nueva página del texto (beta)
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
Introduction
The area of Chiapas, geodynamically controlled by the interaction of the North American, Caribbean, and Cocos plates (Guzmán-Speziale et al., 1989; Meneses-Rocha, 2001; Authemayou et al., 2011; Witt et al., 2012; Molina-Garza et al., 2015), can be considered as a complex geological province of southern Mexico. The Chiapas Massif Complex (CMC) is the “root” of an active continental margin formed during the late Permian to Triassic and actually represents the crystalline basement of the southern Maya block within the North American plate (Figure 1), as it was interpreted by several authors (e.g., Dengo, 1985; Meneses-Rocha, 1985, 2001; Schaaf et al., 2002; Weber et al., 2005, 2007). Numerous field observations, petrogenetic, as well as high- to medium-temperature geochronological studies were performed in the CMC to understand its origin and to reconstruct its geological history for Paleozoic to early Mesozoic times (Damon et al., 1981; Torres et al., 1999; Schaaf et al., 2002; Weber et al., 2005, 2006, 2007, 2008, 2009; Estrada-Carmona et al., 2009, 2012; Pompa-Mera, 2009; González-Guzmán et al., 2016). Recent studies by González-Guzmán et al. (2016) and Cisneros de León et al. (2017) indicate that the geodynamic evolution of the CMC, as well as of the whole Maya block, started, at least, in the late Mesoproterozoic. The CMC, therefore, may include abundant signals (e.g., crystallization or cooling ages) of either thermo-tectonic or magmatic events that occurred in southern Mexico, which can be used to better reconstruct the tectonic history of southern North America.
Figure 1 Regional geodynamic setting of southern Mesoamerica and simplified geologic map of southern Mexico (modified from Martínez-Amador et al., 2005; Weber et al., 2008; Estrada-Carmona et al., 2012; Torres-de León et al., 2012; Solari et al., 2013; Cisneros de León et al., 2017). CMC-Chiapas Massif Complex; GCC-Guichicovi Complex; JCF-Jocotán-Chamelecón fault; MMT-Maya Mountains; MTF-Motagua fault; OAX-Oaxacan Complex; SCT-Sierra de los Cuchumatanes; PLF-Polochic fault; PMFS-Polochic-Motagua Fault System; TMFS-Tuxtla-Malpaso Fault System; TSZ-Tonalá Shear Zone.
Despite of its strategic locality (i.e., in southern Maya block and very close to the Chortís block; Figure 1), no detailed and systematic low-temperature thermochronological studies have been carried out in the CMC to detect relatively younger tectonic events that occurred from Mesozoic to present. Such studies, including primarily apatite fission-track (AFT) and zircon fission-track analyses and helium dating, are particularly useful to understand the evolution of the uppermost continental crust that may be controlled by rifting, collisional, diagenetic, low- to medium-grade metamorphic, hydrothermal, or denudation processes, among others (e.g., Ehlers and Farley, 2003; Reiners et al., 2005; Reiners and Brandon, 2006; Mark et al., 2014). A few AFT results from the central CMC, ranging from ~40 to ~25 Ma (Ratschbacher et al., 2009), were interpreted as indicating a rapid cooling of the CMC in the Oligocene (more than 2 km of exhumation and erosion at ~30 ± 3 Ma), a cooling signal that could have been triggered by the hypothetical arrival of the Chortís block in front of the Gulf of Tehuantepec (e.g., Ratschbacher et al., 2009; Witt et al., 2012). Other AFT results, reported by Witt et al. (2012), apparently dismiss the posibility of a rapid cooling of the CMC during the Oligocene. AFT analysis-derived time-temperature modeling for two sandstones sampled from the Middle(?) Jurassic Todos Santos Formation in the northern CMC area, as well as provenance studies performed in some Cenozoic units from the Sierra de Chiapas (including studies of Witt et al., 2012) suggest that in the CMC not only the Oligocene, but most of the Cenozoic was a tectonically quiescent period (approximately from Eocene to middle Miocene; Abdullin et al., 2016a, 2016b). Thus, up to now, somewhat controversial interpretations about the Cenozoic evolution of the CMC are proposed in these pioneering thermochronological studies. The Cenozoic tectonics of the CMC is of a particular interest because many authors (e.g., Schaaf et al., 1995; Pindell et al., 2006; Ratschbacher et al., 2009; Torres-de León et al., 2012; Witt et al., 2012) suggested an Oligocene arrival of the Chortís block to the Gulf of Tehuantepec region (i.e., as a collision of the Chortís and Maya continental blocks;), which should have been imprinted in the tectonic history of the CMC as a rapid cooling period as well as by ubiquitous stratigraphic records within Oligocene units from the inner parts of the Sierra de Chiapas (e.g., the presence of granitic conglomerates derived from the CMC rocks). Despite the controversy surrounding the Oligocene thermal history, all these AFT data detected a middle to late Miocene thermo-tectonic event that affected almost the entire territory of Chiapas (Ratschbacher et al., 2009; Witt et al., 2012; Abdullin et al., 2016a), which is essentially well-known as the Chiapanecan orogeny (e.g., Sánchez-Montes de Oca, 1979, 2006; Meneses-Rocha, 2001; Padilla y Sánchez, 2007; Guzmán-Speziale, 2010). A Late Cretaceous to late Paleocene-early Eocene orogenic event was proposed previously for Chiapas by several authors (Gutiérrez-Gil, 1956; Sánchez-Montes de Oca, 1979, 2006; Carfantan, 1981, 1985; Moravec, 1983; Meneses-Rocha, 1985, 1991, 2001), a hypothesis that seems to be confirmed by recent AFT analyses performed by Abdullin et al. (2016a).
In this work, new AFT ages are reported and a brief review of the stratigraphy of Chiapas is presented as well, with the purpose to reconstruct the complex tectonic evolution of the Chiapas Massif Compex from Mesozoic to present times. To extract ages of tectonic or magmatic events that occurred during the geological history of the CMC, basement rocks of the CMC and Mesozoic sandstones from the Oxfordian to Aptian San Ricardo Formation were analyzed by in situ AFT technique.
Study area
The present study is focused on the north-central part of the CMC region (Figure 1). The study area displays large exposures of the CMC (Figure 2; basement rocks are not differentiated in the map). The CMC predominantly includes late Permian-Triassic calc-alkaline intermediate to felsic plutonic rocks, most of which are strongly deformed while others are metamorphosed (Damon et al., 1981; Torres et al., 1999; Schaaf et al., 2002; Weber et al., 2005, 2007; Estrada-Carmona et al., 2009).
Figure 2 Simplified geological map of the northern Chiapas Massif area (modified after Martínez-Amador et al., 2005; Weber et al., 2008; Molina-Garza et al., 2015) showing the location of sampling sites for in situ AFT analysis. 1 - Quaternary sediments; 2 - Miocene volcanic and volcanoclastic lithologies; 3 - Miocene granites; 4 - Campanian to Maastrichtian Ocozocoautla Formation (gravel to marly sandstones, shales with marls, and limestones); 5 - Albian-Santonian Sierra Madre Formation (limestones and dolomites); 6 - Oxfordian to Aptian San Ricardo Formation (carbonates, marls, shales, and sandstones); 7 - Late Triassic (?) to Oxfordian Todos Santos Formation (conglomerates, volcanoclastic sediments, and fine-grained to coarse-grained sandstones); 8 - Chiapas Massif Complex (undifferentiated, mostly includes late Permian-Early Triassic granitoids and metamorphic rocks); 9 - strike-slip faults; 10 - Tonalá Shear Zone; 11 - samples analyzed by in situ AFT technique (this study); 12 - sandstones from Todos Santos Formation that were analyzed previously by AFT thermochronology (Abdullin et al., 2016a).
In this area, the base of the outcropping sedimentary cover is seemingly represented by red beds of the Todos Santos Formation (Figure 2), which unconformably overlies the CMC, and includes, from bottom to top, conglomerates, volcanoclastic materials, fine-grained, lacustrine and fluvial, medium- to coarse-grained (sub-) arkosic sandstones and conglomerates (Meneses-Rocha, 1985, 2001; Blair, 1987; Godínez-Urban et al., 2011). Its variable thickness, from 250 m to 1,350 m, can be related to a graben-like geometry derived from extensional processes associated with continental rifting (Sánchez-Montes de Oca, 1979, 2006; Meneses-Rocha, 1985, 2001; Blair, 1987; Padilla y Sánchez, 2007; Godínez-Urban et al., 2011). Because the Todos Santos Formation is continental, it is very difficult to determine its exact age. Nonetheless, most authors argue that these red beds were probably deposited in the Middle Jurassic (Quezada-Muñetón, 1983; Meneses-Rocha, 1985, 2001; Mandujano-Velásquez, 1996; Sánchez et al., 2004; Sánchez-Montes de Oca, 2006; Godínez-Urban et al., 2011; Abdullin et al., 2016a, 2016b).
The San Ricardo Formation conformably overlies the Todos Santos Formation (Figure 2). It includes three members, Oxfordian to Kimmeridgian carbonates, Tithonian marls with shales, and Berriasian to Aptian fine-grained sandstones, and its total thickness varies from 500 to 1400 m (Quezada-Muñetón, 1983; Mandujano-Velásquez, 1996; Sánchez et al., 2004). The lowermost beds were interpreted as a transgressive sequence above the Todos Santos Formation (Quezada-Muñetón, 1983; Meneses-Rocha, 2001; Sánchez-Montes de Oca, 2006). The Albian-Santonian Sierra Madre Formation (1,000-2,500 m thick) is mainly composed of limestones and dolomites deposited on a marine platform during a period of tectonic stability (e.g., Rosales-Domínguez et al., 1997; Rosales-Domínguez, 1998). The 300 to 800 m-thick Campanian-Maastrichtian Ocozocoautla Formation, which was deposited in a shoreline setting, consists, from base to top, of gravel, coarse-grained to marly sandstones, alternating thin layers of shales and marls, and limestones (Mandujano-Velásquez, 1996; Rosales-Domínguez, 1998). The high-energy siliciclastic sediments at the base of this unit could announce the beginning of a Late Cretaceous-early Paleogene orogeny that destroyed some parts of the Albian to Santonian platform (e.g., Meneses-Rocha, 1985, 1991, 2001; Abdullin et al., 2016a). The denudation of the westernmost areas of Chiapas during the Late Cretaceous-early Paleogene orogenic event is also well registered in the Campanian-Maastrichtian Méndez Formation and the Paleocene Soyaló, Lutitas Nanchital, and Conglomerado Uzpanapa formations, all of which contain abundant fragments of calcareous rocks derived from the Albian-Santonian carbonate platform (Meneses-Rocha, 1985; Quezada-Muñetón, 1987; Sánchez-Montes de Oca, 2006). Cenozoic sedimentary rocks were not recognized in the study area (Figure 2). The Cenozoic lithostratigraphic column of Chiapas is dominantly siliciclastic in composition with minor shallow water deposits locally associated with carbonate to mixed siliciclastic-carbonate platforms, all of which crop out in the central, northern, and eastern portions of the Sierra de Chiapas (Meneses-Rocha, 1985, 2001; Quezada-Muñetón, 1987; Sánchez et al., 2004; Sánchez-Montes de Oca, 2006; Padilla y Sánchez, 2007; Witt et al., 2012). During the middle to late Miocene, igneous activity took place along the CMC (e.g., Molina-Garza et al., 2015) and generated ubiquitous intrusions along the Tonalá shear zone (Figure 1) in the SW as well as the emplacement of some volcanic rocks and volcanoclastic deposits in the NE (Figure 2).
To solve the objectives of the present study, a total of seven rocks were sampled: five basement samples from the CMC and two fine-grained sandstones from the Berriasian-Aptian sandy member of the San Ricardo Formation (Figure 2). Basement rocks (P-1, P-2, P-3, P-4 and J-1) were collected principally to obtain relatively recent (i.e., Cenozoic) AFT ages in the CMC. Based on LA-ICP-MS zircon chronology, P-1, P-2, P-3, and P-4 were identified as late Permian (Lopingian) igneous and metaigneous samples, whereas J-1 yielded a Sinemurian crystallization age of 196 ± 2 (2σ) Ma (Ortega-Obregón et al., 2016). Two detrital samples were obtained from the San Ricardo Formation (SR-1 and SR-2). Due to the fact that they were mostly derived from the CMC and include apparently unreset apatites (Abdullin et al., 2016b), those sediments can be very useful to indirectly detect any signals of older cooling events, which apparently took place within the CMC prior to the deposition of this unit. Coordinates of rock samples and other details may be consulted in the Supplementary Appendix A1.
Analytical procedures
Heavy minerals were concentrated from the two narrow grain-size fractions of 60-125 and 125-180 µm using conventional techniques like crushing, sieving, WilfleyTM table, and FrantzTM separator. Nearly 300 apatite grains, extracted from each sample under a binocular microscope, were mounted with EpoFixTM (Struers) in a 2.5-cm-diameter plastic ring, and then were sequentially polished with SiC MicroCutTM Buehler sandpaper P-1500 (~13 µm) and P-2500 (~8 µm) and alumina suspensions (3, 1, 0.3, and 0.1 µm) over MicroClothTM Buehler pads. Polished apatites were etched in 5.5 M HNO3 at 21 °C for 20 s to reveal 238U spontaneous fission tracks (i.e., radiation damages produced by expulsed “daughter nuclides”) generated by the natural fission of 238U atoms (“parent nuclides”), a chemical protocol commonly used in the sample preparation for AFT analyses (e.g., Wagner and Van den haute, 1992; Donelick et al., 2005). Fission track counting and confined track and etch pit length measurements were performed using OlympusTM BX-51 and LeicaTM DM microscopes with “dry” objectives, both upgraded with high-resolution digital cameras and an image processing software. In the case of basement samples, only clear apatite crystals with homogeneously distributed tracks were selected for AFT dating, whereas apatite crystals from siliciclastic samples were selected randomly. Detrital apatite with many visible inclusions, which hinder adequate observation of tracks, as well as those with a strong track density zoning (with an evident zonation of U) were excluded from AFT dating, because these may yield highly biased single-grain AFT ages using the LA-ICP-MS-based technique (Hasebe and Arai, 2007; Abdullin et al., 2014, 2016a; Liu et al., 2014; Vermeesch, 2017). In some apatite grains, spontaneous track densities (ρ S) values were obtained from small areas, ~4,000 to ~5,500 µm2 (for single-spot micro-sampling by LA-ICP-MS), whereas in larger crystals track counting was performed within areas from ~8,000 to ~18,000 µm2 (for multi-spot analyses; from two to four spots per grain). For most apatite samples dated, chlorine contents were measured simultaneously by LA-ICP-MS to calibrate annealing properties of fission tracks. Additionally, in some detrital apatites from the San Ricardo Formation, D-par proposed by Raymond Donelick was also used as a main kinetic parameter (i.e., the mean etch pit length measured parallel to the crystallographic C-axis of apatite using reflected light on a polished and etched surface; see Donelick et al., 2005).
Sample preparation was performed in the Taller de Molienda and Taller de Laminación, while optical studies and LA-ICP-MS analyses were carried out in the Laboratorio de Estudios Isotópicos (LEI Lab), all at Centro de Geociencias (CGEO), Universidad Nacional Autónoma de México (UNAM), Campus Juriquilla. The micro-sampling with LA-ICP-MS was carefully performed within the areas observed previously to determine the ρ S values (for instance, see fig. 2 in Vermeesch, 2017 or fig. 3 of Abdullin et al., 2016a). Raw data were reduced off-line using the IoliteTM 3.4 program (Paton et al., 2011). The results for measured isotopes using the NIST612 glass (Pearce et al., 1997) were normalized using 43Ca as an internal standard and taking an average CaO content, according to Lesnov (2012), for Cl-rich apatites of 53% (J-1 rock sample only) and for typical F-apatites of 55% (the rest of the samples studied; see details below). Fragments of Durango F-apatite from the Cerro de Mercado mine were analyzed during the same sessions of track counting and LA-ICP-MS measurements. This famous mineral, with known age and more or less well-controlled chemical composition, was used to control the quality of AFT results obtained and as a primary standard for chlorine measurements in unknown apatites (taking a value of ~0.43 wt.% for Cl content; Goldoff et al., 2012; Chew et al., 2014). Single-grain AFT ages were calculated using the equations proposed by Hasebe et al. (2004, 2009), whereas 1σ-errors were obtained according to Abdullin et al. (2014, 2016a). Detailed information on our AFT and LA-ICP-MS experiments (including number of tracks counted, D-par values, 238U and Cl concentrations, AFT ages, and analytical errors) is given in the Supplementary Appendix A1. After the LA-ICP-MS sessions, apatite crystals already analyzed were re-polished and re-etched to observe their fresh internal sections with the purpose to increase the number of confined tracks tested, following the same sample preparation procedure described above.
The LA-ICP-MS-based in situ fission-track method, fair to say that was introduced in 2000-2005 (Cox et al., 2000; Svojtka and Košler, 2002; Hasebe et al., 2004, 2009; Donelick et al., 2005), is still in the development stage, or rather in a transition phase from the classical external detector method (EDM) to in situ technique (Vermeesch, 2017). After numerous experiments carried out in the past years, a LA-ICP-MS protocol for simultaneous U-Pb and AFT double dating plus multielemental analysis (rare earth elements, Y, Sr, Mn, Mg, Th, U, Cl) was established at the LEI Lab to be employed routinely in most unknown apatites (see details in Table 1).
Table 1 LA-ICP-MS protocol established at the LEI Lab (CGEO, UNAM) for simultaneous U-Pb and AFT double dating plus multielemental analysis (REE, Y, Sr, Mn, Mg, Th, U, and Cl).
| ICP-MS operating conditions | |
| Instrument | Thermo ScientificTM iCAPTM Qc |
| Forward power | 1,450 W |
| Carrier gas flow rate | ~1 L/min (Ar) and 0.35 L/min (He) |
| Auxiliary gas flow rate | ~1 L/min |
| Plasma gas flow rate | ~14 L/min |
| Nitrogen | ~3.5 mL/min |
| Data acquisition parameters | |
| Mode of operating | STD (standard mode) |
| Sampling scheme | -2NIST612-2MAD-1DUR-10apts- |
| Background scanning | 15 s |
| Data acquisition time | 35 s |
| Wash-out time | 15 s |
| Measured isotopes | 43Ca 44Ca 31P 35Cl 26Mg 55Mn 88Sr 89Y 139La 140Ce 141Pr 146Nd 147Sm 153Eu 157Gd 159Tb 163Dy 165Ho 166Er 169Tm 172Yb 175Lu 232Th 238U 204Pb 206Pb 207Pb 208Pb 202Hg (total = 29) |
| Laser ablation system | |
| Ablation cell | RESOlutionTM Laurin Technic S-155 |
| Model of laser | Resonetics RESOlutionTM LPX Pro |
| Wavelength | 193 nm (Excimer ArF) |
| Repetition rate | 4 Hz |
| *Energy density | *4 J/cm2 |
| Mode of sampling | spot diameter of 60 µm |
Note: REE: rare earth elements (La to Lu); MAD: “1st mine Discovery” U-Pb apatite standard from Madagascar; DUR: Durango F-apatite from the Cerro de Mercado mine (Mexico); apts: unknown apatites. (*): Constant laser pulse energy of 4 J/cm2, which was measured on target with a CoherentTM apparatus. For larger apatites (e.g., 400 µm, such as those from some metamorphic rocks or from certain mineral deposits), accuracy and precision in Cl measurements, AFT dating, and U-Pb system can be significantly improved by employing larger spot sizes (e.g., 110 µm or above) as well as longer signal acquisition times (e.g., 55 s).
Results and interpretations
For the Durango apatite used as a reference mineral, “absolute” mean ages of 30 ± 0.6 (1σ) Ma with dispersion of single-grain ages, D = 0% and chi-square probability test, P(χ2) = 100% (number of apatite grains dated, Ngr = 12) and of 29.4 ± 0.7 (1σ) Ma with D = 0% and P(χ2) = 98% (Ngr = 10) were obtained during two separate LA-ICP-MS sessions (grain ages and other details can be consulted in Supplementary Appendix A1). Both these values are slightly younger than its broadly accepted true age of 31.4 ± 0.5 (1σ) Ma (e.g., Green, 1985; McDowell et al., 2005; Solé and Pi, 2005; Chew and Donelick, 2012; Jonckheere et al., 2015). Hence, AFT ages obtained directly from unknown apatites were corrected to be given as “relative” ages based on the value of 31.4 Ma for the Durango apatite standard (following the methodology described in Hasebe et al., 2004, 2009). This mathematical trick is recommended to obtain more reliable AFT ages and is basically equivalent to the ζ calibration (Hasebe et al., 2004, 2013; Donelick et al., 2005; Vermeesch, 2017) used in the conventional EDM (e.g., Hurford and Green, 1983; Wagner and Van den haute, 1992). The central (weighted-mean) AFT ages of the analyzed samples were calculated using the application RadialPlotter 8.3 written by Vermeesch (2009) and displayed as radial plots (Figures 3 and 4) (in statistics, known as Galbraith’s plots; Galbraith, 1988, 1994). To check the precision of chlorine measurements by LA-ICP-MS, the “1st Mine Discovery” apatite from Madagascar, a stunningly beautiful turquoise-colored mineral with a mean 206Pb-238U age of 485-475 Ma (Thomson et al., 2012; Chew et al., 2014), was used. For this apatite, an average
Figure 3 Radial plots displaying the AFT results for the CMC samples. P-1, P-2, P-3, and P-4 are late Permian granitoids/metagranitoids, and J-1 is a Jurassic subvolcanic sample. Ngr - number of apatite grains dated; Ncl - number of grains used for chlorine measurements by LA-ICP-MS; Ntr - number of confined tracks measured; D - dispersion of single-grain ages (%); P(χ2) - chi-square probability test; MTL - mean track length. Note: only TINT-type confined tracks (i.e., track-in-track; see details in Donelick et al., 2005) were tested for MTL measurements.
Figure 4 Radial plots with AFT ages from the Berriasian to Aptian member of the San Ricardo Formation (SR-1 and SR-2, sampled in the uppermost level of the sandy member). Ngr - number of detrital apatite grains dated; D - age dispersion (%); P(χ2) - chi-squared probability test. Binary plots displaying single-grain AFT ages vs.D-par values (or Cl contents) are also presented.
Cl content of 0.20 ± 0.01 (SD) wt.% was obtained in the LEI Lab, which is in line with the value of 0.22 ± 0.01 (SD) wt.% reported by Chew et al. (2014) also employing the LA-ICP-MS technique. Almost all the samples (except J-1) contain common F-apatites according to low levels of Cl (less than 0.2 wt.%; Figure 3) and small etch pits (D-par less than 2 µm; Figures 4). It appears that etch pit size in apatite is not always linked to its halogen content (e.g., Barbarand et al., 2003; our observations in the LEI Lab). In any case, D-par is widely accepted as a very practical semi-quantitative tool for the identification of different apatite specimens (Carlson et al., 1999; Ketcham et al., 1999; Donelick et al., 2005; Lisker et al., 2009). Annealing rates of fission tracks in apatite are apparently controlled by until-cell parameters, which can be “estimated” from the directly measured Cl content, though the content of certain trace elements (e.g., rare earth elements, Sr, Y) should also be taken into account (Barbarand et al., 2003). Additional annealing experiments are highly recommended to better understand the mechanism and kinetics of fission-track annealing in apatites, especially in Cl-rich to Cl end-member varieties.
As shown in Figure 3, three late Permian igneous/meta-igneous samples (P-1, P-2 and P-3), despite their identical apatite compositions and roughly similar elevations, yielded dispersed AFT ages that are within a time span of ca. 50-25 Ma, and two of them apparently failed the chi-squared test. In spite of the large number of apatite sections observed (ca. 500 etched surfaces in each CMC rock sample), a few confined tracks were found to be measured (Ntr < 20 in most CMC samples and Ntr = 36 in J-1; Figure 3). This can be explained by either low U contents (Supplementary Appendix A1) or relatively young cooling ages, both factors that are crucial for generating low number of fission tracks. Mean confined track lengths (MTL) in these apatite groups range from 12.5 ± 1.9 (SD) to 12.7 ± 2.2 (SD) µm (Figure 3), which is common for samples from thermally undisturbed basement rocks (MTL of 12-13 µm with SD of ~2 µm; e.g., Gleadow et al., 1986). This implies that the samples P-1, P-2, and P-3 passed very slowly through the partial annealing zone (PAZ) of the AFT system (110 to 60 °C for F-apatites; Gleadow et al., 1986; Green et al., 1989; Donelick et al., 2005). In this case, their mean AFT ages are actually “apparent cooling ages” and can hardly have any real geological meaning and, consequently, cannot be interpreted as indicating rapid cooling events sensu stricto. In contrast, apatites from samples P-4 and J-1 yielded longer MTL of 14.1 ± 0.8 (SD) and 13.3 ± 1.6 (SD) µm, respectively, and both rock samples undoubtedly passed the chi-square test (Figure 3). Therefore, these two samples, unlike the remaining three, may have cooled quite rapidly through their PAZ and, thus, can really reflect tectonic and magmatic events that occurred in the CMC region. The central AFT age of 17 ± 1 (1σ) Ma obtained from the late Permian sample P-4 can be interpreted as a cooling signal belonging to the middle-late Miocene event. The sample J-1 shows a mean AFT age of 194 ± 4 (1σ) Ma, which is significantly different from that obtained for other samples from the CMC. This age should be taken as the crystallization age for J-1 due to its apparent andesitic composition and high P(χ2) (Figure 3). Besides, the AFT age obtained for J-1 is identical within error to the zircon U-Pb age of 196 ± 2 (2σ) Ma reported for the same sample by Ortega-Obregón et al. (2016). The number of horizontally confined tracks measured in J-1 was high enough to be used in the interpretation of its complex thermal history (see the next section for details).
Most detrital apatite grains from the uppermost part of the Berriasian-Aptian sandy member of the San Ricardo Formation (SR-1 and SR-2) yielded single-grain AFT ages that are greater than its stratigraphic age, and no obvious positive correlation between D-par (or chlorine) and single-grain age was observed (Figure 4). The high amount of fission tracks preserved in apatite from these clastic samples (see U and ρ S in Supplementary Appendix A1) allowed to test many track lengths (58 and 60 measurements in SR-1 and SR-2, respectively; Figure 5). Apatites from SR-1 and SR-2 display short MTL of 12.4 ± 1.1 (SD) and 12 ± 1.1 (SD) µm, respectively. Track lengths with narrow distributions (SD of ~1 µm) and the absence of any bimodality in the histograms (Figure 5) can indicate that fission tracks in these detrital apatites were slightly annealed after deposition (e.g., Armstrong, 2005; Vermeesch et al., 2006). Therefore, these sandstone samples reached the shallowest levels of the PAZ during burial diagenesis (probably, at most, only ~70 °C), which is in agreement with the stratigraphy of the study area where SR-1 and SR-2 are located in higher levels as compared to the dated Todos Santos Formation rocks (Figure 2), for which a maximum burial temperature of 80-90 °C has been proposed on the basis of thermal history modeling (Abdullin et al., 2016a; for more details, see also the next section). These results suggest that detrital apatites from SR-1 and SR-2 were not fully reset for AFT chronometry by burial-related re-heating during diagenesis (Figure 4); hence, their unreset AFT ages (i.e., ages with provenance signals) may be crucial to define indirectly certain cooling episodes occurred in the CMC during the Triassic to Jurassic period.
Figure 5 Observed (measured) fission-track length distributions for SR-1 and SR-2, shown as histograms. Ntr - number of track lengths measured; MTL - mean track length. Note: only TINT-type confined tracks (track-in-track; see Donelick et al., 2005) were tested for MTL measurements.
Identical central AFT ages of 184 ± 5 (1σ) and of 189 ± 5 (1σ) Ma were obtained for SR-1 (Ngr = 35) and SR-2 (Ngr = 39), respectively (Figure 4), which likely have no real geological significance due to their scattered single-grain ages of 253 ± 21 (1σ) to 142 ± 12 (1σ) Ma (see Supplementary Appendix A1) and P(χ2) of 0%. Therefore, it is necessary to decompose these AFT results into groups of component distributions, which may provide useful information about the cooling events that took place in the source areas (e.g., Armstrong, 2005; Donelick et al., 2005; Vermeesch et al., 2006; Lisker et al., 2009). The data were decomposed using RadialPlotter 8.3 (Vermeesch, 2009) based on the statistics of Galbraith and Green (1990). As a result, two equally important AFT age peaks of 223 ± 10 (1σ) to 218 ± 5 (1σ) and of 171 ± 5 (1σ) to 162 ± 4 (1σ) Ma were obtained (Figure 4), which could be interpreted as Triassic and Jurassic cooling signals derived from the CMC source area.
Discussion
Late Triassic cooling event
The oldest AFT age peaks of 223 ± 10 (1σ) and 218 ± 5 (1σ) Ma obtained from the San Ricardo Formation indicate that a Triassic cooling event took place in the CMC, the source area from which the sediments were entirely derived (e.g., Quezada-Muñetón, 1983; Sánchez et al., 2004; Abdullin et al., 2016b). Triassic cooling signals were also obtained directly from the CMC on the basis of numerous medium-temperature thermochronometric studies. For example, Torres et al. (1999) compiled K-Ar biotite ages from 261 ± 10 (2σ) to 239 ± 5 (2σ) Ma, the youngest of which may be interpreted as Triassic cooling ages. Schaaf et al. (2002) also demonstrated, using Rb-Sr ages of 244 ± 12 (2σ) to 214 ± 11 (2σ) Ma obtained for biotite-whole-rock pairs, that an important cooling event occurred in the CMC during the Triassic. These systems in micas have closure temperatures of 350-300 °C (e.g., Faure and Mensing, 2005). Triassic cooling ages were also verified by apatite U-Pb thermochronology tested on some CMC samples that yielded U-Pb ages of ~250-230 Ma (Ortega-Obregón et al., 2016), which display cooling times through ~550-400 °C paleoisotherms (e.g., Cochrane et al., 2014). However, Triassic AFT ages obtained in this study indicate cooling episodes through significantly colder paleoisotherms (taking ~120-100 °C as an effective closure temperature in F-apatites). All these nearly similar cooling ages imply that the CMC was partially uplifted and eroded already during the Late Triassic (at ca. 230-210 Ma). Some detrital apatite grains from the Todos Santos Formation (Figure 2) also yielded Triassic AFT ages (Abdullin et al., 2016a). Hence, the depositional age for the Todos Santos Formation, overlying the CMC, could be extended up to the Triassic (~230 Ma as the age of the lowermost strata in some parts of SE Mexico), as it was previously suggested by several authors (e.g., Salvador, 1987; Padilla y Sánchez, 2007; Pérez-Gutiérrez et al., 2009). The ~230-210 Ma cooling period is coeval with the initial breakup of Pangea during the Late Triassic, which in turn confirms that the CMC, as well as the whole Maya block, were involved in this global-scale geodynamic activity (Michalzik, 1991; Dickinson and Lawton, 2001; Steiner, 2005; Bird and Burke, 2006; Padilla y Sánchez, 2007; Martini and Ortega-Gutiérrez, 2016).
Early Jurassic magmatism
Although Jurassic igneous rocks have not been separately plotted on available geological maps, there is evidence for the existence of Jurassic magmatic activity in the CMC. For example, a Middle-Late Jurassic magmatism (175-140 Ma) was suggested for the CMC area by Damon et al. (1981) on the basis of reset K-Ar biotite ages from late Permian granitoids as well as on the dating of andesite dykes and hydrothermally altered rocks. More recent geochronological data (e.g., Schaaf et al., 2002; Godínez-Urban et al., 2011; Ortega-Obregón et al., 2016; this study), however, indicate that this magmatic event took place in the Early Jurassic (200-180 Ma). Sample J-1 is of volcanic, or rather of subvolcanic, origin. Its AFT age of 194 ± 4 (1σ) Ma (Figure 3), together with an 40Ar/39Ar date of 186 ± 2 (2σ) Ma on hornblende and a zircon U-Pb age of 191 ± 3 (2σ) Ma reported recently for the southern CMC area (see details in Godínez-Urban et al., 2011), as well as a zircon U-Pb age of 196 ± 2 (2σ) Ma for J-1 (Ortega-Obregón et al., 2016) suggest that the main period of the Jurassic magmatic event in the CMC could be reduced to ca. 200-190 Ma. The Early Jurassic magmatism detected within the territory of Chiapas confirms that the Maya block stores signals of the Early-Middle Jurassic Nazas volcanic arc (ca. 200-160 Ma; e.g., Godínez-Urban et al., 2011; Rubio-Cisneros and Lawton, 2011).
Middle Jurassic tectonic event
Both AFT age peaks at 171 ± 5 (1σ) and 162 ± 4 (1σ) Ma obtained for SR-1 and SR-2 samples from the San Ricardo Formation (Figure 4) suggest that a Middle Jurassic cooling event affected the CMC, at least partly. It is necessary to take into account that these cooling peaks may be slightly overestimated due to the presence of some detrital apatites with Early Jurassic AFT ages (Figure 4), which could have been derived from Early Jurassic volcanic sources as well (e.g., Abdullin et al., 2016b). These age peaks, nevertheless, are in line with the age of the strike-slip event along the Sierra de Juárez mylonitic belt (southern Mexico), which was determined as 169 ± 2 (2σ) Ma based on muscovite 40Ar/39Ar ages (Alaniz-Alvarez et al., 1996). Recently, Martini et al. (2016) also identified a Middle Jurassic exhumation signal of 172-162 Ma from the Matanza fault to the north of the late Paleozoic Totoltepec pluton (southern Mexico). Within the Chiapas region, an important AFT age peak of 163 ± 3 (1σ) Ma was obtained from non- to slightly annealed apatites from four Todos Santos Formation sandstones (Figure 2) studied by Abdullin et al. (2016a). This AFT age peak as well as the AFT peaks of 171 ± 5 (1σ) and 162 ± 4 (1σ) Ma obtained for the San Ricardo Formation (this study) belong to the same Middle Jurassic tectonic event that can be identified elsewhere in southern Mexico (e.g., Alaniz-Alvarez et al., 1996; Martini et al., 2016). The Middle Jurassic cooling period with an approximate age of 170-160 Ma was triggered by continental rifting at the beginning of the opening of the Gulf of Mexico (Alaniz-Alvarez et al., 1996; Bird et al., 2005; Bird and Burke, 2006; Godínez-Urban et al., 2011). In other words, Jurassic cooling ages derive from extensional processes that affected the southern margins of the Gulf of Mexico. This is registered as highly variable thicknesses of Triassic(?) to Jurassic red beds, which can be linked to graben-type geometries essentially related to extensional processes associated with continental rifting during the opening of the Gulf of Mexico (Meneses-Rocha, 1985, 1991, 2001; Blair, 1987; Sánchez-Montes de Oca, 2006; Padilla y Sánchez, 2007; Witt et al., 2012; Martini and Ortega-Gutiérrez, 2016).
Late Cretaceous-Paleocene orogeny
During the Late Jurassic period, a marine transgression took place in the territories now occupied by Chiapas and the Yucatán platform, between which the Late Jurassic to Late Cretaceous basins of southeastern Mexico began to form (e.g., Meneses-Rocha, 1985, 1991, 2001; Sánchez-Montes de Oca, 2006; Padilla y Sánchez, 2007). The well-exposed base of the Oxfordian-Aptian San Ricardo Formation is an important register of this regional transgression (Quezada-Muñetón, 1983; Blair, 1987; Sánchez et al., 2004). The early stage of the development of these basins continued until the Albian-Santonian period, during which an extensive carbonate platform was established (the Sierra Madre Formation), well preserved in the Sierra Madre Oriental as well as in the Sierra de Chiapas (Rosales-Domínguez et al., 1997; Rosales-Domínguez, 1998; Meneses-Rocha, 2001; Sánchez-Montes de Oca, 2006; Fitz-Diaz et al., 2017). Previous thermal history modeling of the study area (Figure 6a; Abdullin et al., 2016a) indicates that burial-related heating of the Todos Santos Formation during its long-term diagenesis, and particularly during the development of the Cretaceous platform, was sufficient to reset the AFT system in the majority of detrital apatites. A Late Cretaceous-early Paleogene orogeny, previously proposed for the area of Chiapas on the basis of stratigraphic and structural studies (e.g., Gutiérrez-Gil, 1956; Sánchez-Montes de Oca, 1979, 2006; Carfantan, 1981, 1985; Moravec, 1983; Meneses-Rocha, 1985, 1991, 2001; Burkart et al., 1987), has been confirmed by AFT thermochronology applied for the Todos Santos Formation sandstones (Figure 6a; Abdullin et al., 2016a). All the aforementioned authors related this orogenic event to the Laramide orogeny. Nonetheless, in order to avoid any confusion, it may be more correct to name the event occurred in Chiapas as the Laramide sensu lato, a term proposed by Garduño-Martínez et al. (2015). Both best-fit curves on AFT analysis-derived thermal models with high values of goodness-of-fit between the measured and model ages (GOF) and goodness-of-fit between the measured and model confined track lengths (GOF K-S) (Figure 6a) suggest that this thermo-tectonic activity affected the CMC area during the Late Cretaceous to Paleocene (~85-80 to 60-55 Ma). However, as it was already proposed in previous works (Meneses-Rocha, 1985, 1991, 2001; Padilla y Sánchez, 2007), the deformation style of this tectonic event in Chiapas may be different if compared to its roughly contemporaneous analogue from central and northern Mexico (e.g., Garduño-Martínez et al., 2015; Fitz-Diaz et al., 2017). Finally, there is abundant evidence that the CMC area was repeatedly uplifted and eroded during the Laramide sensu lato (i.e., during the Late Cretaceous-Paleocene) and, as result, conglomerates and sandstones at the base of the sedimentary cover in the Gulf of Tehuantepec (Figure 1) were deposited in the Late Cretaceous (e.g., Sánchez-Barreda, 1981; Pedrazzini et al., 1982; Meneses-Rocha, 2001; Sánchez-Montes de Oca, 2006; Morán-Zenteno et al., 2009). In other words, Late Cretaceous-Holocene clastic sediments of the Gulf of Tehuantepec were partially derived from the CMC.
Figure 6 (a) Thermal modeling results for samples T-2 and T-4 (see location in Figure 2) from the Todos Santos Formation (taken from Abdullin et al., 2016a). Note that only best-fit curves are presented. The simulation of time-temperature histories was performed using the HeFTy 1.8.0 software of Ketcham (2005) based on the annealing model proposed by Ketcham et al. (2007). Modeling was carried out for a defined population of detrital apatites dated and confined tracks tested with D-par values of less than 1.9 µm, which mainly correspond to F-apatites (etching protocol: 5.5 M HNO3 at 21 °C for 20 s). Ngr - number of dated grains with D-par < 1.9 µm; Ntr - number of tracks measured in apatites with D-par < 1.9 µm; Age GOF - goodness-of-fit between the measured and model ages; GOF (K-S) - goodness-of-fit between the measured and model confined track lengths; (K-S) is the Kolmogorov-Smirnov test; PAZ - partial annealing zone for F-apatites (110 to 60 °C; Gleadow et al., 1986; Green et al., 1989; Donelick et al., 2005). (b) Predictive modeling of the thermal history of Cl-rich apatites from sample J-1 based on the results from TS-2 and TS-4. PAZ was hypothetically assumed as ca. 130-70 °C (see text for more details). The number of tested tracks in J-1 was too low to construct a HeFTy-based time-temperature history, because around 100 to 150 track length measurements are recommended for samples that display complex thermal histories (e.g., exhumation-burial-exhumation, like J-1). Note: only TINT-type confined tracks (i.e., track-in-track) were tested for MTL measurements.
Due to a relatively close localization of sample J-1 to the Todos Santos Formation sandstones already analyzed by AFT thermochronology (Figure 2), its thermal history can be constructed as a predictive model on the basis of modeling results previously reported for the Todos Santos Formation (Figures 6a and 6b; Abdullin et al., 2016a); though, it should be noted that sample J-1 was collected from a stratigraphycally lower level than those from the Todos Santos Formation. It is well known that fission tracks in Cl-apatite specimens are significantly more resistant to thermal annealing as compared with tracks from typical F-apatites (Carlson et al., 1999; Ketcham et al., 1999; Barbarand et al., 2003; Donelick et al., 2005). For example, for apatites displaying D-par ≥ 3 µm or with Cl ≥ 3 wt.%, Donelick et al. (2005) postulated that fission tracks begin to anneal at ~90 °C and experience total annealing at temperatures greater than 160 °C. The Early Jurassic sample J-1 is characterized by Cl-rich apatites with Cl contents of 1.58 ± 0.11 (SD) wt.% (Figures 3 and 6b). This means that, first of all, these apatites have elevated closure temperatures (either PAZ or effective closure temperatures) and, secondly, the PAZ in J-1-derived apatites should be slightly wider and “hotter” as compared to the PAZ of 110-60 °C in common F-apatites (Gleadow et al., 1986; Green et al., 1989; Donelick et al., 2005). According to this premise, for J-1 we can hypothetically assume a PAZ of ~130-70 °C. For this reason, burial-related heating was not sufficient to reset the AFT ages in J-1 (J-1 reached only the uppermost part of the PAZ, as shown in Figure 6b) but the temperature was high enough to partially anneal some fission tracks, finally yielding a bimodal length distribution with MTL of 13.3 ± 1.6 (SD) µm (Figures 3 and 6b), which is quite common for rock samples that experienced complex thermal histories (Gleadow et al., 1986; Wagner and Van den haute, 1992; Donelick et al., 2005; Vermeesch et al., 2006; Lisker et al., 2009). The bimodality in the track length distribution in apatite from J-1 was formed during the development of the Cretaceous platform in southeastern Mexico.
Cenozoic tectonic evolution
The Oligocene rapid cooling event, which could have been triggered by the arrival of the Chortís block in front of the Gulf of Tehuantepec and, consequently, by its collision with the Maya block (Ratschbacher et al., 2009; Witt et al., 2012), was not identified in the northern CMC region. The AFT data from CMC samples P-1, P-2, and P-3 (Figures 2 and 3), as well as two thermal histories simulated for the Todos Santos Formation (Abdullin et al., 2016a; see best-fit curves in Figure 6a), suggest that the Eocene-early Miocene was likely a period of tectonic calm in the north-central CMC area. Recent provenance studies performed in some Cenozoic units from the internal Sierra de Chiapas (Witt et al., 2012; Abdullin et al., 2016b) support this hypothesis, because Eocene clastic sediments were definitely not derived from the CMC source area. Meneses-Rocha (1985, 2001) proposed that Eocene-Oligocene sediments in Chiapas were mainly derived from Central Guatemala, a hypothesis that was in part confirmed by provenance analyses performed on the Ypresian-Lutetian(?) El Bosque Formation (Abdullin et al., 2016b). In a similar way, we can speculate that Oligocene clastic materials of Chiapas were predominantly derived from southern source areas, which is also supported by an apparent south-to-north marine regression during this time (Meneses-Rocha, 1985, 2001; Quezada-Muñetón, 1987). The Oligocene column of the Sierra de Chiapas is characterized by siliciclastic deep-basin sediments as well as by carbonate and mixed platforms developed in shallow water environments (Quezada-Muñetón, 1987; Meneses-Rocha, 2001; Witt et al., 2012). Most Oligocene sediments of the northern Sierra de Chiapas, therefore, were probably derived from regions located farther south; the Guatemala Suture Zone (Figure 1), also known as the Guatemala Suture Complex (Martens et al., 2012), is a good candidate for the main source region (Meneses-Rocha, 1985, 1991, 2001). Finally, it may be proposed that the CMC has not experienced any significant tectonic uplift during the Eocene to early Miocene period. Oligocene-aged abrupt cooling signals from the southernmost section of the CMC require a strong confirmation in further thermochronological studies.
Rapid cooling signals, corresponding to the middle to late Miocene Chiapanecan orogeny (Sánchez-Montes de Oca, 1979, 2006; Meneses-Rocha, 1985, 2001; Padilla y Sánchez, 2007; Guzmán-Speziale, 2010), were best detected in the thermochronological studies of Witt et al. (2012), based on AFT analyses and apatite (U-Th-Sm)/He dating. Both thermal histories modeled from the Todos Santos Formation also reveal this recent tectonic activity, showing very rapid cooling of apatites from the upper PAZ (65-60 °C) to the surface temperature (Figure 6a; Abdullin et al., 2016a). In this study, such a young cooling age was extracted only for P-4 (17 ± 1 Ma; Figure 3), which was sampled closer to the Tonalá Shear Zone as compared to other analyzed samples (Figure 2). Its cooling signal is roughly coeval with the activation time of the Tonalá Shear Zone, along which many plutons were also emplaced during the middle to late Miocene (e.g., Molina-Garza et al., 2015). Thereby, fast exhumation of the CMC, or rather of the entire territory of Chiapas, started at the middle-late Miocene. This thermo-tectonic event, together with the Pliocene to present magmatism in the Sierra de Chiapas, was important for the development of geological structures as well as of the steep present-day topography (~20 to ~4,100 m a.s.l.) in Chiapas (e.g., Meneses-Rocha, 2001; Witt et al., 2012). Both events, and especially the Chiapanecan orogeny, caused extensive erosion and abundant terrigenous sediment supply from the Chiapas territory to the north into the Tabasco plain (Sánchez-Montes de Oca, 1979, 2006; Quezada-Muñetón, 1987; Meneses-Rocha, 2001; Padilla y Sánchez, 2007). In terms of regional tectonics, the middle-late Miocene Chiapanecan event is most likely controlled by the interaction of the North American, Caribbean, and Cocos plates (Meneses-Rocha, 2001; Padilla y Sánchez, 2007; Guzmán-Speziale, 2010; Authemayou et al., 2011; Witt et al., 2012).
Conclusions and recommendations
The new fission-track ages reported in this study, together with some previous data on the tectonic history of southeastern Mexico, strongly indicate that the CMC, as well as the whole Maya block, have experienced a complex long-term tectonic evolution. For southeastern Mexico, at least five post-Permian tectonic and magmatic events may be proposed: (1) a Late Triassic (230-210 Ma) cooling event, probably linked to the initial breakup of Pangea; (2) Early Jurassic (from 200 to 190-180 Ma) magmatism that is related to the Nazas volcanic arc; (3) a Middle Jurassic tectonic event with an age of 170-160 Ma that was likely triggered by continental rifting at the beginning of the opening of the Gulf of Mexico; (4) a Late Cretaceous-Paleocene orogenic event, which can actually represent the southernmost continuation of the Laramide sensu lato that affected central and northern Mexico; and (5) the middle to late Miocene Chiapanecan orogeny that seems to be tectonically controlled by the interaction of the North American, Caribbean, and Cocos plates.
The tectonic history of the CMC is somewhat "entangled" as result of the debatable results and ideas proposed for its Cenozoic evolution (Ratschbacher et al., 2009; Witt et al., 2012; Abdullin et al., 2016a; this study). For example, Ratschbacher et al. (2009) suggested that the CMC region exhumed rapidly during the Oligocene, likely due to the hypothetical arrival of the Chortís block in front of the Gulf of Tehuantepec area. However, the results of recent studies (Witt et al., 2012; Abdullin et al., 2016a; this study) indicate that, at least, the central and northern portions of the CMC display no abrupt exhumation signals for this period of time. This controversy is due to the absence of systematic and detailed studies along the entire CMC. Additional thermochronological studies are also recommended. These integrated studies should be based on those methods that have closure temperatures below 300 °C such as apatite and zircon fission-track analyses and helium dating, because other techniques with higher closure temperatures will most likely yield either Triassic or Jurassic ages, which are typical for the CMC. Also, these studies should be performed across several sections of the CMC, i.e., in the northern, central, and southern CMC, because the southernmost CMC may store Oligocene cooling ages sensu stricto. Furthermore, these studies can be accompanied by provenance analyses in specific clastic successions, in particular from Cenozoic units that crop out within the central and northern areas of the Sierra de Chiapas.
