. Finally, the more isotropic biotitic granites of the Lipetrén superunit intruded at shallower crustal levels during the exhumation of the Gastre superunit (stage 3 in Fig. 9C).

Note that the fabric observed in the CPB roughly matches the structural grain of northern Patagonia, which has a predominant NW-SE subvertical orientation that has been likely inherited from the pervasive late Paleozoic deformation (von Gosen, 2002, 2003, 2009; Álvarez et al., 2014; Giacosa et al., 2004, 2014, 2017). A similar geometry is seen in other Triassic plutons from northern Patagonia cropping out as far as 300 km north of the CPB, where a NW-SE–trending subvertical fabric, including mylonites, has been observed (e.g., Báez et al., 2016). This crustal structure in northern Patagonia also played a role in controlling the opening of the Jurassic basins during the early stages of the Andean tectonic cycle (e.g., Figari et al., 1994, 2015). Considering these observations, the magma batches that form the CPB could have been emplaced along preexisting fractures, within a previously structured upper crust (e.g., as in Fig. 9). This has been recognized in other localities of northern Patagonia, where a previously acquired crustal structure channeled early Mesozoic igneous intrusions (e.g., Márquez et al., 2011; González et al., 2014).

We envisage that the model depicted in Figure 9 is strongly dependent on the inferred Late Triassic age of the oblique-sinistral mylonites at Puesto Uribe. If the deformation of those mylonites with granitic protolith (Uribe mylonite; Fig. 3A) were of late Paleozoic age (as happens with the late Paleozoic Yancamil Granite, which was deformed in late Paleozoic times as stated by von Gosen and Loske [2004]), then there would not be enough arguments to propose that the same sinistral transpressive deformation lasted protractedly throughout the emplacement of the CPB. Nonetheless, the CPB does show evidence for transpressive deformation, but it could have been produced by a separate event in the Early Triassic whose kinematics would not be clear in light of the present data. However, the results of this study contradict the hypothesis that the CPB has been affected by continental-scale Jurassic dextral shearing through the Gastre fault system. This conclusion arises for three reasons: (1) mylonites with sinistral shear sense predominate in the host rock of the CPB, (2) the mylonite zones that traverse the granitic rocks in the area of Gastre are generally thin, and (3) the NW-SE and NNW-SSE mylonite zones with suspected CPB protoliths (including Uribe mylonite, which is the thickest one) record a sinistral shear sense component (Figs. 3, 4).

The sinistral transpressive syntectonic emplacement of CPB probably took advantage of a preexisting fabric within the structural grain of northern Patagonia, and this conclusion has further implications for the existence of the NW-SE subvertical Gastre fault system. The CPB is located in the type locality of this controversial fault, and its outcrops have been considered as records of the activity of the Gastre fault system (Rapela et al., 1991, 1992; Rapela and Pankhurst, 1992). Figure 1A shows the inferred trace of the Gastre fault system passing along the outcrops of the Late Triassic CPB. Recent paleomagnetic data from the overlying Lonco Trapial Formation (Fig. 1B) signal that no clockwise tectonic rotations took place in the Jurassic throughout the Gastre district (Zaffarana and Somoza, 2012). Other arguments coming from regional geological observations can be used against the postulated Jurassic strike-slip motion through the Gastre fault system between northern and southern Patagonia. For example, Figure 1A shows the outcrops of the basal megasequence in the Cañadón Asfalto Basin (Figari et al., 1994, 2015), composed of the 189–183 Ma (⁴⁰Ar/³⁹Ar on amphiboles, ²⁰⁶Pb/²³⁸U on zircons) Las Leoneras and Lonco Trapial Formations (Zaffarana and Somoza, 2012; Figari et al., 2015). The basal megasequence in the Cañadón Asfalto Basin overlays the Gastre fault system without discernible offset, suggesting no Jurassic strike-slip motion through this fault system. In addition, on the Chilean coast there are almost continuous exposures of Permian to Triassic metamorphic complexes (Glodny et al., 2008; Kato and Godoy, 2015) which continue across the supposed westernmost segment of the Gastre fault system without any dextral offset (only a sinistral deflection is seen at 38°S; Fig. 1A). Furthermore, the trend delineated by outcrops and borehole rocks of the fault-bounded, late Pliensbachian–Toarcian Pampa de Agnia Basin (Lesta et al., 1980; Uliana and Biddle, 1987; Uliana and Legarreta, 1999; Vicente, 2005) traverses the San Jorge Basin without any offset. Another possible major boundary between northern and southern Patagonia has been inferred to exist buried beneath the Mesozoic infill of the San Jorge Basin by Pankhurst et al. (2006) (see Fig. 1A). However, these authors assign a Carboniferous age to this suture, so it would be no longer viable to accommodate the dextral motion of the Malvinas-Falklands islands.

Results from structural studies and AMS measurements show that the internal structure of the CPB is dominated by a pattern of NW-SE steeply dipping magmatic foliations. This pattern is clear in the deepest-formed (∼6–7 km) and oldest Gastre superunit, which is composed of NW-SE–elongated plutons. The Lipetrén superunit bears more equidimensional stocks with a more disperse magmatic fabric. This younger superunit was emplaced at a slightly shallower crustal level (∼3–4 km). The solid-state deformation fabrics of both superunits are parallel to each other. The magnetic lineations of magmatic origin can be shallowly, intermediately, or steeply plunging in both superunits, whereas the tectonic lineations are predominantly either steeply or shallowly plunging (although intermediate plunges are also observed). In both suites, the predominant shape of the magnetic ellipsoid is triaxial to oblate when the fabrics are magmatic, and strongly oblate when the fabrics are tectonic. The combination of oblate ellipsoids with subvertical magmatic and tectonic foliations suggests that the emplacement of the batholith was achieved within a transpressive tectonic regime. The kinematics of this strike-slip zone would be sinistral according to the oblique (sinistral) shear sense detected in the Uribe and Rechene mylonites. The magmatic fabrics of the Lipetrén superunit are interpreted as less controlled by this regional deformation, and would be more related to chamber dynamics.

Paleomagnetic studies (work in progress) and field observations further suggest that the CPB has been constructed by the amalgamation of dike-like magma bodies. The shape of these dikes could have been locally molded against NW-SE regional structures that compose the structural grain of the North Patagonian massif, which in turn seems to be inherited from the pervasive late Paleozoic deformational fabrics. Our findings (together with regional geology data) agree with those of previous studies (von Gosen and Loske, 2004; Zaffarana and Somoza, 2012) in rejecting the hypothesis of the Gastre fault system as a major Jurassic dextral fault zone traversing the Gastre locality and crosscutting Patagonia. The application of the Gastre fault system should be abandoned in models attempting to reconstruct the western Gondwana margin. The application of the Gastre fault system should be abandoned in models attempting to reconstruct the western Gondwana margin in Jurassic times.

Basil Tikoff, Werner von Gosen, and Terry Pavlis are thanked for having greatly improved earlier versions of this manuscript. Pablo Diego González and Raúl Giacosa are thanked for having helped with the interpretation of the data. Noelia Auciello and Dolores Alvarez are thanked for accompanying us in the field. Sabrina Crosta is thanked for her help with the petrographic study of opaque minerals. Flavio Sives is thanked for his help with the hysteresis cycles, IRM (isothermal remanent magnetization), and backfield curves. This contribution has been made in the frame of CONICET PIP 112-200901-00766.

We dedicate this article to Dr. Rubén Somoza. C. Zaffarana was honored with his supervision of her doctoral thesis and the mentoring of this work. He was a wonderful teacher and an amazing person. We know that, after so much work, Rubén enjoys seeing this article published from Heaven. God bless your soul.