Abstract

The structural evolution that accompanied the breakup of Pangea during Jurassic time has been constrained in Mexico only at the regional scale on the basis of global plate tectonics and geometric considerations. According to available regional-scale reconstructions, the Jurassic tectonic evolution of Mexico was characterized by: (1) anticlockwise rotation of the Yucatán block along NNW-trending dextral faults and (2) sinistral block motions along W- to WNW-trending faults, which are geometrically needed to restore southern and central Mexico to the northwest of its present position during early Mesozoic time and avoid the overlap between North and South America in the reconstruction of Pangea. Reports of W- to WNW-trending sinistral faults that were active in Mexico during Jurassic time are presently few, and the existence, extension, and age of some of these structures have been questioned by many authors.

In this work, we present the provenance analysis from a Jurassic clastic succession deposited within the Otlaltepec Basin in southern Mexico. Whole-rock sandstone petrography integrated with chemical analysis of detrital-garnet and U-Pb detrital-zircon geochronology documents that the analyzed stratigraphic record was deposited during rapid exhumation of the Totoltepec pluton along the Matanza fault, which is a W-trending sinistral normal fault that extends along the southern boundary of the Otlaltepec Basin. U-Pb zircon ages and biostratigraphic data bracket the age of the Matanza fault between 163.5 ± 1 and 167.5 ± 4 Ma. This indicates that the Matanza fault was involved in the crustal attenuation that accompanied the breakup of Pangea and that sinistral motion of continental blocks along W-trending structures was taking place in southern Mexico as predicted by global plate tectonic reconstructions.

INTRODUCTION

By Late Triassic time, a major plate reorganization produced the breakup of Pangea, which was the most recent supercontinent assembled on Earth (e.g., Pindell, 1985; Ross and Scotese, 1988; Pindell and Kennan, 2009). The fragmentation of such a continental mass was controlled by a number of lithospheric-scale faults that produced progressive attenuation of the continental lithosphere and the subsequent opening of the Atlantic Ocean, Gulf of Mexico, and other associated subsidiary basins (Figs. 1A and 1B). During breakup of Pangea, Precambrian and Paleozoic rocks that make up the backbone of present-day Mexico were located along the nascent plate boundary between North and South America (Figs. 1A and 1B; Anderson and Schmidt, 1983; Pindell, 1985; Ross and Scotese, 1988; Dickinson and Lawton, 2001). Consequently, the early Mesozoic tectonic history of Mexico was dominated by the development of major faults that produced a complex crustal configuration characterized by subsiding basins bounded by exhuming basement highs (e.g., Martini and Ortega-Gutiérrez, 2016). Unfortunately, major problems arise in reconstructing the geometry and kinematics of the Jurassic fault array in Mexico because of the complex deformational overprinting by shortening and lateral tectonic events during late Mesozoic and Cenozoic time (e.g., Alaniz-Álvarez et al., 1996; Elías-Herrera et al., 2005). As a consequence, the kinematics of Pangea breakup in the Mexican territory has been constrained only at the regional scale on the basis of global plate tectonic reconstructions and geophysical data, most of which are the proprietary information of oil companies and, therefore, not publicly available (e.g., Pindell and Dewey, 1982; Anderson and Schmidt, 1983; Pindell, 1985; Pindell and Kennan, 2009). According to these global plate tectonic reconstructions, NNW-trending dextral faults accommodated the anticlockwise rotation of the Yucatán block to its present position (Fig. 1B), whereas W- to WNW-trending sinistral faults were responsible for the eastward to southeastward motion of south-central Mexico, which needs to be restored west or northwest of its present position during early Mesozoic time in order to avoid the overlap between North and South America in the reconstruction of Pangea (Pindell, 1985). In the past decades, some of the NNW-trending dextral structures that accommodated the Jurassic deformation imposed by North America–South America divergence have been clearly identified in the field in different parts of Mexico (e.g., Alaniz-Álvarez et al., 1996; Campos-Madrigal et al., 2013). On the other hand, the role of southeastern block motions in Mexico during Pangea breakup has been underestimated because reports of W- to WNW-trending sinistral faults that were active during Jurassic time are few (Anderson and Schmidt, 1983; Martiny et al., 2012), and the existence, extension, and age of some of these structures have been questioned by many authors (e.g., the Mojave-Sonora megashear in Fig. 1B; Iriondo et al., 2005; Gray et al., 2008).

In this work, we present the provenance analysis of a Jurassic clastic succession deposited within the Otlaltepec Basin, southern Mexico (Fig. 2). We show how whole-rock sandstone petrography integrated with the chemical analysis of detrital-garnet and U-Pb detrital-zircon geochronology can be successfully used to reconstruct the Middle Jurassic activity of the W-trending Matanza fault, which is a sinistral normal structure that bounds the Otlaltepec Basin to the south (Fig. 2). Our results suggest that, at least in the studied part of southern Mexico, sinistral motion of blocks along W-trending structures was an important tectonic process that contributed to the lithospheric attenuation during Pangea breakup.

GEOLOGICAL BACKGROUND

By the beginning of Mesozoic time, the Acatlán-Cuicatlán sector of southern Mexico was composed of a mosaic of distinct metamorphic and plutonic assemblages that are grouped into the Oaxacan Complex (Fig. 2; Ortega-Gutiérrez, 1981), Acatlán Complex (Ortega-Gutiérrez, 1981; Keppie et al., 2008a), Ayú Complex (Helbig et al., 2012), and the East Mexico Arc (Dickinson and Lawton, 2001). The crustal attenuation associated with breakup of Pangea produced a number of Jurassic, extensional to transtensional basins, including the Otlaltepec Basin (Fig. 2), that were directly developed on pre-Mesozoic metamorphic assemblages (e.g., Martini and Ortega-Gutiérrez, 2016). During the Jurassic, these basins were the site of deposition of continental to marine clastic successions that were sourced by pre-Mesozoic basement assemblages and, locally, by Lower–Middle Jurassic synrift volcanic rocks that were emplaced during the lithospheric extension that accompanied the breakup of Pangea (e.g., Martini and Ortega-Gutiérrez, 2016).

Oaxacan Complex

The Oaxacan Complex is made up of early Neoproterozoic–Mesoproterozoic granulite-facies metaplutonic and metasedimentary assemblages, which are exposed between two major NNW-trending, multi-reactivated tectonic structures that are the Caltepec and Sierra de Juárez mylonitic shear zones (Fig. 2; Ortega-Gutiérrez et al., 1995; Alaniz-Álvarez et al., 1996; Elías-Herrera et al., 2005). The most representative lithologies of the Oaxacan Complex are meta-anorthosite, garnetiferous orthogneiss, and charnockite intruding metalutite and calc-silicate rocks (Ortega-Gutiérrez et al., 1995). U-Pb ages obtained by zircons from the Oaxacan Complex range between ca. 880 and ca. 1400 Ma (Fig. 3A), with only few grains displaying older ages up to 1775 Ma (Keppie et al., 2003; Solari et al., 2003; Solari et al., 2014).

Acatlán Complex

The Acatlán Complex crops out west of the Caltepec shear zone (Fig. 2), along which it was sutured with the Oaxacan Complex during a late Paleozoic transpressional collision (Elías-Herrera et al., 2005). Upper Carboniferous–Lower Permian fluvial deposits of the Matzitzi Formation are the oldest overlapping succession that covers the suture belt between the Acatlán and Oaxacan Complexes (Fig. 2; Silva, 1970; Weber et al., 1987; Centeno-García et al., 2009). The Acatlán Complex is composed of Paleozoic, greenschist- to amphibolite-facies metasedimentary and metaplutonic assemblages, as well as their equivalent metamorphosed under blueschist-facies conditions (Ortega-Gutiérrez, 1981; Keppie et al., 2008a). The metaplutonic assemblage is composed of metagranite and minor mafic dikes that yielded U-Pb zircon ages between ca. 420 and ca. 480 Ma, as well as subordinate ages up to ca. 1600 Ma (Fig. 3B; Talavera-Mendoza et al., 2005; Keppie et al., 2008b). Metasedimentary rocks are quartz-rich metasandstone and metalutite. Available geochronologic data from these metasedimentary rocks show significant differences in zircon populations, even between samples that are lithologically similar and were collected at a relative short distance to each other (Talavera-Mendoza et al., 2005; Morales-Gámez et al., 2008; Kirsch et al., 2012). For that reason, a detailed cartographic subdivision of these metasedimentary rocks based on their detrital-zircon age signatures is presently not possible. In general, all analyzed samples ubiquitously contain zircons with ages in the range of ca. 880–1400 Ma and few grains with older ages up to ca. 2600 Ma (Fig. 3C; Talavera-Mendoza et al., 2005; Morales-Gámez et al., 2008; Kirsch et al., 2012). In addition, some samples yielded variable amounts of zircons with ages in the ranges of ca. 420–700 and ca. 270–350 Ma, prevailing in some cases over the ca. 880–1400 Ma grains (Fig. 3D; Talavera-Mendoza et al., 2005; Kirsch et al., 2012).

Ayú Complex

The Ayú Complex is exposed 20 km south of the Otlaltepec Basin (Fig. 2) and is composed of amphibolite-facies metalutite, metasandstone, and minor metabasalt that partly underwent migmatization between ca. 160 and ca. 171 Ma and subsequent metamorphic retrogression under greenschist-facies (Helbig et al., 2012). The age of protholith of the Ayú Complex is controversial. Originally, these rocks were tentatively considered as a part of the Paleozoic Acatlán Complex based on lithological and structural affinities (Ortega-Gutiérrez, 1978). Subsequently, Helbig et al. (2012) proposed a Triassic–Early Jurassic age based on U-Pb geochronology of detrital zircons and introduced the name Ayú Complex in order to differentiate these lower Mesozoic rocks from the Paleozoic Acatlán Complex. However, as underlined by Helbig et al. (2012), Triassic and Lower Jurassic zircons in the Ayú Complex are very few and most of them returned discordant ages. The great majority of detrital zircons from metasedimentary rocks of the Ayú Complex yielded ages between ca. 250 and ca. 1200 Ma, with fewer grains with ages up to ca. 2500 Ma (Fig. 3E; Talavera-Mendoza et al., 2005; Helbig et al., 2012).

East Mexico Arc

The East Mexico Arc in the Acatlán-Cuicatlán sector of southern Mexico is represented by two upper Paleozoic, deformed intrusive bodies that are the Totoltepec and Cozahuico plutons (Fig. 2; Elías-Herrera et al., 2005; Kirsch et al., 2012). In order of decreasing proportion, the Totoltepec pluton is composed of tonalite, diorite, granite, granodiorite, monzogranite, and subordinate hornblende-rich gabbro and is dominated by a main zircon population with ages between ca. 280 and ca. 310 Ma (Fig. 3F; Kirsch et al., 2012). Structural and barometric data suggest that the Totoltepec pluton is a syntectonic intrusive body that was rapidly exhumed to the surface by the end of Early Permian time (Kirsch et al., 2013). The Cozahuico pluton is a sheet-like granitic to granodioritic body exposed 50 km east of the Otlaltepec Basin (Fig. 2). Several lines of evidence indicate that this pluton is a syntectonic intrusive body emplaced along the Permian Caltepec shear zone (Elías-Herrera et al., 2005). Rocks of the Cozahuico pluton are characterized by a main U-Pb zircon age population of ca. 260–280 Ma and subordinate grains with discordant ages between ca. 700 and ca. 1200 Ma (Fig. 3G; Elías-Herrera et al., 2005). Barometric data suggest that the Cozahuico pluton was rapidly exhumed to the surface by Middle Permian time (Elías-Herrera et al., 2005).

Otlaltepec Basin

The Otlaltepec Basin is exposed in the surroundings of Santo Tomás Otlaltepec (Figs. 2 and 4) and is composed of a more than 2000-m-thick clastic sedimentary succession. Along the southeastern boundary of the Otlaltepec Basin, this clastic succession is unconformably deposited on the Acatlán Complex and Totoltepec pluton (Fig. 4). To the southwest, the clastic succession is faulted against the Totoltepec pluton along the W-trending, sinistral normal Matanza fault (Fig. 4). The western boundary of the Otlaltepec Basin is represented by a NNE-trending normal fault along which the clastic succession is juxtaposed to metasedimentary rocks of the Acatlán Complex (Fig. 4). The northern and eastern boundaries of the basin are not exposed. The stratigraphy and internal architecture of the Otlaltepec Basin are poorly defined. Based on Morán-Zenteno et al. (1993) and Verde-Ramírez (2016), the clastic stratigraphic record of the Otlaltepec Basin can be subdivided into four units that are, from the base to the top: the Tianguistengo, Piedra Hueca, Otlaltepec, and Magdalena Formations. These units are separated by regional-scale angular unconformities and are composed of fluvial to littoral deposits, some of which contain Jurassic flora (Ramos-Leal, 1989; Morán-Zenteno et al., 1993; Cruz-Cruz, 2012). Unfortunately, the exact location and stratigraphic level of the studied fossil specimens have not been reported by previous authors; therefore, it is difficult to establish a precise stratigraphic age for the clastic units exposed in the Otlaltepec Basin. Lower Cretaceous, shallow-marine limestones of the Coyotepec Formation overlie the continental succession of the Otlaltepec Basin (Fig. 4; Morán-Zenteno et al., 1993). The contact between the Coyotepec Formation and underlying continental deposits is presently a major detachment fault that produced intense shearing and recrystallization of the lowermost stratigraphic part of the calcareous succession. Consequently, a precise biostratigraphic age of the initial marine transgression in the Otlaltepec Basin is difficult to establish. However, in the Ayuquila Basin, 20 km south of the Otlaltepec Basin (Fig. 2), Middle Jurassic continental deposits are transitionally overlain by the Chimeco Limestone, whose paleontologic age constrains the onset of the marine incursion in this sector of southern Mexico to the Oxfordian (Campos-Madrigal et al., 2013).

METHODS

Field Work

In order to determine the lithofacies associations that compose each stratigraphic unit and interpret the corresponding depositional environments, we measured three stratigraphic columns extending from the Piedra Hueca to the Otlaltepec Formations. Columns were measured along the Magdalena, Xiotillo, and Piedra Hueca Creeks (Fig. 4). Lithofacies used in this work are according to Miall (2006) and are presented in Table 1. Paleocurrent data were collected in cross-bedded deposits along the stratigraphic columns, with the aim of reconstructing the pattern of the fluvial drainage and improving the provenance analysis of the studied sedimentary units. Paleocurrent data were corrected for bedding dip and are presented in Supplemental File A1.

Whole-Rock Sandstone Petrography

We selected 24 and 16 thin sections of medium- and coarse-grained sandstones representative of the Piedra Hueca and Otlaltepec Formations, respectively, for whole-rock compositional analysis. Sample locations are shown in the stratigraphic columns of Figure 5. Between 296 and 489 points were counted for each sample by using the Gazzi-Dickinson method (Gazzi, 1966; Dickinson, 1970). Thin sections were stained with potassium rhodizonate for easy recognition of potassium feldspar. Grain parameters (Zuffa, 1985; Garzanti and Vezzoli, 2003) are defined in Table 2, whereas raw and recalculated point-counting data are presented in Table 3.

Mineral Chemistry

The chemical composition of 14 and 6 detrital garnets from the Piedra Hueca and Otlaltepec Formations, respectively, has been determined in order to improve our provenance analysis. Garnet grains were analyzed with an electron probe microanalyzer JEOL Superprobe JXA-8900R at Laboratorio Universitario de Petrología of Universidad Nacional Autónoma de México (UNAM), as well as by laser ablation–inductively coupled plasma mass spectrometry (LA-ICPMS) at Laboratorio de Estudios Isotópicos (LEI), Centro de Geociencias, UNAM, using matrix-matched garnets (Jarosewich et al., 1980) for major elements. Results are presented in Supplemental File B2, whereas sample locations are given in the stratigraphic columns of Figure 5. Considering that the Ayú Complex represents a potential source for the clastic succession studied in this work and that the garnet chemical composition for this complex was not reported in the previous literature, we also analyzed 39 garnet grains extracted from a garnetiferous metasandstone exposed in the vicinity of Ayú (Fig. 2).

U-Pb Geochronology

We selected six sandstone samples from the Piedra Hueca and Otlaltepec Formations for U-Pb detrital-zircon geochronology by LA-ICPMS. Zircons from the analyzed samples were selected randomly in order to avoid bias toward certain age populations. U-Pb analyses were performed at LEI, Centro de Geociencias, UNAM. Individual zircon ages were obtained with a Resolution M-50 Excimer laser, operating at a 193 nm wavelength and coupled to a Thermo ICap Qc quadrupole ICPMS. Details of the analytical methodology and analytical results are given in Supplemental File C3. Location of samples is given in the stratigraphic columns of Figure 5. According to Spencer et al. (2016), we used the 207Pb/206Pb age for zircons older than 1.5 Ga and the 206Pb/238U age for those younger. Errors are reported to the 2σ level. Representation of statistical distribution for visualization of data was made with a kernel density estimator (KDE) (e.g., Vermeesch, 2012). The maximum depositional age of the analyzed samples is determined on the basis of the youngest single detrital-zircon grain (Dickinson and Gehrels, 2009). Because of the small size (∼40–30 μm) of these youngest zircons, it was not possible to perform multiple analyses within the same textural domain as recommended by Spencer et al. (2016). In the present literature, many authors have suggested that discordant ages should be rejected because they may be representative of disturbance of the isotopic system following the zircon crystallization (e.g., Harris et al., 2004; Gehrels, 2011). As discussed in Spencer et al. (2016), typical discordance filters vary from 1% to 30%, although 10% is the commonly used and accepted filter. However, considering that the adoption of a discordance filter is arbitrary and it can dramatically affect the way in which data are interpreted (Nemchin and Cawood, 2005; Spencer et al., 2016), we decided to use all the obtained data for the construction of KDE plots. Such a decision resides also in the fact that, in the analyzed samples, zircons that returned >10% discordant ages are subordinate, representing between 5% and 1% of the total analyzed grains. Moreover, these few grains yielded ages that overlap at 2σ with <10% discordant ages. Therefore, >10% discordant ages do not introduce potentially meaningless age ranges in the statistical age distribution obtained for each sample.

FIELD OBSERVATIONS

Piedra Hueca Formation

The Piedra Hueca Formation dominantly consists of superposed, meter- to decameter-thick, fining-upward successions bounded by irregular to channelized erosion surfaces (Fig. 5). The lower part of these fining-upward successions is composed of poorly sorted, clast-supported conglomeratic deposits with trough cross bedding and imbrication (lithofacies Gt; Fig. 5). Conglomeratic deposits are less than 1-m-thick and 1- to 10-m-wide, scoop-shaped lenses that cut into each other both laterally and vertically. In order of decreasing abundance, clasts in these conglomeratic deposits are composed of phaneritic quartzofeldspathic rocks, commonly displaying a well-developed foliation defined by ribbons of quartz (Fig. 6A), sandstone and lutite, quartzite, and minor quartz-muscovite schist. Deposits of the Gt lithofacies grade into cross-bedded, conglomeratic to medium-grained sandstones (St and Sp; Fig. 6B), and, eventually, ripple cross-laminated sandstones (Sr) draped by horizontally laminated, very fine grained sandstone, siltstone, and mudstone (Fl). Locally, matrix-supported, poorly sorted, massive conglomerates (Gmm) of less than 1.5 m in thickness are interbedded with cross-bedded conglomerates. Sharp non-erosional bases are typical of these massive conglomerates, suggesting that they were formed by high-strength debris flows (e.g., Miall, 2006). Additionally, in the Piedra Hueca Creek, centimeter- to decimeter-thick, flash-flood deposits composed of tabular-shaped, laterally continuous coarse-grained sandstone with upper-plane lamination (Sh) are locally alternated with cross-bedded lithofacies (Fig. 5). Along the Piedra Hueca Creek, paleocurrent directions obtained from cross-bedded deposits are dominantly directed toward the NW quadrant (Fig. 5). On the other hand, along the Xiotillo and Magdalena Creeks, our measurements indicate a main direction of sedimentary transport to the NE, with fewer paleocurrent data directed to the W and NW (Fig. 5).

Otlaltepec Formation

The Otlaltepec Formation overlies in angular unconformity the Piedra Hueca Formation (Fig. 6C). It is composed of superposed decimeter- to meter-thick, fining-upward successions bounded by flat to channelized erosion surfaces (Fig. 5). Fining-upward successions consist of a lag conglomerate (Gt) that grades upward into trough to planar cross-bedded, conglomeratic to medium-grained sandstone deposits (St and Sp; Fig. 6D), which are locally overlain by fine-grained sandstone beds with ripple cross lamination (Sr) and drapes of horizontally laminated siltstone and mudstone (Fl; Fig. 5). Conglomeratic deposits are dominantly composed of fragments of quartz-rich to quartzofeldspathic phaneritic rocks that ubiquitously display a well-developed foliation defined by elongated quartz. Development of paleosols in the upper part of each fining-upward succession is common. Paleosols range from few decimeters to few meters in thickness and are composed of gray-green to dark-purple, partly silicified, nodular and mottled mudstone, siltstone, and sandstone that progressively grade downward into unweathered sandstone lithofacies (Figs. 6E and 6F). Paleosols contain root traces oriented perpendicular to bedding and abundant bioturbation marks. Along the Magdalena Creek, paleocurrent data obtained from cross-bedded deposits indicate a dominant direction of sedimentary transport to the W and NW, whereas along the Xiotillo Creek, paleocurrents are directed to the N and NE (Fig. 5).

SANDSTONE PETROGRAPHY

Piedra Hueca Formation

Twenty-four medium- to coarse-grained sandstone samples representative of the Piedra Hueca Formation were collected from the Piedra Hueca, Xiotillo, and Magdalena Creeks (Fig. 5). Point-counting results are plotted in QtFL (Garzanti, 2016), QmPK (Dickinson and Suczek, 1979), and LmLsLv (Garzanti et al., 2001) diagrams (Figs. 7A–7C). Analyzed sandstones vary in composition from feldspathoquartzose to quartzofeldspathic (Fig. 7A). In order of decreasing abundance, they are composed of: monocrystalline quartz (47.2%–75.9% of the total framework grains), mesoperthitic and microcline K-feldspar (19.6%–34.8%; Fig. 8A) dominating over plagioclase (2.4%–17.6%; Fig. 7B), volcanic lithic grains with felsitic, microlitic, and lathwork textures (0.6%–8.3%), low-grade greenschist-facies metapsammitic lithic grains (0%–3.5%), sedimentary lithic fragments (Figs. 7C and 8B; 0%–1.8%), heavy minerals such as apatite, zircon, rutile, and garnet (0%–1.3%), and polycrystalline quartz (0%–0.3%). Quartz and feldspar in sandstones from the Piedra Hueca Formation are largely included in polycrystalline phaneritic grains, most of which display clear evidence of solid-state deformation such as grain boundary migration, subgrain domains, undulatory or sector extinction, and shape-preferred orientation (Fig. 8C). In addition to phaneritic clasts derived from deformed rock sources, some samples contain quartzofeldspathic sandstone grains.

Otlaltepec Formation

We collected 16 medium- to coarse-grained sandstone samples representative of the Otlaltepec Formation from the Xiotillo and Magdalena Creeks (Fig. 5). Analyzed samples vary in composition from feldspathoquartzose sandstones to quartzoarenites (Fig. 7A). In order of abundance, sandstones from the Otlaltepec Formation are composed of: monocrystalline quartz (53.1%–98.8%), plagioclase (0%–43.3%) dominating over K-feldspar with mesoperthitic structures (Fig. 8D) and microcline twinning (0%–2.4%), volcanic lithic grains with felsitic texture (0%–2.8%), low-grade greenschist-facies metapsammitic lithic grains (0%–1.3%), polycrystalline quartz (0%–0.7%), and heavy minerals (0%–0.5%) such as zircon, rutile, and garnet. As with the samples from the Piedra Hueca Formation, quartz and feldspar in sandstones from the Otlaltepec Formation are in large part included in polycrystalline phaneritic grains that display evidence of solid-state deformation (Fig. 8E).

GARNET COMPOSITION

Piedra Hueca Formation

Fourteen garnet grains from sandstones of the Piedra Hueca Formation were selected for chemical determinations (Fig. 5). Analyzed grains can be subdivided in two groups. Group 1 is characterized by high almandine (65%–81%), moderate pyrope (14%–22%), and low spessartine (2%–7%) and grossular (1%–10%) contents (Fig. 9). These garnets roughly overlap in composition with garnets from granulite-facies metamorphic rocks of the Oaxacan Complex (Fig. 9; Solari et al., 2004; Schulze, 2011). Group 2 is dominated by garnets with high almandine (64%–75%), moderate grossular (12%–18%), low to moderate pyrope (6%–15%), and low spessartine (2%–6%) contents (Fig. 9). Detrital garnets of group 2 are similar in composition to garnets in high-pressure rocks of the Acatlán Complex (Fig. 9; Ramos-Arias et al., 2011; Keppie et al., 2012; Galaz et al., 2013).

Otlaltepec Formation

Six garnet grains from sandstones of the Otlaltepec Formation were selected for chemical determinations (Fig. 5). Analyzed grains display high almandine (67%–74%), moderate pyrope (17%–29%), and low grossular (3%–9%) and spessartine (1%–4%) contents (Fig. 9). Detrital-garnet grains from the Otlaltepec Formation overlap in composition with metamorphic garnets in granulite-facies rocks from the Oaxacan Complex (Fig. 9; Solari et al., 2004; Schulze, 2011).

Ayú Complex

Analyzed grains from the Ayú Complex are dominated by the almandine component (70%–76%), with moderate contents of spessartine (10%–20%), and low pyrope (6%–11%) and grossular (3%–4%; Fig. 9). Metamorphic garnets from the Ayú Complex define a compositional cluster that does not overlap with the detrital garnets from the Piedra Hueca and Otlaltepec Formations (Fig. 9).

U-Pb GEOCHRONOLOGIC RESULTS

Piedra Hueca Formation

We selected three sandstone samples from the Piedra Hueca Formation for U-Pb detrital-zircon geochronology (Fig. 5). Under cathodoluminescence (CL), all analyzed zircons display oscillatory and sector zoning that are typical of magmatic origin (Connelly, 2001; Corfu et al., 2003). We analyzed between 98 and 100 zircon grains for each sample. Only four, three, and three grains from samples PH1.1, PH1.7, and Xi1.4, respectively, yielded ages with a percentage of discordia >10 (Fig. 10 and Supplemental File C [see footnote 3]). Ninety-two, ninety-seven, and ninety-three percent of the total analyzed grains from samples PH1.1, PH1.7, and Xi1.4, respectively, yielded ages in the range of ca. 875–1400 Ma and fewer ages up to ca. 1800 Ma (Figs. 11A–11C). Only in sample PH1.1, two zircons returned middle Neoproterozoic ages of 789 ± 19 and 829 ± 26 Ma (Fig. 11A). Early Permian (ca. 281–292 Ma), Middle Permian (ca. 261–263 Ma), Early–Middle Triassic (ca. 238–252 Ma), and Early Jurassic (ca. 184–201 Ma) ages are subordinate in the analyzed samples and have been obtained by a number of three to seven grains in each sample (Figs. 11A–11C). The youngest zircon grains in samples PH1.1, PH1.7, and Xi1.4 returned ages of 192 ± 5, 184 ± 4, and 191 ± 11 Ma, respectively, indicating an Early Jurassic maximum depositional age for the Piedra Hueca Formation.

Otlaltepec Formation

Samples SDT2.5, SDT2.8, and Xi2.7 were collected from the Otlaltepec Formation (Fig. 5) and were selected for U-Pb detrital-zircon geochronology. Under CL, zircons display typical magmatic structures such as concentric oscillatory zoning, in some cases developed around xenocrystic cores. We analyzed between 98 and 100 zircon grains for each sample. Only four, four, and one grains for samples SDT2.5, SDT2.8, and Xi 2.7, respectively, yielded percentages of discordia >10 (Fig. 10 and Supplemental File C [see footnote 3]). In the three samples, more than 93% of the analyzed grains yielded ages in the ranges of ca. 260–312 and ca. 940–1400 Ma (Figs. 11D–11F). However, while sample SDT2.5 is characterized by a higher amount of ca. 260–312 Ma grains (67% of the total analyzed crystals), samples SDT2.8 and Xi2.7 are dominated by ca. 940–1400 Ma grains (79% and 88%, respectively; Figs. 11D–11F). In addition to ca. 260–312 and ca. 940–1400 Ma grains, two and five zircons in samples SDT2.8 and Xi2.7, respectively, returned ages between ca. 1430 and ca. 2450 Ma. Subordinate zircon grains (zero to seven grains in each sample) yield ages of ca. 240, ca. 330, ca. 470, and ca. 535 (Figs. 11D–11F). Only one zircon in sample SDT2.5 yielded a Jurassic age of 167 ± 4 Ma (Fig. 11D). Considering that this age is concordant and it has been returned by an inclusion-free zircon grain showing a typical magmatic zoning, we take it as the maximum depositional age for the Otlaltepec Formation.

DISCUSSION

Interpretation of the Depositional Environments

The Piedra Hueca Formation dominantly consists of fining-upward successions that are composed, from the base to the top, of lithofacies Gt, St, and Sp, eventually grading upward into lithofacies Sr and Fl. The superposition of these fining-upward successions records cyclical variations in energy of the fluvial drainage. Following Miall (2006), Gt lithofacies are interpreted as gravel bars deposited by high-energy traction currents during flood events, whereas St, Sp, and Sr lithofacies record the progressive decrease in the flow energy, culminating with the deposition of suspended load represented by lithofacies Fl. In the study area, Gt lithofacies consists of scoop-shaped conglomeratic bodies that cut into each other both laterally and vertically, reflecting a depositional setting dominated by unstable and laterally migrating channels (Miall, 2006). Mobile channels cutting gravel bars have been typically described in present-day braided fluvial systems and alluvial fans (Miall, 2006; Reading, 2009).

Differing from the Piedra Hueca Formation, the Otlaltepec Formation is dominated by rhythmically alternating conglomerate to sandstone deposits emplaced by traction currents (Gt, St, Sp, Sr, and Fl lithofacies) and paleosols developed during prolonged periods of non-deposition in a subaerial environment. Such a lithofacies assemblage is typical of modern overbank environments, which are intermittently flooded during high-water stages and subaerially exposed during low-water periods, favoring pedogenesis and the development of paleosols (Farrell, 1987; Miall, 2006; Reading, 2009).

Sandstone Provenance

Piedra Hueca Formation

Framework grains in sandstones from the Piedra Hueca Formation indicate that this unit was dominantly derived from deformed quartzofeldspathic phaneritic rocks. Deformed phaneritic rocks of quartzofeldspathic composition are exposed in the Oaxacan, Acatlán, and Ayú Complexes (Ortega-Gutiérrez, 1981; Ortega-Gutiérrez et al., 1995; Helbig et al., 2012) and largely comprise plutons of the East Mexico Arc (Elías-Herrera et al., 2005; Kirsch et al., 2012). However, the ubiquitous occurrence of mesoperthitic K-feldspar in sandstones from the Piedra Hueca Formation suggests that this unit was dominantly derived from the Oaxacan Complex. In fact, mesoperthite is a K-feldspar/plagioclase intergrowth that forms at temperatures greater than 800 °C (Barker, 1998), and, in southern Mexico, it has been documented exclusively in granulite-facies metamorphic rocks of the Oaxacan Complex (Mora and Valley, 1985; Keppie et al., 2003). Derivation of the Piedra Hueca Formation from the Oaxacan Complex is also suggested by paleocurrent data, many of which indicate directions of sedimentary transport toward the W and NW (Figs. 5 and 12A). N- to NE-directed paleocurrent directions obtained along the Magdalena and Xiotillo Creeks may represent distributary channels that branched off downstream from the NW-trending main channel or local divergence of the fluvial stream around an overbank area. Predominant derivation of the Piedra Hueca Formation from the Oaxacan Complex is also supported by the ubiquitous occurrence of zircons with ages in the range of ca. 875–1400 Ma (Figs. 11A–11C). Our chemical analyses on detrital garnets also support, in part, such a provenance scenario. Some garnets of group 1 extracted from sandstones of the Piedra Hueca Formation roughly overlap in composition with metamorphic garnets in granulite-facies orthogneisses of the Oaxacan Complex (Fig. 9). The origin of detrital garnets that do not overlap the compositional cluster of the Oaxacan Complex is difficult to interpret. This is because available chemical data for garnet from the Oaxacan Complex are presently limited to two sites located south of the area depicted in Figure 2 (Solari et al., 2004; Schulze, 2011). Therefore, a more extensive chemical characterization of metamorphic garnets from southern Mexico is required in order to assess if some of the group 1 detrital garnets do not overlap the compositional cluster of the Oaxacan Complex because of the presently limited information or because these grains were derived from alternative sources.

Minor detrital contributions from syntectonic granitoids of the East Mexico Arc are documented by zircons that returned ages in the ranges of ca. 261–263 and ca. 281–292 Ma (Figs. 11A and 11C), which coincide with the age ranges documented for the Totoltepec (Fig. 3F; Kirsch et al., 2012) and Cozahuico (Fig. 3G; Elías-Herrera et al., 2005) plutons, respectively. However, considering that these Early and Middle Permian zircons are subordinate in the analyzed sandstones (zero to four grains in each sample), we suggest that the Totoltepec and Cozahuico plutons were not an important source of detritus for the Piedra Hueca Formation.

In addition to quartzofeldspathic clasts derived from deformed phaneritic rocks, sandstones of the Piedra Hueca Formation also contain sedimentary, volcanic, and greenschist-facies metapsammitic lithic grains (Fig. 7C). Greenschist-facies metapsammitic rocks are exposed in the Acatlán and Ayú Complexes (Fig. 2; Keppie et al., 2008a; Helbig et al., 2012). Considering that detrital garnets in the Piedra Hueca Formation are significantly different in composition from metamorphic garnets in the Ayú Complex (Fig. 9), we suggest that this latter complex was not a potential source of detritus. This interpretation confirms available barometric data, which indicate that the Ayú Complex would have not reached the surface before Early Cretaceous time (Helbig et al., 2012). Based on this consideration, we propose that metapsammitic lithic grains are derived from the Acatlán Complex. This is reasonable considering that a NW-trending fluvial stream with origin in the Oaxacan Complex needs to flow across greenschist-facies metasedimentary rocks of the eastern Acatlán Complex to finally discharge detritus in the Otlaltepec Basin (Fig. 12A). Derivation of metasedimentary grains also from the western part of the Acatlán Complex is suggested by the composition of group 2 detrital garnets. In fact, ∼15 km west of the Otlaltepec Basin, greenschist-facies metasedimentary assemblages of the Acatlán Complex are tectonically juxtaposed against high-pressure rocks (Figs. 2 and 12A), which contain metamorphic garnets that are similar in composition to some of group 2 detrital garnets in the Piedra Hueca Formation (Fig. 9).

Volcanic lithic grains are subordinate in sandstones from the Piedra Hueca Formation. Based on the occurrence of seven zircons with ages between ca. 184 and ca. 201 Ma in the analyzed samples from this unit, we suggest that these volcanic grains are likely sourced by Early–Middle Jurassic synrift volcanic rocks that were locally emplaced in eastern Mexico during the lithospheric attenuation that accompanied the breakup of Pangea (e.g., Martini and Ortega-Gutiérrez, 2016). Some of these volcanic rocks are exposed along the southeastern boundary of the Otlaltepec Basin (Fig. 12A) and are represented by andesitic to basaltic dikes that yielded zircons with ages between ca. 165 and ca. 193 Ma (Kirsch et al., 2014). Possible detrital contributions from distal volcanic centers that produced explosive eruptions are not discarded.

The origin of sedimentary lithic grains in the Piedra Hueca Formation is difficult to assess. Sedimentary units that were exposed in the Acatlán-Cuicatlán area during the Early–Middle Jurassic time and represent potential sources for these sedimentary grains are the upper Paleozoic Matzitzi Formation (Centeno-García et al., 2009), which is presently exposed ∼50 km to the east of the Otlaltepec Basin (Figs. 2 and 12A), and the lower Mesozoic Tianguistengo Formation, which is exposed within the Otlaltepec Basin and unconformably underlies the Piedra Hueca Formation. Unfortunately, quantitative petrographic data and U-Pb zircon geochronology are presently not available for these units. Therefore, provenance of sedimentary grains in the Piedra Hueca Formation remains an issue that needs to be addressed in future works.

Otlaltepec Formation

Like the Piedra Hueca Formation, sandstones from the Otlaltepec Formation are dominantly derived from deformed phaneritic rocks with a quartzofeldspathic to quartzitic composition. The occurrence of mesoperthitic K-feldspar in some sandstones from the Otlaltepec Formation suggests that granulite-facies rocks of the Oaxacan Complex represent a source of detritus for this unit. This is supported by the ubiquitous occurrence of ca. 940–1400 Ma zircon grains (Figs. 11D–11F) and detrital garnets that, in part, overlap in composition with metamorphic garnets in the Oaxacan Complex (Fig. 9). Along the Magdalena Creek, paleocurrent data indicate a main direction of sedimentary transport to the WNW (Fig. 5), which is consistent with derivation from the Oaxacan Complex (Fig. 12B). However, along the Xiotillo Creek, paleocurrents are dominantly directed to the N (Fig. 5). Undoubtedly, a more exhaustive set of paleocurrent data is required to reconstruct in detail the pattern of the drainage associated to the Otlaltepec Formation. Here, we speculate that N-directed paleocurrents may reflect local divergence of the fluvial stream around an overbank area or a distributary channel that branched off downstream from the WNW-trending main fluvial flow.

In addition to the ca. 940–1400 Ma zircons, sandstones from the Otlaltepec Formation also contain zircons that returned ages in the range of ca. 260–312 Ma (Figs. 11D–11F). Zircons with ages between ca. 260 and ca. 280 Ma are likely derived from the Cozahuico pluton (Fig. 3G), whereas those with ages in the range of ca. 280–312 Ma may have been sourced by the Totoltepec pluton (Fig. 3F). Some greenschist-facies metasedimentary rocks of the Acatlán Complex also contain zircons with an age range that overlaps with that of the Totoltepec pluton (Fig. 3D; Kirsch et al., 2012). However, considering that greenschist-facies metamorphic lithic grains are subordinate in sandstones from the Otlaltepec Formation, the high amount of ca. 280–312 Ma zircons likely suggests derivation from phaneritic rocks of the Totoltepec pluton. Like sandstones from the Piedra Hueca Formation, ca. 260–280 Ma zircons in the Otlaltepec Formation are subordinate (one to nine grains in each sample). On the contrary, ca. 280–312 Ma zircons are ubiquitous at least in some of the analyzed samples from the Otlaltepec Formation (59, 18, and four grains in samples SDT2.5, SDT2.8, and Xi2.7, respectively; Figs. 11D–11F). This indicates that the Totoltepec was an important source of detritus for at least part of the Otlaltepec Formation (Fig. 12B).

Subordinate framework components in the Otlaltepec Formation are represented by felsitic volcanic and greenschist-facies metapsammitic lithic grains. Like the Piedra Hueca Formation, these latter fragments may be derived from greenschist-facies metasedimentary rocks exposed in the eastern part of the Acatlán Complex (Fig. 12B). Detrital contributions from metasedimentary rocks of the Acatlán Complex are also suggested by few zircons with ages of ca. 330 and ca. 530 Ma, which are not represented in the Oaxacan Complex and East Mexico Arc granitoids.

Volcanic lithic grains may be derived from Early–Middle Jurassic synrift volcanic rocks that were emplaced in eastern Mexico during Pangea breakup. The age of 167 ± 4 Ma returned by one zircon in sample SDT2.5 supports such a hypothesis.

Tectonic Implications

Although metamorphic rocks of the Oaxacan and Acatlán Complexes represent main common sources for the Piedra Hueca and Otlaltepec Formations, major differences in detrital contributions from other pre-Jurassic basement rocks can be documented between these two units. In particular, ca. 280–312 Ma zircons interpreted to be sourced by the Totoltepec pluton are subordinate in the Piedra Hueca Formation (Figs. 11A–11C) but are ubiquitous in the lower stratigraphic part of the Otlaltepec Formation and progressively decrease in abundance toward its upper stratigraphic levels (Figs. 11D–11F). According to available barometric data, the Totoltepec pluton was rapidly exhumed at the surface by the end of Early Permian time (Kirsch et al., 2013). However, the subordinate amount of ca. 280–312 Ma zircons in the Piedra Hueca Formation suggests that, during the deposition of this latter unit, the Totoltepec pluton did not supply detritus into the Otlaltepec Basin. The Tianguistengo Formation, which is the oldest stratigraphic unit within the Otlaltepec Basin, locally overlays in unconformity the Totoltepec pluton. Therefore, we infer that the Totoltepec pluton was extensively covered by sedimentary rocks of the Tianguistengo Formation during the deposition of the Piedra Hueca Formation (Fig. 12A). In light of this scenario, the sudden enrichment in ca. 280–312 Ma zircons in the lower part of the Otlaltepec Formation indicates that, after being buried beneath the Tianguistengo Formation, the Totoltepec pluton was rapidly exhumed to the surface for a second time, forming a local topographic high that supplied detritus into the Otlaltepec Basin (Fig. 12B). The progressive depletion in ca. 280–312 Ma zircons toward the upper part of the Otlatepec Formation is interpreted as recording the end of the exhumation process. The Totoltepec pluton is presently exposed in the footwall of the Matanza fault, which is a main sinistral normal structure that extends for at least 15 km with an E-W orientation (Figs. 4 and 12B). We suggest that this structure was responsible for the second event of exhumation of the Totoltepec pluton (Fig. 12B). Based on this interpretation, the depositional age of the Otlaltepec Formation constrains the time of sinistral normal displacements along the Matanza fault. The integration of our U-Pb geochronologic ages with previously reported biostratigraphic data brackets the deposition of the Otlaltepec Formation between 167.5 ± 4 Ma, the age of the youngest concordant zircon in this unit (this work), and the Oxfordian (163.5 ± 1 Ma; Gradstein et al., 2012), the age of the initial marine transgression in the analyzed sector of southern Mexico (Campos-Madrigal et al., 2013). In light of these data, the provenance analysis presented in this work suggests that sinistral displacements of continental blocks along the W-trending Matanza fault was taking place in southern Mexico during Pangea breakup. This observation supports previous regional-scale reconstructions of Pangea, which infer that southern Mexico was located to the northwest of its present position during Early and Middle Jurassic time (Pindell, 1985; Pindell and Kennan, 2009). Whether the Matanza fault is an isolated segment that only had local significance in the regional tectonic evolution or it was part of a wide zone of crustal weakness remains unclear at this time. Additional work is required to recognize evidence of displacements along other W-trending, sinistral fault segments in the Jurassic stratigraphic record of southern Mexico. The approach that we presented in this study is novel in Mexico, and it offers some perspective for future studies on sandstone provenance aimed to reconstruct in detail the kinematics of fragmentation of Pangea in the Mesoamerican region.

CONCLUSION

Data presented in this work document a major change in sedimentary provenance within the stratigraphic record of the Otlaltepec Basin. Whole-rock petrography integrated with U-Pb zircon geochronology, garnet compositional data, and paleocurrent directions indicates that metamorphic rocks from the Oaxacan and Acatlán Complexes are main sources for the Piedra Hueca Formation. Detrital contributions from the Totoltepec and Cozahuico plutons, as well as from volcanic rocks emplaced during Pangea breakup, were subordinate during the deposition of this unit. On the contrary, the provenance analysis of sandstone from the Otlaltepec Formation indicates that the base of this unit was dominantly sourced by the Totoltepec pluton, with minor detrital contributions from the Oaxacan and Acatlán Complexes, the Cozahuico pluton, and volcanic rocks emplaced during Pangea breakup. We interpret this provenance change to reflect the rapid exhumation of the Totoltepec pluton along the W-trending, sinistral normal Matanza fault. According to this interpretation, the depositional age of the Otlaltepec Formation constrains the time of sinistral normal displacements along the Matanza fault between 163.5 ± 1 and 167.5 ± 4 Ma. This indicates that the Matanza fault was involved in the crustal attenuation that accompanied the breakup of Pangea and that sinistral motion of continental blocks along W-trending structures was taking place in southern Mexico as predicted by global plate tectonic reconstructions.

This research was funded by Programa de Apoyo a Proyectos de Investigación e Innovación Tecnológica (PAPIIT) grant IA100214 to Michelangelo Martini. C. Ortega-Obregón (Centro de Geociencias, UNAM) is acknowledged for maintaining the mass spectrometer facilities at LEI and for performing the U-Pb and garnet LA-ICPMS analyses. Carlos Linares (Instituto de Geofísica, UNAM) performed microprobe analysis of garnets at Laboratorio Universitario de Petrología. Joaquín Aparicio (Instituto de Geología, UNAM), and Juan Tomáz Vázquez (Centro de Geociencias, UNAM) prepared thin sections. We thank Elena Centeno-García for the constant discussion on the tectonostratigraphic evolution of southern Mexico during the Jurassic. We also thank Chris Spencer, Glenn Sharman, James Pindell, and two anonymous reviewers for their helpful and constructive comments that greatly improved the manuscript.

1Supplemental File A. Paleocurrent data. Please visit http://dx.doi.org/10.1130/GES01366.S1 or the full-text article on www.gsapubs.org to view Supplemental File A.
2Supplemental File B. Results of garnet grains analyzed with an electron probe microanalyzer as well as by laser ablation inductively-coupled plasma mass spectrometry using matrix-matched garnets for major elements. Please visit http://dx.doi.org/10.1130/GES01366.S2 or the full-text article on www.gsapubs.org to view Supplemental File B.
3Supplemental File C. Details of analytical methodology and analytical results for individual zircon ages. Please visit http://dx.doi.org/10.1130/GES01366.S3 or the full-text article on www.gsapubs.org to view Supplemental File C.