The Altar complex at Cerros La Batellera and Cerro San Judas (to be referred to collectively as “Cerros La Batellera”) is part of a ~2–3-km-thick sequence derived from sandstone, conglomerate and rare limestone (García y Barragán, 1992; Jacques-Ayala et al., 1990; García y Barragán et al., 1998). The rocks exhibit southwest-dipping foliation (generally bedding-parallel) and southwest-plunging lineation. Clasts in the conglomerate are stretched to varying degrees dependent on strength relative to the matrix (Figure 3a). Metamorphic grade decreases from greenschist facies in the west to incipiently recrystallized in the east. Intensity of deformation also decreases from west to east. The boundary between the moreand less-highly-deformed and metamorphosed parts of the complex is gradational, but, for simplicity, is depicted as a discrete contact in Figure 2.

East of the map area, the lesser-deformed/metamorphosed portion of the Altar complex is in contact with rocks interpreted as El Chanate Group, although the stratigraphy does not match in detail with that of the type El Chanate Group of the Sierra El Chanate (García y Barragán et al., 1998). This contact has been interpreted as either depositional or a thrust fault (Jacques-Ayala et al., 1990; García y Barragán et al., 1998; Nourse, 2001). The fault interpretation appears most plausible considering the evidence presented here that the Altar complex is a metamorphosed and deformed equivalent of El Chanate Group (see also Barth et al., 2008; Jacques-Ayala et al., 2009). Even farther east, El Chanate Group is in depositional contact with the Bisbee Group. Rocks of the Bisbee Group are also exposed to the north of Cerros La Batellera in the upper plate of a low-angle normal fault of inferred Miocene age.

Figure 2 Figure 2 Geologic map of the study area (modified after García y Barragán, 1992). Numbered circles, squares and diamonds on the map show locations of K-Ar samples from Damon et al. (1962) and Hayama et al. (1984) and 40Ar/39Ar and detrital zircon samples from this study. The legend shows the ages and errors (2σ) for the K-Ar and 40Ar/39Ar ages and sample numbers for the zircon samples. Note that some previous publications have applied the name “Cerros El Amol” to hills northeast of Highway 2 and east of Cerros La Batellera. In contrast, we follow INEGI (2001) in applying the name “Cerro El Amol” to the small set of hills 3–5 km southeast of Altar.

The Carnero complex of Cerro Carnero and Cerro El Amol (to be referred to collectively as “Cerro Carnero”) was derived from sandstone and minor conglomerate, mudstone and limestone. The rocks exhibit a well-developed eastto northeast-dipping foliation and east-trending lineation (Figures 2 and 3b) (Damon et al., 1962; Hayama et al., 1984; Jacques-Ayala et al., 1990; García y Barragán et al., 1998). A thin band of black phyllite is present at the north end of the Carnero complex. Mauel et al. (2011) suggested that the phyllite was derived from marine deposits correlative with the Upper Jurassic Cucurpe Formation. A Jurassic age for the protolith of the Carnero complex is consistent with the detrital zircon results described below.

The Carnero complex is intruded by the Rancho Herradura granodiorite (Figure 3c) and the Carnero leucogranite. The granodiorite is foliated and exhibits a gently west-plunging lineation parallel to lineation in the metasedimentary rocks. Asymmetric (“S-C”) fabrics record top-west extension.

The cooling and exhumation history of the area is constrained by a limited number of K-Ar ages. Damon et al. (1962) reported a biotite K-Ar age of 58 ± 3 Ma from the Altar complex of Cerros La Batellera 2 km southeast of Altar, whereas Hayama et al. (1984) obtained a significantly younger biotite K-Ar age of 16.9 ± 0.6 Ma nearby (Figure 2). For Cerro Carnero, Hayama et al. (1984) reported K-Ar ages for the metasedimentary rocks of 55 ± 3 Ma for hornblende and 16.2 ± 0.5 Ma and 16.1 ± 0.5 Ma for muscovite. These workers also obtained biotite K-Ar ages of 14.8 ± 0.5 Ma and 16.6 ± 0.4 Ma for the Rancho Herradura granodiorite and a K-Ar age for fine-grained white mica (sericite) of 16 ± 0.4 Ma for the Carnero leucogranite. We consider that the Paleogene K-Ar ages reflect cooling following Laramide underthrusting of the protoliths. Thermal resetting during Laramide magmatism is also possible, although we deem this option less likely, as no Laramide plutons are exposed near the dated localities. The ~17–15 Ma ages presumably reflect a second phase of cooling associated with detachment faulting and/or conductive cooling following intrusion of the Rancho Herradura granodiorite and Carnero leucogranite.

We analyzed eight samples for detrital zircon, seven from the Altar complex and one from the Carnero complex (Figure 2; see Table A1 of the Electronic Supplement for coordinates of all sample localities). The Altar samples include five of metasandstone (662, 664, 666, 669 and CJ13-15), the (metamorphosed) sand matrix of a metaconglomerate (663) and a quartzite clast from the metaconglomerate (663A). The Carnero sample is a metasandstone (668).

All detrital samples except CJ13-15 were collected and processed during an early phase of the study. We crushed approximately 1 kg of each sample using a jaw crusher and disc mill. For the metaconglomerate clast, only fragments from the interior were used to avoid matrix material adhering to the outer surface. The crushed material was sieved and ~50 ml volumetrically of the fraction finer than 150 μm was processed in heavy liquids (tetrabromoethane followed by methylene iodide) after removing iron filings with a magnet. The final sink was further cleaned with a Frantz® magnetic separator. We obtained a few hundred or more zircon grains for all samples except sample 669, which yielded relatively few grains, most of which were quite small. As a result, we were able to obtain only five ages from this sample.

Figure 3 Figure 3 (a) Metaconglomerate within Altar complex at Cerros La Batellera. Site of sample 663. (b) Carnero complex at Cerro Carnero. Site of sample 668. (c) Granodiorite of Ranch Herradura at Cerro Carnero. Site of sample 667.

Polished epoxy mounts of the above samples were analyzed at the Arizona Laserchron Center, University of Arizona, using a 193-nm excimer laser and a GVI Isoprobe multicollector-inductively coupled plasma-mass spectrometer (MC-ICP-MS) following the procedures of Gehrels et al. (2008). The ions ²⁰⁶Pb, ²⁰⁷Pb, ²⁰⁸Pb, ²³²Th and ²³⁸U were measured with Faraday collectors. A channeltron multiplier was used for ²⁰⁴Pb. Sri Lanka zircon (SL2; 563.5 Ma ± 3.2 Ma 2σ) was used as the primary standard (Gehrels et al., 2008). Common lead was corrected using the ²⁰⁴Pb method. For unknowns, we targeted grain interiors to minimize the effects of lead loss during metamorphism. We avoided visible inclusions and fractures, but otherwise selected grains randomly, without using cathodoluminescence (CL) imagery. Analytical results are presented in Table A2 of the Electronic Supplement.

We use the ²⁰⁶Pb/²³⁸U age for grains younger than 800 Ma and the ²⁰⁷Pb/²⁰⁶Pb age for grains with ²⁰⁶Pb/²³⁸U ages >800 Ma. Discordance is relatively minor among the analyses with ²⁰⁶Pb/²³⁸U ages > 800 Ma. For this age group, maximum normal discordance is 25%, with a median of 3%; the maximum reverse discordance is 3%, with a median of 1%. None of these ages were excluded. For ages <800 Ma, particularly Mesozoic ages, evaluation of discordance can be problematic owing to potential large errors associated with ²⁰⁷Pb/²³⁵U ages obtained by ICPMS analysis. For these analyses, we evaluated discordance individually for each result in the context of its analytical errors.

Because of the low yield from sample 669, we later collected and processed a second sample (CJ13-15) from the same location (Cerro San Judas, near the shrine to Saint Jude; Figure 2). The sample was crushed in the laboratory and heavy minerals were separated using a Wilfley table and a Frantz® magnetic separator. Zircon grains were extracted from the residual non-magnetic fraction, randomly selected under a binocular microscope, mounted in epoxy resin and polished. Intra-grain compositional zoning was identified by CL. Twenty-two zircons were analyzed using a Resonetics Resolution M050 excimer laser workstation coupled to a Thermo ICapQc ICP-MS at the Laboratorio de Estudios Isotópicos (LEI), Centro de Geociencias, Universidad Nacional Autónoma de México, following the methodology reported in Solari et al. (2010). Analytical results are presented in Table A3 of the Electronic Supplement.

Zircon from the Rancho Herradura granodiorite from Cerro Carnero was extracted during processing of the initial set of detrital samples. Grains were mounted in epoxy and polished prior to examination using CL. The CL images were used to guide selection of analysis points in zoned zircon grains. Uranium and Pb concentrations and isotopic ratios were measured on the SHRIMP-RG ion microprobe at Stanford University using analytical and data analysis procedures described in Barth and Wooden (2006). Isotopic ratios were standardized against Braintree Complex zircon R33 (419 Ma; Black et al., 2004). Analytical results are presented in Table A4 of the Electronic Supplement.

Material generated during crushing of sample CJ13-15 for zircon was used to obtain separates of muscovite, biotite and plagioclase for ⁴⁰Ar/³⁹Ar analysis. The separates were irradiated at the University of McMaster reactor. Analyses were performed at Queen’s University, Canada. Step heating was conducted using a Merchantek MIR 10-30 CO2 laser. A facetted lens was used to diffuse the beam over 3 mm pits. Ions were measured using an MAP-216 mass spectrometer. The J values and errors for samples were determined by polynomial fit to replicate analyses of a flux monitor (biotite GA1550, 98.5 Ma) at measured positions along the length of the irradiation capsule. Analytical results results are presented in Tables A5-A7 of the Electronic Supplement.

We report sample ages as total gas ages. In our view, these provide the best estimate for the time of bulk closure of argon diffusion (Jacobson et al., 2002).

Cerros La Batellera

Probability density plots and cumulative probability plots for the individual samples of the Altar complex from Cerros La Batellera are shown in Figure 4. A plot of the combined age results for the metasandstone and metaconglomerate matrix samples is presented in Figure 5a. Detrital zircon ages for the metasandstone and metaconglomerate matrix samples from Cerros La Batellera range from Archean (2708 Ma) to latest Cretaceous (72 Ma) (Figures 4a–4f and 5a). Five of the six samples include zircon of latest Early Cretaceous and/or Late Cretaceous age. Two samples include zircon as young as Campanian, five grains with ages of 81–72 Ma in sample 662 and one grain with an age of 77 Ma in sample 663. A weighted mean calculation using Isoplot (Ludwig, 2003) suggests that the six latest Cretaceous grains could all be part of a single population with an average age of 77.5 ± 2.5 Ma (Figure 6). Alternatively, averaging only the three youngest grains (e.g., Dickinson and Gehrels, 2009a) yields a statistically-equivalent weighted mean age of 76.5 ± 4.4 Ma. We take the average of the six grains as the more conservative estimate of the maximum depositional age for the Cerros La Batellera samples.

Grains younger than 285 Ma comprise 72% of the total for the metasandstone and conglomerate matrix samples and are inferred to be derived from Cordilleran magmatic arcs (cf. Dickinson and Gehrels, 2009b; González-León et al., 2009, 2017; Mauel et al., 2011; Peryam et al., 2012; Lawton and Molina Garza, 2014):

The latest Cretaceous zircon could have been derived from multiple sources in Sonora, including ranges near Cerros La Batellera (McDowell et al., 2001; Iriondo et al., 2005; Jacques-Ayala et al., 2009; Roldán-Quintana et al., 2009; González-León et al., 2011, 2017; OrtegaGutiérrez et al., 2014).

Most pre-Permian (“pre-arc”) zircon ages in the above samples fall within three age peaks centered near 1.7, 1.4 and 1.1 Ga, the first being the most prominent (Figures 4a–4f and 5a). Igneous and metamorphic complexes with ages of ~1.7 and ~1.4 Ga are widespread in the southwestern United States (U.S.) and Sonora (Barth et al., 2000, 2001a, 2001b; Anderson and Silver, 2005; Farmer et al., 2005; Iriondo et al., 2004; Nourse et al., 2005). Those with ages of ~1.7 Ga dominate over ~1.4 Ga units, consistent with the relative sizes of these two peaks in the Altar samples. In contrast, basement rocks with ages of ~1.1 Ga are not common in the southwestern U.S. and Sonora (Iriondo et al., 2004; Anderson and Silver, 2005). On the other hand, ~1.1 Ga detrital zircon is a significant component of Neoproterozoic and Cambrian cratonal and miogeoclinal quartz arenites from this part of the Cordillera (Gehrels et al., 1995, 2011; Gehrels and Stewart, 1998; Stewart et al., 2001; Gehrels and Pecha, 2014). Hence, we suspect that much, if not most, of the ~1.1 Ga zircon in the Altar complex of Cerros La Batellera is recycled from local miogeoclinal sequences (Figure 1). Accordingly, some of the ~1.7 and ~1.4 Ga zircon in La Batellera samples may also be recycled. However, the ratio of ~1.7 Ga to ~1.1 Ga zircon in our samples exceeds that typically found in Neoproterozoic–Cambrian sedimentary rocks. This suggests that ~1.7 Ga bedrock contributed directly to the protolith of the Altar complex. Potential source areas are present in the Caborca block immediately south of the map area of Figure 1.

Figure 4 Figure 4 Relative and cumulative probability plots of detrital zircon ages from the Altar complex at Cerros La Batellera and the Carnero complex at Cerro Carnero. All samples are metasandstone, except sample 663, which is from the sand matrix of a metaconglomerate, and sample 663A, which is a single quartzite clast from that conglomerate. Note age-axis scale break at 300 Ma. For relative probability plots, vertical scales differ to left and right of the age-axis scale break such that equal area represents equal probability throughout the graph. Line color in the cumulative probability plot corresponds to the fill color in relative probability plots. Location of samples shown in Figure 2. n: number of analyses

Figure 5 Figure 5 Composite detrital zircon age distributions for metasandstone and conglomerate matrix from the Altar complex at Cerros La Batellera (samples 662, 663, 664, 666, 669, CJ13-15) and sandstone from El Chanate Group of Sierra El Chanate (Jacques-Ayala et al., 2009). Scale break and color-coding of relative versus cumulative probability plots as in Figure 4.

Some uncertainty in the interpretation of the Cerros La Batellera section is introduced by the results from sample 669, collected from Cerro San Judas near the shrine to St. Jude. This sample differs from all the others from Cerros La Batellera in that it yielded only Jurassic ages. The dataset, however, is small—only five analyses. As noted in the analytical methods section, zircons obtained from sample 669 were smaller on average than those from the other samples. Only the larger grains could be analyzed, which may have biased the results. In addition, Cretaceous zircon is not abundant in any of the samples of Altar complex from Cerros La Batellera (Figures 4 and 5a) and could have been missed in sample 669 owing to the small number of analyses. Furthermore, Cretaceous zircon is present in sample CJ13-15 from the same locality (Figure 4f). Considering these factors, we conclude that sample 669 likely has the same Late Cretaceous depositional age as the other Cerros La Batellera samples, despite the failure to yield any zircon younger than Jurassic.

In contrast to the dominant arc sources for the sand component of the Altar complex, the 70 ages from the quartzite clast (sample 663A) are all Precambrian—68 Proterozoic (minimum age of 1060 ± 37 Ma) and two Archean (Figure 4g). The Proterozoic ages define the same three age peaks (~1.7, ~1.4 and ~1.1 Ga) as in the sands, although in more nearly equal proportions than in the latter (Figure 4g versus Figure 5a). The age distribution for sample 663A is typical of that for the Neoproterozoic and Cambrian sedimentary sequences from Sonora and adjacent areas (Gehrels and Stewart, 1998; Stewart et al., 2001; Gehrels and Pecha, 2014). Miogeoclinal rocks of this age crop out in ranges west and south of Caborca and comprise a potential source for the quartzite clasts in the Altar complex (Figure 1). The original clasts may have been quartz arenite that was converted to quartzite during Laramide deformation and metamorphism.

Lithologic and map relations led Jacques-Ayala et al. (1990), García y Barragán et al. (1998) and Jacques-Ayala (1999) to conclude that the protolith of the Altar complex at Cerros La Batellera was either correlative with El Chanate Group or deposited above it. Comparison of composite age distributions for the Altar complex at Cerros La Batellera (this study) and El Chanate Group of the Sierra El Chanate (Jacques-Ayala et al., 2009) favors the former (Figure 5). This conclusion is somewhat speculative, because application of the KolmogorovSmirnoff (K-S) test to the two groups yields a P value of 0.005, which implies that the two groups were not derived from the same source (cf. Dickinson and Gehrels, 2009b). Nonetheless, we note multiple similarities between the two datasets: (1) Dominance of Jurassic zircon in each, albeit skewed to older ages in El Chanate Group. (2) Similar proportions of ~1.7, ~1.4 and ~1.1 Ga populations. (3) Comparable small numbers of Permian–Triassic zircons. (4) Similar distribution of Cretaceous ages, including similar youngest ages (~75 Ma in both groups). These relations lead us to conclude that the protoliths of the Altar complex of Cerros La Batellera and El Chanate Group are correlatives, despite the negative K-S result. We suggest that the low P value derives from the small sample sizes, minor local variation in source area and/or systematic analytical differences between the Altar complex ages, which were obtained by ICP-MS methods, and El Chanate results, which were determined by ion microprobe. Further work is needed to resolve this issue.

Figure 6 Figure 6 Weighted mean plot of six youngest zircon ages from the Altar complex of Cerros La Batellera. Diagram constructed using Isoplot Version 3.75 (Ludwig, 2003).

Based on the high abundance of quartzite clasts, Jacques-Ayala (1999) and Jacques-Ayala et al. (2009) proposed that the Altar complex at Cerros La Batellera correlates with the Pozo Duro Formation, the oldest unit of El Chanate Group. Alternatively, the 77.5 ± 2.5 Ma maximum depositional age of the Cerros La Batellera samples may indicate a correlation with the Escalante Formation, the youngest part of El Chanate Group. The Escalante Formation has a maximum depositional age of 75 ± 2 Ma based on two zircon ages derived from the upper part of the formation (Jacques-Ayala et al. (2009). Correlation of the Altar complex of Cerros La Batellera with the Escalante Formation has also been proposed by González-León et al. (2017). Nonetheless, the maximum depositional ages of both the Altar and El Chanate samples are not well constrained, as most of the detrital zircon ages are pre-Late Cretaceous (Figure 5). In other words, even those samples that have so far failed to yield zircon younger than late Early or early Late Cretaceous could, in fact, have Campanian depositional ages. This ambiguity reinforces the need for additional analyses.

Cerro Carnero

We determined detrital zircon ages for one sample of metasandstone from the Carnero complex (sample 668; Figure 4h). In contrast to the samples from the Altar complex, the Carnero sample includes little arc-derived zircon. Another distinctive aspect of the Carnero sample is the high fraction of Grenville-age zircon (1310–912 age group of Dickinson et al., 2012) relative to the ~1.7 and ~1.4 Ga peaks. In addition, the Carnero sample includes abundant Neoproterozoic (~600 Ma) and Paleozoic (~400 Ma) zircon, age groups not observed in the samples of Altar complex.

The strong Grenville, Neoproterozoic and Paleozoic components of the Carnero sample suggest affinities with the Jurassic eolian (erg) sandstones of the Colorado Plateau, U.S. (Figure 7b; Dickinson and Gehrels, 2009b). Erg-type sands have been documented in Sonora; e.g. in the Middle Jurassic Rancho San Martín Formation near Cucurpe (Leggett, 2009; Mauel et al., 2011). The Rancho San Martín Formation and our sample of Carnero complex exhibit similar detrital zircon age spectra, including similar maximum depositional ages (~180–170 Ma) (Figures 7a and 7c; Leggett, 2009; Mauel et al., 2011). In detail, the Carnero complex and Rancho San Martín Formation display some contrasts, such as fewer Neoproterozoic grains in the latter. However, variability of this scale is typical within the Jurassic eolianites (Dickinson and Gehrels, 2009b).

Figure 7 Figure 7 Comparison of detrital zircon age results from the Carnero complex of Cerro Carnero (sample 668 of Figure 4) with selected Jurassic and Cretaceous formations in Sonora and the southwestern U.S. exhibiting an erg-type detrital zircon signature (cf. Dickinson and Gehrels, 2009b). Data sources: (a) this study; (b) Dickinson and Gehrels (2009b); (c) Leggett (2009); (d) Mauel et al. (2011); (e) Clinkscales and Lawton (2015). Scale break and color-coding of relative versus cumulative probability plots as in Figure 4. N: number of samples, n: number of zircon ages.

The zircon age results for the Carnero sample are consistent with a Middle Jurassic protolith age, but do not rule out a younger age. For example, the basal part of the Upper Jurassic Cucurpe Formation of the Cucurpe area exhibits a strong erg signature, attributed to reworking of eolianite layers in the underlying Rancho San Martín Formation (Figure 7d; Mauel et al., 2011). The basal Cucurpe Formation also includes Late Jurassic zircon, confirming its young age relative to the Rancho San Martín Formation. However, the Late Jurassic zircon is thought to be locally derived (Mauel et al., 2011), and need not be present in all parts of the Cucurpe Formation derived by recycling of eolianites sands.

A strong erg signature has also been observed in the Cintura Formation of the Bisbee Group in southern Arizona, U.S. (Dickinson et al., 2009) and in the Mojado Formation, a correlative of the Cintura Formation in southwestern New Mexico, U.S., (Clinkscales and Lawton, 2015). The Mojado example is particularly instructive. Clinkscales and Lawton (2015) obtained 92 zircon ages for one sample of the Mojado Formation. The pre-Permian grains exhibit an age distribution remarkably like that of the Colorado Plateau Jurassic ergs (Figures 7b and 7e). The Mojado sample also includes a modest number of Triassic and Jurassic arc-derived grains. However, none of the ages are younger than early Late Jurassic, despite the well-constrained Albian depositional age.

Erg-style detrital zircon age patterns have also been found locally in Upper Cretaceous strata in Sonora (Jacques-Ayala and Garcia y Barragán, 2018). Nonetheless, most Upper Cretaceous sandstones from Sonora that have been analyzed for detrital zircon include only minor numbers of erg-derived grains (Jacques-Ayala et al., 2009; GonzálezLeón et al., 2017; this study).

In summary, correlation of the Carnero sample with the Middle Jurassic Rancho San Martín Formation is the most straightforward interpretation, although correlation with the Cucurpe Formation, Bisbee Group or even El Chanate Group is also permissible, with the last option least likely.

Rancho Herradura Granodiorite

Zircon grains extracted from the Rancho Herradura granodiorite (Sample 667) are euhedral to subhedral prismatic and up to 200 microns long. About half the grains contain recognizable cores. We analyzed nine grains. Seven yielded Miocene ages, which we interpret as synmagmatic. Two older ages (56 and 946 Ma) are considered premagmatic. The ²⁰⁶Pb/²³⁸U ages for the synmagmatic zircons range from 23 to 19 Ma and form an array along concordia on a Tera-Wasserburg diagram (Figure 8). We interpret this age range to indicate varying degrees of lead loss, perhaps associated with extensional deformation of the granodiorite. The four oldest grains, which presumably underwent the least lead loss, yield a mean age of 21 ± 1 Ma (95% confidence level). However, even these grains may have lost some Pb, so the true age of the granodiorite could be closer to the upper end of the range of analyses; i.e., ~22 Ma. On the other hand, Nourse et al. (2007) inferred a zircon U-Pb age of 21.2 ± 0.2 (95% confidence level) for this same intrusion, also using the SHRIMP-RG ion microprobe at Stanford University. Considering the two datasets together, we conclude that the age of the granodiorite is ~22–21 Ma.

The 40Ar/39Ar results vary in degree of interpretability (Figure 9). The muscovite spectrum for sample CJ13-15 is the least complicated, with a profile indicative of Ar loss in the low-temperature steps leading to a moderately well-developed plateau for the high-temperature steps. The total gas age is 45.1 ± 0.5 Ma.

The biotite age spectrum is less regular. Nonetheless, even the oldest steps are younger than those on the plateau of the muscovite spectrum, consistent with the lower inferred closure temperature for biotite than muscovite (~350 ± 50 °C and ~400 ± 50 °C, respectively; McDougall and Harrison, 1999). The total gas age is 36.0 ± 0.8 Ma.

The plagioclase sample exhibits a U-shaped spectrum indicative of excess radiogenic argon (McDougall and Harrison, 1999). The youngest steps yield ages >50 Ma; i.e. older than the oldest steps for muscovite and biotite. This confirms that all steps have been impacted by excess radiogenic argon, as the closure temperature for plagioclase (and thus age) is expected to be equal to or less than those for muscovite and biotite (McDougall and Harrison, 1999). We do not believe that a geologically meaningful age can be determined for this sample and do not consider it further.

The muscovite and biotite 40Ar/39Ar ages fall in a gap between clusters at 58–55 Ma and 17–15 Ma defined by the K-Ar analyses of Damon et al. (1962) and Hayama et al. (1984) (Figure 2). It is difficult to compare 40Ar/39Ar and K-Ar ages directly. Nonetheless, our results are consistent with the evidence from the K-Ar ages for an extended period of cooling from early to mid-Cenozoic time.

Our detrital zircon analyses confirm the view that much, or all, of the Altar complex at Cerros La Batellera was derived from an Upper Cretaceous protolith likely correlative with El Chanate Group (e.g., Jacques-Ayala et al., 1990, 2009; García y Barragán et al., 1998; Jacques-Ayala, 1999, 2000) and contradict the interpretation that the entire protolith of El Batamote belt is correlative with the Glance Conglomerate (e.g., Nourse, 2001). Some parts of La Batellera section are no older than 77.5 ± 2.5 Ma, which may indicate correlation with the Escalante Formation, the uppermost unit of El Chanate Group (González-León et al., 2017).

Some fraction of El Batamote belt, however, must be derived from units older than Late Cretaceous. The most direct line of evidence is provided by the latest Middle Jurassic (164 Ma) tuff (?) located northwest of Altar (Mauel, 2008; Mauel et al., 2011). An age as old as Middle Jurassic is also suggested by the erg-style detrital zircon age spectrum of the sample from the Carnero complex. Alternatively, the Carnero sample could be younger than Middle Jurassic (above), although the most-plausible options nonetheless require that the protolith be pre-Late Cretaceous and thus older than the Altar complex. Similarly, the distinctive lithologic composition of the black phyllite at the north end of the Carnero complex body is consistent with, but not proof of, correlation with the Middle Jurassic Cucurpe Formation (Mauel et al., 2011).

Figure 8 Figure 8 Tera–Wasserburg concordia plot of zircon ages from the Rancho Herradura granodiorite (sample 667).

It is apparent from various issues discussed herein that further work is needed to more fully characterize the protolith of El Batamote belt, including additional mapping, petrography, whole-rock geochemical analysis and detrital zircon studies throughout its geographic extent. Nonetheless, based on the available data, we hypothesize that El Batamote belt was derived from a range of Middle Jurassic to Upper Cretaceous supracrustal sequences that were present in the Caborca– Altar area at the time of the Laramide orogeny (see also García y Barragán and Jacques-Ayala, 2010). We envision that the various units were deformed and metamorphosed together beneath a now-hidden Laramide thrust sheet (or sheets). The youngest strata involved in this thrusting (Escalante Formation and parts of the Altar complex) may have been synorogenic deposits shed from the thrust belt as it advanced northeastward toward the Caborca–Altar area (Jacques-Ayala et al., 1990, 2009; García y Barragán et al., 1998; Jacques-Ayala, 1999, 2000). This style of tectonism is analogous to that proposed for much of the Cretaceous–Paleogene Mexican orogen (Fitz-Díaz et al., 2018). Application to the Caborca–Altar area, however, is more speculative, because of the lack of definitive evidence for major thrust faults, either bounding or within El Batamote belt. The inferred thrust fault(s) could be (1) buried beneath basin fill southwest of the exposures of El Batamote belt (Jacques-Ayala and Garcia y Barragán, 2015, 2018); (2) present within the complex, but difficult to recognize because of the lack of good stratigraphic control; and/or (3) overprinted by Miocene detachment faults (Nourse, 2001). Late Cretaceous thrust faults have been recognized within the Sierra La Vibora of the Caborca block southwest of Altar (De Jong et al., 1988; Jacques-Ayala et al., 1990) and could be part of the proposed thrust system.

According to Nourse (1995; 2001) and Anderson and Nourse (2005), the protolith of El Batamote belt was deposited in a northwestsoutheast–elongated transtensional basin along the Late Jurassic Mojave-Sonora megashear. Our results, however, indicate that much of the protolith is too young to be related to the megashear. Instead, we propose that the northwest-southeast trend of the belt is the consequence of Late Cretaceous northeast-southwest shortening followed by Miocene extension along approximately the same azimuth. We cannot rule out that some components of the protolith of El Batamote belt were deposited in northwest-southeast–trending basins (e.g. The Altar-Cucurpe basin of Mauel et al., 2011). Nonetheless, we consider it unlikely that the present linear outcrop geometry of El Batamote belt is a fundamental signature of Late Jurassic transtension. This conclusion does not by itself disprove the megashear concept, but it greatly diminishes the importance of El Batamote belt as an argument in favor of the megashear.

The contractional event that caused deformation and metamorphism of El Batamote belt and folding of the Bisbee and El Chanate Groups in the Caborca–Altar area must be at least partly younger than the ~78–75 Ma maximum depositional ages of the Altar complex and El Chanate Group, as well as the 72 ± 1 Ma El Charro volcanic complex, which was folded along with El Chanate Group (Jacques-Ayala, 1999). The end of Laramide deformation in this region is constrained by eruption of the post-tectonic San Jacinto volcanics rocks at 51 ± 2 Ma (Jacques-Ayala, 1999, 2000). In addition, deformation was likely over, or at least waning, by the early Eocene based on the biotite K-Ar age of 58 ± 3 Ma (Damon et al., 1962) for Altar complex at Cerros La Batellera. However, our ⁴⁰Ar/³⁹Ar ages of 45.1 ± 0.5 and 36.0 ± 0.8 Ma for muscovite and biotite, respectively, indicate that at least some parts of the section remained at temperatures of 400–350 °C through most of the Eocene.

Figure 9 Figure 9 40Ar/39Ar release spectra for Altar complex sample CJ13-15.

Deformation similar in style and age to that of the Caborca–Altar area is widespread through northern Sonora and adjoining areas; e.g., near Quitovac, Imuris, Arizpe and Naco in Sonora (Iriondo et al., 2005; González-León et al., 2017) and in the Quitobaquito Hills to Baboquivari Mountains of south-central Arizona (Haxel et al., 1984; Goodwin and Haxel, 1990; Calmus and Sosson, 1995). Common attributes of these areas include: (1) deformation involved the full thickness of Mesozoic sedimentary and volcanic rocks; (2) Upper Cretaceous strata (Cabullona Group, Cocóspera Formation, Fort Crittenden Formation) were deposited in basins bounded by Laramide thrust and high-angle reverse faults; (3) greenschist facies metamorphism, southwest-dipping cleavage and stretched-pebble conglomerates are locally well developed; (4) magmatism generally outlasted deformation; and (5) both magmatism and deformation appear to have migrated northeastward with time. Among these areas, González-León et al. (2017) highlighted similarities between the stratigraphic, structural and age relations for the Cocóspera basin near Imuris and Arizpe with those for the Altar complex and El Chanate Group. The Cocóspera Formation was thrust beneath the Bisbee Group along the Late Cretaceous San Antonio fault. González-León et al. (2017) interpreted the San Antonio fault as a regional Laramide tectonic front that extended westward to the Caborca–Altar area. The San Antonio fault would thus be equivalent to the Altar thrust of Figure 1 (the “Altar sole fault” of Jacques-Ayala and García y Barragán, 2015).

The orogenic system represented by El Batamote belt may have an analog in the McCoy Mountains Formation of southwesternmost Arizona and southeasternmost California, 400 km northwest of the Caborca–Altar area (Harding and Coney, 1985; Barth et al., 2004; Spencer et al., 2011). The McCoy Mountains Formation includes sandstone, siltstone and conglomerate that, like the inferred protoliths of El Batamote belt, range from Middle or Upper Jurassic to Upper Cretaceous. The lower to middle parts of the McCoy Mountains Formation have been linked to the Bisbee Group, including the Upper Jurassic-Lower Cretaceous Glance Conglomerate (Dickinson and Lawton, 2001). The upper part of the McCoy Mountains Formation, in turn, resembles the upper part of the Altar complex, both in terms of its Late Cretaceous depositional age and the abundance of conglomerate. All parts of the McCoy Mountains Formation were affected by Late Cretaceous deformation. These similarities further highlight the extent and continuity of the Laramide orogen in the southwestern U.S. and northwestern Mexico.

Laramide deformation probably accounts for the main structures and fabrics in El Batamote belt and the Bisbee and El Chanate Groups of the Caborca–Altar area (see also Nourse, 2001). Nonetheless, topwest extensional fabrics (Jacques-Ayala et al., 1990) in the ~22–21 Ma Rancho Herradura granodiorite at Cerro Carnero provide evidence of an important overprint during Miocene detachment faulting. Biotite and white mica K-Ar ages clustered at 17–15 Ma may indicate that the main phase of extension postdated intrusion of the granodiorite by several million years. In any case, middle Cenozoic extension in the Caborca–Altar is likely part of a much broader metamorphic corecomplex event (Nourse, et al., 1994; Wong et al., 2010). The difficulty in recognizing Laramide thrust faults is probably due in large measure to overprinting by middle Cenozoic extensional structures.