Combining shallow and deep geophysical information: the Yehuin–Chepelmut Fault Zone in the Magallanes fold and thrust belt (Tierra del Fuego, Argentina)

Abstract

The outer Fuegian fold and thrust belt is the deformed sector of the Austral-Magallanes Basin, which underwent a contractional phase since the Late Cretaceous and a strike-slip phase during the Neogene. The area hosts two Neogene basins (Yehuin and Chepelmut) which have been mainly developed by the influence of wrench deformation, and were subsequently modified by glacial activity of the Ewan and Fuego glacier lobes of the Fagnano palaeoglacier. Geophysical surveys using Electrical Resistivity Tomography, onland shallow seismic surveys, and magnetometry, have been combined with outcrop analysis and deep seismic lines to recognize both shallow and intermediate deep structures related to the lacustrine basins. Four sets of faults with different deformation history were involved in the basin genesis. Comparison with the Deseado Fault Zone indicates similarities of the structural features between the two zones. The stress field interpreted as responsible for the origin of the studied basins is analogous to the one in the Deseado valley, located in the western Tierra del Fuego. Therefore, a transcurrent fault zone, here named Yehuin–Chepelmut Fault Zone, is proposed as part of the diffuse boundary between South America and Scotia plates.

Introduction

Corazón de la Isla is a provincial reserve located in the southernmost Andes, in Isla Grande de Tierra del Fuego (TdF; Fig. 1), where two main lakes—Yehuin and Chepelmut—are located. This area underwent several tectonic phases since the Late Jurassic, and was also affected by repeated glacial events until the Holocene (see Menichetti et al. 2008; Coronato et al. 2009, and references therein). The area lies about 12 km to the north of Lago Fagnano, the largest freshwater lake of TdF, which is interpreted as a pull-apart basin developed within the main deformation zone of the Magallanes-Fagnano Fault System (MFFS), which marks the South America–Scotia transform boundary (Lodolo et al. 2002, 2003; Tassone et al. 2005, 2011; Esteban et al. 2014, among others). This boundary is a complex system of fault arrays (Fig. 1) along which the wrench deformation is partitioned in different segments.

Fig. 1

Structural provinces in Isla Grande de TdF. The Magallanes foreland, external and internal FTB are shown. The black dashed boxes bound the studied areas

The main geological structures in the studied zone are oriented WNW–ESE, and the Yehuin and Chepelmut lakes have their major axis along this trend. Both lakes are settled in a hilly area in the Fuegian steppe in a low-altitude zone comprised between 50 and 900 m above sea level in the northeastern side of Sierra de Beauvoir. The geomorphology in this central area of the Isla Grande de TdF derives mostly from the glacial activity of the palaeoglacier Fagnano and its diffluent lobes, fed from the easternmost part of the Cordillera Darwin and Patagonian ice field (Coronato et al. 2009; Fig. 1).

Some works, mainly of geophysical nature, have been performed in other Fuegian lakes, i.e., Lago Fagnano in the last decades to reconstruct its main structures (Lippai et al. 2004; Zanolla et al. 2011; Waldmann et al. 2008, 2010a, b; 2011; Esteban et al. 2014), the small Lago Udaeta for a surface morphology (Onorato et al. 2017) and Lago Roca for bathymetry (Lodolo et al. 2010). The Yehuin and Chepelmut lakes have been investigated recently with the objectives of analyzing the sedimentary infill and the basin evolution in the framework of the regional geodynamics (Lozano et al. 2018a, b).

Here, we present new information derived from seismic surveys, ERT and magnetometry with the aim of analyzing the structural setting both at shallow and deep levels of the Yehuin–Chepelmut Fault Zone (ZFYC), a strike-slip tectonic lineament running in the central area of Corazón de la Isla. This lineament is structurally compared with the transcurrent Deseado Fault Zone (DFZ) in the Deseado valley (Kepeis 1994b; Palotti 2013; Palotti et al. 2013) to demonstrate similarities between kinematic indicators and fault patterns, and to investigate a possible common genesis as part of the diffuse Scotia–South America plate boundary.

Considering the complex geologic setting of the area, and given the lack of well-exposed outcrops, geophysical methods become an essential tool to achieve the objectives proposed for this work. These include the characterization of the shallow subsurface structure in the area of Corazón de la Isla by means of electrical resistivity tomography and reflection seismic lines, linked with the outcrop analysis. The shallow geophysical data are then combined with deep seismic data to propose a plausible tectonic model. This paper highlights the complex geometries of one of the segments that compose the diffuse boundary between Scotia and South America plates.

Tectonic setting and geological history

Mesozoic and Cenozoic of the Fuegian Andes

The current tectonic setting of the area of TdF is the result of long geological history that starts from the Mesozoic. The stratigraphic succession is mainly composed by silicic volcanoclastic rocks of the Late Jurassic Lemaire Formation, belonging to the extensional back-arc Rocas Verdes basin developed along the southern Patagonian and Fuegian continental margin (Dalziel et al. 1974; Suárez and Pettigrew 1976; Hanson and Wilson 1991; Calderón et al. 2007). The Lower Cretaceous is composed of the Beauvoir and Yahgan formations, and represents the upper part of the sedimentary infill of the Rocas Verdes basin (Olivero and Martinioni 2001; Menichetti et al. 2008). In the Late Cretaceous, the closure and inversion of the marginal basin was followed by an uplift of the Cordillera and the emplacement of plutonic rocks (Katz 1972; Suárez and Pettigrew 1976; Winslow 1981). Since the Albian to the Cenomanian, the basin developed as a foreland basin, while deformation propagated towards the NE, forming the Magallanes fold and thrust belt (FTB) (Winslow 1982; Biddle et al. 1986; Ramos 1989; Wilson 1991; Alvarez-Marrón et al. 1993; Kraemer 1993; Klepeis 1994a; Diraison et al. 2000).

During the Paleocene–Early Eocene, an extensional period characterized the entire island of TdF (Dalziel and Brown 1989; Galeazzi 1998; Ghiglione et al. 2008, 2010). Extensional structures were recognized near Canal del Beagle (Dalziel and Brown 1989) and in the western Malvinas Basin, offshore of TdF (Galeazzi 1998; Baristeas et al. 2013). Later, during the Late Eocene, the tectonic regime changed to a compressive period and the propagation of the FTB (Ghiglione 2016).

Since the late Oligocene–Early Miocene, the tectonics shifted to a mainly strike-slip regime with two main events in the southern margin of the South America: The formation of the Scotia Plate after the opening of the Drake Passage at 30 Ma (Lawver et al. 1985; Tassone et al. 2005, 2008; Lodolo et al. 2006; Esteban et al. 2012; Verard et al. 2012; Eagles and Jokat 2014); and the collision of the Chile ridge with South America and the oblique subduction beneath it, at 14 Ma (Cande and Leslie 1986). A left lateral motion, represented by the MFFS (Klepeis and Austin 1997; Diraison et al. 2000; Lodolo et al. 2003; Ghiglione and Ramos 2005; Esteban et al. 2014), continued all through the Cenozoic (Cunningham 1993; Klepeis 1994b; Cunningham et al. 1995). The MFFS is an example of strike-slip displacement from a plate bounding strike-slip fault zone that is diffusely distributed within a pre-existing fold–thrust belt (Betka et al. 2016). The Paleogene conditions were transpressional; while in the Neogene, they appear to have become transtensional (Cunningham 1995; Klepeis and Austin 1997). The associated structures are mainly transtensional with extensional faults arrays and pull-apart basins along the main wrench faults (Lodolo et al. 2003; Menichetti et al. 2008).

Deseado Fault Zone

The Deseado Fault Zone is one of the major strands of the MFFS and the South America–Scotia plate boundary (Klepeis 1994b; Menichetti et al. 2008; Palotti et al. 2013). It is located 15 km north of the MFFS, in the Chilean territory of TdF (Fig. 1). This zone is characterized by strike-slip and oblique-slip faults that are superimposed on contractional structures that affected Meso-Cenozoic rocks of the Magallanes foreland FTB.

The physiographic expression of the DFZ includes a 3.5-km-wide, linear, east-southeast trending intermountain valley which is covered by Quaternary glacio-fluvial and alluvial sediments and an elongated lake called Lago Deseado (Fig. 2). The exposures of the Deseado valley show a variety of faults which indicate the geometry and sense of motion of individual faults within the DFZ (Fig. 3). The orientation and sense of movement of the faults represent a Riedel shear geometry from a left-lateral model (Palotti 2013). In the north of the valley, two sets of sub-vertical, secondary strike-slip and oblique-slip faults exist. The first fault set is composed of two populations of E–NE (Fig. 2a) and W–NW (Fig. 2b) striking, en echelon, sinistral faults that displace thrust faults and related folds (Fig. 3b; Klepeis 1994b). The first population is synthetic and is related with the secondary R Riedel shear; the second population corresponds to the synthetic P shear. The second set of faults includes minor, dextral strike-slip and oblique-slip faults (Fig. 2c). These ones represent the antithetic R′ secondary system, which is not very clear (Palotti et al. 2013).

Fig. 2

Geologic map of the western area of Lago Deseado, in Chile (Location in Fig. 1). Figure modified from Palotti (2013) and Palotti et al. (2013). The faults are analyzed as Riedel shear structures according to a left-lateral strike-slip model. N is the number of faults; MP is the maximum percentage; R, P, R′ and T are the Riedel secondary shear structures according to the model. Within the map, a few focal mechanisms for sinistral faults are shown. The inset box shows a rose diagram of morpho-structural lineaments mapped from remote sensing with the main trends of morphological features, highlighted with different colors and labels. Below, three stereograms with the fault planes, striations and poles of measured faults in the region are shown: (A), (B) and (C) are faults represented by the Riedel shear geometry: R (faults dipping NW–SE 60° to 90°; rakes 20°) and P (faults dipping SSW 50° to 70°) correspond to two groups of sinistral faults, and R’–P’ (subvertical faults dipping predominantly SW, with minor NW and SE dipping faults) corresponds to dextral faults

Fig. 3

Outcrops of Lago Deseado. The location waypoint is shown in the lower right corner of each line drawing. a Normal fault cutting the Quaternary (Q) glacio-lacustrine deposits in the southern side of Lago Deseado. The blue box shows the attitude of the measured fault plane. b A complex shear zone which shows younger strike-slip and older thrust faults affecting the Upper Cretaceous, Cerro Matrero Formation (CM); the stereogram shows the attitudes of the fault planes with the slip direction

The studied structures in the DFZ are predominantly sinistral strike-slip and normal faults that reactivate and cut the compressive structures developed during the Andean orogenesis (Palotti et al. 2013). However, the length and exact orientation of the slip vector for the fault zone are unknown (Klepeis 1994b).

The regional and geological data suggest that the activity of the MFFS and DFZ near Mount Hope (Fig. 1) area, with respect to the motion in sinistral shear zones, occurred dominantly since 30 Ma. The occurrence of pre-30 Ma strike-slip activity south of Mount Hope (Cunningham 1993) could indicate that the locus of sinistral shear within southernmost South America stepped northward since late Mesozoic to early Cenozoic time (Klepeis 1994b).

The relation between Lago Deseado and the DFZ is not fully clear at the time. Most of the deformation associated with the left-lateral strike-slip tectonics is overprinting Mesozoic rocks of the inner Magallanes fold and thrust belt. However, recent episodes of strike-slip faults activity have left evidence in Quaternary sediments (Fig. 3a), similar to the deformation observed in the glaciolacustrine sediments east of Lago Fagnano (Esteban et al. 2014). Fault scarps associated with neo-tectonics in the Deseado valley have been recognized from aerial photos (Winslow 1982), morphologic lineaments (Klepeis 1994b), and by outcrop observations (Menichetti et al. 2008). Localized extension in the DFZ is also suggested by the presence of sag ponds in the valley (Palotti 2013).

Geology of Corazón de la Isla

The study area exposes a succession of Lower Cretaceous to Eocene units along WNW–ESE-oriented belts (Fig. 4; Buatois and Camacho 1993; Malumián and Olivero 2006; Olivero and Malumián 2008; Martinioni et al. 2013). The Beauvoir Formation outcrops at the northern margin of Lago Fagnano, in Sierra de Beauvoir. This Lower Cretaceous unit represents part of the infill of the former back-arc Rocas Verdes marginal basin and is composed of slates, mudstones and sandstones of deep marine environment (Olivero and Martinioni 2001; Martinioni et al. 2013). The Upper Cretaceous formations exposed in the area are Arroyo Castorera, Río Rodriguez and Policarpo formations. These units are composed of mudstones with an upward increase of coarse sand material and represent the transition to the Magallanes–Austral foreland basin in the Late Cretaceous. They are interpreted as turbiditic deposits that progressively accumulated in the front of the rising Fuegian Andes (Martinioni et al. 2013). In the central area, the Paleocene Tres Amigos Formation crops out. It is made of conglomerates, sandstones and siltstones from fan delta deposits. Eocene Leticia and Cerro Colorado formations (both from La Despedida Group) crop out in the central and northeastern zone. They are composed of yellow mudstones and green sandstones from coastal environment (Olivero and Malumián 1999; Malumián and Olivero 2006).

Fig. 4

adapted from Buatois and Camacho (1993), Coronato et al. (2008a, b), Menichetti et al. (2008), Martinioni et al. (2013), Torres Carbonell et al. (2013) and Esteban et al. (2014). The geophysical profiles (red lines) are marked with numbers: (1) Profile GE-EN (Fig. 8a), M-EN (Fig. 8a) and S-EN (Fig. 7); (2) GE-EU (Fig. 8b), M-EU (Fig. 8b) and S-EU (Fig. 7); (3) Profile GE-WY (Fig. 8c), M-WY (Fig. 8c); (4) Yehuin Seismic Line 13 (Fig. 11); (5) Chepelmut Seismic Line 03 (Fig. 12); (6) Seismic line 18,108 (Fig. 6). References for lakes: LY Lago Yehuin, LC Lago Chepelmut, LYk Lago Yakush, LE Laguna Esperanza, LF Lago Fagnano. The stereoplot locations are indicated in red numbers, and correspond to outcrops where fault planes were measured. At the right side, the chrono-stratigraphic chart, modified from Menichetti et al. (2008). The stratigraphy is based on Olivero and Malumián (2008) and Martinioni et al. (2013). The inset box shows a rose diagram of mapped lineaments from high-resolution DEM, with the principal strain trends

Geologic map of the Corazón de la Isla zone (Location in Fig. 1). The bathymetry of the main lakes of the reserve is shown with contour levels. The stratigraphy is displayed in the right lateral table. The map was

Yehuin and Chepelmut lakes

Yehuin–Chepelmut lakes are the two larger fresh water bodies within the northern Fuegian steppe. From a structural perspective, these lacustrine basins are located in a roughly rectangular area (Fig. 4) with WNW–ESE northern and southern boundaries defined by NE verging thrusts formed during Mesozoic contractional tectonics. The eastern and western boundaries, in turn, correspond to NE striking extensional faults of early Jurassic origin (Ghiglione et al. 2013) later reactivated by the Cenozoic transtensional regime. This broadly rectangular area is mostly covered by Quaternary deposits, except for a thin string of Eocene rocks (Leticia and Cerro Colorado formations) at the north; whereas, a belt of Upper Cretaceous and Paleocene rocks bounds the southern margin of Lago Yehuin. A WNW–ESE somewhat segmented Paleocene ridge separates Chepelmut and Yehuin lakes. The thrust sheet of Cerro Shenolsh is the western expression of this ridge; its eastwards extension, the Upper Cretaceous Cerro Nahuin, is also segmented. The whole area is located in the outer Fuegian FTB.

Lago Yehuin

Lago Yehuin is an elongated basin (Fig. 4) which occupies a compartmented structural depression. Two sub-basins were recognized in the lake: a western sub-basin, 7.5 km in length with a maximum depth of 118 m and an eastern sub-basin, of 7.2-km length and a maximum depth of 80 m. Both sub-basins are limited by a set of normal faults which overprint NE-verging thrusts. The basin was generated by several normal faults arranged as a negative flower-like structure that cut across the Cretaceous thrust faults.

The deposits within the lake were grouped into three seismic units: a lower unit (Yh-LU) with wedged geometry interpreted as mass flow deposits; an intermediate unit (Yh-IU) of glacier origin distributed in the entire basin and with a maximum thickness up to 120 m; an upper unit (Yh-UU) of lacustrine depositional setting with a mean thickness of 10 m which drapes the entire basin (Lozano et al. 2018a).

Lago Chepelmut

Lago Chepelmut is a sub-rectangular basin (Fig. 4) with a maximum depth of almost 40 m and filled with sediments grouped in two units: a lower unit (Ch-LU) of glacier origin, 80-m thick and mounded geometry distributed in three mounds; an upper unit (Ch-UU) of lacustrine depositional setting with a maximum thickness of 40 m and an average of 30 m. A few normal faults were recognized in the basement that affect also the sedimentary infill. However, no major structures have been recognized in the substratum (Lozano et al. 2018b).

Glaciations and geomorphology

The whole TdF landscape was strongly modeled by glacier activity which reached its maximum extension during the Last Glacial Maximum (LGM; Caldenius 1932; Lozano et al. 2018a, b and references there in). In the central part of the TdF, the glaciers filled the deep valley shaped by the active tectonics of the MFFS; while in the northern part of the FTB, the glaciers spread in the hilly and plain territory, transporting large amount of sediments including large boulders. The Fagnano glacier was the main ice tongue which flowed eastwards from the Cordillera Darwin ice-sheet during the LGM (Coronato et al. 2009). Fuegian and Ewan glacier lobes—the two main glacier lobes of the Fagnano glacier—flowed northwards and northeastward, passing through the Yehuin and Chepelmut basins. Four moraine arcs were deposited in the Fuego river valley and three moraine arcs in the Ewan river valley during glacial retreat (Fig. 4).

After retreatment of the glacier lobes, Lago Chepelmut and Lago Yehuin, together with Lago Fagnano, were included in a moraine-dammed lake known as “Paleolago Fueguino”, which drained its waters to the Atlantic Ocean. This lake progressively lowered its water level, as recognized in the sedimentary record of Lago Chepelmut (Lozano et al. 2018b).

The latest exposed moraine related to the Fuego ice lobe is the Yehuin moraine—whose age can be correlated with the stage C or D of the LGM—dated between 25.2 and 23.1 kyr B.P. (stage B from McCulloch et al. 2005, dated with terrestrial cosmogenic nuclide ¹⁰Be; Coronato et al. 2008a). The youngest moraine arc of Lago Chepelmut (Chepelmut moraine) is also correlated with the same stages, C or D of the LGM (Coronato et al. 2008b). In both lakes, submerged moraine ridges were recognized. In Lago Yehuin, the moraine ridge is interpreted as a deposit of the glacier retreatment along the path of the Fuego glacier lobe trough the Fuego river valley. In Lago Chepelmut, the submerged glacier deposits are interpreted as deposits of the Ewan glacier lobe during its retreatment along the Ewan river valley (Lozano et al. 2018a, b).

Quaternary deposits cover a significant proportion of the area. Their origin is mostly glacial, but a significant part is composed by alluvial sediments (Buatois and Camacho 1993). Three terrace levels were recognized in the Fuego river valley (Coronato et al. 2008a); while, three–four terrace levels have been reported in the Ewan river valley (Coronato et al. 2008b).

Data acquisition—methodology

Almost 3 km of high-resolution resistivity data arranged in three 2D electrical resistivity tomography (ERT) profiles were acquired in the surrounding area of the two lakes. These profiles were correlated with magnetometric profiles and further analyzed with seismic surveys within the Lago Chepelmut and Lago Yehuin to bring a comprehensive approach to the shallow architecture of the area and the relation between the structures in the basins and their sedimentation.

The selected position of the surveyed profiles (Fig. 4) was defined with the aim of deciphering the onland continuation of the submerged structures, such as the normal faults interpreted within the lacustrine basins (Lozano et al. 2018a, b), and the evidence of a transcurrent fault zone in this sector of the island.

Electrical resistivity tomographies

ERT acquisition

The ERT method is a quick and non-invasive geophysical method widely applied to obtain 2D, high-resolution imaging of the resistivity subsurface pattern. The method has been demonstrated to be useful to determine the location of faults and to study the structural setting of geological sedimentary basins in a variety of environments (Suzuki et al. 2000; Fazzito et al. 2009; Tassone et al. 2010, 2011, among others). Therefore, ERT yields important information which allows studying of the sedimentary and depositional variables and their relation with the genesis.

The field survey consisted in the acquisition of three ERT (Fig. 4). The disposition of these profiles was chosen along a NE strike, cutting across the strike of the main structures of the area: the NE verging thrust faults and the normal faults of WNW strike already recognized (Lozano et al. 2018a). The measurements of the ERT were performed with a Syscal R1 Plus S with resistivity-meter system which operates automatically once the geometrical parameters (array type, electrode spacing, depth level) are set.

The resistivity method was applied with a Wenner–Schlumberger array using 48 electrodes connected to a 480-m-long multi-core cable that is divided in four sections and has 48 channels. The total number of quadripoles for this array is 546. An electrode spacing of 10 m has been chosen; the time of current injection was 1 s and the number of stacks per measurement was between 3 and 6. All the profiles have a length of 950 m and were obtained by the “roll-along method”, which consists of transferring the first 120 m cable section to the end of the profile and repeat the measurement.

The processing of the resistivity pseudosection was accomplished with the RES2DINV program of Geotomo Software (Loke, 2004). A data filtering phase and an inversion phase were applied to the data obtained from the measurement. The filtering phase consisted on the elimination of low-quality points, characterized by high Q values (> 3), being Q the maximum standard deviation measurement.

The inversion phase consisted in the application of a smoothness-constrained least-squares inversion method. This method attempts to find an idealized mathematical representation of the resistivity distribution of a subsurface section that best fits the measured pseudosection. The discrepancy between calculated and measured resistivity data is expressed by the root means square (RMS) as a percentage. The acquired profiles have been modeled using a robust inversion, where resistivity distribution is quite homogeneous and boundaries within the bodies are sharp.

Resistivity, lithology and structure relationships

The resistivity model of pseudosections was then interpreted considering that the distinct lithologies have different rock resistivity responses. The range of resistivities for different lithologies and sediments depends on many factors. The resistivity is strongly influenced by the degree of fracturing and weathering, the position of the water table and the concentration of dissolved salts in their pore waters (Loke 2004; Palacky 1983). For example, granitoid rocks tend to be more resistive than volcanic and volcaniclastic rocks and that metamorphic rocks. Glacial till and glaciofluvial deposits are usually resistive to poorly conductive, with values of 50–10,000 Ωm; glaciolacustrine deposits with high content of clays tend to be more conductive (5–100 Ωm; Palacky 1983).

Brittle structures such as joints and faults can be recognized as resistivity drops. The highly fractured zones enhance weathering and water percolation along them, changing the electrical properties in comparison with the unaltered surrounding rock (Berge 2014; Caputo et al. 2003; Fazzito et al. 2009; Stepancikova et al. 2011; Tassone et al. 2010). Also, faulting might be evidenced as a sharp resistivity contrast as a result from the different materials of distinct electrical responses, or even by flexures in the resistivity fields produced by offsets within a unit along the fault plane (Colella et al. 2004; Galli et al. 2013; Suzuki et al. 2000).

Magnetometric survey

The analysis of the magnetic anomalies can show changes in the magnetic properties of the rocks in a certain area, such as the magnetic susceptibility or remanence. The use of this method in combination with the ERT method may yield a more complete view of the subsurface of the studied profiles. The correlation of the values of electrical resistivity with a variation in the magnetic susceptibility can be attributed to lithologic variations or a structural arrangement that can overlay different types of lithologies. The magnetic data were acquired from all GPS-fixed point stations with a Geometrics proton magnetometer in March 2016. Three profiles of 1 km each and 10 m of spacing between points were acquired, two located to the E of Lago Chepelmut and one to the W of Lago Yehuin (Fig. 4). The data were corrected for diurnal variations with the information from a base station located in the Islas Malvinas (observatory of Port Stanley, British Geological Survey) for the acquisition period. The international geomagnetic reference field (IGRF) was removed from the surveyed magnetic data, so the residual magnetic anomalies were due solely to the geology and structure in the subsurface. The processing of the data and modeling of the subsurface geometry and susceptibility was accomplished with the software Encom Model Vision v10.0, which employs an algorithm proposed by Won and Bevis (1987) to calculate the magnetic anomaly of a n-sides polygon in a 2D space.

Seismic survey

Seismic-reflection data provide a 2D and 3D view of the subsurface of an area, and it is particularly useful in those sectors where outcrops are not so well exposed and the vegetation and Quaternary deposits widely cover the area. Within lakes, it has been demonstrated to be a powerful method to view the basement geometry, sedimentary thickness, stratigraphy and structure (Lippai et al. 2004; Lodolo et al. 2010; Waldmann et al. 2008, 2011; Esteban et al. 2014; Lozano et al. 2018a, b).

Three sets of seismic data are presented in this work, with different locations, extension and depth ranges. The first one are regional seismic lines (Fig. 1) that cover a length of almost 50 km and 5 s of record length. These consist of two 2D reflection seismic lines, 18108 and 38174-18 (provided by the Secretaria de Energía de la Nación) that were acquired during the early 80s onshore of TdF and are located along a SW–NE transect near the Fuego river valley in the central region of the island. The display and interpretation of these surveys were made with the IHS Kingdom Software 2015.

The second set are two high-resolution seismic profiles that were acquired along a N–S transect of almost 2 km to the east of Lago Chepelmut (Fig. 4). The acquisition was conducted with a Geode Geometrics DZ device and the software Seismodule Controller v9.1. The geometry adopted in both profiles was the deployment of 96 geophones with a spacing of 10 m. Seismic energy was provided by a manually operated 10-kg hammer hitting on a heavy steel plate placed on the floor. Three hits were stacked at the midpoints of every pair of consecutive receivers. The data were acquired with a 0.7-s time window and a 0.00025-s sampling interval.

The overall signal–noise ratio was low, with no records showing conspicuous reflection signals for times beyond 0.3 s. In addition, ground roll velocity was very low (< 350 m/s). This resulted in severely aliased surface wave signals, precluding their attenuation by means of multichannel filters in f-k or τ-p domain. The complete processing flow consists of refraction static corrections, spiking deconvolution, frequency band pass filtering, automatic gain control, top and bottom surgical mutes (to remove refractions and ground roll energy, respectively), normal move-out correction and stacking. The velocity analysis was performed using the Constant Velocity Stack method.

For the refraction statics, a horizontal datum was selected to coincide with the lowest surface elevation of each line, which is 20-m deeper for the northern line. Deconvolution parameters and mute schedules were picked for each field record, while line-wide frequency and gain parameters were determined by performing broad sweeps and evaluating their effect on the final stacked sections. A 30–90-Hz band pass window and a 0.2-s AGC window were selected in this manner. The velocity analysis yielded a single velocity for each line (2000 and 1650 m/s for the southern and the northern line, respectively). Finally, a time to depth conversion was applied.

In the third data set, we show two seismic lines surveyed in Lago Yehuin (seismic line 13) and Lago Chepelmut (seismic line 03). The methodology employed for the seismic survey in both lakes is described in Lozano et al. (2018a, b).

Geology of Corazón de la Isla area

Field data

The stratigraphic succession in the area is deformed by episodes of faulting which include thrusting from the Cretaceous compressive stage and later development of normal faults (new and reactivated older basement structures) by the Neogene wrenching stage (Lozano et al. 2018a). Some normal faults are exposed in the scarce outcrops at the western area, near Estancia Los Cerros (Fig. 5a). These faults generally have minor Riedel fractures near the main shear zone. In the eastern area, near Estancia Ushuaia, some thrust faults affect the Upper Cretaceous Policarpo Formation (Fig. 5b). In both cases, the brittle deformation is evident by the presence of fault gouge. Another shear zone outcrops on Cerro Cabras (Fig. 5c); Riedel fractures are filled with fault gouge which highlight the main deformation zone.

Fig. 5

Outcrops in the area of Corazón de la Isla (see locations in Fig. 4). a Outcrop near Estancia Los Cerros showing a normal fault affecting Tres Amigos Formation (TA). b Outcrop near Estancia Ushuaia. A shear zone can be recognized affecting Policarpo Formation (P) with an almost E–W strike. c A shear zone of N–S strike on Policarpo Formation (P), showing Riedel fractures and fault gouge. D. Five fracture sets affecting Policarpo Formation (P) exposures in the southern margin of Lago Yehuin. The cross-cut relations are displayed with colors. Measured structures are represented in the stereoplots of Fig. 4 labeled with the corresponding figure number

The most notable features in the outcrops are the fractures distribution showing a complex array in the mudstones and sandstones of the Upper Cretaceous and Lower Paleogene formations (Fig. 4, 1–6). The outcrop at the southern margin of Lago Yehuin (number 4 in the Fig. 4, Fig. 5d) shows 5 fractures sets (C1–C5) with different relative timing and cross-cut relations. C1 is almost NE (~ 53°) and has a major length continuity along the affected mudstone; it is slightly displaced by C2, which is W–E and shows a left-lateral kinematics (~ 95°; Fig. 5d). C3 (~ 73°) is older than C2 and C1, and is ENE. C4 (~ 156°) and C5 (~ 135°) are small joints without continuity against the other fractures sets.

Subsurface geology from geophysical survey

Regional seismic lines

This survey, located in the western part of the studied area, provides a regional framework for the deep structure (location in Figs. 1, 4). The section is the combination of two seismic profiles (lines 18108 and 38174-18, Fig. 6) and images the presence of two main zones: the northern area shows well-defined and continuous reflectors (particularly the ones located between 1.0 and 2.0 TWT s) and a southern area, with more discontinuous reflectors, evidencing a more significant degree of deformation than the northern area. The thrust front has been marked as the limit of the orogenic wedge in the foreland direction, and represents the boundary between the external FTB and the foreland area. The external FTB is characterized by a detachment level and a few thrust-related folds.

Fig. 6

Regional seismic line located at the western of Lago Yehuin (location in Figs. 1, 4). Above, seismic lines 18108 and 38174-18. Below, interpreted structures and seismostratigraphic units. The gray and red boxes show the projected locations of the seismic profiles of Figs. 11 and 7, which highlight the shallow strike-slip structures

Shallow seismic survey

Shallow penetration profiles were carried out in the eastern of Lago Chepelmut (profiles EN and EU; labeled 1 and 2 in Fig. 4). Here, the landscape is characterized by a smooth relief of mounds composed of glacial deposits. The two profiles are located in the outwash plain of the Ewan ice lobe (Lozano et al. 2018a, b).

The combined seismic profile (S-EU and S-EN; Fig. 7a) shows at 0.025 s a shallow and well-defined reflector mainly seen in the northern part. It is continuous and almost horizontal, and overlies a series of north-dipping reflectors with high reflectivity and continuity which at the 1500-m position becomes discontinuous and seems to continue at shallow levels. These reflector arrangements, in the northern side, continue at deeper levels. At approximately 0.25 s, the reflectors become chaotic.

Fig. 7

a Seismic sections of the eastern side of Yehuin–Chepelmut lakes: S-EU at the left; S-EN at the right. b Depth-converted seismic sections, with the interpreted discontinuities and faults. Location in Fig. 4. The dashed line in the a shows the extent of the b

The southern half of the profile shows a pattern of horizontal, continuous reflectors, with a high-amplitude reflector located at 0.10 s. A sub-vertical (south dipping) discontinuity is located at the position of 200 m. Locally, some small, vertical to sub-vertical discontinuities involve the reflectors located at 0.025 s and also affect the uppermost reflectors. The depth-converted seismic sections (Fig. 7b) show in the northern side, clustered sub-vertical lineaments bounding a horst-like structure; while in the southern side, the lineaments are vertical and sparse, and affect a parallel to sub-parallel layer arrangement of the reflectors.

ERT

The obtained pseudosections and the inversion models derived from the ERTs display low RMS error (< 5% in all the cases). Therefore, the models are consistent with the measurements, and a reliable interpretation may be drawn from the ERT profiles.

The GE-EN profile (Fig. 4) is 950-m long and displays a variation of resistivity values mainly vertically and, in a lesser amount, horizontally. This profile can be divided into three zones (Fig. 8a) according to their resistivity distribution: The upper zone has a variable thickness of 35–55 m and exhibits a depth gradient of resistivity from ~ 550 to ~ 200 Ωm; the second zone exhibits low resistivity values, between 20 and 80 Ωm and is located deeper than 35 m in the southern section and 55 m in the central section; the third zone involves a high resistivity area located deeper in the southern part, reaching values of more than 2000 Ωm. However, this zone may be an artifact produced by edge effects.

Fig. 8

Inverse models for the ERT transects, from top to base: GE-EN, GE-EU and GE-WY. Location in Fig. 4 with the labels 1, 2 and 3, respectively. Above each ERT inverse model, the calculated pseudo-section is shown; below, the magnetic anomaly profiles (with diurnal variation and IGRF corrections) is displayed. The gaps in each curve are lost data attributed to an erroneous measuring from the device

The GE-EU profile is 950-m long and shows a variation of resistivity values mainly vertically and in lesser grade horizontally. Two zones can be distinguished (Fig. 8b), based on the distribution of resistivity values. The upper zone has a resistivity of 200 to 400 Ωm which is laterally continuous and reaches a depth of 25 m. It overlies a lower resistivity zone with values between 30 and 80 Ωm. Three mostly vertical discontinuities are recognized within this zone, which are located at 300, 500 and 720 m, respectively: the first and second are composed of a more conductive lithology, while the third one is characterized by a more resistive lithology.

The GE-WY profile is 950-m long and is located to the W of Lago Yehuin, over the Yehuin Moraine. It exhibits resistivity variations both vertically and horizontally (Fig. 8c). The shallowest part of the section (between 0 and 15 m) has resistivity values from 100 to 130 Ωm. From 15 to 65 m, the resistivity values range from 130 to 350 Ωm. This zone displays some discontinuities with lower resistivity values located at 100, 220, 680 and 820 m. The deepest part of the section has the lower values of resistivity, reaching 45 Ωm. At 60 m depth, there is an abrupt change of the resistivity value from 25 to almost 200 Ωm.

Magnetic survey

The eastern profiles of magnetic anomalies are M-EN and M-EU. The profile M-EN (Fig. 8a) shows average values that varies between − 65 and – 58 nT. Two negative peaks of 18 nT are located between 740 and 800 m. The profile M-EU (Fig. 8b) shows average values similar to M-EN, between − 65 and – 58 nT. At 300 m, there is a positive peak of almost 30 nT, and other of 20 nT is located at 960 m. In both profiles, the slope of the anomaly increases northwards.

The western profile (M-WY; Fig. 8c) shows a more variable behavior of the magnetic anomaly. The general tendency is located between − 70 and – 80 nT, but is mainly recognized in the northern part of the profile (from 500 to almost 900 m). At least four negative peaks are observed at 100, 120, 300 and 920 m and with an amplitude of 40, 2, 55 30 nT, respectively. A few positive, minor peaks are recognized at 140, 340, 360 and 400 with an average anomaly of no more than 20 nT.

Lineament analysis

A lineament plot for the studied area has been developed by the use of high-resolution DEMs (ALOS-PALSAR; 10 m resolution; Rosenqvist et al. 2007). The regional lineaments are shown in Fig. 1 and correspond to 612 measurements of linear features such as valleys, fault scarps and sag ponds, with lengths greater than 2 km. The orientation of these lineaments is plotted as azimuths in rose diagrams of 10° intervals. The criteria to divide orientation trends are based on the presence of peaks or intervals with high amount of samples. The number of lineaments for the Deseado area is 42 (Fig. 2): the majority of the measurements (32.2%; blue) corresponds to the trend between 100° and 110°, while other minor peaks include the trends 80°–90° (16.1%; green), 40°–50° (9.7%; magenta) and 110°–120° (16.1%; red). Another minor peak of 0°–10° (6.4%) is also recognized. In the area of Corazón de la Isla (Fig. 4), the sample number of lineaments is 189: the major trend of lineaments corresponds to the 100°–110° interval (14.6%; blue). A minor peak of a lineament trend between 30° and 40° can be recognized (4.4%; magenta). Trends between 80° and 90° (12.4%; green), 90° and 100° (13.1%; blue) and 110° and 120° (10.9%; red) are represented with a high amount of samples.

Discussion

To facilitate the description and interpretation of the surveyed zone, we analyze two profiles. The first profile, EN–EU (labeled 1 and 2 in Fig. 4), which involves the ERT survey GE-EN and GE-EU, superimposed with the seismic lines S-EU and S-EN and the magnetometry profiles, M-EU and M-EN (Fig. 9). The second profile, WY (labeled 3 in Fig. 4), is located at the western side of Lago Yehuin, and correlated with the magnetometry profile M-WY (Fig. 10).

Fig. 9

Correlation among the eastern magnetometry profiles, interpreted ERT and the land seismic survey. Red lines correspond to highlighted seismic reflectors; black lines are interpreted normal faults. There is a good correlation between the few faults of the ERT and some of the seismic record. The small basin in the northern side is observed in both profiles; however, the magnetic anomaly only shows three small local anomalies which can be attributed to shallow rock boulders

Fig. 10

Correlation between the western profiles of magnetometry and the interpreted ERT. The green curve, at the center, shows the modeled magnetic anomaly interpreted as due to shallow boulders of low susceptibility (brown polygons below the green curve). These boulders can be included in the gravel fraction of the interpreted ERT. Some normal vertical faults were interpreted in the lower part of the ERT

Profile EN–EU: interpretation and correlation between ERT and seismic data

Several similarities between the main structures of ERT and shallow seismic sections are seen in the geometry and distribution of the deposits along the profiles. The seismic survey reveals the number and extent of the normal faults, while the ERT allows to recognize the different lithologies. The combination of these two methods allows to image the subsurface geology, complementing the data in depth with the use of the regional seismic section. The interpreted lithology and structure of the N–S cross-section EN–EU located to the E of Lago Chepelmut (Fig. 9) shows a resistive narrow, shallow layer as the modern sand deposits based on its low resistivity values and the surface geology in the area where the electrodes were nailed to Quaternary deposits. The position of the water table is estimated at 20-m depth (based on the low resistivity values) and probably is located within the same sandy deposits. The most resistive layer in depth was interpreted as a gravel deposit, and based on its position below the water table, was assigned to wet gravels. It is probable that the gravel consists only of small lens accumulated within the widespread sandy deposits.

The assignment of units in the seismic record uses the ERT interpretation of lithologies, geometry and thickness of the units, in addition with the characteristics of the seismic reflectors (i.e., continuity, amplitude intensity and strength). The deposit distribution shows a layered pattern with a trench-like structure in the northern side. The gravels and dry sands can be interpreted as the package of mostly chaotic reflectors; while, the wet deposits show more continuous and intense reflectors, which are located at depths deeper than 80 m.

Detailed comparison between ERT and seismic discontinuities confidently allowed to identify several faults. At the position of 180 m, the more resistivity sector led us to interpret a fault that affects the subsurface rocks. The location of this fault can be correlated with a discontinuity recognized in the seismic survey, where a sub-vertical fault plane is observed disrupting the layered reflectors (Figs. 7 and 8b; inset 1). At 740 m, the abrupt step in the resistive shallow part of the ERT section may suggest the presence of another fault which causes the displacement of the southern sector (Figs. 7 and 8a; inset 2). At the position of 1600 m, a clear change of the conductive domain from shallow zones to deeper zones is interpreted as a normal fault, also recognized in the seismic survey (Figs. 7 and 8a; inset 3).

The seismic survey reveals with more detail the presence of a few fault planes interpreted as normal faults in the northern side, and several other lineaments of a minor scale in the southern side. These faults are found at a depth of more than 50 m and might be caused by the fracturing of the consolidated sediment. Therefore, the presence of freshwater leads to an increase of the electrical conductivity in the subsurface, which is observed in the ERT surveys (Fig. 8). The geometry of the seismic reflectors displays in the northern side a basin analogous to that observed in the ERT section. In the northern side of the profile, the dip of the reflectors steepens over the fault plane located at 740 m (Figs. 7 and 8a; inset 2). This type of feature can be interpreted as a half-graben structures, filled with syn-tectonic deposits developed as growth strata. In this case, the sedimentary infill consists of dry sands from the outwash area beyond the Chepelmut moraine. The sedimentary composition of the outwash plain is widely variable depending on the glacier till and the fluvial processes involved. Typically, in an outwash plain with a braided fluvial system, the composition varies from massive gravel deposits, gravel bars, stratified and cross-bedded sands (Walker and James 1992).

The magnetic profile shows in general constant values. Only a few small anomalies are located in three sectors; however, these are restricted to small intervals of less than 20 nT, and may be due to accumulation of gravel with a higher amount of magnetic minerals.

Profile WY: interpretation of ERT and magnetometry

The profile WY located at the western margin of Lago Yehuin (Fig. 10) consists in an ERT survey and a magnetometry transect. The interpretation of the lithology in the subsurface is based on the resistivity domains versus depth (see Bran et al. 2018; Loke 2000, 2004). The shallowest part of the section is composed of a layer of sandstones, more conductive than the underlying one. At the northern side, a more conductive vertical zone is recognized, which is inferred as a vertical fault due to the increasing presence of freshwater through the fracturing of the consolidated sediment (Figs. 8c and 11; inset 4).

Fig. 11

Above, uninterpreted N–S high-resolution single-channel seismic section from the Lago Yehuin. Below, interpreted sketch with the reflector configuration and faults. The depth is given in two-way traveltime (TWT) and converted to sub-lake level depth (m) based on a P-wave velocity of 1432 m/s for water; the depth conversion for this figure was made only as a qualitative measure. The vertical exaggeration is shown in the lower left side of the interpreted section. The black lines are interpreted normal faults which cut across the lowermost seismic facies and the sedimentary infill. Lago Yehuin is characterized by three seismic units: Yh-LU, Yh-IU and Yh-UU

The composition of the recessional moraines ranges in grain size from silt to gravel and large boulders; these large variation in grain sizes and the chaotic distribution result in a heterogeneous accumulation of different types of sediment which can contain variable proportion of magnetic minerals. Therefore, the measured magnetic response can show small magnetic anomalies affecting a relatively homogeneous profile. The modeled magnetic anomaly shows the presence of shallow, small bodies of low magnetic susceptibility located at the main anomaly peaks. These small bodies can correspond to large boulders hosted in sandy deposits (see Sambuelli et al. 2011). The difference between eastern and western magnetic profiles (Fig. 8) is probably due to their location in two different types of deposits: in M-WY, the profile is located over the moraine deposits; while in the eastern of the Chepelmut (M-EN and M-EU), the profile runs through the outwash glacier area. Both environments have different composition of sediments, grain size and distribution of facies. In the eastern margin of Lago Fagnano, glaciofluvial deposits are composed by sedimentary breccia which includes boulders of siltstone, sandstone and fossil peat. In other locations, laminated clayey–sandy silts overlie gravel beds, and yellowish gray, fine sands with planar-parallel bedding are also recognized (Bujaleksy et al. 1997).

Structure of the Corazón de la Isla

The position and orientation of the shallow geophysical profiles were designed aiming to analyze one specific element of the general tectonic setting observed in the regional seismic section (Fig. 6): a transcurrent zone in the Corazón de la Isla area here named ZFYC. For an integrated view of the area, two seismic sections of Yehuin–Chepelmut lakes (Lozano et al. 2018a, b) are analyzed and compared to get a structural frame of the tectonics of the area (Figs. 11 and 12). Figure 13 shows a summary of the main characteristics and structures in the analyzed areas of Corazón de la Isla, and a comparative profile of the two lakes and their sedimentary infill. Thick packages of glacier till are hosted at the western and eastern sides of Yehuin–Chepelmut basins, respectively. These bodies are related to the two recessional moraines from Ewan and Fuego glacier lobes (Yehuin and Chepelmut moraines) in the lake margins. In Lago Yehuin western and eastern sub-basins, several normal faults were recognized affecting the lake substratum and part of the sedimentary infill. In Lago Chepelmut, only a few vertical to sub-vertical faults affect the basement. The age of the glacier deposits, in all cases, corresponds with the latest stages of the LGM, while the lacustrine deposits are younger. In the case of Lago Chepelmut, the upper sequence of the lacustrine record was considered younger than 7.8 ky B.P. (Coronato et al. 2008b).

Fig. 12

Above, uninterpreted N–S high-resolution single-channel seismic section from the Lago Chepelmut. Depth is given in two-way traveltime (TWT) and was converted to sub-lake depth level (m) using a P-wave velocity of 1432 m/s. Dashed lines are interpreted normal faults; colored lines bound the lower and upper units

Fig. 13

Comparative chart of the main characteristics and structure of both Yehuin–Chepelmut lakes, and the neighboring areas where the eastern and western geophysical profiles are located. The location of the schematic profiles of both lakes is displayed with red lines in the upper-right corner inset

Figure 11 is a seismic line from the western sub-basin of Lago Yehuin. The lake substratum shows a trench-like morphology with stepped slopes bounding the northern and southern margins, generating a basin with a sedimentary thickness of tens of meters. A few WNW–ESE normal faults involve the substratum of the western sub-basin, and bound the basement at both margins (Lozano et al. 2018a). The same structure can be directly correlated with the ERT profile WY (located 3 km to the W of seismic line 13 and almost parallel to it; Figs. 8c and 11; inset 4). However, in the ERT section, only the lower part of the glacial till is affected by a few normal faults. This observed difference could be due to the resolution ranges of both geophysical methods.

In Lago Chepelmut (Fig. 12), the sedimentary infill seems to be affected by several faults; many of them are low-angle faults related with gravity collapse. However, some of them affect the basement and are vertical to sub-vertical structures (Lozano et al. 2018b). These structures and those of Lago Yehuin share a common orientation. These structures can be correlated with the extensional faults of the EN–EU profile, which displace the glacial till, but do not reach the superficial layer of the ground.

For a complete and integrated structural approach, the regional seismic section (Fig. 6) is a key part to define the tectonic setting and the major faults. In this seismic section, the strike-slip faults were only inferred by a discontinuity zone that bounds a narrow E–W area near Los Cerros syncline. Therefore, the shallow geophysical profiles display the ZFYC with a higher detail by the interpretation of vertical to sub-vertical normal faults that were later mapped. The Fig. 14a shows the interpreted structures, based on the recognized structures on land and in both Yehuin and Chepelmut basins (Lozano et al. 2018a, b).

Fig. 14

a Structural map of the Corazón de la Isla provincial reserve. The map shows the fault arrangement elaborated from the integration of the seismic surveys and ERT profiles. The structure consists of youngest normal faults of WNW–ESE strike that break older thrust faults. These faults are nucleated within the ZFYC. The transfer faults of NE–SW strike correspond to ancestral Jurassic transfer structures and were reactivated by the Neogene tectonics (Ghiglione et al. 2013). b Rose diagram showing the full set of lineations which includes the Lago Deseado, Corazón de la Isla zones (Figs. 2, 4) and the area between them (lineaments from Fig. 1). c Schematic model for the relative timing of the four group of faults, showing three distinct episodes; numbers indicate the type of fault represented in the map

Four main groups of faults are recognized in Corazon de la Isla area:

• The oldest one includes the Jurassic faults mapped by Ghiglione et al. (2013). It was proposed that Fuego and Ewan river valleys have a main structural control by some of these faults, in particular, the NE–SW transfer faults (Lozano et al. 2018a).

• The second group of faults includes the WNW–ESE thrust faults related to the Late Cretaceous compressive stage (Fig. 5b), which are clearly recognized in the deep seismic line by major thrust and fold-related thrusts (Fig. 6). 56 Ma ago, the location of the thrust front of the FTB was nearly at the southwestern margin of the Lago Yehuin (see Torres Carbonell and Dimieri 2013).

• The third group of faults corresponds to a strike-slip set, which is related with the ZFYC. The location of these structures was inferred by the position of the normal faults recognized in Lago Yehuin in both sub-basins (Lozano et al. 2018a) and the location in the southern part of the regional seismic line, close to the Los Cerros Syncline (Fig. 6).

• The fourth group of faults includes normal faults and they can be divided into two sub-groups (Fig. 14c).

  • One of them corresponds to the NE–SW normal faults of the Jurassic transfer faults, but with the reactivated normal slip due to the strike-slips tectonics (Fig. 14c, set 4); the inversion of these faults was facilitated by the orthogonal orientation of the fault array against the stress vector from the MFFS. Studies about the active tectonics in TdF (Mendoza et al. 2011, 2015) indicate that there are components of deformation which define two deformation styles: in the eastern side of the island, the stretching is characterized by a NW–SE extensional component and a SW–NE subordinate contraction component in the area of Yehuin–Chepelmut lakes (Mendoza et al. 2011, 2015). Therefore, the reactivated NE striking faults can be related with these measured strain axes.

  • The other sub-group is composed of normal faults of WNW–ESE strike, located between the strike-slip faults in Corazón de la Isla zone (Fig. 14c, set 4).

The regional morphologic lineation plot obtained through the mapping (Fig. 1) of the Lago Deseado and Corazón de la Isla zones, and the area between them, is shown in Fig. 14b. The trends are in agreement with the main trends in both studied areas (Figs. 2, 4) and with the Riedel shear model for a left-lateral deformation (Atmaoui et al. 2006). In both, the WNW–ESE structures (red trend) may be attributed to the valleys and mountains developed during the compressive stage of the Magallanes FTB. The second trend is related with the main displacement zone of the MFFS and is W–E (blue trend). It has a marked presence in the two areas, but mainly in Corazón de la Isla (Fig. 4). In Lago Deseado zone, the E–W linear traces of the principal faults of the DFZ were observed near the southern margin of the lake (Klepeis 1994b). A third main trend (magenta trend) of NE–SW direction is also recognized in the rose diagram of both areas. This trend could be related with the T tension fractures of the Riedel model (Atmaoui et al. 2006). The fourth one includes the lineaments of ENE–WSW (green trend) which can be assigned to the R synthetic Riedel shears. Considering that the theoretical difference between the main displacement zone with the R shears is 15°, and the thrusts of the FTB are WNW–ESE structures, the distribution for the three trends (blue, red and green trends) could overlap, as a consequence of the high amount of dispersed lineaments samples along the interval between 80° and 120°. The lineaments patterns in the two studied areas show an evident correlation between the main trends (rose diagrams in Figs. 2 and 4). Therefore, the similar morphologic expressions in both areas are interpreted as a result of the same structural processes.

This proposition is reinforced by striking similarities of outcrop scale structures. The fractures distribution in the southern margin of Lago Yehuin (Fig. 4, stereoplot 4) show 5 sets. C2, of E-W direction, matches with the strike of the strike-slip faults of the third group observed in Fig. 14. The minor left-lateral slip observed for C2 (Fig. 5d) is compatible with the left-lateral interpreted kinematics for the third group of faults. C1 can be related with the T tensile fractures of the Riedel model, in which the joints are developed parallel with the major principal stress direction (σ1) and at 45° of the main displacement zone (difference between C1 and C2 is 42°). The small fractures of C3, following the model, can be related with the R shear fractures (difference between C3 and C2 is 22°). The older sets, C4 and C5, do not match with the Riedel model of left-lateral slip. However, these two sets could have been developed during the older compressive stage. In the Lago Deseado valley, the outcrops of Cerro Matrero display a series of thrust faults affected by a shear zone of almost WNW-ESE strike (Fig. 3a, b). A similar fault scenario is found in Corazon de la Isla area where young left lateral faults (Figs. 4, 6d) and shear zones (Fig. 5b) overprint compressive structures in the Upper Cretaceous formations. It is important to remark that the DFZ is carved up in Mesozoic outcrops from the inner FTB, while the ZFYC is developed along Mesozoic-Cenozoic rocks from the external FTB. The difference between the structural settings and lithological conditions in both areas could have controlled the strain in two different ways. In the DFZ, the strike-slip faults are distributed along a straight line, while in the Yehuin–Chepelmut lakes, the deformation of the ZFYC is partitioned into the series of faults previous descripted, a possible consequence of a more diffuse strain.

The variety of geophysical methods used for studying the subsurface of the surrounding area of the lakes, and their combination, result in a wide and complete picture of the lithology and shallow structure. These methods support evidence for two distinct types of processes: the deep part, associated with the deformation of strike-slip tectonics; the shallow undeformed section of Quaternary age, with the main influence of the glacier activity in the zone. The lack of outcrops near the Yehuin–Chepelmut lakes in Corazón de la Isla and the widespread thick glacial deposits covering the entire area makes difficult to recognize exposed structures as in the Lago Deseado area.

The timing of the wrench deformation in Corazón de la Isla is difficult to establish. In the regional framework, the movement across the Magallanes fault zone suggest that the displacement has been distributed diffusely across fault zones on the continent since 30 Ma (Klepeis 1994b). However, the onset of major strike-slip faulting in TdF is considered to be not older than the late Miocene (Torres Carbonell et al. 2014). The presence of a till deposit in the outwash plain observed in the profile EN–EU (Fig. 9) with a syn-tectonic signature is strong evidence to support a recent activity of the fault zone in the Corazón de la Isla zone. This pulse of activity may be bracketed during the stages C or D of the LGM (Fig. 13), this is, 21.7–17.5 kyr B.P. (see McCulloch et al. 2005) or younger.

The deformation associated with the strike-slip faults in Corazón de la Isla affects the glacier deposits within the Yehuin–Chepelmut basins and the onland moraines and outwash plain beyond Chepelmut moraine. However, this deformation is covered by an undeformed layer of sediments, indicating present inactivity of the ZFYC in this area.

Earthquake records and monitoring in the island shows a great variety of seismic events in the area that comprises earthquakes of low to medium magnitude, with the majority of the events between magnitudes of 2 to 4 and the 50% located in the uppermost 10 km (see Buffoni et al. 2009; Sabbione et al. 2007, 2016). It is not discarded that the seismicity may have triggered the normal faulting and gravity collapse recognized in the sedimentary infill.

Although Deseado and Corazón de la Isla areas are separated by around 50 km, the linear traces of the main faults from the DFZ and the ZFYC are traceable from E to W along incipient valleys and other minor morphologic elements (Lozano et al. 2018a). A possible structural link between both systems can be proposed as a part of one sole system with continuity between both zones. However, there is not enough data and further evidence to support or decline the proposition. Independently, the presence of both fault zones strengthens the model in which the diffuse boundary between Scotia and South America plates with the left-lateral strike-slip tectonics distributes the deformation along a main deformation zone in Lago Fagnano and other secondary zones. In this context, it is evident that the ZFYC and its kinematics has a main role in the development of the lacustrine basins and in the glaciers paths for the ice-lobes that carved the landscape (Lozano et al. 2018b).

Conclusions

• The comparison between Lago Deseado and Corazón de la Isla areas was made to verify the structural and kinematic connection between the DFZ and the wrench deformation of the ZFYC in the area of the Yehuin–Chepelmut lakes. Given the widespread glacial till that covers the area and the lack of pre-Quaternary outcrops and the widespread forest cover, geophysical methods become the main means to recognize the structures associated with that deformation zone. The general setting was given by two regional seismic lines which provided the deep overall deformation styles in which the shallow geophysical surveys were performed. The newly acquired geophysical data (ERT, magnetometric and seismic profiles) cover a widespread area of the Corazón de la Isla. These data were compared with information from the subsurface within the lakes and their surroundings. The nature of the geophysical methods provided closer insights into the buried shallow structure, lithology, thickness and architecture of the sedimentary infill.

• The WNW-ESE extensional faults recognized in the Yehuin–Chepelmut basins were correlated with the interpreted structures from the ERT and onland seismic profiles in the western and eastern areas of the lakes. Therefore, the eastwards partial continuation of the Chepelmut basin could be defined, as well as the activity of WNW-ESE faults affecting the Yehuin moraine (Fig. 14).

• These WNW-ESE extensional faults are part of a major deformation zone composed by E-W faults that cut across the area and is interpreted as a satellite structure of the MFFS, the ZFYC. The activity of these left-lateral structures led to a reactivation of older Jurassic NE-SW transfer faults. We suggest that the formation of both lakes was mostly controlled by Neogene wrench kinematics, linked with the left-lateral transtensional strike-slip tectonics. However, the ice-carving by the glaciers during the LGM was another key element in the final shaping of the lacustrine basins.

• Although no outcrops of glacial till were observed to be directly affected by normal faults in Corazón de la Isla, the ERT and onland seismic profiles revealed vertical to sub-vertical faults which affect the Pleistocene glacier till. In Lago Chepelmut, these faults also affect the glacial sedimentary infill; in Lago Yehuin, however, the resolution of the seismic lines or the sediment characteristics did not allow to recognize them affecting the uppermost sedimentary infill of the basin.

• The lineament analysis shows consistent patterns with the Riedel shear model for a left-lateral deformation in Lago Deseado and Corazón de la Isla zones. Two main trends can be recognized and correlated with the main displacement zone of the MFFS and with the Magallanes FTB. A third, minor trend could be related with the T tension fractures of the Riedel model, and a fourth minor trend, close to the main displacement zone, is associated with the R synthetic shears.

• The analysis of fractures in the southern margin of Lago Yehuin fits well with a Riedel model for left-lateral deformation. In this model, two sets of shears are linked with main displacement zone and the R synthetic shears, while a third set are joints that matches with the T tensile fractures. The other two fractures sets are older, probably developed during the Late Cretaceous compressive stage.

• The recognition of the broad ZFYC in Corazon de la Isla area with a similar, but a more diffuse, strain field as DFZ in western TdF contributes to establish the diffuse plate boundary that comprises the MFFS. In this scenario, part of the deformation is distributed along the main deformation zone of Lago Fagnano and other minor part along the DFZ and the ZFYC.