Introduction

Orogenic eclogites record key evidence of subduction-related processes and serve as important markers to identify paleo-subduction zones1,2,3. This type of eclogites occurs in two different convergent-margin regimes, namely the Pacific- and collision-type orogens. Pacific-type eclogites form along colder geotherms compared to collision-type eclogites, reach slightly lower peak pressures (2.0–2.3 GPa) than collision-type eclogites (> 2.3 GPa), and exhume at different rates than collision-type eclogites3,4. Therefore, constraining the pressure and temperature (P–T) evolution of eclogites is crucial for characterizing orogens, and specifically, for quantifying their burial—exhumation cycle.

The Acatlán Complex, southeastern Mexico (Fig. 1), exposes Paleozoic metamorphic rocks of various grades, including high-pressure (HP) lithologies such as chloritoid–rutile–phengite micaschists, garnet–rutile–phengite orthogneisses, garnet–epidote blueschists, and amphibole-phengite eclogites5. Eclogites in the Acatlán Complex represent the deepest subducted portions of the complex. A recent study6 reported maximum subduction depths of ~ 70 km—i.e. ~ 22–38 km higher than previously suggested for the Complex7,8,9,10,11,12, questioning accepted geodynamic models and exhumation rates for the Acatlán Complex13,14.

Figure 1
figure 1

(a) Tectonostratigraphic terranes of southern Mexico. (b) Geological map of the Acatlán Complex. The green stars represent eclogite localities with the Piaxtla Suite (approximate locations). 152—Ortega-Gutiérrez7; MI6 and MP3—Meza-Figueroa et al.8; ACA7, ACA8, and RAC148—Vega-Granillo et al.10; EC-1—Hernández-Uribe et al.6; TQF—Tetla-Quicayán Fault; CF—Caltepec Fault. Modified from Hernández-Uribe et al.6.

Full size image

In this contribution, I re-evaluate the eclogite-facies metamorphism in the Acatlán Complex. For this, I calculated eclogite P–T conditions from different localities within the Acatlán Complex to evaluate whether the complex indeed experienced deeper subduction than previously thought or that this finding represents a single deep exposure within the Acatlán Complex. Finally, I discuss the implications of these results for the subduction and exhumation rates of the Acatlán Complex during the Paleozoic as well as for eclogite thermobarometry.

Geological context

The Acatlán Complex, southern Mexico, is the largest exposure (~ 10,500 km2) of Paleozoic metamorphic rocks in Mexico, and one of the two localities in the country where eclogites and eclogite-facies rocks have been described so far5,15.

The Acatlán Complex is a fault-bounded crystalline basement, bounded to the north by the Trans-Mexican Volcanic belt (Cenozoic), to the east by the Oaxacan Complex (Mesoproterozoic), to the south by the Xolapa Complex (Mesozoic), and to the by with the Guerrero-Morelos platform (Mesozoic)15,16,17,18 (Fig. 1). Since the work of Ortega-Gutiérrez15, the tectonothermal evolution and stratigraphy of the polymetamorphic Acatlán Complex have been the subject of debate5,19,20,21,22. Complications arise as the Acatlán Complex may record distinct orogenic cycles associated with opening and closure of the Iapetus, Rheic and Paleo-Pacific oceans20,21,22,23 leading to different subdivisions of the Complex, obscuring our understanding of its evolution.

This study focuses on the Piaxtla Suite (Fig. 1)—the HP belt of the Acatlán Complex where eclogite-facies metamorphism has been described. Details of the subdivision of the Acatlán Complex are described elsewhere5,19,20,21,22. The Piaxtla Suite is a N–S trending HP belt comprised of metasediments, metabasites, metagranitoids, and serpentinized ultramafic bodies, which record variable metamorphic grade5,12,15,23,24 (Fig. 1). Different areas within the Piaxtla Suite record peak blueschist- to eclogite-facies conditions, and exhumation through the amphibolite and greenschist facies5,6,7,8,10,11,15.

Acatlán eclogites

Eclogites sensu lato (the word eclogite here refers broadly to variably retrogressed mafic rocks with garnet, omphacitic clinopyroxene, and rutile) in the Piaxtla Suite crops out near the towns of Mimilulco, Las Minas, Santa Cruz Organal-San Francisco de Asís, and Piaxtla (Fig. 1). Eclogites display a peak mineral assemblage of Grt + Omp + Rt ± Ph ± Czo/Zo ± Qz ± Amp6,7,8,9,10,14; this assemblage is replaced by amphibolite- (Amp + Pl + Ttn ± Czo/Zo) and greenschist-facies (Amp ± Chl) retrograde assemblages during exhumation6,7,8,9,10,14. Existing P–T conditions calculated for eclogites (including variably retrogressed samples) vary depending on the area within the Piaxtla Suite. Previous works suggest P–T conditions of 1.1–1.5 GPa and 560 ± 60 °C in the area of Mimilulco8. In the Santa Cruz Organal area, eclogites record 1.5–1.7 GPa and 768–830 °C10, and in the Piaxtla area P–T conditions of 1.1–1.3 GPa and 491–609 °C7,8,10. A new eclogite locality was reported by Hernández-Uribe et al.6 west of Piaxtla, and considered to record peak metamorphic conditions of ~ 2.2 GPa and ~ 690 °C. Other authors have reported “eclogite-facies” P–T conditions in amphibolites9,11, that lack omphacitic clinopyroxene, garnet, and/or rutile, and in non-mafic lithologies such as rutile-bearing mica schists12 and HP granitoids5,23.

Lu–Hf garnet–whole-rock geochronology from an amphibolitized eclogites in the Piaxtla area indicates that a single eclogite-facies metamorphism of the suite took place c. 351–353 Ma14, consistent with: (a) a Sm–Nd garnet–whole rock age of c. 388 ± 44 Ma25 (sample near town of Xayacatlán); (b) U–Pb zircon ages in retrogressed eclogites and amphibolites (interpreted as former eclogites) in the Piaxtla26 and San Francisco de Asís areas9; and (c) with amphibole 40Ar/39Ar ages of c. 342–344 from an amphibolite (“retrogressed eclogite” from the area of Piaxtla)13 and of c.336 ± 6 Ma of an eclogite from the Piaxtla area10. Other studies, however, suggest that the Acatlán Complex records more than one eclogite facies event10,19,22. An amphibole 40Ar/39Ar age of c. 430 ± 5 Ma reported from a retrogressed eclogite in the Santa Cruz Organal area was interpreted to date eclogite-facies metamorphism, followed by a c. 374 Ma 40Ar/39Ar phengite age in the same sample, interpreted to date cooling during exhumation10; however, the interpretation of the 40Ar/39Ar ages is disputed, as the samples display complex Ar spectra suggesting Ar excess or inheritance14.

Thermobarometry of Acatlán eclogites

To re-explore the eclogite-facies P–T conditions of the Piaxtla Suite in the Acatlán Complex, I applied thermobarometric methods suitable for eclogite-facies mafic rocks. There are four published studies (to my knowledge) with available compositions of garnet, omphacitic clinopyroxene, and phengite in the Piaxtla Suite, which result in three eclogite localities in the Acatlán Complex (Fig. 1). These include the Piaxtla area with samples 1527, MP38, ACA710, ACA810, and EC-16. Towards the north, localities include Mimilulco, with sample MI68 and Santa Cruz Organal, with eclogite RAC14810. The study of Middleton et al.9, from the San Francisco de Asís area (near Santa Cruz Organal; Fig. 1), was not included here because the published clinopyroxene (inclusion in amphibole) composition is a non-omphacitic clinopyroxene.

Pressures were calculated using the garnet–clinopyroxene–phengite barometer of Ravna and Terry27. Temperatures were obtained using the garnet–clinopyroxene thermometer calibrations of Ravna28, Nakamura29, and Sudholz et al.30. While some of these thermobarometers have been available for several decades, their application for the Acatlán eclogites is still novel. For internal consistency, the reported results in the text and in Fig. 2 correspond to the calculated mean of the intersections between the Ravna and Terry27 barometer and Ravna28 thermometer (Table 1). Temperatures obtained from the Nakamura29 and Sudholz et al.30 calibrations are also given in Table 1. Details of the methodology and related uncertainties are provided in the Methods section.

Figure 2
figure 2

Pressure and temperature (P–T) conditions of Piaxtla Suite eclogites. (a) Piaxtla area and (b) Mimilulco and Santa Cruz Organal areas. The colored stars represent the mean P–T conditions and the error bars indicate the ± 0.2 GPa and ± 60 °C uncertainty associated to the thermobarometric methods (see Methods). The shaded colored rectangles with the sample number in italics correspond to the previously calculated P–T conditions for the same samples used in this study. The colors of the stars and boxes refer to: 152 (green)—Ortega-Gutiérrez7; MP3 and MI6 (blue)—Meza-Figueroa et al.8; ACA7, ACA8, and RAC148 (orange)—Vega-Granillo et al.10; EC-1 (yellow)—Hernández-Uribe et al.6.

Full size image
Table 1 Calculated P⎼T conditions of Acatlán eclogites.
Full size table

Different eclogites from the Piaxtla area yield similar P–T conditions (Fig. 2a). Eclogite MP3 yields P–T conditions of 1.97 GPa and 555 °C, eclogites ACA7 and ACA8 yields conditions of 1.83 GPa and 519 °C and 1.94 GPa and 565 °C, respectively. Eclogite 152 yields P–T conditions of 1.94 GPa and 468 °C (Fig. 2a). Eclogite EC-1, from the new locality west to the area of Piaxtla, yields 2.3 GPa and 675 °C (Fig. 2a), the greatest of all the Acatlán Complex.

In contrast to the Piaxtla eclogites, the Mimilulco and Santa Cruz Organal eclogites yield contrasting temperature estimates (Fig. 2b); unfortunately, there are no phengite analyses from either localities, thus no new pressure estimates were calculated. The Mimilulco eclogite MI6 yields a garnet–clinopyroxene temperature of 511 °C at 2.0 GPa (only one garnet–clinopyroxene pair available8). By contrast, the Santa Cruz Organal eclogite RAC148 yields a temperature of 870 °C at 2.0 GPa (Fig. 2b).

Discussion

Comparison with previous studies

Temperatures calculated in this work are similar to previously published estimates for eclogites in different parts of the Acatlán Complex6,7,8,10 (Fig. 2). However, our barometric calculations show substantially—and systematically—higher pressures than previously calculated across the Acatlán Complex (Fig. 2). For instance, in the Piaxtla area, the calculated pressure is ~ 2.0 GPa for four different samples (eclogite 152, MP3 ACA7, and ACA8; Fig. 2) whereas previous studies7,8,10 suggested that the eclogite-facies metamorphic event in this area occurred at ~ 1.1–1.5 GPa; such estimates are different even when considering the ± 0.2 GPa uncertainty related to the barometer (see Methods).

Unfortunately, the lack of phengite in the eclogites from the Mimilulco and Santa Cruz Organal areas precluded the recalculation of new pressures with the approach used in this work. Yet, there is no reason for why the 1.1–1.5 GPa for the Mimilulco eclogite and 1.5–1.7 GPa for the Santa Cruz Organal eclogite could not be higher than previously reported. For example, Meza-Figueroa et al.8 suggest that the Mimilulco eclogite (MI6) was metamorphosed at the same P–T conditions than eclogite MP3 from Piaxtla8; thus the Mimilulco eclogite could also record pressures of ~ 2.0 GPa. By contrast, it is more challenging to infer a pressure for the Santa Cruz Organal eclogite, as P–T conditions are not available for other eclogites near the area. Previous work in the San Francisco de Asís area (relatively near Santa Cruz Organal; Fig. 1) estimated pressures > 1.6 GPa9, similar to that calculated by Vega-Granillo et al.10 for the Santa Cruz Organal eclogite (1.5–1.7 GPa), but with contrasting temperatures (650–750 °C9 vs 768–830 °C10). The temperature mismatch between these studies may be explained either by the fact that eclogites in both localities experienced different P–T conditions or due to the use of non-omphacitic clinopyroxene for the thermobarometry9. The fact that the Santa Cruz Organal eclogite records the highest temperature in the Acatlán Complex (Fig. 2 and Table 1) may suggest that the pressure could be different to other eclogites in the Piaxtla Suite as well.

A recent petrologic study combining phase-equilibrium modeling and Zr-in-rutile thermometry for an eclogite from a new locality west of Piaxtla area obtained conditions of ~ 2.2 GPa and ~ 690 °C6. Our thermobarometric calculations for the same eclogite sample (EC-1) yield similar P–T conditions than previously calculated (Fig. 2; Table 1). Importantly, the agreement between the previous P–T calculations from Hernández-Uribe et al.6 with the conditions obtained here, further support our findings for the other eclogites in different parts of the Acatlán Complex.

Implications for the geodynamic evolution

The difference between the previously reported pressures for all the complex and the contrastingly higher pressure in the new eclogite locality was interpreted to be an artifact due to differences in thermobarometric methods6. However, here, I obtained similar pressures from different parts of the Piaxtla Suite of the Acatlán Complex using conventional thermobarometric approaches (Fig. 2). Therefore, I argue that these new results indicate systematic deeper subduction than previously thought. If no errors are considered, the calculated pressures of ~ 1.9–2.3 GPa, and corresponding inferred depths (63–75 km; see methods for pressure-to-depth conversion) suggest that different areas within the complex record slightly different depths. On the other hand, if the uncertainties in the calculations are considered (± 0.2 GPa, ± 6–7 km), then the calculated pressures and inferred depths in this work converge suggesting the complex reached a similar depth during subduction.

Temperatures from the Piaxtla and Mimilulco areas are the same considering the ± 60–100 °C uncertainty related to the thermometric calculations. However, the temperature calculated here and in a previous work10 for the Santa Cruz Organal eclogite indicate that this area records the highest temperature of any rocks in the Acatlán Complex (Fig. 2). The differences in calculated temperatures could suggest different locations of the proto-Acatlán Complex within the subducting slab (i.e., hotter towards the slab top vs colder towards the bottom). Regardless of the temperature interpretation, the greatest depths obtained here situates the subducting proto-Acatlán Complex deeper than previously hypothesizes by all the tectonic models for the region.

The greater pressures–depths calculated here for the eclogites imply a faster subduction–exhumation cycle for the Acatlán Complex. Simple tectonic-rate calculations for the Acatlán Complex were obtained by using: (a) the youngest depositional ages of the sediments above the mafic oceanic crust; (b) the greatest depth reached during subduction and related eclogite-facies age (considering a single event); as well as (c) the depth and time of exhumation. For calculating the burial rate, I use the youngest detrital zircon in a metapsammite in the Piaxtla Suite with an age of c. 365 Ma interpreted to represent the youngest depositional limit13. Coupled with the depth from this work (i.e. 75 km) and an eclogite-facies age of c. 353 Ma (Lu–Hf garnet–whole-rock13), I obtained a linear burial rate of ~ 6.3 mm/yr, well within the estimates of convergence rates of tectonic plates in subduction zones31. Furthermore, the eclogite-facies data obtained here combined with muscovite 40Ar/39Ar cooling age of c.334 Ma in a retrogressed eclogite13 with amphibolite-facies P–T conditions of ~ 0.6 GPa8 equates to an exhumation rate of ~ 2.8 mm/yr, similar to other HP terranes worldwide32. In summary, these simple calculations indicate it took the proto-Acatlán Complex ~ 12 Myr to subduct to ~ 75 km depth, and ~ 19 Myr to return to crustal depths, resulting in a subduction–exhumation cycle of ~ 31 Myr. These calculations contrast and thus challenge models for the Acatlán Complex with slower subduction and exhumation rates. For example, a previous burial rate of 2.7 mm/yr13 and an exhumation rate of 2.4 mm/yr13 are 3.6 mm/yr and 0.8 mm/yr slower, respectively, than the ones calculated here. The discrepancy between the calculated burial–exhumation cycle may be explained by the input data, as the burial and exhumation rates are strongly dependent in the timing of both the formation of the eclogite protolith and the exhumation to crustal depths. However, regardless of these data, the thermobarometric and new depth calculations obtained in this work would unequivocally result in faster tectonic rates.

Implications for eclogite thermobarometry

The results presented here indicate that for the Acatlán eclogites, conventional thermobarometric methods, phase-equilibrium modeling6, and Zr-in-rutile thermometry6 yield consistent P–T conditions (Fig. 2a). While the uncertainties related to conventional thermobarometric methods (see Methods section) are considerably larger than the ones from other methods33, I argue that in relatively well-equilibrated rocks, P–T estimates should be similar. Thus, as shown by other studies2,34,35, the obtention of reliable P–T data needs to involve the application of different thermobarometric methods.

The temperatures obtained using different calibrations are the same including uncertainties (Table 1). The calibration from Sudhloz et al.30 yields the highest temperatures compared to the other calibrations, whereas the Nakamura29 calibration tend to yield the lowest temperature of all the thermometers (Table 1). Importantly, the Sudhloz et al.30 calibration was parametrized from high-temperature experiments of mantle lithologies, potentially explaining why such temperatures are the highest. Yet, the mineral compositions from Acatlán eclogites are within the ranges recommended for that calibration (Supplementary Tables S1–S3). Further, from all the calibrations used here, only the Sudhloz et al.30 thermometer includes a correction for the jadeite content in clinopyroxene, which is key for yielding reliable temperatures for subduction-related eclogites36,37.

The comparison between the calculated temperatures in this work and the study of Hernández-Uribe et al.6 (which used Zr-in-rutile thermometry) indicates that the Sudhloz et al.30 thermometer yields almost identical temperatures than the Zr-in-rutile thermometer (695 °C6 vs 688 °C). By contrast, such Zr-in-rutile temperatures differ the most from the Nakamura29 temperature (695 °C6 vs 643 °C). Therefore, the comparison between independent approaches seems to suggest that the Sudhloz et al.30 calibration may yield more reliable temperatures for subduction-related eclogites than the other Fe–Mg garnet–clinopyroxene thermometers.

Methods

Barometric calculations were done using the garnet–clinopyroxene–phengite barometer with the calibration of Ravna and Terry27. The garnet–clinopyroxene–phengite barometer relies in the net transfer reaction between garnet, clinopyroxene, and phengite (mineral abbreviations follow Warr38):

$$ 6Di + 3Ms = 2Grs + 1Prp + 3Cel $$
(1)

where the equilibrium constant (K1) of this reaction can be expressed as:

$$ K_{1} = \frac{{(a_{Prp}^{Grt} )(a_{Grs}^{Grt} )^{2} (a_{Cel}^{Ph} )^{3} }}{{(a_{Di}^{Cpx} )^{6} (a_{Ms}^{Ph} )^{3} }} $$
(2)

Temperatures were calculated using the garnet–clinopyroxene thermometer using the Ravna28, Nakamura29, and Sudholz et al.30 calibrations. The garnet–clinopyroxene thermometer relies on the exchange of Fe2+ and Mg between garnet and clinopyroxene. The equilibrium Fe2+–Mg distribution coefficient (KD) can be expressed as:

$$ K_{D} = \frac{{\left( {\frac{{Fe^{2 + } }}{Mg}} \right)^{Grt} }}{{\left( {\frac{{Fe^{2 + } }}{Mg}} \right)^{Cpx} }} $$
(3)

Uncertainties related to the conventional thermobarometers applied here are commonly quoted to be ± 0.2 GPa for the barometer27,39,40 and ± 60 °C for the thermometer27,28,37. For the latter, temperatures can be up ± 100 °C due to the Fe3+ estimation in clinopyroxene41. In this work, Fe3+ in clinopyroxene calculated with the following relation: Fe3+ = Na–Al–Cr. This Fe3+ recalculation scheme was used instead of the stochiometric Fe3+ as the latter resulted in unrealistic lower garnet–clinopyroxene temperatures (< 350 °C).

For the P–T calculations, published garnet, clinopyroxene, and phengite chemical compositions6,7,8,10 were used from eclogites distributed along different portions of the Piaxtla Suite within the Acatlán Complex (Fig. 1; Supplementary Tables S1–S3). When provided, available petrological context in the publications (e.g., rims vs core and/or interpretations of peak vs retrograde), were considered for the P–T calculations. For samples MP38, MI68, and EC-16, all the mineralogical data come from such papers. Data for eclogite 1527 was partially published by Ortega-Gutiérrez7; the complete analyses were kindly provided by the author. Similarly, data for ACA7, ACA8, and RAC148 was partially published by Vega-Granillo et al.10. Complete analyses were obtained from that author’s doctoral dissertation. In this case, we only picked a pair of each mineral for the P–T calculations. All mineralogical analyses used for the thermobarometric calculations are given in Tables S1–S3 in the Supplementary Material.

Pressure-to-depth conversion uses a layered model assuming a total crustal thickness of 30 km, where the upper crust is 20 km and has a density of 2.8 g/cm3, and where the lower crust is 10 km and has a density of 2.9 g/cm3. The crust is followed by an upper mantle with density of 3.3 g/cm3. A greater crustal thickness results in greater subduction depths, whereas changes in the considered densities would have minor effects on the overall calculated depth. Tectonic overpressure was not considered in our pressure-to-depth calculations but deviation from lithostatic pressure is likely within the order of the depth uncertainty related to the barometric method used here42.