Abstract

The timing and spatial extent of mid-Cenozoic ignimbrite flare-up volcanism of the Sierra Madre Occidental silicic large igneous province of Mexico in relation to crustal extension is relatively unknown. Extension in the Sierra Madre Occidental has been variably interpreted to have preceded, postdated, or begun during Early Oligocene flare-up volcanism of the silicic large igneous province. New geologic mapping, zircon U-Pb laser ablation–inductively coupled plasma–mass spectrometry dating, modal analysis, and geochemical data from the Guazapares Mining District region along the western edge of the northern Sierra Madre Occidental silicic large igneous province have identified three informal synextensional formations. The ca. 27.5 Ma Parajes formation is an ∼1-km-thick succession composed primarily of welded to nonwelded silicic outflow ignimbrite sheets erupted from distant sources. The 27–24.5 Ma Témoris formation is interpreted as an andesitic volcanic center composed of locally erupted mafic to intermediate composition lavas and associated intrusions, with interbedded andesite-clast fluvial and debris flow deposits, and an upper section of thin distal silicic outflow ignimbrites. The 24.5–23 Ma Sierra Guazapares formation is composed of silicic vent facies ignimbrites to proximal ignimbrites, lavas, plugs, dome-collapse deposits, and fluvially or debris flow–reworked equivalents. These three formations record (1) the accumulation of outflow ignimbrite sheets, presumably erupted from calderas mapped ∼50–100 km east of the study area that were active during the Early Oligocene pulse of the mid-Cenozoic ignimbrite flare-up; (2) development of an andesitic volcanic field in the study area, likely related to rocks of the Southern Cordillera basaltic andesite province that were intermittently erupted across all of the northern Sierra Madre Occidental toward the end of and following the Early Oligocene ignimbrite pulse; and (3) the initiation of explosive and effusive silicic fissure magmatism in the study area during the Early Miocene pulse of the mid-Cenozoic ignimbrite flare-up.

The main geologic structures identified in the Guazapares Mining District region are NNW–trending normal faults, with an estimated minimum of 20% total horizontal extension. Normal faults were active during deposition of all three formations (Parajes, Témoris, and Sierra Guazapares), and bound half-graben basins that show evidence of synvolcanic extension (e.g., growth strata) during deposition. Normal faulting began by ca. 27.5 Ma during deposition of the youngest ignimbrites of the Parajes formation, concurrent with the end of the Early Oligocene silicic ignimbrite pulse to the east and before magmatism began in the study area. In addition, preexisting normal faults localized andesitic volcanic vents of the Témoris formation and silicic vents of the Sierra Guazapares formation, and some faults were reactivated during, as well as after, deposition of these formations.

We interpret extensional faulting and magmatism in the Guazapares Mining District region to be part of a regional-scale Middle Eocene to Early Miocene southwestward migration of active volcanism and crustal extension in the northern Sierra Madre Occidental. We show that extension accompanied silicic volcanism in the Guazapares region, and overlapped with the peak of mid-Cenozoic ignimbrite flare-up in the Sierra Madre Occidental; this supports the interpretation that there is a relationship between lithospheric extension and silicic large igneous province magmatism.

INTRODUCTION

Silicic large igneous provinces are significant in the geologic record due to their unusually extensive areal coverage (>100,000 km2), large volumes (>250,000 km3), and potential to induce environmental change (e.g., Bryan, 2007; Cather et al., 2009; Jicha et al., 2009; Bryan and Ferrari, 2013). Compositions within silicic large igneous provinces range from basalt to high-silica rhyolite, but are volumetrically dominated (>80%) by dacite-rhyolite compositions, with >75% of the total magmatic volume emplaced during short duration (∼1–5 Myr) pulses over a maximum province lifespan of ∼50 Myr (Bryan, 2007; Bryan and Ernst, 2008). Previous studies suggest that silicic large igneous provinces may be characteristic of continental regions undergoing broad lithospheric extension and typically initiate as prerifting magmatic events (Bryan et al., 2002; Bryan, 2007; Best et al., 2013; Bryan and Ferrari, 2013). Therefore, determining the timing of extensional deformation in relation to magmatism is an important consideration toward understanding silicic large igneous province processes, as crustal extension is suggested as one mechanism that favors the generation of large silicic magma volumes (Hildreth, 1981; Wark, 1991; Hanson and Glazner, 1995) as well as very large magnitude explosive silicic eruptions (Aguirre-Díaz and Labarthe-Hernández, 2003; Costa et al., 2011).

The Sierra Madre Occidental of western Mexico is the third largest silicic large igneous province of the Phanerozoic and is the largest and best-preserved of the Cenozoic (Fig. 1; Bryan, 2007; Ferrari et al., 2007). It extends for ∼1200 km south from the U.S.-Mexico border to the Trans-Mexican Volcanic Belt, forming a high plateau with an average elevation >2000 m, consisting primarily of Oligocene to Early Miocene ignimbrites that cover an estimated area of 300,000–400,000 km2 with an average thickness of 1 km (McDowell and Keizer, 1977; McDowell and Clabaugh, 1979; Aguirre-Díaz and Labarthe-Hernández, 2003). The volcanism of the Sierra Madre Occidental silicic large igneous province is contemporaneous with, and is considered part of, the extensive mid-Cenozoic ignimbrite flare-up that affected much of the southwestern North American Cordillera from the Middle Eocene to Late Miocene (e.g., Coney, 1978; Armstrong and Ward, 1991; Ward, 1991; Ferrari et al., 2002; Lipman, 2007; Cather et al., 2009; Henry et al., 2010; Best et al., 2013). The core of the Sierra Madre Occidental is relatively unextended in comparison to the surrounding Late Oligocene to Miocene extensional belts of the southern Basin and Range to the east and the Gulf Extensional Province to the west (Fig. 1; Nieto-Samaniego et al., 1999; Henry and Aranda-Gómez, 2000). Rocks related to the silicic large igneous province extend beyond the Sierra Madre Occidental proper (Fig. 1) to the Mesa Central and parts of the southern Basin and Range in eastern Chihuahua and Durango (Gunderson et al., 1986; Aguirre-Díaz and McDowell, 1991, 1993), as well as southwesternmost mainland Mexico and Baja California Sur (Umhoefer et al., 2001; Ferrari et al., 2002).

A large part of the Sierra Madre Occidental remains unmapped and undated (>90%; Swanson et al., 2006). Previous work in the Sierra Madre Occidental has been primarily restricted to the southern region of the igneous province (e.g., Nieto-Samaniego et al., 1999; Ferrari et al., 2002), the vicinity of the Mazatlán–Durango highway in the central region (e.g., McDowell and Keizer, 1977; McDowell and Clabaugh, 1979; Henry and Fredrikson, 1987), and the areas around the Hermosillo–Chihuahua City highway and the Tomóchic–Creel road in the northern region (e.g., Swanson, 1977; Swanson and McDowell, 1984, 1985; Wark et al., 1990; Cochemé and Demant, 1991; Wark, 1991; McDowell and Mauger, 1994; Albrecht and Goldstein, 2000; Swanson et al., 2006; McDowell, 2007; McDowell and McIntosh, 2012) (Fig. 1). As a result, the age relationships between ignimbrite flare-up volcanism and crustal extension remain unclear. Previous studies have suggested that significant crustal extension in the region did not occur until after the peak of large volume ignimbrite flare-up volcanism, which was inferred to have occurred between ca. 32 and 28 Ma (Early Oligocene; e.g., McDowell and Clabaugh, 1979; Wark et al., 1990; McDowell and Mauger, 1994; Gans, 1997; Grijalva-Noriega and Roldán-Quintana, 1998). However, other studies have inferred that initial regional extension is recorded by the onset of large volume Early Oligocene ignimbrite flare-up volcanism (e.g., Aguirre-Díaz and McDowell, 1993), or that extensional deformation began before the flare-up (e.g., Dreier, 1984; Ferrari et al., 2007). Uncertainty regarding the timing of extension relative to ignimbrite flare-up volcanism is also a problem in the Basin and Range of the western U.S., where previous studies have inferred that extension preceded, postdated, or began during ignimbrite flare-up volcanism (e.g., Gans et al., 1989; Best and Christiansen, 1991; Axen et al., 1993; Best et al., 2013).

The Guazapares Mining District region of western Chihuahua, Mexico, is located ∼250 km southwest of Chihuahua City in the northern Sierra Madre Occidental (Fig. 1). The excellent rock exposure and topographic relief in this previously unmapped area make it ideal for studying the relationships between silicic large igneous province volcanism and crustal extension. In this paper we show that extension preceded the onset of magmatism in the study area. We demonstrate that extension was active in the study area during deposition of ca. 27.5 Ma outflow ignimbrites, presumably derived from calderas of similar ages identified to the north and east by other workers. Extension continued during growth of a ca. 27–24.5 Ma andesitic volcanic center in the study area, followed by continued extension during ca. 24.5–23 Ma silicic flare-up magmatism in the study area. This study shows how extensional structures controlled the siting of the andesitic and silicic volcanic vents and shallow-level intrusions. This study also shows that the onset of extension in the study area overlaps with the end of peak Oligocene silicic magmatism to the east, and that extension in the study area preceded and coincided with a second peak of magmatism in the Miocene, which is represented in the study area. Last, we show that our data support the interpretation that silicic flare-up magmatism swept southwestward with time, due to rollback and/or removal of the slab that was subducting beneath western Mexico.

GEOLOGIC SETTING

Previous regional-scale studies in the Sierra Madre Occidental subdivided volcanic rocks into: (1) the Late Cretaceous to Eocene Lower Volcanic Complex of dominantly andesitic composition; (2) the Eocene to Early Miocene Upper Volcanic Supergroup of dominantly silicic composition; and (3) the Early Oligocene to Early Miocene basaltic andesite volcanic rocks of the Southern Cordillera basaltic andesite province (McDowell and Keizer, 1977; Cameron et al., 1989; Ferrari et al., 2007). The Lower Volcanic Complex is believed to underlie most of the Upper Volcanic Supergroup (Aguirre-Díaz and McDowell, 1991; Ferrari et al., 2007), although the thick ignimbrite cover of the Upper Volcanic Supergroup obscures much of the geologic relationships between these two subdivisions in most areas. The volcanic rocks of the Lower Volcanic Complex generally consist of intermediate composition lavas and lesser silicic tuffs, and are interpreted as the products of normal steady-state (i.e., non–flare-up-style) continental subduction-related magmatism broadly contemporaneous with the Laramide orogeny in western North America (McDowell and Keizer, 1977; McDowell et al., 2001).

The ∼1-km-thick Upper Volcanic Supergroup broadly refers to the products of large-volume flare-up–style (i.e., high output rate and large eruptive volumes) silicic magmatism, also known as the mid-Cenozoic ignimbrite flare-up, and defines the extent of the Sierra Madre Occidental silicic large igneous province (McDowell and Keizer, 1977; Bryan, 2007; Ferrari et al., 2007). The Upper Volcanic Supergroup is composed of Eocene to Early Miocene silicic ignimbrites, lavas, and intrusions, and lesser intermediate to mafic lavas (McDowell and Keizer, 1977; McDowell and Clabaugh, 1979; Aguirre-Díaz and McDowell, 1991, 1993; Ferrari et al., 2002, 2007; McDowell, 2007). The large volume of silicic ignimbrites and high output rate suggest multiple caldera and fissure sources for these volcanic deposits (e.g., Swanson and McDowell, 1984; Aguirre-Díaz and Labarthe-Hernández, 2003; Swanson et al., 2006; McDowell, 2007). Ferrari et al. (2002, 2007) proposed that there were at least two main pulses of large volume silicic ignimbrite flare-up volcanism in the Sierra Madre Occidental during the mid-Cenozoic, one during the Early Oligocene (ca. 32–28 Ma) and another during the Early Miocene (ca. 24–20 Ma). The Early Oligocene ignimbrite pulse is inferred to have occurred throughout the Sierra Madre Occidental, while the Early Miocene ignimbrite pulse was inferred to be volumetrically more significant in the southern Sierra Madre Occidental and less abundant, with more mafic compositions, in the north (Ferrari et al., 2002, 2007; Bryan et al., 2013). The Early Oligocene pulse is estimated to have contributed at least half to three-quarters (>200,000 km3) of the erupted volume of the Upper Volcanic Supergroup, but at least 50,000–100,000 km3 was erupted during the Early Miocene pulse (Cather et al., 2009; Bryan et al., 2013). McDowell and McIntosh (2012) suggested that most ignimbrites in the northern and central Sierra Madre Occidental were erupted during discrete time intervals (36–33.5 Ma and 31.5–28 Ma). In addition, an older Eocene pulse of ignimbrite eruptions between 46 and 42 Ma is only recognized along the eastern margin of the Sierra Madre Occidental, and an interval of ca. 24 Ma ignimbrite eruptions that coincides with the Early Miocene pulse of Ferrari et al. (2002, 2007) is observed in the western regions of the igneous province (McDowell and McIntosh, 2012), west of our study area.

During the final stages of and after each silicic ignimbrite pulse of the Upper Volcanic Supergroup, basaltic andesite lavas were intermittently erupted across all of the northern Sierra Madre Occidental (Ferrari et al., 2007). In the northern part of the Sierra Madre Occidental these rocks were generally considered part of the Southern Cordillera basaltic andesite province (Cameron et al., 1989) with ages ranging from 33 to 17.6 Ma, although they mostly are Oligocene (Cameron et al., 1989, and references therein; Ferrari et al., 2007). The rocks of the Southern Cordillera basaltic andesite province have been interpreted as magmatism recording the initiation of crustal extension across the region (e.g., Cameron et al., 1989; Cochemé and Demant, 1991; Gans, 1997; McDowell et al., 1997; González León et al., 2000; Ferrari et al., 2007).

Several prior studies recognized significant crustal extension in the Sierra Madre Occidental immediately following the Early Oligocene ignimbrite pulse of the Upper Volcanic Supergroup (e.g., McDowell and Clabaugh, 1979; Wark et al., 1990; McDowell and Mauger, 1994; Gans, 1997; Grijalva-Noriega and Roldán-Quintana, 1998). The earliest evidence of extensional faulting in the northern Sierra Madre Occidental is found in central Chihuahua (younger than 29 Ma), immediately following the Early Oligocene ignimbrite pulse (McDowell and Mauger, 1994). In east-central Sonora, the earliest age of crustal extension is possibly as old as 27 Ma and synvolcanic deposition in many normal-fault basins was active by 24 Ma, following the peak of Early Oligocene ignimbrite flare-up volcanism (Gans, 1997; McDowell et al., 1997; Gans et al., 2003). However, extension in the Sierra Madre Occidental may have begun as early as the Eocene, prior to the eruption of the Early Oligocene ignimbrite pulse, based on the orientation and age of epithermal vein deposits (Dreier, 1984) and a moderate angular unconformity between the Lower Volcanic Complex and Upper Volcanic Supergroup (e.g., Ferrari et al., 2007). Direct evidence of Early Eocene (pre–Upper Volcanic Supergroup) extensional faulting is observed in the Mesa Central region to the east of the core of the southern Sierra Madre Occidental and includes a moderate angular unconformity within continental clastic and andesitic volcanic sequences and subvolcanic intrusions along normal faults (Aranda-Gómez and McDowell, 1998; Aguillón-Robles et al., 2009; Tristán-González et al., 2009), as well as ca. 32 Ma synvolcanic normal faults that were active until ca. 24 Ma (Aguirre-Díaz and McDowell, 1993; Luhr et al., 2001). However, Eocene-age extensional faulting has not been documented in the Sierra Madre Occidental proper.

The Guazapares Mining District of western Chihuahua is located at the western edge of the relatively unextended core of the northern Sierra Madre Occidental, at the boundary with the highly extended Gulf Extensional Province (Fig. 1). Previous geologic studies in this ∼300 km2 region were restricted to regional 1:50,000 and 1:250,000 geologic mapping by the Mexican Geological Survey (Minjárez Sosa et al., 2002; Ramírez Tello and Garcia Peralta, 2004) and mining company reports (e.g., Roy et al., 2008; Wood and Durgin, 2009; Gustin, 2011, 2012). On these maps and reports, Paleocene–Eocene Lower Volcanic Complex andesitic rocks were inferred to underlie the Oligocene Upper Volcanic Supergroup silicic ignimbrites, but we show here that these rocks (which we informally refer to as the Témoris formation) are both underlain and overlain by silicic ignimbrites, and therefore cannot be assigned to the Lower Volcanic Complex. Prior to this study there were no geochronological data from the Guazapares Mining District region and the closest reported dates were from Upper Volcanic Supergroup ignimbrites ∼50 km to the northeast near Divisadero (ca. 30 Ma; Swanson et al., 2006).

LITHOLOGY AND STRATIGRAPHY

New geologic mapping in the Guazapares Mining District region (Figs. 2, 3, and 4; Supplemental Figure 1,1) provides the basis for the subdivision of three informally named formations described in the following (from oldest to youngest): (1) the Parajes formation, consisting mainly of silicic outflow ignimbrites; (2) the Témoris formation, composed mainly of mafic to intermediate composition lavas and intrusions; and (3) the Sierra Guazapares formation, consisting of silicic vent-proximal ignimbrites, lavas, and subvolcanic intrusions (Fig. 5).

The volcanic and volcaniclastic terminologies used in this paper are those of Fisher and Schmincke (1984), Fisher and Smith (1991), and Sigurdsson et al. (2000). Following Fisher and Schmincke (1984), volcaniclastic refers to all fragmental rocks made dominantly of volcanic detritus; these include (1) pyroclastic fragmental deposits, inferred to have been directly fed from an eruption, e.g., pyroclastic fall, ignimbrites, autoclastic flow breccias; (2) reworked fragmental deposits, inferred to result from downslope reworking of unconsolidated eruption-fed fragmental deposits, e.g., block-and-ash-flow deposits commonly pass downslope into debris flow and fluvial deposits; and (3) epiclastic deposits, made of volcanic fragments inferred to have been derived from erosion of preexisting rock. When the distinctions cannot be made, the general term volcaniclastic is applied. Delicate pyroclastic detritus such as pumice, shards, or euhedral crystals cannot be derived from erosion of preexisting rock, so their presence in fluvial or debris flow deposits indicates that at least some of the deposit consists of reworked pyroclastic material, indicating broadly coeval explosive volcanism. Similarly, if a debris flow deposit is dominated by one volcanic clast type, it can be inferred to record reworking of a block-and-ash-flow deposit or flow breccia. However, the presence of a broad range of volcanic clast types is not proof of an epiclastic origin, because a wide variety of volcanic clast types can become incorporated into an eruption-triggered debris flow; in that case, a distinction between reworked and epiclastic cannot be made, and the deposit is simply a volcaniclastic debris flow deposit. Debris-flow deposits with blocks of welded ignimbrite, however, cannot be derived by any downslope reworking process known in outflow ignimbrite fields, and instead likely record erosion of preexisting rocks, so those can be classified as epiclastic (note that intracaldera ignimbrites commonly have blocks of welded ignimbrite cannibalized from the caldera wall during ongoing collapse; see discussion in Schermer and Busby, 1994).

The three formations in the Guazapares Mining District region are subdivided into 30 distinct lithologic units by outcrop and thin section characteristics, mineralogy, chemical composition, and inferred volcanic or sedimentary processes (Fig. 3C; Table 1). These lithologic units include volcanic rocks (e.g., lavas, ignimbrites), volcaniclastic rocks (e.g., sandstone, conglomerate, breccia), and hypabyssal intrusions (e.g., plugs, dikes). Modal point-count analyses were carried out for 39 samples, chosen to represent most of the volcanic and hypabyssal map units (Fig. 6). Reconnaissance whole-rock geochemical analyses were performed on 15 relatively unaltered samples of volcanic rock and hypabyssal intrusions from the Témoris and Sierra Guazapares formations (Fig. 7; Table 2).

Parajes Formation

The Parajes formation is primarily exposed in the eastern part of the study area; continuous stratigraphic sequences are found in the vicinity of Rancho de Santiago (Fig. 3A). The base of this formation is not exposed in the study area. The formation is composed of seven lithologically distinct silicic ignimbrites, with lesser locally interbedded sandstone, conglomerate, and reworked tuff (Figs. 6 and 8; Table 1). Individual ignimbrites are informally named in this study, and are distinguished based on phenocryst assemblages and outcrop characteristics such as degree of welding, weathering style, color, and percentage and type of pumice and/or fiamme and lithic fragments (Fig. 6; Table 1).

Description

Each ignimbrite of the Parajes formation has a densely welded to partially welded lower part that passes upward into a less welded to nonwelded top (Fig. 8A), forming a single cooling unit, as well as a single flow unit with normal coarse-tail grading of lithic fragments and inverse coarse-tail grading of pumice. Where the bases of ignimbrites are exposed, 0.5–2-m-thick basal vitrophyres are present. The ignimbrites are generally crystal poor to crystal moderate (<20%), with a dacitic phenocryst assemblage (no chemical analyses were done) consisting primarily of plagioclase and pyroxene phenocrysts, with minor amounts of hornblende, biotite, and quartz in some ignimbrites; sanidine is lacking in all of the ignimbrites of the Parajes formation (Fig. 6). The thickness of individual ignimbrites range from ∼20 to ∼210 m; the total thickness of the Parajes formation is ∼1 km (Table 1). Some ignimbrites appear to thicken due to ponding in paleotopographic lows (e.g., Rancho de Santiago [Tpr] and KM [Tpk] ignimbrites); ponded thicknesses are 2.5 times greater than nonponded parts of the same ignimbrite (Figs. 3A and 4; Table 1).

Each ignimbrite of the Parajes formation has distinguishing outcrop and/or compositional characteristics, described in ascending stratigraphic order (Fig. 6; Table 1). The Chepe, Ericicuchi, and Portero ignimbrites form the oldest continuous stratigraphic sequence, which is only found on the southwest (footwall) side of the Chapotillo fault in the Guazapares Mining District region (Fig. 3A). The Chepe ignimbrite (Tpc) is the only crystal-rich (∼30%) ignimbrite in the study area, with embayed quartz and biotite phenocrysts to 2 mm in diameter. The Ericicuchi ignimbrite (Tpe) has dark gray fiamme to 1 cm in length, typically with orange rims, and it has a mafic phenocryst assemblage that includes pyroxene, hornblende, and biotite. The Portero ignimbrite (Tpp) is characterized by a pink groundmass with eutaxitic texture in the densely welded lower portion, dark reddish-gray fiamme to 30 cm in length, and trace quartz phenocrysts.

The Puerto Blanco, Rancho de Santiago, KM, and Traza ignimbrites form a second, younger continuous stratigraphic sequence that is only found on the northeast (hanging wall) side of the Chapotillo fault (Fig. 3A); the depositional relationship between the two stratigraphic sequences on either side of the fault is not known, but is considered younger than the previously described sequence on the footwall based on the sense of fault offset (Fig. 4) and inferred regional correlations (described in the Discussion following). The base of the Puerto Blanco ignimbrite (Tpb) is not exposed; however, the exposed portion of its lower part, as well as its upper part, are nonwelded, with a welded middle. The Puerto Blanco ignimbrite (Tpb) has the greatest amount and size of lithic fragments (10%–40%, to 5 cm) compared to the other ignimbrites of the Parajes formation, with normal coarse-tail grading and upsection decrease in lithic fragments (from ∼40% to 10%); it also shows an upsection increase in phenocrysts (from <5% to 20%) and an upsection increase in fiamme, which are distinctively yellow. The Rancho de Santiago ignimbrite (Tpr) is similar in appearance and composition to the Portero ignimbrite (Tpp) described above, but has gray fiamme with dark gray rims (Fig. 8B); these are generally 3 cm (to 1 m) in length. It has a 2-m-thick basal vitrophyre at the contact with the underlying Puerto Blanco ignimbrite. The KM ignimbrite (Tpk) is similar to the underlying Rancho de Santiago ignimbrite (Tpr), but is distinguished by the presence of a brownish-red, ∼10-m-thick, crystal-poor (<5%) lower welded section and an overall lower lithic fragment content (5%–10%). The youngest unit of the Parajes formation is the Traza ignimbrite (Tpt), which is similar in appearance to both the Chepe and Puerto Blanco ignimbrites, but is distinguished by having gray fiamme and a moderate crystal content (20%) with trace quartz and no biotite.

Sedimentary rocks occur locally between ignimbrite units. An ∼150-m-thick sequence of reworked tuff and cross-bedded sandstone with fragments of tuff and pumice (Tps) is between the Rancho de Santiago ignimbrite (Tpr) and KM ignimbrite (Tpk) southwest of the Arroyo Hondo–Puerto Blanco fault (Figs. 3A and 4). Also present at this stratigraphic interval in the Mesa de Cristal area east of Rancho de Santiago (Supplemental Fig. 1 [see footnote 1]) is a monomictic matrix-supported pebble to cobble conglomerate with welded ignimbrite clasts similar in appearance to ignimbrites of the Parajes formation (Fig. 8C). In addition, a thin (<1 m) layer of fine- to medium-grained sandstone is present along the contact between the Ericicuchi ignimbrite (Tpe) and Portero ignimbrites (Tpp) (Fig. 8D).

Interpretation

The Parajes formation represents medial facies of silicic outflow ignimbrite sheets, based on the sheet-like geometry of the flow units, the moderate thicknesses of flow units (each <∼200 m thick, locally thicker where ponded by paleotopography), the presence of welding textures and vitrophyres, and the lack of associated lithic lag breccias. No caldera or vent-proximal lithofacies have been identified for these outflow ignimbrites, so the locations of their sources are not known. However, lithic fragments and fiamme within in the Rancho de Santiago ignimbrite (Tpr) increase in size eastward, suggesting that the source for this ignimbrite is located toward this direction. Based on flow thicknesses and degree of welding relative to distance from the source recorded in large-volume silicic ignimbrites in the western U.S. (e.g., Smith, 1960; Lipman, 2007), the ignimbrites of the Parajes formation were likely erupted from calderas located within 50–100 km. The large size and concentration of lithic fragments within the Puerto Blanco ignimbrite (Tpb) are suggestive of a somewhat closer source.

Sedimentary rocks (Tps) interbedded with the ignimbrites of the Parajes formation record both erosion of welded units and reworking of unconsolidated pyroclastic debris, with deposition by fluvial and debris flow processes (Figs. 8C, 8D). The debris flow deposits are massive, poorly sorted matrix-supported conglomerates, while fluvial sandstones and fluvially reworked tuffs have trough cross-bedding, normal grading, and well-sorted granule conglomerate lenses. The clasts in these sedimentary rocks are predominantly silicic volcanic fragments, including welded and nonwelded tuff and pumice (e.g., Fig. 8C); there are no andesitic volcanic fragments in these rocks. This suggests that the Parajes formation ignimbrites were uplifted and partly eroded prior to deposition of overlying andesitic rocks of the Témoris formation.

Témoris Formation

The Témoris formation overlies the Parajes formation in angular unconformity, and is best exposed in the central and western portions of the study area in the vicinity of Puerto La Cruz and Guazapares (Fig. 3). This formation is primarily composed of mafic to intermediate composition lavas (flow-banded and/or vesicular) and hypabyssal intrusions, intercalated conglomerates, breccias, and sandstones dominated by mafic to intermediate volcanic lithic fragments, and lesser thin silicic nonwelded ignimbrites and reworked silicic tuff (Figs. 6, 7, and 9; Tables 1 and 2). This formation has undergone mild hematitic and propylitic alteration, with infilling of vesicles and autoclastic flow breccia interstices with zeolite minerals; the most intense alteration is in the rocks within the Guazapares fault zone (Fig. 3B).

Description

The basal deposits of the Témoris formation consist of sandstones with silicic tuff fragments (Ttss), matrix- to clast-supported breccias with welded silicic ignimbrite boulders (Ttdi, Ttdt; Figs. 9A, 9B), and lesser interbedded silicic tuffs (Ttt). The welded ignimbrite clasts were derived from the underlying Parajes formation, indicating continued erosion of this formation. One ignimbrite of the Parajes formation (Portero ignimbrite, Tpp), located east of Ericicuchi near 12R 775504E 3024974N (Universal Transverse Mercator coordinates, North American Datum 1927; Fig. 3A), contains clastic dikes directly below the Parajes–Témoris formation contact. These dikes are composed of overlying Témoris formation sandstone that infills fissures formed in the top of the Portero ignimbrite.

The Témoris formation is subdivided into three sections based on volcanic rock compositions and types (Figs. 6, 7, and 10; Tables 1 and 2; Supplemental Table 1,2). These subdivisions have gradational contacts and consist of (1) a lower section of pyroxene-plagioclase ± olivine-bearing amygdaloidal basalt, basaltic andesite, and andesite lavas and autoclastic flow breccias (Ttba, Ttb; Figs. 9C, 9D); (2) a middle section of pyroxene-plagioclase–bearing flow-banded andesite lavas (Tta; Fig. 9E); and (3) an upper section of several thin (<5-m-thick) primary and reworked rhyolite ignimbrites (Tti; Figs. 9F, 9G); this upper section is only locally preserved beneath the angular unconformity with the overlying Sierra Guazapares formation. Conglomerates, breccias, and sandstones with well-sorted gravel lenses and trough cross-bedding are interbedded with and laterally interfinger with all of the volcanic rocks listed above (Figs. 9B, 9F–9K, and 10; Table 1). These volcaniclastic deposits contain detritus similar in composition to the interstratified lavas and ignimbrites: amygdaloidal and flow-banded basaltic andesite to andesite clasts dominate the lower and middle sections of the Témoris formation (Ttda, Ttsa, Ttdt), while the upper section of the Témoris formation has mixed mafic-intermediate and silicic volcanic clasts, including pumice fragments in tuffaceous sandstones and tuffaceous conglomerates (Ttss, Ttds, Tti). Lavas and autoclastic flow breccias locally infill channels incised into the underlying sedimentary rock (Fig. 9E), and wet sediment-magma interactions (peperitic) are locally observed where lavas were apparently emplaced over wet sand (Fig. 9L).

In the area around Témoris, the Témoris formation thickens from ∼100–400 m to >700 m (Fig. 3; Supplemental Fig. 1 [see footnote 1]). There, basalt to andesite lavas of the lower and middle sections of the Témoris formation are heavily hematite stained and are complexly intruded by numerous andesitic dikes and aphyric hypabyssal rocks (Ttv; Table 1).

Interpretation

The rocks of the Témoris formation are interpreted as the products of vent to proximal mafic to intermediate composition magmatism and distal silicic ignimbrite volcanism. Deposition in a terrestrial environment, likely part of alluvial fan systems (e.g., Kelly and Olsen, 1993; Blair and McPherson, 1994; Hampton and Horton, 2007; Murray et al., 2010), is indicated by interstratified matrix-supported debris flow breccias and conglomerates (Ttda, Ttds, Ttdi), clast-supported avalanche and/or talus breccias (Ttdt), well-sorted stratified and cross-bedded fluvial sandstones and conglomerates (Ttas, Ttss), and some lavas infilling fluvial channels and forming peperites within them (Fig. 9). The composition of fragments in the fluvial and debris flow deposits is similar to that of the interstratified volcanic rocks, indicating intrabasinal reworking of eruptive products. The upper section of rhyolite ignimbrites in the Témoris formation likely erupted from distal sources, because they are thin and nonwelded, with a high proportion of interstratified fluvially reworked tuff (Tti; Figs. 9F, 9G; Table 1).

We interpret the Témoris area to be the site of an andesitic volcanic center in the Témoris formation, based on dramatic thickening of the lava section, abundant plugs and dikes, and increased alteration (Fig. 3; Supplemental Fig. 1 [see footnote 1]). This andesitic volcanic center, roughly defined by map unit Ttv, greatly thickens toward its subvolcanic intrusion–dominated core located along the ridge east of Témoris, with a minimum volume of 9 km3 based on the mapped area and exposed thickness (Fig. 3; Supplemental Fig. 1 [see footnote 1]). A feeder dike emanating from the volcanic center can be traced upward into an andesitic lava flow in the Puerto La Cruz area (Fig. 10). In addition, andesitic dikes crosscut rocks of the Témoris formation away from the volcanic center, locally along faults.

Sierra Guazapares Formation

The Sierra Guazapares formation comprises much of the central and northwestern part of the study area, with best exposures located along the NS–trending ridge east of Guazapares (Fig. 3; Supplemental Fig. 1 [see footnote 1]). This formation is composed of plagioclase-biotite ± quartz ± sanidine–bearing rhyolitic ignimbrites, rhyolite lavas, flow-banded rhyolite hypabyssal intrusions, and lesser silicic volcaniclastic deposits (Figs. 3, 6, 7, and 11; Table 1). The Sierra Guazapares formation is flat-lying to gently dipping (<10°) and overlies the Témoris formation in low to moderate angular unconformity (Fig. 12A). The Sierra Guazapares formation is >200 m thick; it is not known how much of the formation is preserved, because the top is eroded. The formation locally infills lows cut into older stratigraphic units, recording paleotopography produced by erosion or faulting.

Description

The dominant lithofacies of the Sierra Guazapares formation is massive to stratified nonwelded to partially welded rhyolite ignimbrites (Tst; Fig. 11A; Table 1). Locally, these ignimbrites show evidence of reworking, including sorting and rounding of lithic, pumice, and crystal fragments, stratification and cut-and-fill structures, and small- to medium-scale cross-lamination (Fig. 11B).

Very large scale cross-bedded rhyolitic ignimbrites (Tsxi) form a distinctive lithofacies of the Sierra Guazapares formation (Fig. 11C, 11D; Table 1). These deposits are mainly restricted to a linear belt ∼11 km long and 3 km wide within and immediately adjacent to the Guazapares fault zone–La Palmera fault (Fig. 3B), and laterally grade away from this linear belt into massive to stratified ignimbrites (Tst; Figs. 3 and 5; Supplemental Fig. 1 [see footnote 1]). The very large scale cross-bedded ignimbrites have average set heights of ∼5 m; some are as great as ∼20 m (Figs. 11C, 11D). The cross-bedding in these ignimbrites is defined by alternating lithic-rich (>50%) and lithic-poor (<30%) layers (Fig. 11D). The lithic fragments are very coarse grained, with blocks to 50 cm in diameter; these are dominantly mafic to intermediate volcanic rocks likely derived from the underlying Témoris formation (Fig. 11E). The matrix of the very large scale cross-bedded ignimbrites is an unsorted mixture of angular pumice, euhedral crystals, and glass shards, and the very large scale cross-beds lack internal laminations, sorting, or other fine-scale sedimentary structures indicative of reworking by water.

Rhyolite lavas (Tsl) and hypabyssal intrusions (Tsi, Tsiw, Tsib) occur in the same linear belt along the Guazapares fault zone–La Palmera fault as the very large scale cross-bedded ignimbrites, and also occur along additional NNW–striking faults in the region (Figs. 2 and 3; Supplemental Fig. 1 [see footnote 1]). The silicic hypabyssal intrusions (Fig. 11F) are typically plugs with related dikes that intrude the ignimbrites (Tst, Tsxi) of the Sierra Guazapares formation, and some of the plugs pass continuously upward into rhyolite lavas (Tsl) (Fig. 12B). The rhyolite lavas typically overlie the ignimbrites, but are locally interstratified (Fig. 11G).

In addition to silicic ignimbrites, lavas, and plugs, the Sierra Guazapares formation also includes a volcaniclastic unit (Tsv) in the Monte Cristo area (Fig. 3B). This unit includes a rhyolitic dome-collapse breccia associated with a rhyolite dome complex (Tsiw) that overlies and interfingers with normal graded sandstones, mudstones with soft-sediment deformation features, and moderately to poorly sorted sandstone with trough cross-bedding and cut-and-fill structures (Table 1; Supplemental Data File3).

Interpretation

We interpret the very large scale cross-bedded ignimbrites to be vent-proximal lag breccias deposited from energetic, turbulent pyroclastic density currents erupted during several events from a major fissure vent along the Guazapares fault zone–La Palmera fault (Figs. 2 and 3). Their linear map distribution indicates they were erupted from fissure vents, rather than a central vent, and likely formed coarse-grained ramparts. Interstratified silicic lavas and plugs are concentrated along either side of the same fault zone, in the same linear map distribution, supporting the interpretation that the Guazapares fault zone–La Palmera fault controlled the siting of an 11-km-long silicic fissure vent.

The very large scale cross-bedded ignimbrites (Tsxi) represent a gradation between the pyroclastic surge and pyroclastic flow end members of pyroclastic density current classification (e.g., Fisher and Schmincke, 1984; Branney and Kokelaar, 2002). The abundant very coarse-grained lithic layers in these cross-bedded ignimbrites are similar to lithic lag breccias described from other vent to proximal ignimbrites (e.g., Fisher and Schmincke, 1984; Carey, 1991; Freundt et al., 2000; Branney and Kokelaar, 2002). The angularity of the lithic components and their derivation from the underlying Témoris formation suggests that they were fragmented and incorporated into the pumice-rich pyroclastic material as it ascended through the vent. However, the very large scale cross-stratification is unusual for ignimbrite lithic lag breccias. Very large scale cross-bedding has been described in vent to proximal ignimbrites in other localities, including Mount St. Helens (e.g., Rowley et al., 1985), Tenerife (e.g., Brown and Branney, 2004), Santorini (e.g., Gertisser et al., 2009), and Volcán Villarrica, Chile (e.g., Silva Parejas et al., 2010); however, these cross-bedded ignimbrites are generally dominated by ash- to lapilli-sized material and do not contain the large lithic blocks such as in the very large scale cross-bedded ignimbrite (Tsxi) described here.

Given their coarse-grained nature and large-scale cross-stratification, the very large scale cross-bedded ignimbrites (Tsxi) suggest deposition from highly energetic low-concentration pyroclastic flows in a vent to proximal setting, due to the high amount of turbulent energy required to produce these very large bedforms while transporting the large lithic fragments (e.g., Wright et al., 1981; Carey, 1991; Branney and Kokelaar, 2002). The gradational lateral transition from very large scale cross-bedded ignimbrites (Tsxi) into massive to stratified ignimbrites (Tst) within 1–2 km of the Guazapares fault zone–La Palmera fault (Fig. 3; Supplemental Fig. 1 [see footnote 1]) suggests decreased turbulence and an increased pyroclastic sedimentation rate farther from the vent.

Lithostratigraphic Summary

The three informal formations defined in the Guazapares Mining District region represent three distinct volcanic episodes:

  1. The Parajes formation consists of welded to nonwelded silicic outflow ignimbrite sheets that were erupted from caldera sources within 50–100 km of the study area, with intercalated volcaniclastic rocks derived from erosion of these ignimbrites.

  2. The lower and middle Témoris formation consists dominantly of locally erupted mafic to intermediate composition lavas and associated subvolcanic intrusions, including an andesitic center in the area around Témoris, as well as fault-controlled dikes that likely fed flows outside the main center. The lower and middle Témoris formation also contains interstratified volcaniclastic fluvial and debris flow deposits. Detritus at the base of the formation that was derived from the underlying Parajes formation silicic ignimbrites records erosion of that formation, perhaps along fault scarps. In contrast, the andesitic detritus that dominates higher in the section could record resedimentation of primary eruptive products, such as the collapsing fronts of lavas, or block-and-ash flows or tephras, although erosion of constructional volcanic features or fault scarps is also probable, particularly for polymictic deposits. The distal thin nonwelded silicic ignimbrites and sedimentary rocks of the upper section of the Témoris formation record waning of local mafic to intermediate volcanism prior to the onset of local silicic volcanism, and indicate continuing or recurring silicic ignimbrite-forming eruptions from distant sources.

  3. The Sierra Guazapares formation records the local eruption of silicic volcanic rocks within the Guazapares Mining District region. These include ignimbrites with vent facies lithic lag breccias that formed very large scale cross-beds along either side of an 11-km-long fault-controlled fissure, which also controlled the emplacement of silicic plugs and eruption of silicic lavas. The Sierra Guazapares formation also includes silicic dome-collapse breccias and interstratified silicic lavas and volcaniclastic rocks that interfinger with lacustrine deposits preserved in a half-graben basin.

GEOLOGIC STRUCTURES AND BASIN DEVELOPMENT

The main geologic structures in the Guazapares Mining District region are primarily NNW–trending normal faults, including the Guazapares fault zone and faults to the northeast of Témoris (Figs. 2, 3, and 4; Supplemental Fig. 1 [see footnote 1]). The Guazapares fault zone extends from Témoris northward to the Monte Cristo mining claim, and is a complex system of NNW–striking normal faults with numerous splays that dip both east and west, with several changes of fault dip polarity along strike (Fig. 3B; Supplemental Fig. 1 [see footnote 1]; Supplemental Data File [see footnote 3]). This fault zone hosts the majority of mineralization within the mining district (e.g., Gustin, 2012). The normal faults located northeast of Témoris have significant vertical offset and bound half-graben basins (Figs. 3A and 4). Although many of the half graben faults die out upsection, making them relatively easy to recognize, faults of the Guazapares fault zone were reactivated many times, and cut all formations (Figs. 2 and 5), making their earlier history more difficult to document.

Synvolcanic Half-Graben Basins

Several normal faults bound half-graben basins in the Guazapares Mining District region, including the NNW–striking, west-dipping Arroyo Hondo–Puerto Blanco, La Palmera, and Agujerado faults; the NNE–striking, west-dipping Rancho de Santiago fault; and the NNW–striking, east-dipping Sangre de Cristo fault (Figs. 3 and 4). In general, these half-graben basins contain sedimentary and volcanic deposits that thicken and/or coarsen toward basin-bounding normal faults, which either terminate at the fault or thin onto the footwall, indicating synextensional deposition (Fig. 4). Angular unconformities occur between each of the formations, and fanning dips (e.g., Fig. 12A) indicate synextensional deposition, with the Parajes and Témoris formations dipping more steeply than the gently dipping to flat-lying Sierra Guazapares formation.

The upper part of the Parajes formation (younger than the Puerto Blanco ignimbrite [Tpb]) was likely deposited into synvolcanic extensional basins, based on the variable thicknesses of individual outflow ignimbrite sheets and distribution of interbedded sedimentary rocks across faults. Evidence for synextensional deposition includes (1) the presence of reworked tuff, sandstone, and conglomerate (Tps) above the Rancho de Santiago ignimbrite (Tpr) within the half-graben basin adjacent to Arroyo Hondo–Puerto Blanco fault and in the Mesa de Cristal area, which thicken toward and terminate at faults and are not present on the footwall blocks, and (2) thickening of the Rancho de Santiago ignimbrite (Tpr) within the half-graben basin bounded by the Arroyo Hondo–Puerto Blanco fault (∼200 m thick), relative to the ∼80 m thickness on the footwall block (Figs. 3A, 4, and C8C; Supplemental Fig. 1 [see footnote 1]).

Synextensional deposition of the Témoris formation is evident in the three half-graben basins bounded by the La Palmera, Agujerado, and Rancho de Santiago–Arroyo Hondo–Puerto Blanco faults (Figs. 3A and 4). In these basins, the Témoris formation is deposited in angular unconformity on the more steeply dipping Parajes formation, and the thickness and average grain size of sedimentary deposits increases dramatically eastward toward each of the basin-bounding normal faults (Fig. 4). In the half-graben bounded by the Agujerado fault, a coarse-grained debris flow (Ttds) deposited proximal to the basin-bounding fault interfingers basinward with finer grained sandstone and siltstone (Ttss; Figs. 3A and 4B).

The largest of the three synvolcanic half-grabens of the Témoris formation is the Rancho de Santiago basin, which is unique in that it developed as a half-graben bounded by two west-dipping normal faults on the eastern side of the basin; the southernmost fault is the NNE–striking Rancho de Santiago fault, which is crosscut on the north end by the NNW–striking Arroyo Hondo–Puerto Blanco fault (Fig. 3A). In this basin, a clast-supported breccia (Ttdt) containing large (to 4 m) intermediate volcanic and lesser silicic ignimbrite rock fragments, as well as slide blocks of fractured but intact sedimentary strata to 15 m thick and 20 m long, is adjacent to the Rancho de Santiago fault (Figs. 3A, 4C, and 9B; Table 1). This breccia is interpreted as talus and avalanche deposits that were shed from the uplifted footwall fault scarps directly into the half-graben basin to the west.

Synvolcanic extension during emplacement of the Sierra Guazapares formation is recorded by silicic dome-collapse deposits, reworked tuffs, and fluvial-lacustrine deposits (Tsv) preserved within the half-graben basin bounded by the Sangre de Cristo fault in the Monte Cristo mining claim at the northern mapped end of the Guazapares fault zone (Fig. 3B; Table 1; Supplemental Data File [see footnote 3]). In this basin, a rhyolitic breccia thickens and coarsens toward the Sangre de Cristo fault and interfingers basinward with basal lacustrine sedimentary rocks. Additional evidence of synvolcanic extension in this basin includes the development of a normal fault within the hanging-wall block of the Sangre de Cristo fault that provided a conduit for a small silicic plug and coulee (Tsl) to intrude and flow over the actively depositing volcaniclastic unit (Tsv; Fig. 3B; Supplemental Data File [see footnote 3]).

Relative Timing and Amount of Extensional Deformation

Extensional deformation in the Guazapares Mining District region was concurrent with deposition of at least the upper part of the Parajes formation, the Témoris formation, and the Sierra Guazapares formation, with continued extension following deposition of the Sierra Guazapares formation. Pre–Sierra Guazapares formation extension is suggested by the low to moderate angular unconformities between the Témoris formation and the underlying Parajes formation and the overlying Sierra Guazapares formation (Fig. 12A). Older normal faults that offset the Parajes and Témoris formations localized the vents and silicic plugs of the Sierra Guazapares formation, which utilized these preexisting structures as pathways for magma accent (e.g., La Palmera and La Escalera faults, Guazapares fault zone; Figs. 2, 3, and 12B; Supplemental Fig. 1 [see footnote 1]). In addition, unfaulted Sierra Guazapares formation lavas bury some faults that offset the Parajes and Témoris formations (Figs. 2 and 3).

Further evidence of pre–Sierra Guazapares formation extension includes greater fault offsets of the older formations compared to offset of the Sierra Guazapares formation (Figs. 3A and 4). The minimum vertical displacement of the base of the Témoris formation across the Ericicuchi fault is >300 m, ∼110 m across the Agujerado fault, and >450 m across the La Palmera fault (Fig. 4). In comparison, these faults offset the Sierra Guazapares formation to a lesser degree: the base of the Sierra Guazapares formation is only offset ∼60 m across the Ericicuchi fault, ∼30 m across the Agujerado fault, and ∼100 m across the La Palmera fault (Fig. 4). This shows that a significant amount of extensional deformation (at least 350 m vertical displacement) occurred prior to the eruption of the Sierra Guazapares formation.

A minimum of 20% total horizontal extension is estimated in the Guazapares Mining District region (for the area shown in Fig. 4), based on the vertical displacement of stratigraphic units across normal faults. This amount of extension is significantly lower than that of the Gulf Extensional Province to the west in Sonora, where ∼90% extension is estimated to have occurred (Gans, 1997). The structural style also differs between these two areas; high-angle normal faults are found in the Guazapares Mining District region, while highly extended core complexes are located in Sonora (e.g., Gans, 1997; Wong et al., 2010).

Although not directly quantifiable, several faults within the Guazapares Mining District region appear to accommodate considerable amounts of deformation based solely on the juxtaposition of stratigraphic units. The La Palmera fault has significant vertical offset (over 450 m) based on the offset of the Parajes–Témoris formation contact; the Parajes formation is exposed on the footwall, but is not exposed in the hanging wall, which suggests that it is deeply buried beneath Témoris formation deposits there (Figs. 2 and 4). A distinct lithologic boundary in the Parajes formation occurs across the Chapotillo fault, as the younger outflow ignimbrite sheets in the hanging wall of this fault are not exposed on the footwall to the southwest (Fig. 3A). Postdepositional drag folding related to normal fault deformation is observed in the Témoris formation adjacent to many of the NNW–striking faults with significant offset (e.g., La Palmera, Agujerado, and La Escalera faults; Figs. 3A and 4); the underlying Parajes formation has small-scale normal faulting to accommodate this deformation.

AGE CONSTRAINTS

Methodology

We report new U-Pb zircon ages from each of the three informally defined formations, providing constraints on the age of the previously undated volcanic rocks of the Guazapares Mining District region. Laser ablation–inductively coupled plasma–mass spectrometry (LA–ICP-MS) U-Pb analyses were performed at the Laboratorio de Estudios Isotópicos, Centro de Geociencias, Universidad Nacional Autónoma de México on zircons separated from 13 silicic rock samples (Fig. 13; Table 3; Supplemental Table 2,4). The zircons were hand-picked under binocular microscope, mounted in an epoxy cast, polished, and imaged by cathodoluminescence (CL). The zircons selected for U-Pb geochronology were analyzed following the procedure reported by Solari et al. (2010), employing a Resonetics M050 excimer laser ablation workstation coupled to a Thermo XSeries II ICP-MS. Based on CL imaging, one ablation site was selected on each zircon analyzed, located either in the middle, near rim, or core of the crystal (Supplemental Fig. 2,5). The Plešovice standard zircon (ca. 337 Ma; Sláma et al., 2008) was used as a bracketing standard, interdispersed and measured after every five unknown zircons. The observed uncertainties of the 206Pb/238U, 207Pb/206Pb, and 208Pb/232Th ratios, during the different sessions in which the current samples were analyzed, as measured on the Plešovice standard zircon, were 0.65%, 1.0%, and 1.1%, respectively. These values are quadratically propagated to the quoted uncertainties of the unknown zircons, to take into account the heterogeneities of the natural standard zircon. A second standard (NIST 610) is used to recalculate the elemental concentrations for each zircon, measured together with the isotopes of interest for U-Pb geochronology. The common Pb correction cannot be performed measuring the 204Pb isotope with the current setup; common Pb is evaluated using the 207Pb/206Pb ratio, graphing the results using Tera-Wasserburg diagrams (Tera and Wasserburg, 1972). If a correction is needed, the algebraic method of Andersen (2002) is used. Filters are then applied to reduce outliers: largely discordant analyses (e.g., >50% discordant) and those with >4% 1σ error on the corrected 206Pb/238U ratio are eliminated. A further screening is applied to check for possible microscopic inclusions of minerals other than zircons that could have been inadvertently hit during the analysis. This screening is performed during data reduction, employing a script written in R (UPb.age; Solari and Tanner, 2011). Additional screenings are performed, checking for analyses with high P and light rare earth elements, which could be indicative of apatite inclusions, and those few analyses that present high concentrations of U and Th (generally >1000 ppm), which could yield to a Pb loss and a consequent discordant or, in any case, younger and geologically meaningless ages.

Concordia plots, probability density distribution and histogram plots, mean age, and age-error calculations were performed using Isoplot v. 3.70 (Ludwig, 2008). The mean 206Pb/238U age is especially useful for the Tertiary ages presented here, because the 207Pb measurement is problematic in these young zircons and the consequent uncertainty on the 207Pb/206Pb ratio is not a good indicator of geologically meaningful discordance. In Tertiary zircons, it is also common to observe scattering of the mean 206Pb/238U ages that yields MSWD (mean square of weighted deviates) values that are largely >1, an indication that a mixed age population possibly exists. An example of this scenario was presented by Bryan et al. (2008). In order to recognize possible different age components in samples that showed an initial MSWD of >3, the deconvolution method, based on the mixture modeling method of Sambridge and Compston (1994), was implemented in Isoplot.

When two mixture components are recognized, their respective mean 206Pb/238U ages are plotted together with errors and recalculated MSWD. The mean 206Pb/238U age of the older mixture component in a sample represents the crystallization age of inherited zircons within the host magma, while the younger mean 206Pb/238U age population represents the phenocryst crystallization age of the sample. This youngest age population of each sample is interpreted as the preferred eruption or emplacement age of the rock, as it is consistent (within error) with stratigraphic relationships in the study area. Age results are presented in the following and summarized in Figure 13 and Table 3; detailed analytical data are given in Supplemental Table 2 (see footnote 4).

Parajes Formation

Three samples were dated from the Parajes formation (Figs. 13A–13C; Table 1), including two samples from the Ericicuchi ignimbrite (Tpe) and one sample from the Puerto Blanco ignimbrite (Tpb). Sample BM100306–1 is Ericicuchi ignimbrite (Tpe), which has separated zircons that are bipyramidal to short and stubby, and to 220 µm in length. Under CL, the zircons show uniform areas with limited luminescence; in a few cases, oscillatory zoning is present around possible inherited cores. The U-Pb geochronological analysis, as well as the screening and filtering, shows the presence of inherited cores that are slightly discordant but older than 40 Ma. Most of the analyzed, nearly concordant crystals range from ca. 26 Ma to 31 Ma (Fig. 13A). Two zircon age populations can be distinguished: the oldest population has a mean age of 29.59 ± 0.33 Ma (n = 22, MSWD = 1.6), whereas the youngest has a mean age of 27.55 ± 0.33 Ma (n = 6, MSWD = 1.04). A second sample from the Ericicuchi ignimbrite (sample BM100306–3; Fig. 13B) yielded fewer and smaller zircon crystals (to 180 µm in length) that are euhedral to subhedral with a prevalence of stubby morphologies with short pyramidal terminations. These zircons show bright CL zoning around darker cores. The U-Pb geochronology for this sample also revealed the presence of inherited cores of ca. 76, 50, and 38 Ma and two main zircon age populations consisting of an older grouping having a mean age of 29.01 ± 0.32 Ma (n = 16, MSWD = 1.6) and a younger grouping with a mean age of 27.04 ± 0.74 Ma (n = 6, MSWD = 2.5). The third sample of the Parajes formation is from the Puerto Blanco ignimbrite (Tpb; sample BM100306–6; Fig. 13C). The zircons in this sample are larger (to 260 µm in length) with elongated shapes that are mostly prismatic with well-developed pyramids. Under CL they show evident bright rims developed outside darker zones. The dated zircons define a homogeneous group with few outliers and have a mean age of 27.58 ± 0.26 Ma (n = 31, MSWD = 2.6).

Témoris Formation

Three samples of silicic tuffs from basal and upper sections of the Témoris formation were dated (Figs. 13D–13F; Table 3). Sample BM100305–4 was collected from the basal section of the Témoris formation (Fig. 13D) and has zircons to 300 µm in length that are prismatic and elongated. Under CL, the zircons are characterized by darker cores surrounded by bright zones. Despite similar crystal morphologies, two zircon age populations are identified; the oldest group has a mean age of 29.73 ± 0.70 Ma (n = 11, MSWD = 2.1), whereas the youngest mean age is 27.27 ± 0.33 Ma (n = 18, MSWD = 1.7). Sample BM100304–5 was collected from the upper section of the Témoris formation (Tti; Fig. 13E) and has zircons that are indistinguishable in size, morphology, and CL imaging from those of the previous sample. Two well-constrained zircon age populations are also defined; the oldest has a mean age of 26.07 ± 0.25 Ma (n = 20, MSWD = 1.5), whereas the youngest mean age is 24.58 ± 0.19 Ma (n = 12, MSWD = 0.96). Sample BM100305–1 is from the uppermost section of the Témoris formation (Tti), ∼35 m below the Sierra Guazapares formation contact (Figs. 10 and 13F). Its zircons are also prismatic and very elongated, although they are somewhat smaller (to 200 µm in length) in this sample. Under CL, the zircons are also characterized by darker cores surrounded by bright zones. U-Pb analyses identified two zircon age populations in this sample; the oldest group yields a mean age of 25.58 ± 0.29 Ma (n = 17, MSWD = 1.6), whereas the youngest group yields a mean age of 24.14 ± 0.25 Ma (n = 10, MSWD = 0.49).

Sierra Guazapares Formation

Seven samples from the various lithologies of the Sierra Guazapares formation were chosen for U-Pb geochronology (Figs. 13G–13M; Table 3). Sample BM100304–4 was collected from the very large scale cross-bedded ignimbrite unit (Tsxi; Fig. 13G). It has somewhat small (to 150 µm) euhedral zircons that range in shape from prismatic to stubby and bipyramidal morphologies. CL imaging is not different from the previously described samples, although cores are not as evident as in other samples. The dated zircons define only one coherent group, in which the mean age is 24.66 ± 0.24 Ma (n = 19, MSWD = 1.3).

Three rhyolite lava (Tsl) samples were analyzed. Sample BM100307–1 (Fig. 13H) is characterized by prismatic euhedral zircons (to 300 µm in length) with the same CL characteristics as those previously described. Two zircon age groups are also defined; the oldest group yields a mean age of 25.78 ± 0.27 Ma (n = 17, MSWD = 1.7), whereas the mean age of the youngest group is 23.72 ± 0.22 Ma (n = 5, MSWD = 0.42). Sample BM100305–2 (Fig. 13I) also has prismatic zircons (to 200 µm in length) with most showing oscillatory zoning. U-Pb zircon dating of this sample defines two age populations; the oldest group with a mean age of 25.69 ± 0.32 Ma (n = 23, MSWD = 1.5) and the youngest group with a mean age of 23.92 ± 0.29 Ma (n = 8, MSWD = 0.94). Sample BM100304–1 (Fig. 13J) was collected from a small lava in the Monte Cristo area (Fig. 3B). It also has prismatic zircons (to 180 µm in length), although in this sample they are somewhat more needle shaped. Two age populations are identified in this sample; the oldest has a mean age of 25.07 ± 0.24 Ma (n = 22, MSWD = 1.5), whereas a few grains define the youngest group with a mean age of 22.94 ± 0.25 Ma (n = 3, MSWD = 0.18).

Two samples collected from rhyolite plugs (Tsi) were analyzed. Sample BM080717–3 (Fig. 13K) is characterized by stubby to bipyramidal zircons (to 150 µm in length) that are sector zoned under CL. Its U-Pb dating yields only one age group, with a mean age of 25.03 ± 0.31 Ma (n = 18, MSWD = 1.7). The zircons belonging to the sample BM100305–3 (Fig. 13L) are prismatic and large (to 340 µm in length). The U-Pb dating yields a homogeneous age group, with a mean age of 24.61 ± 0.22 Ma (n = 23, MSWD = 1.5).

Sample BM100304–2 was collected from a rhyolitic breccia locally exposed in the Monte Cristo area (Tsv; Figs. 3B and 13M). Its zircons are prismatic and large (to 250 µm in length). Two age populations are recognized; the oldest group has a mean age of 25.76 ± 0.45 Ma (n = 9, MSWD = 1.9), whereas the youngest group has a mean age of 24.17 ± 0.17 Ma (n = 24, MSWD = 1.6).

Age Interpretations

Previous dating of silicic volcanic rocks in the Sierra Madre Occidental using zircon U-Pb LA-ICP-MS showed that zircon ages are occasionally older (to 1–4 Myr) than the ages obtained from the same rocks using K/Ar and 40Ar/39Ar dating methods (Bryan et al., 2008). The older zircon ages in their study are attributed to the presence of antecrysts, a term proposed by Charlier et al. (2004) to describe crystals that predate the crystallization and eruption of a host magma, but formed during an earlier phase of related magmatism. In a region of long-lived magmatism like the Sierra Madre Occidental, the antecryst ages could predate the phenocryst age by more than 10 Myr, making it difficult to distinguish antecrysts from xenocrysts (Bryan et al., 2008). In addition, the occurrence of antecrysts tends to be greater in the younger silicic volcanic rocks of a sequence, when the probability of remelting partially molten or solidified upper crustal rocks formed during a preceding magmatic phase is higher (Bryan et al., 2008).

The presence of antecrysts in a zircon population for a sample will tend to produce initial MSWD values much greater than unity and probability density function curves of zircon ages that are positively skewed and asymmetric, and/or have broad, bimodal, or polymodal peaks. In comparison, a well-defined unimodal peak likely indicates the crystallization age of phenocrysts with limited antecrysts, which is a close approximation to the eruption age of the host magma (Charlier et al., 2004; Bryan et al., 2008).

The ages obtained for most of the samples dated for this study in the Guazapares Mining District region suggest the presence of antecrysts in the zircon population. The probability density function curves tend to be positively skewed and asymmetric, and several have broad or bimodal peaks (Fig. 13; Supplemental Table 2 [see footnote 4]). The oldest zircon population in a sample represents the crystallization age of antecrysts, which generally correspond to zircons with crystal core to middle ablation sites. In comparison, the youngest zircon population indicates the age of phenocryst crystallization and typically represents the zircons with middle to near-rim ablation sites. The antecryst age populations in these samples tend to be ∼1.5–2 Myr older than the phenocryst age populations (Table 3); antecryst ages tend to cluster around 29.5 Ma for samples from the Parajes formation and 25.5 Ma for samples from the overlying Témoris and Sierra Guazapares formations.

DISCUSSION

Volcanic and Tectonic Evolution

The new geologic mapping and geochronology presented in this study show that the three informal formations in the Guazapares Mining District region (Fig. 5) record Late Oligocene to Early Miocene synextensional volcanic activity during the mid-Cenozoic ignimbrite flare-up in the northern Sierra Madre Occidental: (1) the synextensional deposition of outflow ignimbrite sheets (Parajes formation) ca. 27.5 Ma, which were likely erupted from calderas ∼50–100 km from the study area; these overlap in time with the end of peak ignimbrite flare-up volcanism to the east; (2) synextensional growth of an andesitic volcanic center (Témoris formation) between ca. 27 Ma and ca. 24.5 Ma; and (3) synextensional silicic fissure magmatism (Sierra Guazapares formation), including vent facies ignimbrites, lavas, and intrusions, between ca. 24.5 and ca. 23 Ma (Fig. 14).

Stratigraphic and structural evidence show that the outflow ignimbrite sheets of the Parajes formation younger than the 27.58 ± 0.26 Ma Puerto Blanco ignimbrite (Tpb) were deposited in a developing half-graben basin (Fig. 14A). It is uncertain whether the older outflow ignimbrite sheets in the formation (older than 27.5 Ma) were deposited in half-graben basins. The Parajes formation was tilted by extension and partly eroded from normal fault footwalls prior to and during deposition of the overlying Témoris formation (Figs. 4, 8C, and 9A).

The ca. 27–24.5 Ma Témoris formation records the onset of magmatism in the area, which was primarily andesitic, with compositions ranging from basalt to andesite (Fig. 7). Like the Parajes formation, the Témoris formation was deposited in synvolcanic half-graben basins (Fig. 14B). Fluvial and debris flow processes developed alluvial fan systems that prograded into the half-grabens to become interbedded with andesitic lavas. At least some of these alluvial deposits were likely eroded from andesitic lavas exposed in uplifted normal fault footwall blocks, although some of the detritus could also have been reworked from unconsolidated primary volcanic fragmental eruptive products (Fig. 14B). Normal faults in the study area control the siting of some vents of the Témoris formation, including andesitic feeder dikes along normal faults and the andesitic volcanic center (Ttv) in the area around Témoris, which is located at the southern projection of the Guazapares fault zone (Figs. 2 and 3; Supplemental Fig. 1 [see footnote 1]). The presence of distal silicic ignimbrites (Tti) in the uppermost part of the mafic to andesitic Témoris formation, below the silicic ignimbrite-dominated Sierra Guazapares formation (Figs. 5, 10, and 12A) records a hiatus between local andesitic and silicic magmatism in the region, modified by extension, tilting, and erosion, to produce an angular unconformity.

The ca. 24.5–23 Ma Sierra Guazapares formation records the onset of silicic magmatism within the Guazapares Mining District region. Based on composition and geochronology (Figs. 6, 7, and 13; Table 3), the vent to proximal facies along the Guazapares fault zone–La Palmera fault records several eruption events of high-energy explosive volcanism that resulted in deposition of very large scale cross-bedded ignimbrites with lag breccias (Tsxi) in a wedge that defines a linear, fault-controlled fissure-type vent system (Figs. 3 and 14C). The eruptive style of each event of the Sierra Guazapares formation likely transitioned into effusive volcanism, with the emplacement of rhyolite plugs along the fissures and the deposition of related rhyolite lavas over the ignimbrites (e.g., Fig. 12B). This sequence of fissure-fed ignimbrites and effusive lava and plugs is similar to the fissure ignimbrite eruption model proposed by Aguirre-Díaz and Labarthe-Hernández (2003) to explain the origin of large volume silicic ignimbrites and related effusive volcanic deposits in other extended regions of the Sierra Madre Occidental. Their model suggests that during crustal extension, a volatile-rich silicic magma chamber reaches high crustal levels and encounters preexisting normal faults that provide a conduit for magma ascent. Magma decompression follows, resulting in an explosive eruption event with deposition of proximal pyroclastic volcanic facies adjacent to the fault-controlled vents; silicic lava domes and dikes follow the pyroclastic rocks and close the vents as the magma becomes depleted of volatiles (e.g., Aguirre-Díaz and Labarthe-Hernández, 2003). Each explosive and effusive volcanic event of the Sierra Guazapares formation may have progressed in a fashion similar to this fissure ignimbrite eruption model proposed by Aguirre-Díaz and Labarthe-Hernández (2003), with several silicic magma chambers interacting at high crustal levels with the Guazapares fault zone–La Palmera fault to develop a fissure-vent system. Further mapping is needed in the region to determine whether the fissure continues to the south of Témoris, where resistant silicic intrusions are obvious from a distance (Fig. 2; Supplemental Fig. 1 [see footnote 1]).

Regional Correlations

New stratigraphic and geochronologic data presented in this study indicate that mafic to intermediate volcanic rocks in the study area are not related to the Lower Volcanic Complex as proposed by previous workers (e.g., Ramírez Tello and Garcia Peralta, 2004; Roy et al., 2008; Wood and Durgin, 2009; Gustin, 2011, 2012). The Témoris formation instead represents a period of mafic to intermediate volcanism that occurred between two ignimbrite pulses of the mid-Cenozoic ignimbrite flare-up in the northern Sierra Madre Occidental and preceded local silicic ignimbrite flare-up magmatism in the study area.

The ca. 27.5 Ma Parajes formation is interpreted as medial welded to nonwelded silicic outflow ignimbrite sheets erupted at the end of the Early Oligocene pulse of the mid-Cenozoic ignimbrite flare-up in the Sierra Madre Occidental (ca. 36–27 Ma; Ferrari et al., 2007; Cather et al., 2009; McDowell and McIntosh, 2012), based on the similar eruption ages and physical characteristics to ignimbrite sequences described elsewhere in the region (e.g., Swanson et al., 2006; McDowell, 2007, and references therein). Possible sources for the outflow ignimbrites of the Parajes formation include (1) vent to proximal volcanic facies of similar ages previously identified ∼100 km toward the north and northeast near Basaseachic and Tomóchic (e.g., McDowell, 2007, and references therein; McDowell and McIntosh, 2012), and (2) several calderas identified <50 km to the north, south, and east of the Guazapares Mining District region (e.g., Ferrari et al., 2007, and references therein) (Fig. 15).

Based on phenocryst assemblages and an eruption age older than 27.5 Ma, the oldest flow unit of the Parajes formation, the Chepe ignimbrite (Tpc; Table 1), is tentatively correlated with the regionally extensive Divisadero tuff of Swanson et al. (2006). The Divisadero tuff is distinctive for its crystal-rich nature (to ∼40% phenocrysts) of mostly large (to 4 mm) grains of plagioclase and deeply embayed quartz. It is highly variable in thickness (∼10–300 m) and has multiple cooling units with densely welded red-brown interiors that grade upward to poorly welded white tops (Swanson et al., 2006). We sampled the upper Divisadero tuff near Divisadero, southwest of Creel (sample DIV-2; Fig. 6), to compare it with the Chepe ignimbrite (Tpc) of this study. Both have a very similar crystal-rich nature with large plagioclase, biotite, and embayed quartz phenocrysts, and the Chepe ignimbrite, like the Divisadero tuff, is densely welded. However, further investigation is needed to confirm this regional correlation, such as pumice and zircon geochemistry, and U-Pb zircon geochronology on the Divisadero tuff, which was previously dated by Swanson et al. (2006) using the K-Ar method as 29.9 ± 0.7 and 29.8 ± 0.5 Ma (±1σ errors). The Divisadero tuff extends from San Juanito to Divisadero for a length of ∼60 km (Swanson et al., 2006); our tentative correlation would expand the extent of the Divisadero tuff an additional ∼75 km southwest, to a total length of ∼135 km (Fig. 1).

In several localities in the northern Sierra Madre Occidental, mafic to intermediate composition volcanism followed the large-volume eruptions of the Early Oligocene ignimbrite pulse (the Southern Cordillera basaltic andesite province); the Témoris formation in the Guazapares Mining District region may be related to this period of mafic to intermediate composition volcanism. The mafic to intermediate composition volcanic rocks in other parts of the Sierra Madre Occidental are roughly coeval with or slightly younger than the ca. 27–24.5 Ma Témoris formation. In addition, the composition of Témoris formation rocks is similar to those of the Southern Cordillera basaltic andesite province (Fig. 7).

The age of the ca. 24.5–23 Ma Sierra Guazapares formation generally coincides with the onset of the regional Early Miocene (ca. 24–20 Ma) ignimbrite pulse of the mid-Cenozoic ignimbrite flare-up (e.g., Ferrari et al., 2002, 2007; McDowell and McIntosh, 2012; Bryan et al., 2013). Although the Early Miocene ignimbrite pulse is volumetrically significant in the southern Sierra Madre Occidental (Ferrari et al., 2002, 2007), in the northern and central Sierra Madre Occidental this ignimbrite pulse was previously thought to be less abundant and restricted to the westernmost part of the silicic large igneous province (Ferrari et al., 2007; McDowell and McIntosh, 2012; Bryan et al., 2013). The Sierra Guazapares formation thus represents a previously unrecognized part of the Early Miocene ignimbrite pulse that may have been more widespread, east of the area where rocks erupted during this pulse were previously recognized in the northern Sierra Madre Occidental.

Regional Timing of Volcanism and Extension

Previous studies have interpreted that a transition from andesitic arc magmatism in a compressional (Laramide) stress regime accompanying rapid plate convergence (Lower Volcanic Complex) to silicic ignimbrite flare-up magmatism in an extensional stress regime (Upper Volcanic Supergroup) was the result of decreased convergence between the Farallon and North American plates beginning in the Late Eocene ca. 40 Ma (Wark et al., 1990; Aguirre-Díaz and McDowell, 1991; Ward, 1991; Wark, 1991; Grijalva-Noriega and Roldán-Quintana, 1998; Ferrari et al., 2007). After the end of the Laramide orogeny in Mexico (Late Eocene), the Farallon plate was removed from the base of the North American plate by either steepening (slab rollback) and possible detachment of the deeper part of the subducted slab (e.g., Ferrari et al., 2007; Henry et al., 2010; Best et al., 2013; Busby, 2013), or through the development of a slab window (e.g., Wong et al., 2010). Based on the available age distribution of volcanic rocks in the southwestern U.S. and the Sierra Madre Occidental, the locus of magmatism is inferred to have migrated eastward (inboard) from the trench in Cretaceous to Eocene time, followed by a general southwestward migration of the arc-front magmatism toward the trench commencing by ca. 40 Ma in response to these Farallon–North American plate interactions (e.g., Coney and Reynolds, 1977; Damon et al., 1981; Ferrari et al., 1999; Gans et al., 2003; Ferrari et al., 2007; Henry et al., 2010; Wong et al., 2010; McDowell and McIntosh, 2012; Bryan et al., 2013; Busby, 2013). This plate tectonic interpretation is similar to space-time models of mid-Cenozoic volcanism proposed in the western U.S. (e.g., Coney and Reynolds, 1977; Damon et al., 1981; Gans et al., 1989; Best and Christiansen, 1991; Christiansen and Yates, 1992; Axen et al., 1993; Humphreys, 1995; Dickinson, 2002, 2006; Henry et al., 2010; Best et al., 2013; Busby, 2013). However, at a more detailed level this age trend shows greater complexity, as the Early Oligocene pulse of the ignimbrite flare-up occurred in a wide belt throughout the entire Sierra Madre Occidental at essentially the same age without internal migration patterns, and volcanism reappears in the rear-arc east of the arc front in the Middle to Late Miocene (Ferrari et al., 2007; Bryan et al., 2013).

The timing of the onset of extension relative to southwestward-migrating volcanism in the Sierra Madre Occidental has been poorly constrained, due at least in part to sparse map data. At the regional scale, the onset of extension possibly migrated episodically from east to west along the entire Sierra Madre Occidental, roughly corresponding to the southwestward migration of the arc front toward the trench; however, in detail volcanism in a given area may be preextensional, synextensional, or postextensional (Ferrari et al., 2007). Although no direct evidence has been found for Eocene extension in the eastern Sierra Madre Occidental proper, there is evidence of an initial episode of extensional faulting during the Early Eocene in the Mesa Central region to the east of the southern Sierra Madre Occidental (Aranda-Gómez and McDowell, 1998; Aguillón-Robles et al., 2009; Tristán-González et al., 2009) and at its easternmost boundary east of Durango during the Early Oligocene (32.3–30.6 Ma; Luhr et al., 2001), east of the unextended core. The earliest initiation of upper crustal extension that developed regionally is inferred to have occurred ca. 30 Ma, marked by the widespread eruption of the Southern Cordillera basaltic andesite province (Cameron et al., 1989). The timing of this event immediately followed the peak of ignimbrite flare-up volcanism of the Early Oligocene pulse and coincided with a decline in silicic explosive volcanism (Bryan and Ferrari, 2013). Following this regional event, extensional deformation generally became focused in the Gulf Extensional Province to the west of the unextended core of the Sierra Madre Occidental and the timing of initial extensional deformation appears to have migrated westward with time in this region (Fig. 15; Gans, 1997; Gans et al., 2003).

Our new geologic mapping and geochronological data from the Guazapares Mining District region is broadly consistent with the interpretations that the inception of volcanism and extension generally migrated southwestward with time across the Sierra Madre Occidental. The Late Oligocene age (ca. 27 Ma) of initial local volcanism in the study area is younger than Late Eocene to Early Eocene volcanism to the northeast, and older than to coeval with Late Oligocene to Early Miocene volcanism to the west (Fig. 15). Our data clearly show that extension in the study area not only preceded local mafic to intermediate volcanism ca. 27 Ma and local silicic ignimbrite flare-up magmatism during the Early Miocene pulse ca. 24.5 Ma, but also overlapped in time with the end of the Early Oligocene pulse of the ignimbrite flare-up in the northern Sierra Madre Occidental, which occurred ∼50–150 km to the north and east ca. 32–28 Ma (Fig. 15). The Late Oligocene age (ca. 27.5 to after 23 Ma) of extension in the Guazapares Mining District region is slightly older than to roughly coeval with the onset of extension farther west in Sonora (Fig. 15), where sedimentation in fault-bound grabens and rapid footwall cooling of core complexes also began at the end of the Oligocene to Early Miocene (Gans, 1997; McDowell et al., 1997; Wong et al., 2010).

Extensional Effects on Volcanism

Although much more mapping and dating are needed, we suggest that widespread crustal extension in northwestern Mexico may have played an important role in the later stages of magmatic development of the Sierra Madre Occidental silicic large igneous province. As magmatism migrated southwestward during the Late Oligocene to Miocene toward the Gulf of California, previously extended or currently extending crust likely influenced the composition of melts and promoted the localization of volcanic vents along favorable structures. Extension has been inferred to favor the generation and storage of melt (e.g., Hildreth, 1981; McKenzie and Bickle, 1988; White and McKenzie, 1989; Wark, 1991; Hanson and Glazner, 1995), and crustal thinning and active normal faulting is inferred to promote the ascent of basaltic magma, which results in crustal melting and the formation of silicic magma compositions (e.g., Johnson and Grunder, 2000; Ferrari et al., 2010). Although much of the pre–30 Ma volcanic and structural relationships are unclear, the inferred relationship between lithospheric extension and magmatism is supported by the synextensional nature of the of the Late Oligocene to Early Miocene mid-Cenozoic ignimbrite flare-up in the Sierra Madre Occidental silicic large igneous province, as observed in the Guazapares Mining District region.

CONCLUSIONS

New geologic mapping and zircon U-Pb LA-ICP-MS ages indicate that the Late Oligocene to Early Miocene rocks of the Guazapares Mining District region record synextensional volcanism in the northern Sierra Madre Occidental. Three informal formations are recognized: (1) the Parajes formation, consisting of silicic outflow ignimbrite sheets erupted from distant sources by ca. 27.5 Ma, during the end of the Early Oligocene pulse of the mid-Cenozoic ignimbrite flare-up; (2) the ca. 27–24.5 Ma Témoris formation, comprising locally erupted mafic to intermediate composition volcanic rocks, including an andesitic volcanic center; and (3) the ca. 24.5–23 Ma Sierra Guazapares formation, consisting of vent to proximal silicic ignimbrites, lavas, and plugs erupted by fissure magmatism during the onset of the Early Miocene pulse of the mid-Cenozoic ignimbrite flare-up.

The main geologic structures in the Guazapares Mining District region are NNW–trending normal faults, several of which bound synvolcanic half-graben basins that began to form by the time of deposition of the upper part of the Parajes formation, and continued to develop during deposition of the Témoris and Sierra Guazapares formations. Much of the crustal extension occurred prior to the eruption of the Sierra Guazapares formation, with the earliest evidence of crustal extension by ca. 27.5 Ma. A minimum of 20% total horizontal extension is estimated in the Guazapares Mining District region. Preexisting extensional structures controlled the localization of andesitic and silicic volcanic vents and shallow level intrusions of the Témoris and Sierra Guazapares formations. The age of volcanism and extensional faulting in the Guazapares Mining District region generally corresponds to regional models inferring a post-Eocene southwestward migration of volcanism and crustal extension in the northern Sierra Madre Occidental.

In summary, this study presents direct evidence that crustal extension occurred in the western part of the northern Sierra Madre Occidental during the end of the Early Oligocene pulse of the ignimbrite flare-up. Extension in the Guazapares Mining District region preceded and continued during the onset of local magmatism, consisting first of mafic to andesitic magmatism, followed by silicic magmatism related to the Early Miocene pulse of the ignimbrite flare-up. Regional crustal extension in northwestern Mexico may have played an important role in the magmatic development of the Sierra Madre Occidental silicic large igneous province during the mid-Cenozoic ignimbrite flare-up, promoting the generation of silicic and intermediate magmas and the localization of volcanic eruptions along favorable preexisting geologic structures.

We thank Paramount Gold & Silver Corp., Larry Segerstrom, Danny Sims, and Dana Durgin for financial, logistical, and intellectual contributions, and Denis Norton for arranging this support. Additional financial support was provided by a UC Mexus (University of California Institute for Mexico and the United States) grant (2009-2010) to Busby and Elena Centeno-García, National Science Foundation grant EAR-1019559 to Busby, and a Geological Society of America student research grant to Murray. Dana Roeber Murray, Jordan Lewis, Adrienne Kentner, and Angeles Verde-Ramírez assisted in the field. Carlos Ortega-Obregón and Ofelia Pérez-Arvizu (Universidad Nacional Autónoma de México, UNAM) assisted with laser ablation–inductively coupled plasma–mass spectrometry analyses. Rufino Lozano Santa Cruz (UNAM) assisted with whole-rock geochemical analyses. We thank Graham Andrews for a detailed informal review. Constructive formal reviews by Scott Bryan and an anonymous reviewer and comments by Carol Frost and Keith Putirka helped improve the manuscript.

1Supplemental Figure 1. Geologic map of Guazapares Mining District region, northern Sierra Madre Occidental, Chihuahua, Mexico. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00862.S1 or the full-text article on www.gsapubs.org to view Supplemental Figure 1.
2Supplemental Table 1. Modal point-count analyses. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00862.S2 or the full-text article on www.gsapubs.org to view Supplemental Table 1.
3Supplemental Data File. Mining claims of the Guazapares fault zone. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00862.S3 or the full-text article on www.gsapubs.org to view the Supplemental Data File.
4Supplemental Table 2. Zircon U-Pb laser ablation–inductively coupled plasma–mass spectrometry analytical results. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00862.S4 or the full-text article on www.gsapubs.org to view Supplemental Table 2.
5Supplemental Figure 2. Cathodoluminescence images of zircons from U-Pb laser ablation–inductively coupled plasma–mass spectrometry analyses. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00862.S5 or the full-text article on www.gsapubs.org to view Supplemental Figure 2.