Sie sind auf Seite 1von 21

Tectonophysics 319 (2000) 129149 www.elsevier.

com/ locate/tecto

Contrasting nature of deformation along an intra-arc shear zone, the Liquin eOfqui fault zone, southern Chilean Andes
Cembrano a, *, Elizabeth Schermer b, Alain Lavenu c, Alejandro Sanhueza a Jose
a, Universidad de Chile, Casilla 13518 Correo 21, Santiago, Chile a Departamento de Geolog b Geology Department, Western Washington University, Bellingham, Washington, DC 98225, USA c Institut Franc veloppement en Coope ration (ORSTOM), Casilla 53390 Correo Central, ais de Recherche Scientique pour le De Santiago 1, Chile y Departamento de Geolog a, Universidad de Chile, Santiago, Chile Received 1 April 1999; accepted for publication 22 November 1999

Abstract The Liquin eOfquifault zone (LOFZ ) cuts the Patagonian batholith and the modern volcanic arc of southern Chile for ca 1000 km. The rock fabric along three transects of the LOFZ combined with new and published geochronological data suggest marked dierences in the nature and timing of deformation along strike. In the Liquin e transect (39S), a 1 km wide, northeast-striking subvertical mylonitic zone shows north-plunging stretching lineations. This mylonitic zone has been juxtaposed by high-angle reverse faulting against a nearly undeformed Miocene granitoid. Metamorphic assemblages and microstructures in the mylonites indicate greenschist facies conditions and sinistral reverse displacement. Deformation pre-dates a 1002 Ma undeformed porphyritic dike (hornblende 40Ar39Ar mean age). In the Reloncav transect (4142S), deformation in Cretaceous and Miocene plutons is predominantly brittle. Kinematic analysis of two fault populations yields compressional and dextral strike-slip stress regimes, interpreted as late Miocene in age. In the Hornopire n transect (4243S), a 4 km wide mylonitic zone, developed in plutons and wallrocks, shows subhorizontal stretching lineations and dextral displacement. A single fault population overprinting the mylonites supports a dextral strike-slip stress regime. Available UPb, KAr and 40Ar39Ar dates on deformed Cenozoic plutons and wallrock range from 9 to 13 Ma on hornblende and from 6 to 3 Ma on biotite. Microstructures and mineral assemblages indicate that the youngest ductile fabrics in the plutons formed at greenschist facies, similar to the biotite Ar closure temperature. Subparallel magmatic and solid-state fabrics combined with geochronology suggest that dextral displacement was continuous during emplacement and cooling of the plutons. Dextral-oblique subduction of the Nazca plate beneath western South America has driven long-term intra-arc deformation at the southern Andes plate boundary zone; ridge collision, in turn, has favored dextral displacement at the leading edge of the continent since the Pliocene 2000 Elsevier Science B.V. All rights reserved.
Keywords: Andes; Cenozoic; faults; magmatic arc; tectonics

1. Introduction Regional-scale fault zones spatially associated with ancient and present-day magmatic arcs pro* Corresponding author. Fax: +56-2-6963050. Cembrano) E-mail address: jcembran@cec.uchile.cl (Jose

vide a unique opportunity to address both tectonic and magmatic phenomena on a crustal scale (e.g. Beck, 1983; Saint Blanquat et al., 1998). A study of the nature and timing of both ductile and brittle deformation along intra-arc fault zones may provide a valuable insight into deformation partitioning along and across convergent margins. Many

0040-1951/00/$ - see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0 04 0 - 1 95 1 ( 9 9 ) 00 3 2 1- 2

130

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

examples of ancient and active intra-arc faults have been reported (see Fitch, 1972; Dewey, 1980; Beck, 1983, 1991; Jarrard, 1986; Busby-Spera and Saleeby, 1990; McCarey, 1992). However, welldocumented cases are surprisingly rare (e.g. Scheuber and Reutter, 1992; Tiko and Teyssier, 1992; Brown et al., 1993; Tiko and Greene, 1997). One reason for this scarcity may be that most available evidence for intra-arc faulting comes from earthquake fault plane solutions (e.g. McCarey, 1992), while eld studies of long-lived intra-arc shear zones have been complicated by plutonic and volcanic activity which partially obliterates evidence of faulting. One way to address the complexity of the deformation is to study along-strike variations in the nature and timing of the deformation using systematic eld and microstructural studies and geochronological data. In this paper, we present structural and geochronological data from three transects across the Liquin eOfqui fault zone of the southern Chilean Andes. These data constrain part of the long-term kinematic history of the fault zone and indicate the variation in timing and style of deformation in dierent transects. The Liquin eOfqui fault zone is marked by a set of north-northeast-trending lineaments, faults and ductile shear zones that parallel the magmatic arc from near the Nazca South AmericaAntarctica triple junction northward for ca 1000 km (Fig. 1) (Herve , 1977; Herve and Thiele, 1987; Cembrano et al., 1996). Several workers have claimed that the Liquin eOfqui fault zone has played a major role in deformation and location of the magmatic activity during the Cenozoic (Parada et al., 1987; Dewey and Lamb, 1992; Pankhurst et al., 1992). Recently, speculative models for the Liquin eOfqui fault zone kinematics have been proposed on the basis of paleomagnetic data (Garc a et al., 1988; Cembrano et al., 1992; Beck et al., 1993; Rojas et al., 1994) and upon regional considerations of the Liquin eOfqui fault zone geometry and geology (Herve , 1994; Cembrano et al., 1996). However, until now, few structural and geochronological data were available to constrain the nature and timing of deformation along the fault zone. Thus, its signicance in partitioning oblique subduction of the Nazca plate has been poorly understood.

2. Tectonic setting of the Liquin eOfqui fault zone The Cenozoic geodynamic setting of the southern Chilean Andes is well constrained, showing relatively steady right-oblique subduction of the Farallon (Nazca) plate beneath South America since 48 Ma with the exception of nearly orthogonal convergence from 26 to 20 Ma (Pardo-Casas and Molnar, 1987). At present, the angle of obliquity of the NazcaSouth America plate convergence vector with respect to the orthogonal to the trench is ~26 for southern Chile (Jarrard, 1986). The slab dip is approximately 16 (Jarrard, 1986), and the age of the subducting Nazca plate decreases from ~25 Ma at 38S to virtually 0 Ma at 46S, where the Chile Ridge is currently subducting (Herron et al., 1981). The limited seismic data available suggest that the Chilean forearc between 39 and 46S is currently undergoing trench-orthogonal shortening; the volcanic arc is absorbing a small trench-parallel component (Chinn and Isacks, 1983; Cifuentes, 1989; Barrientos and Acevedo, 1992; Dewey and Lamb, 1992; Murdie, 1994). Thus, oblique subduction has been considered to be a driving mechanism for the long-term right-lateral displacement along the Liquin eOfqui fault zone (Herve , 1977; Beck, 1988; Cembrano et al., 1996). Other authors have emphasized the indenter eect arising from subduction of the Chile Ridge at the southern termination of the Liquin e Ofqui fault zone (Forsythe and Nelson, 1985; Nelson et al., 1994).

3. Regional geology of the Liquin eOfqui fault zone The Meso-Cenozoic North Patagonian batholith forms most of the southern Chilean Andes and consists of heterogeneously deformed granodioritic to tonalitic plutonic rocks and minor undeformed two-mica granite and leucogranite (Parada et al., 1987; Herve et al., 1993a). The batholith is made up of a Late JurassicEarly Cretaceous belt to the west, a central relatively narrow (30 km wide) Mio-Pliocene belt, and an eastern midCretaceous belt (Pankhurst and Herve , 1994). Available KAr, 40Ar39Ar, and RbSr radiomet-

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

131

Fig. 1. Regional scale geometry of the Liquin eOfqui fault zone and geotectonic setting of the southern Chilean Andes. Relative NazcaSouth America plate motion vector was highly dextral-oblique to the trench between 49 and 26 Ma, nearly orthogonal between 26 and 20 Ma, and slightly dextral-oblique after 20 Ma (modied from Pardo-Casas and Molnar, 1987; Cifuentes, 1989; Cembrano et al., 1996). Transect locations are shown by small boxes across the Liquin eOfqui fault zone.

ric ages of plutonic rocks in the study area are shown in Table 1 and in Figs. 24 (Cembrano, 1990; Carrasco, 1995; Schermer et al., 1995, 1996; Cembrano et al., 1996). The plutonic rocks intrude low-to medium-grade metamorphic rocks making

up an upper Paleozoic accretionary prism (Herve , 1988; Munizaga et al., 1988; Pankhurst et al., 1992). Recent studies show that at least some of the low-grade wallrocks, particularly some deformed metabasites and dike swarms, may be

132

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

Fig. 2. Regional geology and local structure of the Liquin eOfqui fault zone for the Liquin e region, simplied from Moreno and Parada (1976) and Herve (1977). Locations of available radiometric age determinations (Herve , 1977; Munizaga et al., 1988) are shown with symbols. Triangle: KAr dates on biotite, inverted triangle: KAr whole rock dates.

much younger than previously thought and spatially associated with transtension during midTertiary times (Herve et al., 1993b, 1995). The regional-scale geometry of the Liquin e Ofqui fault zone is characterized by two major north-northeast-trending segments joined by a series of en e chelon northeast -trending lineaments at a right step (Herve and Thiele, 1987; Herve , 1994). The resulting arrangement has been interpreted as a strike-slip duplex (Cembrano et al., 1996) similar to those described in Woodcock and Fisher (1986) (Fig. 1). Along the LOFZ, both the North Patagonian batholith and metamorphic wallrocks have centimeter- to meter-wide highstrain zones showing penetrative foliations and a poorly to moderately well-dened stretching linea-

tion. Superposed on the ductile fabric, an anastomosing network of fractures and faults occurs along conspicuous north-trending lineaments. Quaternary volcanoes are aligned along the fault zone, as well as in northeast- and northwesttrending groups oblique to the fault zone (Cembrano and Moreno, 1994; Lo pez-Escobar et al., 1995).

4. Structural transects Three transects were mapped across the LOFZ at Liquin and Hornopire n ( Fig. 1). e, Reloncav Geologic maps of these areas are shown in Figs. 2 , 1977, 1979) 4, respectively. Earlier workers (Herve

J. Cembrano et al. / Tectonophysics 319 (2000) 129149 Table 1 Previously published geochronological data from the three geological transects Location Liquin e Liquin e Liquin e Reloncav Reloncav Reloncav Reloncav Reloncav Reloncav Reloncav Reloncav Reloncav Hornopire n Hornopire n n Hornopire Hornopire n Hornopire n Hornopire n Hornopire n Hornopire n Hornopire n n Hornopire Hornopire n Hornopire n n Hornopire Hornopire n Hornopire n Rock type Dacytic dyke Dacytic dyke Granodiorite Tonalite Diorite Diorite Tonalite Tonalite Tonalite Tonalite Granodiorite Granodiorite Granodiorite Granodiorite Granodiorite Tonalite Tonalite Tonalite Tonalite Diorite Diorite Granodiorite Tonalite Mica schist Mica schist Tonalite Diorite Material Whole rock Biotite Biotite Biotite Biotite Biotite Biotite Biotite Biotite Biotite Biotite Biotite Biotite Biotite Hornblende Biotite Hornblende Whole rock Zircon Biotite Hornblende Biotite Biotite Biotite Muscovite Biotite Hornblende Method KAr ArAr, ArAr, ArAr, ArAr, ArAr, ArAr, ArAr, ArAr, ArAr, KAr KAr ArAr, ArAr, ArAr, ArAr, ArAr, RbSr UPb ArAr, ArAr, ArAr, ArAr, ArAr, ArAr, ArAr, ArAr, Age (Ma) 291 48 15 124 119 113 1210.3 1150.7 1140.5 1150.5 10.7 11.5 6.60.3 3.60.3 8.70.7 6.41.7 13.10.1 4.70.6 9.90.2 8.81.7 8.60.6 3.590.01 3.780.01 7.20.7 10.60.1 3.40.5 114.33 Observations Cross-cuts mylonites Cross-cuts low strain mylonites Non-deformed, faulted Non-deformed, faulted Non-deformed, faulted Non-deformed, faulted Non-deformed, faulted Non-deformed, faulted Non-deformed, faulted Non-deformed, faulted Non-deformed, faulted Non-deformed, faulted Low strain Low strain Low strain Low strain, SC fabrics Low strain, SC fabrics Low strain, SC fabrics Low strain, SC fabrics Non-deformed Non-deformed High strain mylonite High strain mylonite High strain High strain Non-deformed Non-deformed Source

133

steps total fusion total fusion total fusion total fusion total fusion total fusion total fusion total fusion

total total total total total

fusion fusion fusion fusion fusion

total fusion total fusion steps steps total fusion total fusion total fusion total fusion

et al. (1979) Herve Schermer et al. (1996) Munizaga et al. (1988) Carrasco (1995) Carrasco (1995) Carrasco (1995) Carrasco (1995) Carrasco (1995) Carrasco (1995) Carrasco (1995) Carrasco (1995) Carrasco (1995) Cembrano (1990) Cembrano (1990) Cembrano (1990) Cembrano (1990) Cembrano (1990) Pankhurst et al. (1992) Schermer et al. (1996) Cembrano (1990) Cembrano (1990) Schermer et al. (1995) Schermer et al. (1995) Cembrano (1990) Cembrano (1990) Cembrano (1990) Cembrano (1990)

identied and characterized, at the regional scale, the main lineaments comprising the fault zone, but little detailed structural or geochronological work was done to document the kinematics, timing and style of deformation. Dense forest cover prevents detailed mapping of large areas, but the good coastal and roadside exposures analyzed during this study provide the rst detailed kinematic analysis of the LOFZ. We analyzed ductile structures in mylonitic rocks on the mesoscopic and microscopic scales to determine the kinematics and approximate conditions of deformation. Kinematic indicators include SC fabrics (Berthe et al., 1979; White et al., 1980; Platt, 1984; Shimamoto, 1989), asymmetric porphyroclast systems (Passchier and Simpson, 1986), mica sh (Lister and Snoke, 1984), and domino structures (Simpson and Schmid, 1983). Quartz ribbon microstructure (types 1, 2 and 3 of Boullier and Bouchez, 1978,

Simpson, 1985) and feldspar microstructure ( White, 1975; Simpson, 1985, Tullis and Yund, 1987; Fitz-Gerald and Stunitz, 1993) were documented to provide a rough estimate of PT conditions of deformation. These diagnostic microstructures, indicative of dierent physical conditions of deformation, were combined, when possible, with co-existing metamorphic mineral assemblages to better constrain deformation temperatures. Because other important factors, such as strain rate and water content, can also aect rock behavior (e.g. Passchier and Trouw, 1996), our estimates of temperatures by combining microstructure and metamorphic mineral assemblages have been obtained by assuming that the studied mylonitic rocks deformed under roughly similar strain rate conditions. Brittle structures were analyzed using the criteria of Petit (1987). Stress inversion analysis using

134

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

135

the methods of Carey and Brunier (1974) and Carey (1979) was conducted for the areas where numerous faults contained unambiguous kinematic indicators. In this technique, homogeneous populations of faults (faults which are kinematically consistent, with striations compatible with a single stress tensor) were identied by iteration tests based on the following steps: (1) faults having similar strikes, dips, and pitches of striae were grouped together; then a comparison of these dierent groups was done qualitatively with the purpose of evaluating their kinematic compatibility (e.g. dextral and sinistral faults with similar strikes are not compatible); (2) consistent faults sets were regrouped, and a mean stress tensor was calculated. This stress tensor was then applied to the entire population of faults collected at a given site; (3) according to the distribution of the angular deviation (ts) taken less than or equal to 30 between the orientation of the theoretical striae (t) derived from the computed stress tensor and the measured striae (s), a homogeneous population of faults were identied (for a complete description of the method and underlying rationale, see Bott, 1959; Se brier et al., 1985; Lavenu et al., 1995). Stress tensors are classied as compressional, extensional or strike-slip according to the spatial orientation of the principal stress axes (e.g. Ritz and Taboada, 1993). In simple terms, a stress tensor is compressional when s and s are nearly 1 2 horizontal and s is close to vertical; it is exten3 sional when s and s are nearly horizontal and 2 3 s is vertical. A strike-slip stress tensor shows subh1 orizontal s and s axes, whereas s is close to 1 3 2 vertical. In particular, strike-slip stress tensors can be further constrained (dextral vs. sinistral ) by examining the sense of obliquity of s with respect 1 to the overall trend of the crustal-scale deformation zone being investigated. For instance, if s trends 1 northeast, a dextral strike-slip deformation regime is expected for a northsouth-trending regionalscale shear zone.

4.1. Liquin e transect Previous reconnaissance mapping (Herve et al., 1974; Moreno and Parada, 1976; Herve , 1977) shows that the Liquin e region is characterized by two main, north-trending geologic units: the Liquin e granitoid and the Liquin e gneisses ( Fig. 2). The Liquin e granitoid was originally assigned to the Jurassic (Herve , 1977), although recent KAr dating (biotite) suggests a Miocene emplacement age (Munizaga et al., 1988). The Liquin e gneisses have been dated as upper Paleozoic on the basis of a poorly constrained RbSr errorchron (Herve , 1977). The Liquin e granitoid and the Liquin e gneisses are juxtaposed by the Liquin fault (Moreno and e-Reloncav Parada, 1976; Herve , 1977). According to Herve (1977), a 1 km wide zone of cataclasites developed in the granitoid is exposed west of the fault, whereas a 1 km wide mylonitic zone occurs to the east in Paleozoic gneisses. An undeformed dacitic porphyry, dated as 29 Ma ( KAr, whole rock), cuts the mylonites. We obtained a new 40Ar39Ar step-heating date on hornblende grains separated from the cross-cutting dyke. This gives a wellconstrained mean age of 1002 Ma (Fig. 5). We interpret this age as the minimum age of deformation for the mylonites of the Liquin e transect. Another dyke, which cross-cuts a low-strain variety of the mylonites, has recently been dated at 48 Ma (40Ar39Ar, biotite) (Schermer et al., 1996). A mesoscopic-scale conjugate system of northtrending right-separation and east-northeasttrending left-separation mesoscopic faults was identied within and to the east of the mylonitic zone. This observation, together with a predominantly steeply dipping mylonitic foliation, was used to interpret the shear zone as a pre-Oligocene dextral fault zone (Herve , 1977). No stretching lineations or fault striae were reported in these early studies. Our own mapping along the Liquin e transect partly agrees with that of Herve (1977)

Fig. 3. Regional geologic map of the Liquin area, simplied from Thiele et al. (1986), Parada eOfqui fault zone for the Reloncav et al. (1987) and Carrasco (1995). Locations of available radiometric age determinations (Drake et al., 1990, 1991; Carrasco, 1995) are shown with symbols. Solid square: total fusion ArAr dates on biotite; half-lled square: total fusion ArAr dates on hornblende; triangle: KAr dates on biotite, inverted triangle: KAr whole rock dates.

136

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

137

Fig. 5. 40Ar39Ar apparent age spectra for the Liquin e transect cross-cutting dyke (hornblende bulk mineral separate).

and indicates that there is a sharp contact between a brittle, deformed granitoid and high-strain mylonitic rocks formed from the gneisses (Fig. 2). However, in contrast with previous mapping, we found that deformation in the granitoid is weak and only locally developed within a few tens of meters of the contact with the mylonites. Since there is no thermal aureole developed in the mylonites close to the granitoid, the contact between the granitoid and the mylonitized gneisses very likely corresponds to a brittle fault with a signicant east-side-up component of motion, as recognized by Herve (1977). A detailed structural study carried out within the mylonitic zone of the Liquin e area, east of the main lineament of the Liquin eOfqui fault ( Fig. 2), shows that a meter-wide dark green mylonitic strip grades eastward into distinctive meter-wide domains of dark gray porphyroclastfree mylonite. These dark-gray domains alternate with domains of porphyroclastic quartz-feldsparmica-bearing, strongly foliated and poorly lineated mylonite. The penetrative schistosity strikes approximately north and dips steeply to the west ( Figs. 2 and 6a); stretching lineations, dened by

slightly elongated quartz-feldspar and mica aggregates, plunge shallowly to moderately to the north ( Figs. 2 and 6a). Three types of deformed rocks can be distinguished across the Liquin e transect: widely distributed quartz-feldspar-mica mylonites, which include quartz-rich and quartz-poor varieties, and chloritebearing SC mylonites, which are only locally observed. The quartz-feldspar rocks show distinctive microstructures depending upon quartz abundance. Quartz-rich mylonites display a penetrative schistosity resulting from the parallel alignment of type 2 quartz ribbons and muscovite-biotite foliae, which wrap around isolated feldspar porphyroclasts showing both symmetrical and asymmetrical tails. Among the latter, s and d porphyroclast systems are present and exhibit a sinistral shearsense geometry (Fig. 7a). This sense of displacement is consistent with discrete C surfaces dened by ne-grained mica aggregates that are oblique to the schistosity. Quartz-poor mylonites show a less well-developed schistosity as quartz-ribbons are scarce, and globular aggregates of nearly equant quartz grains are more common. Porphyroclasts of feldspar are more widespread than in the quartz-rich mylonites and are locally surrounded by conjugate sets of microshear bands along which ne-grained biotite occurs (Fig. 7b). The overall fabric of these rocks is markedly symmetric in character (sensu Choukroune et al., 1987). The chlorite-bearing mylonites show isolated altered plagioclase porphyroclasts in a very ne-grained matrix of chlorite calcitesericite and residual mac minerals. Well-developed SC fabrics indicate sinistral shear sense. The most prominent brittle fabrics in the Liquin e area occur in the Liquin e granitoid and correspond to a meter-spaced set of fractures and faults striking north-northeast. Some of the faults

Fig. 4. Regional geology and local structure of the Liquin n region, simplied from Herve et al. eOfqui fault zone for the Hornopire (1979) and Cembrano (1990). Locations of available radiometric age determinations (Herve et al., 1979; Drake et al., 1990, 1991; Schermer et al., 1996) are indicated with symbols. Solid square: total fusion ArAr dates on biotite; half-lled square: total fusion ArAr dates on hornblende; diamond: RbSr whole rock isochron; triangle: KAr dates on biotite, inverted triangle: KAr whole rock dates; lled circle: UPb dates on zircon.

138

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

Fig. 6. Lower hemisphere, equal area projections showing the orientation of ductile fabrics from deformed rocks within the Liquin e n (c) (crosses); stretching and mineral lineation plots for Liquin Ofqui fault zone: poles to schistosity for Liquin e (a) and Hornopire e (b) and Hornopire n (d ) (dots).

are high-angle reverse faults; others show evidence of left or right-lateral motion, as indicated by kinematic indicators such as small-scale Riedel shears and congruous steps (e.g. Petit, 1987). However, the limited number and erratic nature of the mesoscopic faults in the Liquin e region prevent a detailed kinematic analysis. Within a few tens of meters west of the Liquin eOfqui fault zone main lineament ( Fig. 2), the Liquin e granitoid displays spaced centimeter-scale fractures that coalesce locally to dene a roughly dened northstriking foliation. 4.2. Reloncav transect Previous studies ( Thiele et al., 1986; Parada et al., 1987; Carrasco, 1995) have shown that northeast-trending granodioritic and tonalitic plutonic units are only slightly ductilely deformed along the continuation of the major lineaments trending south from Liquin e. The plutonic rocks

have yielded 40Ar39Ar ages that concentrated within Early Cretaceous and Miocene times (Drake et al., 1991; Carrasco, 1995) ( Table 1, Fig. 3). Carrasco (1995) interpreted the slightly ductilely deformed Miocene rocks as resulting from high-temperature deformation during plutonic emplacement, with only minor, late emplacement sub-solidus deformation. Our observations that most plutonic units show little evidence of high strain ductile deformation, with the exception of isolated meter-wide septa of annealed mylonitic rock within largely undeformed plutonic rocks, agree with those of previous authors ( Thiele et al., 1986; Carrasco, 1995). On both sides of Estuario Reloncav , centimeterto decameter-scale striated fault planes cut all the plutonic units ranging in age from Cretaceous to Miocene; Quaternary sedimentary and volcanic rocks do not show any evidence of faulting. Most faults are steeply dipping (>60 ), with striae ranging in pitch from subhorizontal to subvertical.

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

139

Fig. 7. Photomicrographs of shear sense indicators from XZ sections of deformed rocks occurring within the Liquin eOfqui fault zone. The scale bar is 1 mm. (a) s-type porphyroclast system of feldspar and coexisting shear bands oblique to the schistosity from mylonites of the Liquin e transect. The sense of shear is sinistral. (b) Conjugate shear bands and symmetric porphyroclasts from low strain mylonites of the Liquin e transect. The microstructure suggests shear zone-orthogonal contraction, as indicated by arrows. (c) Microfaulted plagioclase from deformed plutonic rocks of the Hornopire n transect. The domino structure and shear bands indicate dextral shear. (d) Mica-sh from wallrock pelitic schists of the Hornopire n section, indicating dextral shear sense.

Using the procedure described in Section 4, two distinct homogeneous populations of faults were recognized in the Reloncav region. The rst is represented in Fig. 8a, which includes faults with a predominant strike-slip component, and a few with predominant dip-slip component. Faults whose strike range between 337 and 040 (average 010 ) show right lateral slip-sense; faults striking between 060106 (average around 080 ) are leftlateral; and faults striking west-northwest are reverse-slip. A second diagram (Fig. 8b) shows the second homogeneous population of faults identied; most have a signicant down-dip compo-

nent, although some strike-slip faults are also included. Faults striking 010080 (mainly 6080 ) are dextral-reverse slip. Faults striking 115144 (mostly 120130 ) show sinistral reverse slip. The kinematic analysis of the mesoscopic faults using E.C.G.-Geoldynsoft software (Carey and Brunier, 1974; Carey, 1979) is consistent with a NNE-trending dextral strike-slip deformation zone for the rst population of faults: s =219, 05; 1 s =110, 75; s =310, 14 (Fig. 8a). In contrast, 2 3 a compressional ~EW-trending stress regime is suggested by the second fault population: s =275, 05; s =184, 12; s =25, 77 (Fig. 8b). 1 2 3

140

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

141

4.3. Hornopire n transect In the Hornopire n Region, the Liquin eOfqui fault zone is represented by a 45 km wide zone of heterogeneously deformed granitoids and by low- to medium-grade metamorphic wallrock ( Fig. 4). The plutonic rocks, mainly hornblendebiotite tonalite and granodiorite, have been grouped into informal units on the basis of petrography and poorly mapped eld relations (Fig. 4). The largely inferred contacts between the plutonic units and wallrocks trend roughly northwest (Cembrano, 1990). Plutonic rocks of the Hornopire n region appear to be mainly Cretaceous and Miocene in age. Granodiorite dated by UPb zircon at 135 Ma (Schermer et al., 1996) occurs along the northeast ank of the canal, and diorite with an 40Ar39Ar total fusion age of 114 Ma occurs on Isla Llancahue . The diorite is undeformed, in contrast to other plutonic rocks of the region, possibly due to its low quartz content. A Miocene (9.9 Ma) UPb age was obtained on the Cholgo pluton (Schermer et al., 1996) (Fig. 4), but the emplacement ages of other plutons are uncertain. A RbSr isochron of 4.70.6 Ma has been considered the age of emplacement of a partly gneissic tonalite located within, and to the east of, the LOFZ (Pankhurst et al., 1992) (Fig. 4). However, the fact that the isochron is largely constrained by the aplitic end-members suggests that it combines unrelated plutonic units. Further mapping and dating by the authors support this hypothesis. Abundant Miocene 40Ar39Ar and whole rock KAr dates have been obtained from the Hornopire n region (Fig. 4). The ages on plutonic rocks have been interpreted as emplacement ages related to the migration of intrusive foci toward the LOFZ from Jurassic to Neogene or, alternatively, as the result of dierential uplift and static cooling of granitoids in the vicinity of the LOFZ during the Neogene (Herve et al., 1979). A

third interpretation is that the ages are reset ages resulting from shear heating along the LOFZ (Herve et al., 1979). More recent laser total fusion 40Ar39Ar dates of biotite and hornblende (Cembrano, 1990; Drake et al., 1991) have been interpreted to reect Late Miocene to Pliocene (6 3 Ma) solid-state recrystallization of these minerals during ductile dextral shear deformation (Cembrano, 1992). Two recently obtained 40Ar39Ar step-heating dates on recrystallized biotite from mylonitic samples of the R o Mariquita and Cholgo units are in the range of 3.63.8 Ma. These were interpreted by Schermer et al. (1995) as the age of a greenschist facies ductile deformation event. The metamorphic wallrocks in the Hornopire n region consist of metapelite and metabasite belonging to a Late Paleozoic accretionary prism of regional distribution (Herve , 1988; Pankhurst et al., 1992). These rocks record a progressive west-to-east change in metamorphic grade from greenschist to amphibolite facies as the contact with intrusive rocks is approached (Cembrano, 1992). Pelitic rocks near the contact have yielded a laser total fusion 40Ar39Ar age of 7.20.7 Ma on biotite (Drake et al., 1991). The plutonic bodies display a north-northwest striking, poorly to well-developed subvertical schistosity (Figs. 4 and 6c) dened by the preferred alignment of plagioclase crystals, hornblende-biotite aggregates, and by attened interstitial quartz grains. Whereas, in the Cholgo Unit, a mesoscopic set of centimeter-spaced north-northeast-striking shear bands occurs at an angle of 1525 to the schistosity (Fig. 7c), in the Rio Mariquita Unit, there are spaced meter-wide mylonite zones. Lineations are typically absent in the granitoids, although isolated meter-wide ultra-mylonitic zones show subhorizontal stretching lineations ( Fig. 6d ). More common are subhorizontal striae of biotite akes lying in the shear bands, similar to those

Fig. 8. Lower hemisphere, equal angle projections, showing orientation and kinematic data for mesoscopic faults at Reloncav (a, b) and Hornopire n (c) computed to obtain the local stress tensor. Arrows attached to the fault traces correspond to the measured slip vectors. Thick segments on the fault traces show deviations between the measured (s) and predicted (s) slip vector on each plane. Corresponding histograms show the number of faults (n) as a function of the dierence in pitch between s and s. Convergent large arrows give azimuths of the computed maximum principal stress direction (s ). From the method of Carey and Brunier (1974). 1

142

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

described by Lin and Williams (1992) for granitoids deformed in the brittleductile transition zone. Typical microstructures exhibited in the plutonic rocks include deformation bands and core and mantle structure in quartz, which locally occurs as type 1 and type 2 ribbons (Boullier and Bouchez, 1978; Simpson, 1985). Plagioclase crystals, which in places show a strong alignment, are generally microfaulted, bent, and less commonly recrystallized to a ne-grained aggregate along edges and internal fractures. The preferred alignment of plagioclase could be of an early magmatic origin, overprinted later by sub-parallel solid-state fabrics (e.g. Paterson et al., 1989). Hornblende occurs as porphyroclasts surrounded by biotite, which in turn is extensively recrystallized to a ner-grained aggregate. A conspicuous pattern of north-northeast-striking dextral and east-northeast-striking sinistral microfaults is observed in oriented thin sections. Localized high strain zones exhibit a mylonitic fabric characterized by s-type porphyroclasts of feldspar oating in a matrix of ne-grained recrystallized quartz-muscovite-epidote. Kinematic indicators, such as asymmetrical tails and domino structures, indicate dextral displacement ( Fig. 7c). Metamorphic wallrocks exhibit a more penetrative foliation than the plutonic rocks. Foliation in the metapelites strikes north-northwest and dips moderately to steeply to both west and east (Figs. 4 and 6c). Quartz-feldspar-rich layers and micaceous foliae dene the schistosity. Subhorizontal stretching and mineral lineations are present on the schistosity. Spaced vertical shear bands that strike north-northeast and cut the schistosity occur locally. The metabasites show an anastomosing subvertical foliation, which locally wraps around mesoscopic-scale pods of virtually undeformed rock. Linear fabrics have not been recognized in the metabasic rocks. Pelitic schists exhibit a penetrative foliation dened by discontinuous alternating layers of type-3 quartz ribbons (Boullier and Bouchez, 1978) and biotite+muscovite foliae. Well-preserved mica-sh (Lister and Snoke, 1984) dene C-surfaces along interconnecting trails that are oblique to the quartz ribbon-schistosity, indicating

dextral displacement ( Fig. 7d). The geometry of asymmetrical tails on feldspar porphyroclasts also indicates dextral displacement. In the Hornopire n section, both ductile and brittle deformation are well expressed. Dextral ductile deformation is recorded within plutonic units and metamorphic wallrocks. Core and mantle microstructure in quartz aggregates, type 1 and 2 quartz ribbons (Boullier and Bouchez, 1978), and fracturing and bending of plagioclase all indicate low-to mid-greenschist facies conditions for the latest solid-state deformation of the plutonic rocks. Metamorphic wallrocks exhibit synkinematic amphibolite facies mineral assemblages (quartzcordierite-biotite-sillimanite) and microstructures (type 3 quartz ribbons of Boullier and Bouchez, 1978; Simpson, 1985) but do not show any evidence of retrogression, suggesting that deformation of the wallrocks ended earlier than in the plutonic rocks. The RbSr whole rock isochron age of 4.70.6 Ma has been interpreted as an emplacement age for adjacent coarse-grained tonalite and granodiorite (Pankhurst et al., 1992). However, the locally intense solid-state deformation of these rocks, the dierence in deformation between the tonalite (foliated ) and granodiorite hosting aplite (relatively undeformed), and single-crystal total fusion ages on hornblende and biotite from the tonalite of 136 Ma ( Fig. 4) suggest that there may be two distinct plutons, an older tonalite and younger granodiorite. Detailed studies carried out along the main lineament anking Hornopire n Canal reveal the existence of sharp faults that cut the ductile fabric of the plutonic rocks. Most faults are vertical to subvertical, have prominent subhorizontal striae and vary in length from decimeters to decameters. The majority of these faults group into a single homogeneous population, including two families: the most abundant strikes 000030; the other shows variable strikes between 060 and 080. Both sets of faults display well-developed kinematic indicators on the fault surfaces, which indicate dextral slip for the north-northeast-striking set and sinistral slip for the east-northeast-striking set ( Fig. 8c). A strike-slip deformation regime is obtained when applying the E.C.G. Geodynsoft to

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

143

these fault sets, giving stress axes of s =244, 13; 1 s =57, 77; s =154, 02 (Fig. 8c). Considering 2 3 that the LOFZ trends roughly northsouth, and that s strikes northeast, an overall dextral strike1 slip tectonics characterizes the fault zone.

5. Discussion 5.1. Nature and timing of deformation Heterogeneously deformed rock units exposed in three traverses mapped along the LOFZ document a complex history of episodic ductile deformation followed by brittle deformation. In comparing the three transects, we emphasized the similarities and dierences in our detailed observations, which are separated by kilometers of poorly exposed and poorly studied terrain or areas covered by Quaternary volcanic units. However, reconnaissance observations between transects indicate the structural style is similar for tens of kilometers north and south of each transect. Regardless of the detailed reasons for along strike continuity (or lack thereof ) in kinematics, style and timing of deformation, our new data indicate that important new distinctions can be made between the dierent parts of the LOFZ. Detailed consideration of timing of deformation must await further geochronological and thermochronological study. The 2 km wide mylonitic belt occurring in the northernmost end of the Liquin eOfqui fault zone records an important episode of ductile deformation that aected Paleozoic gneisses. Our recently obtained 40Ar39Ar date of 1002 Ma in amphibole phenocrysts from an undeformed dyke, which cross-cuts the mylonites in the center of the Liquin e shear zone, suggests that deformation in this region is pre-late Cretaceous. Kinematic indicators suggest sinistral-reverse displacement in most domains. Coexisting symmetric fabrics, such as globular quartz porphyroclasts and conjugate shear bands suggest that at least part of the deformation was coaxial within discrete domains (sensu Choukroune et al., 1987). The interpretation of local coaxial ow from symmetric fabrics, however, may not always be straightforward (e.g. Passchier and Trouw, 1996).

The coexistence of sinistral-reverse displacement and shortening across-strike in the rocks from Liquin e may be the result of a bulk sinistral transpressional deformation that was partitioned into orogen-parallel oblique-slip and across-strike shortening in the anastomosing fashion described by Bell (1981) and Bell et al. (1986). Alternatively, two discrete episodes of deformation may have occurred, but no evidence of cross-cutting structures was observed. Diagnostic microstructures such as type-2 quartz ribbons (Boullier and Bouchez, 1978; Simpson, 1985) and plagioclase crystals with internal fractures and partly recrystallized rims in the rocks of the Liquin e region suggest that deformation occurred under mid-to upper-greenschist metamorphic facies conditions (i.e. temperatures in the range of 300350C; Simpson, 1985). The metamorphic mineral assemblage present in these rocks (quartz-biotite-muscovite) is compatible with the metamorphic conditions suggested from the microstructures. Brittle overprinting on the early ductile fabric is poorly developed in Liquin e and does not show the consistent geometry and kinematics necessary to obtain a local stress tensor. Some 200 km south of Liquin e, in the Reloncav region, ductile deformation is almost absent, while brittle deformation is very conspicuous along north-trending lineaments. Faults cut isotropic plutonic rock ranging in age from mid-Cretaceous to Miocene. North-northeast-striking dextral and east-northeast-striking sinistral faults along with west-northwest-striking reverse faults are consistent with a horizontal maximum compressional stress direction trending northeast associated with an overall dextral strike-slip tectonics. Another fault population of east-northeast-striking dextralreverse and west-northwest striking sinistral reverse faults appears to have been the result of eastwest compression. The time relationship between these two brittle deformational events is not clear, as cross-cutting relations of distinctive striae are not observed on single fault surfaces. However, the absolute timing of the EW compressional event is likely late Miocene or later as similar faults are observed in both Cretaceous and Miocene rocks. However, the strike-slip stress

144

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

tensor identied in Reloncav is very similar to that obtained in Hornopire n, which may indicate that both regions record the same post ~3 Ma deformational event. The presence of parallel magmatic and subsolidus fabrics in the Cholgo unit suggests that dextral displacement started during emplacement or early cooling of the plutonic rocks and continued during cooling and unroong of the batholith. Single crystal total fusion 40Ar39Ar ages of biotite from plutonic rock ranging from 8.7 to 3.6 Ma (Drake et al., 1991; Cembrano et al., 1996; Schermer et al., 1996) may reect recrystallization of these minerals during solid-state deformation and/or cooling following deformation. The 7.2 Ma 40Ar39Ar biotite age from synkinematic biotite in wallrock schist represents cooling following wallrock deformation and appears to be coeval with intrusion and/or subsequent cooling and deformation of the plutonic rocks. The predominance of hornblende ages from 9 to 13 Ma and biotite ages from 3 to 6 Ma suggests that ductile deformation began in the late Miocene, shortly after emplacement of the Cholgo Unit at 9.9 Ma, and progressed during cooling from 500 to 300C during late the MiocenePliocene. Combined structural and chronological data suggest that the Cholgo Unit emplaced syntectonically and then cooled in a regional dextral strike-slip displacement regime. However, more eld and thermochronological data are needed to test this hypothesis. The brittle deformation in the Hornopire n region overprints the ductile fabric, forming a population of north-northeast-striking subvertical dextral faults and less common east-northeaststriking sinistral faults. As observed in the Reloncav area, the geometry and kinematics are consistent with a dextral strike-slip deformation regime. In contrast to the Reloncav transect, however, the eastwest compressional event is not represented, showing that this event may have been local in character or possibly older and not recorded in the younger plutonic rocks of Hornopire n. The brittle deformation in the Hornopire n region certainly took place after the mid-Pliocene since the faults cut rocks with 40Ar39Ar ages as young as 3.3 Ma. In fact, this xes a maximum age for the brittle deformation

of the rock as the closure temperature for biotite is around 325C, which is slightly higher than the temperature at which quartz begins to exhibit ductile behavior. The absence of ductile deformation in the Reloncav segment may reect any of the following scenarios: (1) a dierent level of exposure due to dierential uplift along the fault zone; (2) the lack of widespread pluton emplacement and concomitant thermal weakening during deformation; and (3) higher strain rates and/or lower uid supply than that of the Hornopire n region. 5.2. Geodynamic signicance of the Liquin eOfqui fault zone The Liquin eOfqui fault zone has been regarded as an intra-arc strike-slip fault zone resulting either from oblique subduction (Herve , 1977; Beck, 1988) or from the indenter eect of the Chile Ridge at the southern termination of the fault system ( Forsythe and Nelson, 1985; Nelson et al., 1994). Although both models seem plausible, previous workers had limited structural data to relate plate motions to the actual deformation of the overriding plate. Kinematic data from the Liquin eOfqui fault zone obtained in this study can be used to assess the inuence of these parameters in the tectonics of an active continental margin. 5.2.1. Oblique subduction Most authors studying margin-parallel faulting along oblique subduction zones (Fitch, 1972; Beck, 1983, 1991; Jarrard, 1986; McCarey, 1992; Beck et al., 1993) agree that there are several factors controlling the formation of a forearc sliver. These include the nature of the overriding plate (continental versus oceanic), angle of obliquity, convergence rate, and interplate coupling. The importance of these factors can be assessed for the LOFZ. The best-constrained period of ductile deformation reported here occurred during the Late Miocene to Pliocene, when dextral displacement appears to have dominated the tectonic setting of the magmatic arc. Plate reconstructions show a slightly right oblique convergence at a high rate for that time period (Pardo-Casas and

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

145

Molnar, 1987; Engebretson, written communication, 1995). Although the sense of obliquity is consistent with the documented sense of shear, the angle between the convergence vector and the orthogonal to the trench seems to have been too small (~25 ) to account by itself for the strikeslip tectonics within the magmatic arc. In fact, McCarey (1992) has shown that, for present-day oblique convergent margins, slip on the intra-arc strike-slip fault does not occur for obliquities less than 30. For the southern Chilean Andes, we propose that strong intraplate coupling resulting from the subduction of young and buoyant oceanic lithosphere north of the NazcaSouth America Antarctica triple junction plus a thermally weak continental crust appear to be the key factors for intra-arc faulting. UPb ages on plutons and 40Ar39Ar ages on mylonites suggest that ductile to near-brittle dextral displacement occurred during cooling from 9 to 3 Ma and may have been continuous into the post-3 Ma brittle dextral fault regime. Since plate motions have not changed substantially after the Pliocene (Pardo-Casas and Molnar, 1987), partitioning of the NazcaSouth America slip vector into forearc shortening and intra-arc dextral strikeslip displacement may have been continuous for several million years. In contrast to MiocenePliocene dextral displacement, pre-Late Cretaceous sinistral-reverse displacement documented in Liquin e is too poorly constrained in time to allow a meaningful correlation with plate kinematics. In any case, plate motion reconstructions are not reliable for the South American plate margin before the Early Cenozoic ( Engebretson, pers. commun. 1995). However, Early Cretaceous sinistral displacement deformation has been found along the intra-arc Atacama fault system in northern Chile (Scheuber and Andriessen, 1990; Brown et al., 1993), raising the possibility of activity along the Liquin eOfqui fault zone extending back into the Mesozoic. If that is the case, the fault zone would be a longlived structure reactivated at dierent times in response to an evolving plate kinematic framework (e.g. White et al., 1986). The short-term instantaneous deformation regime at the Liquin eOfqui

fault zone has been obtained from the dextral strike-slip fault plane solution of two crustal earthquakes at both extremes of the fault system (Chinn and Isacks, 1983; Barrientos and Acevedo, 1992). Although sparse and insucient, these data suggest that the fault zone is currently active as a dextral strike-slip fault absorbing the margin-parallel component of the convergence slip vector (Dewey and Lamb, 1992). More seismic data are needed to better constrain the present-day nature of motion of the Liquin eOfqui fault zone.

5.2.2. Indenter tectonics Forsythe and Nelson (1985) proposed that some motion on the Liquin eOfqui fault zone has been the result of the impingement of the Chile Ridge on the southern Andes continental margin during the last 14 Ma. According to their model, the Chile ridge has acted as an indenter causing an outboard sliver, the Chiloe block, to move northward and become detached from the continent along the Liquin eOfqui fault zone. More recently, Nelson et al. (1994) have assessed the eect of ridge subduction and oblique subduction at the Peru Chile trench through numerical modeling. Their model predicts the trajectory of maximum horizontal stress (s ) and the distribution of vertical Hmax strain along and across the continental margin as a result of the collision of the Chile ridge at its present-day position. According to Nelson et al. (1994), the current geometry and sense of motion of the southern termination of the Liquin eOfqui fault zone seem to be compatible with the predicted orientation of s . Modeled strain patterns also Hmax are consistent with limited available short-term seismic and geodetic data for the southern Andes forearc (Plafker and Savage, 1970; Barrientos and Ward, 1990). However, eld evidence on the nature and timing of long-term and short-term deformation along the entire length of the Liquin eOfqui fault zone is not discussed in detail by Nelson et al. (1994) to test the compatibility of their ridge subduction model with the long-term evolution of the LOFZ. Therefore, although fairly consistent with the short-term deformation, Nelson et al.s model does not address the eect of long-term ridge subduction on deformation along the

146

J. Cembrano et al. / Tectonophysics 319 (2000) 129149

Liquin eOfqui fault zone, particularly at its northern half, which is some 5001000 km away from the present-day triple junction, and may have been unaected by ridge subduction. None the less, the oblique subduction component of their model does explain recent dextral motion along the length of the LOFZ. Consequently, the long-term and steady nature of oblique subduction makes this driving mechanism more likely to explain the overall tectonics of the Liquin eOfqui fault zone rather than the indenter eect of the Chile Ridge, which we see as a second-order mechanism enhancing the PlioceneRecent dextral strike-slip motion along the southern portion of the Liquin eOfqui fault zone.

tion of the Nazca plate and partitioning of deformation in the thermally weakened crust of the continental margin. The dextral system appears to characterize the long- and short-term kinematics of the Liquin eOfqui fault zone, as evidenced by both the structural data presented herein and limited earthquake data.

Acknowledgements Fondecyt Grants 1931096 and 1950497 to J.C., A.L. and E.S. funded most of this research. Part of this paper was written while J.C. was at Dalhousie University under a Killam Doctoral fellowship. Dr Peter Reynolds and Mr Keith Taylor were in charge of the 40Ar39Ar age determination of the dike sample from Liquin e. E.S. acknowledges support from the Bureau of Faculty Research of Western Washington University. The participation of A.L. has been possible by an agreement between the Geology Department of the University of Chile and ORSTOM, France. Fondecyt Grant 1920914 and 1931096 supported A.S. Judith Oliva prepared some of the gures, and Cristina Maureira helped with the manuscript. Discussions and partial revisions of the manuscript by Professors L. Aguirre, M. Beck, R. Burmester, N. Culshaw, F. Herve , R. Jamieson, M.A. Parada and D. Prior contributed to the clarication of the ideas contained in this paper. Journal co-editor T. Engelder, and referees E. Nelson and K. Kepleis are thanked for their thorough reviews and helpful suggestions.

6. Conclusions Structural and geochronological data presented in this paper document a long and complex history of intra-arc deformation in the southern Chilean Andes. The northern half of the Liquin eOfqui fault zone exhibits dierent styles and times of deformation in three transects. The northernmost, Liquin e region experienced sinistral ductile deformation prior to 100 Ma and later faulting of uncertain kinematics. The central, Reloncav segment shows no clear evidence of ductile deformation from Cretaceous to Miocene time, but brittle fault fabrics are well developed in plutons of both mid-Cretaceous and late Miocene ages. Inversion of fault slip data for the stress tensor suggests dextral strike-slip and compressional deformation regimes. The southern, Hornopire n segment exhibits evidence for both ductile and brittle deformations. Kinematic data from both brittle and ductile structures suggest dextral displacement; published isotopic ages suggest ductile deformation during the late MiocenePliocene (133 Ma), with brittle displacement occurring after 3 Ma. The older sinistral deformation seen in the northernmost segment is of uncertain signicance but may be related to a period of margin parallel sinistral deformation in the Early Cretaceous coeval with that of the Atacama fault system. The dextral structures appear to be related to oblique subduc-

References
Barrientos, S.E., Ward, S.N., 1990. The 1960 Chile earthquake: inversion for slip distribution from surface deformation. Geophys. J. Int. 103, 589598. Barrientos, S.E., Acevedo, P., 1992. Seismological aspects of the 19881989 Lonquimay (Chile) volcanic eruption. J. Volcanol. Geotherm. Res. 53, 7387. Beck, M., 1983. On the mechanism of tectonic transport in zones of oblique subduction. Tectonophysics 93, 111. Beck, M.E., 1988. Analysis of Late JurassicRecent paleomagnetic data from active plate margins of South America. J. S. Am. Earth Sci. 1, 3952.

J. Cembrano et al. / Tectonophysics 319 (2000) 129149 Beck, M.E., 1991. Coastwise transport reconsidered: lateral displacements in oblique subduction zones and tectonic consequences. Phys. Earth Planet. Inter. 68, 18. Beck, M.E., Rojas, C., Cembrano, J., 1993. On the nature of buttressing in margin-parallel strike-fault systems. Geology 21, 755758. Bell, T.H., 1981. Foliation development the contribution, geometry and signicance of progressive, bulk, inhomogeneous shortening. Tectonophysics 75, 273296. Bell, T.H., Rubenach, M.J., Fleming, P.D., 1986. Porphyroblast nucleation growth and dissolution in regional metamorphic rocks as a function of deformation partitioning during foliation development. J. Metamorph. Geol. 4, 3767. Berthe , D., Choukroune, P., Jegouzo, P., 1979. Ortogneiss, mylonite and non-coaxial deformation of granites: the example of the South Armorican shear zone. J. Struct. Geol. 1, 3142. Bott, M.H.P., 1959. The mechanisms of oblique slip faulting. Geological Magazine 96, 109117. Boullier, A.M., Bouchez, J.-L., 1978. Le quartz en rubans dans les mylonites. Soc. ge ol. Fr. Bull. 7, 253262. Brown, M., Diaz, F., Grocott, J., 1993. Displacement history of the Atacama fault system 2500S2700S northern Chile. Geol. Soc. Am. Bull. 105, 11651174. Busby-Spera, C.J., Saleeby, J.B., 1990. Intra-arc strike-slip fault exposed at batholithic levels in the southern Sierra Nevada, California. Geology 18, 255259. Carey, E., 1979. Recherche des directions principales de contraintes associe es au jeu dune population de failles. Rev. Ge ogr. Phys. Ge ol. Dynam. 21 (1), 5766. Carey, E., Brunier, B., 1974. Analyse the orique dun mode ` le me canique e le mentaire applique a ` le tude dune population de failles. C. R. Acad. Sci. Paris D 269, 891894. Carrasco, V., 1995. Geolog a y geoqu mica del Batolito Norpatago nico y rocas volca nicas asociadas a la zona de n. Thesis, falla Liquin eOfqui (41054140L.S.), X Regio Departamento de Geolog a, Univ. de Chile, Santiago. Cembrano, J., 1990. Geolog a del Batolito Norpatago nico y de rcas de su margen occidental: las rocas metamo 4150S4210S. Thesis, Departamento de Geolog a, Universidad de Chile, Santiago. Cembrano, J., 1992. Geologic and paleomagnetic reconnaissance of the LiquineOfqui fault zone in the province of Palena, Chile (4244S ). Masters thesis, Bellingham, Western Washington University. a, Cembrano, J., Beck, M.E., Burmester, R.F., Rojas, C., Garc A., Herve , F., 1992. Paleomagnetism of Lower Cretaceous rocks from east of the Liquin eOfqui fault zone, southern Chile: evidence of small in-situ clockwise rotations. Earth Planet. Sci. Lett. 113, 539551. Cembrano, J., Moreno, H., 1994. Geometria y naturaleza contrastante del volcanismo cuaternario entre los 38 y 46S Dominios compresionales y tensionales en un regimen transcurrente? In Septimo Congreso Geologico Chileno 1, 240244. , F., Lavenu, A., 1996. The LiquineOfqui Cembrano, J., Herve

147

fault zone: a long-lived intra-arc fault system in southern Chile. Tectonophysics 259, 5566. Chinn, D.S., Isacks, B.L., 1983. Accurate source depths and focal mechanisms of shallow earthquakes in western South America and in the New Hebrides island arc. Tectonics 2, 529563. Choukroune, P., Gapais, D., Merle, O., 1987. Shear criteria and structural symmetry. J. Struct. Geol. 9, 525530. Cifuentes, I.L., 1989. The 1960 Chilean Earthquakes. J. Geophys. Res. 94, 665680. Dewey, J.F., 1980. Episodicity, sequence and style at convergent plate boundaries. Geol. Assoc. Can., Spec. Pap. 20, 553573. Dewey, J.F., Lamb, S.H., 1992. Active tectonics of the Andes. Tectonophysics 205, 7995. , F., Munizaga, F., Pankhurst, R., 1990. In: Drake, R., Herve Radioisotopic ages of the North Patagonian batholith and their relation to the Liquin eOfqui Fault Zone, Chile (4042S Lat.), ICOG-7. Geological Society of Australia Abstr., Canberra, pp. 2729. Drake, R., Herve , F., Munizaga, F., Beck, M., 1991. In: Magmatism and the Liquin eOfqui Fault Zone, Chile Sur (4046S. Lat.). Comunicaciones Vol. 42. Departamento de Geolog a Universidad de Chile, Santiago, pp. 6974. Fitch, T.J., 1972. Plate convergence, transcurrent faults and internal deformation adjacent to southeast Asia and the western Pacic. J. Geophys. Res. 77, 44324460. Fitz-Gerald, J.D., Stunitz, H., 1993. Deformation of granitoids at low metamorphic grade. I. Reactions and grain size reduction. Tectonophysics 221, 269297. Forsythe, R.D., Nelson, E., 1985. Geological manifestation of ridge collision: Evidence for the Golfo de Penas, Taitao basin, southern Chile. Tectonics 4, 477495. Garc a, A., Beck, M.E., Burmester, R.F., Herve , F., Munizaga, F., 1988. Paleomagnetic reconnaissance of the region de los Lagos southern Chile and its tectonic implications. Rev. Geol. Chile 15, 1330. Herron, E.M., Cande, S.C., Hall, B.R., 1981. An active spreading center collides with a subduction zone, a geophysical survey of the Chile margin triple junction. Geol. Soc. Am., Mem. 154, 683701. Herve , F., Moreno, H., Parada, M.A., 1974. Granitoids of the Andean Range of Valdivia Province Chile. Pacic Geol. 8, 3945. Herve , F., Araya, E., Fuenzalida, J.L., Solano, A., 1979. Edades tricas y tecto nica neo gena en el sector costero de radiome Chiloe continental, X Regio n. II Congreso Geolo gico Chileno, Actas 1, F1F8. , F., Thiele, R., 1987. Estado de conocimiento de las megHerve afallas en Chile y su signicado tecto nico. Comunicaciones Vol. 38. Departamento de Geolog a Universidad de Chile, Santiago, pp. 6791. Herve , F., 1988. Late Paleozoic subduction and accretion in southern Chile. Episodes 11, 183188. Herve , F., Pankhurst, R.J., Drake, R., Beck, M., Mpodozis, C., 1993. Granite generation and rapid unroong related to

148

J. Cembrano et al. / Tectonophysics 319 (2000) 129149 Passchier, C.W., Simpson, C., 1986. Porphyroclast systems as kinematic indicators. J. Struct. Geol. 8, 831843. Passchier, C.W., Trouw, R.A.J., 1996. Microtectonics. Springer, Berlin, 289 pp. Paterson, S.R., Vernon, R.H., Tobisch, O., 1989. A review of criteria for the identication of magmatic and tectonic foliations in granitoids. J. Struct. Geol. 11, 349363. Petit, J.P., 1987. Criteria for the sense of movement on fault surfaces in brittle rocks. J. Struct. Geol. 9, 597608. Plafker, G., Savage, J.C., 1970. Mechanism of the Chilean earthquakes of May 21 and 22, 1960. Geol. Soc. Am. Bull. 81, 10011030. Platt, J.P., 1984. Secondary cleavages in ductile shear zones. J. Struct. Geol. 6, 439442. Ritz, J.F., Taboada, A., 1993. Revolution stress ellipsoids in britlle tectonics resulting from an uncritical use of inverse methods. Bull. Soc. Ge ol. Fr. 164, 519531. Rojas, C., Beck, M.E., Burmester, R.F., Cembrano, J., Herve , F., 1994. Paleomagnetism of the Mid-Tertiary Ayacara Formation, southern Chile: counterclockwise rotation in a dextral shear zone. J. S. Am. Earth Sci. 7, 4556. Saint Blanquat, M., Tiko, B., Teyssier, C., Vigneresse, J.L., 1998. Transpressional kinematics and magmatic arcs. In: Holdsworth, R.E., Strachan, R.A., Dewey, J.F. (Eds.), Continental Transpressional and Transtensional Tectonics, Geol. Soc., London, Spec. Publ. 135. 327340. Schermer, E.R., Cembrano, J., Sanhueza, A., 1995. Kinematics and timing of intra-arc shear, southern Chile. Geol. S. Am. Abstr. Programs, A409. Schermer, E.R., Cembrano, J., Sanhueza, A., McClelland, W.C., 1996. Geometry, kinematics and timing of intra-arc shear, southern Chile. In: International Geological Congress Proceedings, Beijing, China Vol. 1, 214. Scheuber, E., Andriessen, P.A.M., 1990. The kinematic and geodynamic signicance of the Atacama Fault Zone, northern Chile. J. Struct. Geol. 12, 243257. Scheuber, E., Reutter, K.-J., 1992. Magmatic arc tectonics in the Central Andes between 21 and 25S. Tectonophysics 205, 127140. brier, M., Mercier, J.L., Me gard, F., Laubacher, G., CareySe Gailhardis, E., 1985. Quaternary normal and reverse faulting and the state of stress in the central Andes. Tectonics 4, 739780. Shimamoto, T., 1989. The origin of SC mylonites and a new fault zone model. J. Struct. Geol. 11, 5164. Simpson, C., Schmid, S.H., 1983. An evaluation of criteria to deduce the sense of movement in sheared rocks. Bull. Geol. Soc. Am. 94, 12811288. Simpson, C., 1985. Deformation of granitic rocks across the brittleductile transition. J. Struct. Geol. 5, 503511. Thiele, R., Herve , F., Parada, M.A., Godoy, E., 1986. In: La megafalla Liquin eOfqui en ordo Reloncavi (41301/2S), Chile, Comunicaciones, Departamento de Geolog a Universidad de Chile Vol. 37, 3144. Tiko, B., Teyssier, C., 1992. Crustal-scale, en e chelon P-Shear tensional bridges: A possible solution to the batholithic room problem. Geology 20, 927930.

strike-slip faulting Aysen Chile. Earth Planet. Sci. Lett. 120, 375386. Herve , F., 1994. The southern Andes between 39 and 44S latitude: the geological signature of a transpressive tectonic regime related to a magmatic arc. In: Reutter, K.-J., Scheuber, E., Wigger, P.J. ( Eds.), Tectonics of the Southern Central Andes. pp. 243248. Herve , F., Pankhurst, R.J., Drake, R., Beck, M., 1995. Pillow metabasalts in a mid-Tertiary extensional basin adjacent to the Liquin eOfqui fault zone: the Isla Magdalena area Ayse n Chile. J. S. Am. Earth Sci. 8 (1), 3346. Herve , M., 1977. Geolog a del a rea al este de Liquin e, Provincia de Valdivia, X Regio n. Thesis, Depto. de Geolog a, Univ. de Chile. Jarrard, R.D., 1986. Relations among subduction parameters. Rev. Geophys. 24, 217284. Lavenu, A., Noblet, C., Winter, T., 1995. Neogene ongoing tectonics in the Southern Ecuadorian Andes: Analysis of the evolution of the stress eld. J. Struct. Geol. 17, 4758. Lin, S., Williams, P.F., 1992. The origin of ridge-in-groove slickenside striae and associated steps in an SC mylonite. J. Struct. Geol. 14, 315321. Lister, G.S., Snoke, A.W., 1984. SC mylonites. J. Struct. Geol. 6, 617638. Lo pez-Escobar, L., Cembrano, J., Moreno, H., 1995. Geochemistry and tectonics of the Chilean Southern Andes Quaternary volcanism (3746S). Rev. Geol. Chile 22, 219234. McCarey, R., 1992. Oblique plate convergence slip vectors and forearc deformation. J. Geophys. Res. 97, 89058915. Moreno, H., Parada, M.A., 1976. Esquema geolo gico de la Cordillera de los Andes entre los paralelos 39 y 4130S. Actas I Congreso geolo gico chileno (Santiago), A213A226. Munizaga, F., Herve , F., Drake, R., Pankhurst, R., Brook, M., Snelling, N., 1988. Geochronology of the lake region of south central Chile (3942S): Preliminary results. J. S. Am. Earth Sci. 1, 309316. Murdie, R.E., 1994. Seismicity and Neotectonics associated with the subduction of an active ocean ridge transform system Southern Chile. Ph.D. thesis, University of Liverpool. Nelson, E., Forsythe, R., Arit, I., 1994. Ridge collision tectonics in terrane development. J. S. Am. Earth Sci. 7, 271278. Pankhurst, R., Herve , F., Rojas, L., Cembrano, J., 1992. Magmatism and tectonics in continental Chiloe , Chile (42 and 4230S ). Tectonophysics 205, 283294. Pankhurst, R.J., Herve , F., 1994. Granitoid age distribution and emplacement control in the North Patagonian bathilith gico Chileno in Aysen (4447S ). In: 7 Congreso Geolo Vol. II, 14091413. Parada, M.A., Godoy, E., Herve , F., Thiele, R., 1987. Miocene calc-alkaline plutonism in the Chilean southern Andes. Revista Brasileira de Geociencias 17, 450455. Pardo-Casas, F., Molnar, P., 1987. Relative motion of the Nazca (Farallon) and South American plates since Late Cretaceous times. Tectonics 6, 233248.

J. Cembrano et al. / Tectonophysics 319 (2000) 129149 Tiko, B., Greene, D., 1997. Stretching lineation in transpressional shear zones: an example from the Sierra Nevada Batholith, California. J. Struct. Geol. 2, 2939. Tullis, J., Yund, R.A., 1987. Transition from cataclastic ow to dislocation cree of feldspar: Mechanisms and microstructures. Geology 15, 606609. White, S.H., 1975. Tectonic deformation and recrystallization of oligoclase. Contrib. Mineral. Petrol. 50, 287340.

149

White, S.H., Burrows, S.E., Carreras, J., Shaw, N.D., Humpreys, F.J., 1980. On mylonites in ductile shear zones. J. Struct. Geol. 2, 175187. White, S.H., Bretan, P.G., Rutter, E.H., 1986. Fault-zone reactivation: kinematics and mechanisms. Philos. Trans., R. Soc. London A 317, 8197. Woodcock, N.H., Fisher, M., 1986. Strike-slip duplexes. J. Struct. Geol. 8, 725735.