Geological framework

Our study area, located east of Arivechi, has been subjected to several studies of stratigraphy and paleontology and, to a lesser extent, structural geology. We defined two units of Mesozoic age: 1) Cañada de Tarachi Unit, and 2) El Potrero Grande Unit. A third unit is of Oligocene-Miocene age, the Sierra Madre Occidental volcanic sequence. The two Upper Cretaceous units constitute a volcano-sedimentary sequence of more than 6 km thickness. El Potrero Grande Unit (the youngest Cretaceous sequence) consists of conglomerate, sandstone, siltstone, and shale with interbedded andesite and diorite dikes and layers of rhyolitic tuffs. These rocks constitude a column more than 2,400 m thick, but this is merly part of a thicker sequence. The oldest unite include a large number of unusually well-exposed megaclasts of different lithogies and ages. A megaclast is a term used in grain-size schemes after modification of the Udden-Wentworth sedimentary grain-size scale devised by ^{Blair and McPherson (1999)}, who proposed a new fraction in sediment size to account for clasts larger than 4.1m. They proposed four sizes for the megaclast fraction: block (4.1 to 65.5 m), slab (65.5 to 1049 m), monolith (1 to 33.6 km), and megalith (d _I from 33.6 to 1075 km). Megaclasts are associated to gravity sliding, or to vertical movements as described by various authors as olistoliths and olistostromes (^{Teale and Young, 1987}; ^{Heubeck, 1992}).

Megaclasts have also been reported from numerous other tectonic environments; the Pyrenées (^{Johns et al ., 1981}), the Canadian Rockies (^{Cook et al ., 1972}), eastern Australia (^{Conaghan et al ., 1976}), the western US Cordillera (^{Heck and Speed, 1987}), the Apennines (^{Naylor, 1982}; ^{Teale and Young, 1987}), the Avalonían Terrane of New England (^{Bailey et al ., 1989}), Hispaniola (^{Heubeck, 1992}), global recent continental slopes (^{Prior et al ., 1982}), Alaska (^{Harp et al ., 2003}), the Valles-San Luis Potosi Platform (^{Carrasco-Velázquez., 1977}), and north-eastern Sonora (^{McKee and Anderson, 1998}; ^{McKee et al ., 2005}).

Megaclasts (monoliths and blocks) as described by ^{Blair and McPherson (1999)}, have been found and are common in eastern Sonora Upper Cretaceous sequences (^{McKee and Anderson, 1998}; ^{Rodríguez-Castañeda, 2002}; and ^{McKee et al ., 2005}). These megaclasts have been linked to the emergence of a positive land, possibly related to the Late Jurassic Aldama Platform (Ramirez and Acevedo, 1957; ^{Monreal, 1996}; ^{Haenggi, 2002}; ^{Anderson and Nourse, 2005}) (Figure 2). The Aldama Platform is defined as an emerged land oriented NW-SE in north-west Chihuahua and eastern Sonora. Its extension toward north-eastern Sonora could be the Cananea High (^{McKee, 1991}; ^{Rodríguez-Castañeda, 2002}). The Aldama Platform is bounded by faults related to the platform uplift. Uplift has been suggested as a mechanism of megaclast generation. It has been attributed to different causes but mainly to extension by reactivation of old faults and early Cenozoic pluton emplacement (^{Rodríguez-Castañeda, 2002}; Roldán-Quintana, 2002). Epiclastic, pyroclastic and reworked deposits have also been identified in the study area. Flows and pyroclastic rock falls, mass waste deposits and volcaniclastic turbidites are common in the sequence deposited in the Arivechi back-arc basin. Towards the top of the unit the finest sediments increase in proportion, marking the latest stage of the evolution of this type of basin (^{Spalletti, 2006}).

Figure 2 Map showing the paleogeography and structural elements that controlled the geologic evolution in the Upper Cretaceous. The Aldama Platform and its possible continuation towards the north-east of Sonora as the Cananea High were important elements in the evolution of the study area. The black arrows show the occurrence of slipped megaclasts and the gray arrows suggest the possible source of the sediments that constitute the rocky outcrops of the Upper Cretaceous in the study area (square). Megaclast localities: Los Ajos (LA), El Tigre (ET), Magdalena (MA), San Antonio (SA), La Madera (MD), Lampazos (LA), Los Chinos (LC) and Arivechi (AR), msm = Mojave-Sonora megashear, saf = San Antonio fault, laf= Los Ajos fault. Modified from Anderson and Nourse (2005) including Monreal (1996). 

U/Pb Method

The geochronology was carried out at the Centro de Geociencias Isotopic lab, UNAM. The method of dating is as follows. Laser ablation inductively-coupled plasma mass spectrometry (LA-ICPMS) U-Pb analyses were performed at the Laboratorio de Estudios Isotópicos (LEI), Centro de Geociencias, UNAM, employing a 193 nm excimer laser workstation (Resolution M-050) coupled with a Thermo X-ii quadrupole ICPMS. The protocol reported by ^{Solari et al . (2010)} was used, employing a 23 μm analytical spot and the Plešovice zircon (^{Slama et al ., 2008}) as bracketing standard. Time-resolved analyses were then reduced off-line with in-house developed software written in R (^{Solari and Tanner, 2011}), and the output was then imported into Excel, where the concordia as well as age-error calculations were obtained with Isoplot v. 3.70 (^{Ludwig, 2008}).

During the analytical sessions in which the data presented in this paper were measured, the observed uncertainties (1sigma relative standard deviation) on the ²⁰⁶Pb/²³⁸U, ²⁰⁷Pb/²⁰⁶Pb and ²⁰⁸Pb/²³²Th ratios measured on the Plešovice standard zircon were 0.75, 1.1 and 0.95% respectively. These errors were quadratically added to the quoted uncertainties observed on the measured isotopic ratios of the unknown zircons. This last factor takes into account the heterogeneities of the natural standard zircons. Provided that the isotope ²⁰⁴Pb, used to correct for initial common Pb, was not measured (because its tiny signal is swamped by the ²⁰⁴Hg normally present in the He carrier gas), the common Pb was thus evaluated with the ²⁰⁷Pb/²⁰⁶Pb ratio, and all the analyses carefully graphed on ^{Tera and Wasserburg (1972)} diagrams. Correction, if needed, was then performed with the algebraic method of ^{Andersen (2002)}. In figures, tables and results ²⁰⁶Pb/²³⁸U ages are used for zircons < 1.0 Ga, whereas ²⁰⁷Pb/²⁰⁶Pb ages are cited for older grains. The TuffZirc algorithm of ^{Ludwig and Mundil (2002)} was used to calculate the best mean of ²⁰⁶Pb/²³⁸U ages, which is preferred for estimating the apparent age of those young zircons in which the ²⁰⁷Pb signal is low and thus yields imprecise results. The ²⁰⁷Pb/²⁰⁶Pb ages were furthermore considered as minimum ages because of the effect of possible Pb loss. We sampled granitic megaclasts in the Cañada de Tarachi Unit.

⁴⁰Ar/³⁹Ar Method

A K-feldspar mineral separation from a selected rock sample was carried out by standard methods at the mineral separation laboratory at the Estación Regional del Noreste, Instituto de Geología, UNAM. Age analyses were carried out at Queen's University ⁴⁰Ar/³⁹Ar geochronology laboratory, Kingston, Canada. For each mineral separate, ∼5-10 mg samples of material were wrapped in Al-foil and stacked vertically into Al canisters, which were then irradiated in the McMaster University Nuclear Reactor in Hamilton, Canada, with a ⁴⁰Ar/³⁹Ar flux monitor Hb3gr hornblende (1072 ± 11 Ma [2σ]; ^{Roddick, 1983}). Following irradiation, the samples and monitors were placed in small pits, ∼2 mm in diameter, drilled in a Cu sample holder. The holders were placed inside a small, bakeable, stainless steel chamber with a Zn Se viewport connected to an ultra-high vacuum purification system. Monitors were fused in a single step; with a focused New Wave MIR-10 30 W CO₂ laser. For the step heating experiments, the laser beam was focused at different temperatures on mineral separates or whole rock samples in approximately 10 steps. The evolved gases were purified with an SAES C50 getter for ∼5 min. Argon isotopes were measured with a MAP 216 mass spectrometer, with a Bäur Signer source and an electron multiplier. All data were corrected for blanks, atmospheric contamination, and neutron-induced interferences (^{Roddick, 1983}; ^{Onstott and Peacock, 1987}). All errors are reported as ±2σ, unless otherwise noted, and dates were calculated using the decay constants recommended by ^{Steiger and Jäger (1977)}.

Paleostress analysis

The stress inversion method was applied with Win_Tensor software (^{Delvaux, 1993}). The inversion is based on the assumption that slip planes occur in the direction of the maximum resolved shear stress. The slip direction at a fault surface is inferred from the striations' slip-fiber lineations as a product of friction. Also, we determined the type of fault (normal or strike-slip) using drag folds that allowed us to discriminate faulting. Therefore, the strike and dip of the fault surfaces, the orientation of the lineation (pitch) and the sense of movement identified at the fault surface were the data used for the inversion.

The obtained fault slip data were used to compute the four parameters of the reduced stress tensor according to ^{Angelier (1994)}: the principal stress axes were S₁ (maximum compression), S₂ (intermediate compression) and S₃ (minimum compression), and the ratio of principal stress differences is R = (S ₂- S₃ ) (S ₁- S₃ ). The four parameters were determined by successive use of an improved version of the right dihedral method of ^{Angelier and Mechler (1977)} and a four-dimensional numeric rotational optimization using the Tensor computer program of ^{Delvaux (1993)}.

Cañada de Tarachi Unit

The oldest unit, known as the Cañada de Tarachi, a name taken from the arroyo of the same name, is located south-east of Arivechi (Figure 3).The Cañada de Tarachi Unit consists of a sequence of conglomerate, sandstone, siltstone, shale, and interbedded andesitic tuffs, with a significant presence of monoliths (^{Blair and McPherson, 1999}) comprised of Precambrian, Paleozoic and Mesozoic sedimentary and igneous rocks. The conglomerate beds in the sequence are composed of several lithologies located at different levels; some of them form a package over 100 m, but most are metric in thickness (Figure 4). The conglomerates originated as a debris flow. Some of the conglomerates in the lower level are of volcanic origin and rounded shape, well sorted, with clast size ranging between 2 and 10 cm, although a few of them measure 20 cm (Figure 4a and Figure 4b); the conglomerates derived from sedimentary rocks are generally angular in shape and unsorted (Figure 4c and Figure 4d).The whole Cañada de Tarachi Unit, on the other hand, records a strong deformation, which is believed to be synsedimentary and related to the gravitational movements of the megaclasts caused by vertical uplift and sliding, rather than the result of a compressive tectonic deformation.

Figure 4 Conglomerate in Cañada de Tarachi Unit showing variation in composition and poor sorting. Images a and b show a conglomerate consisting of rock clasts derived from volcanic rocks which are more or less spherical in shape. Images c and d show a conglomerate consisting mainly of angular-shaped sedimentary clast rocks in a quartz-rich matrix. Images e and f show a conglomerate exposed to the north of the area in Cerro Zoropuchi (f) and in the range to the right. Here, the conglomerate is entirely composed of both Paleozoic and Lower Cretaceous limestone rock rounded clasts. 

The Precambrian monolith is exposed at Cerro El Palmar (Figure 5a), south of arroyo Cañada de Tarachi. The monolith consists of layers of quartz sandstone and dolomite. It measures 7 km long and 2 km wide. The contact between rocks of the Neoproterozoic and the Cañada de Tarachi Unit are marked by a breccia over 10 m thick.

Figure 5 Megaclasts exposed in the Arivechi region. a) megaclast composed of quartzite and dolomite of Proterozoic age constituting Cerro El Palmar. In the forefront is the conglomerate that, together with the megablock, forms part of the Cañada de Tarachi Unit. dol = dolomite, qtz = quartzite, Ku = volcano-sedimentary rocks. b) Paleozoic limestone megaclast. North view. c) 76 Ma granite block (Ku) which is jammed in Lower Cretaceous rocks (Kl). Note the deformation in the rocks of the Lower Cretaceous (Ki). d) Megaclast of uncertain age embedded in the Cañada de Tarachi Unit. Outcrop along the arroyo Cañada de Tarachi. 

At the base of the monolith, near the contact with the Upper Cretaceous rocks, along the arroyo San Miguel an important finding was the presence of stromatolite fossils, possibly of the Jacutophyton genus, similar to those found in the Caborca region (^{Weber et al ., 1979}) assigned to the Neoproterozoic. The monoliths of Paleozoic rocks (Figure 5b) are 2 km long, and consist of fossiliferous limestone and quartz sandstone which are assigned to the Mississippian by their fossil contents (^{Fernández-Aguirre and Almazán-Vazquez, 1991}). These megablocks can be observed at Cerro Peñasco Blanco and Cerro Las Conchas.

At Cerro Las Conchas, another monolith composed of limestone, shale and sandstone and containing fossils from the Early Cretaceous (^{Fernández-Aguirre and Almazán-Vazquez, 1991}) (Figure 3) is exposed. ^{Bartolini (1993)} suggested that these megaclasts have better correlation with exposed units in western Chihuahua than with rocks of the same age cropping out in Sonora.

We found several megaclasts composed of granite, and from these we selected two for dating. One was collected in the arroyo Cañada de Tarachi and was dated by U/Pb zircon geochronology at 76.00 ± 2 Ma (Campanian, Late Cretaceous, Figure 6a and Figure 6b, Table 1), and a second sample around 100 m distant was collected at the Arivechi-Tarachi road and dated by ⁴⁰Ar/³⁹Ar step-heating of a k-spar at 69.57 ± 0.48 Ma (Figure 6c, Table 2). It was noted that these megaclasts are not only found on top but also within the sequence, in smaller thicknesses of dozens of meters. The fact that these are glided fragments, and because of their size were not previously identified as megaclasts, was the main reason for assuming that the stratigraphy was complicated and incorrectly interpreted by previous authors. Hence, we consider that the exposures of Cerro Las Conchas and the Paleozoic limestones are actually gliding megaclasts that are not "in situ."

Figure 6 U/Pb zircon data from the granite megablock in the Cañada de Tarachi. a) Concordia curve for the A134-12 sample and b) Weighted mean age graph at 76 Ma.c) Ar/Ar data from the Arivechi-Tarachi road. 

Table 1 U-Pb analytical data for zircons from sample in eastern Sonora, Arivechi region. 

Table 2  ⁴⁰Ar/³⁹Ar analytical data from sample in eastern Sonora, Arivechi region. 

No fossils have been identified in the Cañada de Tarachi Unit, apart from those that are contained in the slipped blocks. This statement has genetic significance for the sedimentary environment. The absence of fossils could be interpreted as resulting from the development of these basins inland. In terms of the thickness of this unit, based on the conducted mapping and strike and slip directions, we estimated that it must have a minimum thickness of 2000 m.

Fossil studies carried out on the conglo-merate clasts indicate Paleozoic and Early Cretaceous ages; therefore the Cañada de Tarachi Unit is situated in the Upper Cretaceous sequence. The 76 Ma U/Pb zircon date for the granite block at the Cañada de Tarachi Unit was interpreted as the crystallization date and the 70 Ma ⁴⁰Ar/³⁹Ar date was interpreted as the cooling age, i.e., postdating a Late Cretaceous age and suggesting that the onset of erosion was probably present by that time.

The base of the Cañada de Tarachi Unit is not exposed, but from correlation with other localities it has been interpreted as an angular unconformity unit over older rocks; in addition, the upper contact with the El Potrero Grande Unit is transitional to a younger volcaniclastic sequence.

El Potrero Grande Unit

This unit takes its name from the ranch located south-east of Cerro El Batamote (Figure 3). ^{King (1939)} defines this as the Potrero For-mation, which seems to correspond to the same location, although he assigns it to the Albian. The El Potrero Grande Unit is much more of a volcano-sedimentary sequence than the Cañada de Tarachi Unit. The El Potrero Grande Unit consists of conglomerate, sandstone, siltstone, and shale with interbedded andesite, diorite strata dikes and layers of rhyolitic tuffs. These rocks together form a partial column more than 2,400 m thick (Figure 7). It must be said that we did not measure the entire El Potrero Grande Unit, because it is covered by Tertiary volcanic to the east.

Figure 7 A 2,400 m measured column on a segment of the El Potrero Grande Unit, from the Cerro El Volantín to the Cerro El Batamote. Pictures go from top to bottom, i.e., from a to d. The bold letters in the lithologic column show the approximate location of the picture in the column. Compare the structures observed in the El Potrero Grande Unit with structures recorded in the Cañada de Tarachi Unit in Figure 9. 

The base of the unit consists of a section of conglomerates and volcanic rocks of andesitic composition that includes tuffs, continuing with shale, siltstone and sandstone that form a rhythmic sequence.

Isotopic K/Ar ages in biotite concentrates from volcanic rocks provided dates of 83.4 ± 4.17Ma (^{Pubellier, 1987}), 84 and 86 Ma K-Ar, respectively, for two tuffs in the upper part of the sequence, indicating an Upper Cretaceous age for the sequence (^{Grajales-Nishimura et al ., 1990}). In the Sahuaripa-Natora road we dated a tuff in a similar section 20 km north of the area. The age sample was dated as 76.3 ± 1.98 Ma (U/Pb zircon at Geociencias isotopic Lab), placing it in the Campanian (Upper Cretaceous). We consider this tuff equivalent in age to tuffs in our study area.

The Upper Cretaceous in Arivechi, eastern Sonora was deposited in a deep sedimentary basin that received sediments from an existing volcanic arc and also accumulated sediments from both the exposed north-western and north-eastern basement.

Normal faulting and paleostress reconstruction

To explain the tectonic processes from which the development of these structures originated, a paleostress field reconstruction is offered. The paleostress analysis was performed with measured surface data collected over faults in the Upper Cretaceous rocks along the Arivechi-Tarachi road, in the section between Cerro Las Conchas and Cerro El Volantín and along the arroyo Cañada de Tarachi. The data sites are located in sedimentary rocks as much as in igneous rocks. The displacement over the measured faults is generally from centimeters up to meters, and the faults are normal, although some of them record a strike-slip component. In general, all the measured faults display good striations. The results obtained from the paleostress analysis are in shown in Table 3 and in Figure 10 and Figure 11.

Figure 10 Paleostress reconstructions for the rocks of the Cañada de Tarachi Unit. Circle = S₁; Triangle = S₂; Square = S₃. 

Figure 11 Paleostress reconstructions in the El Potrero Grande Unit. Circle = S₁, triangle = S₂, square = S₃. 

The evolution of the stress fields is shown in structural maps (Figure 12 and Figure 13) with symbols indicating the two horizontal S_Hmax and S_hmin, with black inward arrows for compressive stress and white outward arrows for extensive stress. The length depends on the magnitude of the relative stress in function of the stress relation R and orientation of the principal stress directions. The vertical axis is indicated by an open circle for a compressive regime (vertical extension), by a dot for a strike-slip regime, and by a black circle for an extensive regime (vertical compression). The classifications proposed by ^{Delvaux et al . (1995)} for the stress tensors are: radial / pure / strike-slip extensive, extensive / pure / compressive strike-slip or strike slip / pure / radial compressive based on the relative magnitude of the intermediate axis and given by the stress relation R (Figure 14).

Figure 12 Orientation of the horizontal maximum and minimum stress in the Cañada de Tarachi Unit. The square shows the location of measured sites. Black arrow = S₁, white arrow = S₃. The central circle corresponds to S₁ when black and S₂ when gray. 

Figure 13 Orientation of the horizontal maximum and minimum stress in the El Potrero Grande Unit. The square shows the location of measured sites. Black arrow = S₁, white arrow = S₃. The central circle corresponds to S₁ when black and S₂ when gray. 

Figure 14 Types of stress regimes and their representation on a map. Arrows indicate the azimuth of horizontal stress axes, with their length according to relative stress magnitude, in base of the stress ratio R. White outward arrows indicate extensive stress axes and black inward arrows compressive stress axes. Vertical stress axes are symbolized by a solid circle for extensive regimes (S₁ vertical), a dot for strike-slip regimes (S₂ vertical) or an empty circle for compressive regimes (S₃ vertical). 

The orientation of the stress field was determined with the measured structures which show a clear sense of movement. Twenty-six sites were examined in the studied sector. The sites are located in outcrops of Upper Cretaceous rocks. The results of the inversion (Table 3) and their correlation allow the definition of several paleostress groups. The tensors are grouped according to their similarity, based on the direction of the main axes S_Hmax, S_hmi, the stress regime index and the stratigraphic relation.

The faulted sediments, which belong to the Cañada de Tarachi Unit, dip towards the north-east. Almost all the faults are normal with oblique-slip motions predominating, but dip-slip and strike-slip are present and play an important role. Faults in Figure 13 show that the motion of these faults was induced by extension along a direction of S₃ NW-SE, NE-SW and E-W for dip-slip normal faults; whereas S₁ is in general NNE-WSW and ENE-WSW in relation to strike-slip faults (Figure 10).

On the other hand, at the El Potrero Grande Unit faults are dip-slip, oblique slip normal faults and strike-slip faults. The direction of S₃ is NE-SW for dip-slip normal faults; NE-SW, almost N-S, and NW-SE for the oblique normal faults, and the S₁ directions E-W, NW-SE, and WNW-ESW are related to strike-slip faults (Figure 11).

Evolution

The studies carried out in the area indicated the existence of two Cretaceous sequences differentiated by their composition and stratigraphic position. The observed structures are mesoscopic folds in the Cañada de Tarachi Unit and the development of normal faults and smaller proportion by strike-slip faults.

The Cañada de Tarachi Unit is characterized by intercalations of conglomerate, sandstone, siltstone, rhyolitic tuff and andesite, showing a strong deformation. The geographic extension of Precambrian, Paleozoic and Lower Cretaceous monoliths or remnant rocks suggests that they belong to a basement possibly located towards the east, in the State of Chihuahua. The fold vergence (Figure 8) indicates that the sequence was transported towards the south-west.

The directions of stress that we were able to determine from the Cañada de Tarachi Unit from normal oblique, normal dip-slip and strike-slip faulting are shown in Figures 10 and 12. Local results are given in Figure 12 for both minimum horizontal stress (S₃) and maximum stress (S₁). Extension stress (S₃) is more common than compression stress (S₁), because the normal faults (subverticalS₁) in the section are common whereas strike-slip faults moved when both compression and extension were horizontal (subverticalS₂). Normal faults with different orientations, such as NNW-SSE, NE-SW, E-W, and N-S, show high and medium dips (Figure 10). The faults with an oblique component in the NNW direction have high dips towards the west and were induced by the S₃ extension direction NE-SW; NE oblique faults have medium angle dips towards the south-east and are the result of a N-S and NW-SW S₃; E-W dip-normal faults dip towards the south with intermediate angles, whereas the N-S faults display high angle dips towards the east as a result of NNW S₃ and E-W S₃ directions, respectively (Figure 10).The strike-slip faults with NE-SW and NW-SE orientations characteristically have high angle dips associated with ENE-WSW and NNE-SSW compression (S₁), respectively (Figure 10).

The paleostress tensors are shown in Table 3 and in the structural planes of Figures 12 and 13. The tensors were grouped according to their similarities based on the stress regime and the direction of the stress axis. The paleostress is only recorded in Upper Cretaceous rocks, the reason why chronological ordering of paleostress data depends upon rock comparison with other sites outside the study area.

The Cañada de Tarachi Unit is characterized by an extension regime and, to a lesser degree, by a pure strike-slip regime (Figure 13). The main axis of extension (S₃) is almost horizontal with a NW-SE main direction, and with a smaller proportion in a NE-SW direction. The main axis of compression, corresponding to the strike-slip movements,(S₁) is sub-vertical with a NE-SW direction. In conclusion, the previous analysis shows that at the time there was an extension regime with development of high angle normal faults.

Interpretation of the paleostress results in the El Potrero Grande Unit indicates a stress regime characterized by extension and strike-slip regimes, both well defined (Figure 13). The extension regime displays an S₃ closer to the horizontal in a NE-SW direction (Figure 13). The strike-slip regime is shown by a varying main axis of compression S₁, almost horizontal in sites A41-10 and A51-10, sub-vertical in sites A40-10, A46-10a, A41-10, A48-10a, A56-10 and A65-10, and with almost E-W, NW-SE and N-S directions in site A46-10a. The stress of extension S₃ is horizontal to a N-S, NE-SW direction (Figure 13).