Synextensional dike emplacement across the footwall of a continental core complex, Chemehuevi Mountains, southeastern California

Exposures of Oligocene–Miocene metamorphic core complexes along the Colorado River extensional corridor (CREC; eastern Basin and Range Province, USA) provide a natural laboratory to study the evolution of structures associated with large-magnitude crustal extension (Fig. 1; Howard and John, 1987). These core complexes offer a wealth of information on extensional processes from inception to cessation of large-slip normal faults, often including associated magmatism (e.g., John, 1987b; Lister and Davis, 1989; Campbell-Stone and John, 2002). The original orientation of large-offset low-angle normal faults in the CREC (and elsewhere) remains the subject of significant debate due to the apparent incompatibility with Andersonian mechanics (Axen, 2007). The early evolution of such systems across the brittle-plastic transition are complex and influenced by a spectrum of competing processes, including fracture and brittle faulting, plastic deformation, fluid flow, and commonly magmatism (Parsons and Thompson, 1993; Faulkner et al., 2010). We focus on constraining the timing and style of intrusive magmatism in the early evolution of a metamorphic core complex.

To better understand the evolution of dike emplacement and tectonic extension in low-angle normal fault systems across the brittle-plastic transition, this study documents the syntectonic Miocene Chemehuevi dike swarm, exposed in the Chemehuevi Mountains, southeastern California (Fig. 1). Here we concentrate on the relation between the dike swarm and the gently northeast-dipping Miocene Mohave Wash fault (MWF), a subdetachment hosted in the footwall to the large slip (>18 km) Chemehuevi detachment fault (CDF; John, 1987a). The MWF accommodated only 1–2 km of northeast-directed offset at paleodepths of ∼5 to >10 km across the base of the seismogenic zone, and was passively denuded by the structurally higher Chemehuevi detachment fault, preserving the early-slip deformation history (John, 1987b). Based on geochronologic and thermochronologic constraints from previous studies (John and Foster, 1993; Foster and John, 1999), we compare characteristics of the dike swarm in 5 northwest-striking structural domains perpendicular to the regional extension direction, in an effort to understand dike emplacement over 23 km of exposed footwall during early slip (Fig. 1). We use this documentation along with related studies to summarize a comprehensive tectono-magmatic history of the Chemehuevi Mountains core complex during Miocene extension.

The CREC lies within the southern North American Basin and Range Province (Howard and John, 1987). It is a 50–100-km-wide region of extension that continues from Las Vegas to Mexico, in part along the Colorado River (Fig. 1). The CREC underwent at least ∼100% to as much as 400% northeast-directed Miocene stretching. Extension was accommodated along a system of low-angle normal faults that cut downsection eastward from the headwall breakaway in the Old Woman–Piute Mountains and root under the Hualapai Mountains and Colorado Plateau in Arizona (Fig. 1; Howard and John, 1987; Davis and Lister, 1988; Spencer and Reynolds, 1991). These faults initially cut to paleodepths of 10–15 km and are estimated to have accommodated ∼50 km total unidirectional top-to-the-northeast slip (040°–060°), often juxtaposing volcanic and sedimentary rocks against mid-crustal basement (Howard and John, 1987; Miller and John, 1999; Lister and Davis, 1989). In the center of the CREC, isostatic doming exposes the well-known metamorphic core complexes in the Sacramento, Chemehuevi, and Whipple mountains (Fig. 1), which offer a natural laboratory for studying low-angle normal fault–related processes.

The regionally developed detachment system exposed in the Chemehuevi Mountains comprises northeast-dipping, imbricate low-angle Miocene normal faults (Fig. 1A; Howard and John, 1987; John, 1987a, 1987b). Three low-angle normal faults are exposed for more than 23 km in the downdip direction: the most shallow Devil’s Elbow fault, the regionally developed CDF (>18 km of offset), and the structurally deepest small offset (1–2 km) MWF (John, 1987b). Hornblende, biotite, and alkali feldspar ⁴⁰Ar/³⁹Ar plateau ages combined with apatite fission-track thermochronology constrain the thermal structure of the CDF footwall at initiation of extension 23 ± 1 Ma, from <150° in the southwest to >400 °C in the northeast (Fig. 1), with an initial dip between 15° and 30° northeast (John, 1987a; John and Foster, 1993; Foster and John, 1999). Peak slip rates constrained by geochronology allied with sedimentation patterns in the hanging wall to the CDF range from 7 to 8 mm/yr between 19 and 15 Ma (Miller and John, 1999); the majority of slip was accomplished by ca. 13.9 Ma based on diminished sedimentation rates (Miller and John, 1999), with cessation by 11.6 Ma based on coeval dikes and plugs that cut the CDF (John, 1987b; John and Foster, 1993).

Bedrock in the Chemehuevi Mountains is dominated by crystalline basement with minor sedimentary and volcanic deposits. Mid-crustal crystalline rocks cut by the CDF include heterogeneous Proterozoic amphibolite gneiss and migmatite, isotropic granitic rocks of the Cretaceous Chemehuevi Mountains Plutonic Suite, and Proterozoic, Cretaceous, and Miocene dikes and small plugs (Fig. 1; Howard and John, 1987; John, 1987a, 1987b, 1988). Although the history of the Miocene Chemehuevi dike swarm was not a focus of previous studies, some dikes in the suite are known to be Miocene based on preliminary ⁴⁰Ar/³⁹Ar ages, composition, and morphology, having accommodated minor (≤2%) northeast-directed extension (John and Foster, 1993). The flanks of the range host Oligocene to Miocene volcanic rocks and locally derived Paleogene to Quaternary sedimentary deposits in the hanging wall to both the Chemehuevi and Devil’s Elbow faults (Miller and John, 1999). The volcanic succession includes an ∼500-m-thick basal sequence of mafic and intermediate lavas (lower volcanic and sedimentary section of Sherrod and Nielson, 1993) erupted between ca. 22.0 and 18.78 ± 0.02 Ma (Miller and John, 1999), the 0–60-m-thick 18.78 ± 0.02 Ma Peach Spring Tuff (Miller and John, 1999; Ferguson et al., 2013), and minor rhyolite ash flows and basaltic lavas deposited between 15.54 ± 0.03 and 13.9 ± 0.1 Ma (Miller and John, 1999).

The history of the CREC includes extensive Oligocene–Miocene magmatism during extension. Voluminous extrusive magmatism is documented to have propagated northward through the CREC, mimicked by a similar geographic pattern of brittle normal faulting ∼1–4 m.y. later (Gans et al., 1989; Gans and Bohrson, 1998; Faulds et al., 1999). Following cessation of major fault slip, extrusive magmatism resumed at a significantly reduced level (Howard and John, 1987; Gans and Bohrson, 1998; Miller and John, 1999), highlighting a temporal and geographic pattern characteristic of active rifting, with the bulk of magmatism facilitating brittle failure of the upper and middle crust (Gans et al., 1989). Extrusive magmatism in the CREC also records an evolution in composition during extension; the oldest volcanic deposits comprise a heterogeneous suite of basaltic, intermediate, and minor felsic rocks, whereas the youngest volcanism is basaltic or bimodal in composition (Gans et al., 1989; Gans and Bohrson, 1998; Foster and John, 1999). Although a significant volume of the documented syntectonic magmatism in the CREC is extrusive, plutons and dikes were also emplaced prior to, during, and after peak slip on low-angle normal faults associated with the core complexes (Davis et al., 1982; Campbell and John, 1996; Campbell-Stone et al., 2000; Campbell-Stone and John, 2002; Singleton and Mosher, 2012). Dikes characteristic of the early syntectonic magmatism in the CREC are exposed in the Whipple, Mohave, Chemehuevi, and Sacramento mountains (Howard and John, 1987; John, 1987b; Lister and Davis, 1989; Nielson and Beratan, 1995; Sherrod and Nielson, 1993).

Studies of synextensional dikes associated with metamorphic core complexes can offer a wealth of kinematic information, and have proposed that dikes may influence low-angle normal fault evolution. Dikes commonly intrude as mode I fractures parallel to the greatest principal stress and orthogonal to the least principal stress, highlighting the regional stress field during their emplacement (Anderson, 1951). In extensional stress regimes with the greatest principal stress (σ₁) vertical, dikes have characteristic strikes orthogonal to the extension direction and dip steeply (Anderson, 1951). Dikes can therefore be used as indicators of the magnitude of postemplacement block rotation (Howard, 1991; Wong and Gans, 2008; John and Cheadle, 2010) and to determine fault offset (e.g., John, 1987b; Howard and John, 1987). Dikes with anomalous orientations have invoked discussion related to evolving stress fields during metamorphic core complex development (Spencer, 1985). Dike emplacement has also been proposed to facilitate initiation of low-angle normal faults by stress reorientation, or effective weakening of the host rock. Collectively, dikes may facilitate initiation of a gently dipping fault just above its upper termination or tip, by rotating the greatest principal stress away from vertical (Parsons and Thompson, 1993).

Outcrop and macroscopic investigations documenting relations between the MWF and Chemehuevi dike swarm were completed by measuring structural orientations and dike thickness, and documenting the geometry and macroscopic textures of syntectonic dikes. Volume percentages of dikes were estimated by transects in the field, measuring cumulative dike thickness divided by total transect distance. Representative samples were collected for geochemical and geochronologic analysis to constrain both compositional variation and age (Tables 1 and 2; Supplemental Fig. S1¹).

X-ray fluorescence analyses of 27 dikes for SiO₂, TiO₂, Al₂O₃, FeO, MnO, MgO, Ca₂O, Na₂O, K₂O, P₂O₅, and trace elements Ba, Sr, and Zr were conducted on the Rigaku Supermini200 at Ohio University. Concentrations were determined using the fundamental parameters (FP) method and calibrated against 12 U.S. Geological Survey rock standards. Relative uncertainty (accuracy and precision) for trace elements was <15%.

U-Pb dating and trace element geochemical analyses of zircon from five felsic dikes were performed at the U.S. Geological Survey–Stanford Ion Microprobe Laboratory using sensitive high-resolution ion microprobe–reverse geometry. Recovered zircon range from large, euhedral grains with well-defined oscillatory zoning and rare embayed cores to subhedral or subequant grains with weak oscillatory or irregular zoning. We analyzed 13–29 grains ∼40–400 µm in long dimension from each sample. We note that a significant proportion of the Chemehuevi dike swarm are weakly to moderately altered basalt, trachybasalt, and basaltic trachyandesite, hosting no appropriate minerals for geochronology (i.e., U-Pb or ⁴⁰Ar/³⁹Ar); dated samples are therefore limited to dacite and rhyolite dike compositions. Data reduction for geochronology follows the methods described by Ireland and Williams (2003) using the Microsoft Excel add-in programs Squid 2.5 and Isoplot 4.1 (Ludwig, 2012). Ti-in-zircon temperatures were calculated using measured ⁴⁸Ti and the recalibrated method of Ferry and Watson (2007). Detailed geochronology and geochemistry methods are in Supplemental Data².

We collected and analyzed 27 dikes from southwest to northeast across the core complex footwall that show a broad range in composition including basalt, trachybasalt, basaltic trachyandesite, trachyandesite, trachydacite, dacite, and rhyolite (50.9–80.5 wt% SiO₂; Table 1; Fig. 2). We refer to these compositions as mafic (<52% SiO₂), intermediate (>52% to <63% SiO₂), intermediate-felsic (>63% to <69% SiO₂), and felsic (>69% SiO₂; Fig. 2). All dikes with <56 wt% SiO₂ are calc-alkalic; those with higher silica range from calc-alkalic to alkali-calcic (Fig. S2 [see ^{footnote 1}]). All dikes are magnesian (Fig. S2 [see ^{footnote 1}]); dikes with <64 wt% SiO₂ are metaluminous, and those with >64 wt% SiO₂ are peraluminous (Fig. S2 [see ^{footnote 1}]), and commonly host mica (either biotite and/or muscovite) and rarely garnet. Geographically, mafic to intermediate-felsic dikes are present throughout the footwall to the detachment system; dikes with >70% SiO₂ are characteristic of and restricted to the easternmost domains (4 and 5, Fig. 3).

Major and trace element oxides show typical continuous differentiation trends for compatible and incompatible elements (Harker, 1909). TiO₂, CaO, Al₂O₃, FeO_tot, MgO, and P₂O₅ all decrease as K₂O and Na₂O increase with increasing SiO₂ wt% (Table 1; Figs. 4A–4C; Fig. S3l [see ^{footnote 1}]). Ba remains at similar concentrations over the expanded range of silica concentrations (Fig. 4D); Zr concentrations positively correlate with SiO₂ between 50 and 62 wt% and negatively correlate at >62 wt% SiO₂ (Fig. 4E), and positively correlate with Sr concentrations (Fig. 4F).

The Chemehuevi dike swarm is similar in major element chemistry to Miocene dikes hosted in the adjacent Whipple and Mohave Mountains, and pre–Peach Spring Tuff (PST) volcanic deposits, the Swansea Plutonic Suite in the Buckskin Mountains, the PST, and post-PST volcanic deposits hosted in the Black Mountains (Fig. 4; Figs. S2 and S3 [see ^{footnote 1}]; Pease, 1991; Bryant and Wooden, 2008; Pamukcu et al., 2013; McDowell et al., 2014; Gentry, 2015). Trace element concentrations in Chemehuevi dikes are comparable to coeval regional magmatism, but differ from the PST ignimbrite (Fig. 4: Fig. S3 [see ^{footnote 1}]; Pease, 1991; Bryant and Wooden, 2008; Pamukcu et al., 2013; McDowell et al., 2014; Gentry, 2015).

U-Pb zircon dating from five representative dikes proximal to the MWF in domains 1, 2, 4, and 5 shows that intermediate to felsic dikes of the Chemehuevi dike swarm are Miocene in age (Table 2; Fig. 5). Dated dacite dikes vary in age between 21.45 ± 0.19 Ma (BJ13Ch-1, mylonitic dacite), to 20.94 ± 0.13 Ma (BJ13Ch-3, undeformed biotite hornblende dacite). U concentrations of analyzed zircon range from 63 to 2846 ppm; most grains host abundances of a few hundred parts per million. Errors on calculated ages for each sample range from >1% to <6%, and are shown as 2σ standard deviation (Fig. S4 [see ^{footnote 1}]). Note that none of the dacite dikes host inherited zircon, and mean Ti-in-zircon temperatures span a narrow range from 721 ± 16–748 ± 43 °C. In contrast, rhyolitic dikes have ages from 20.23 ± 0.23 Ma (CG14Ch-123, mylonitic) to 19.59 ± 0.71 Ma (BJ14Ch-4, mylonitic, garnet, two mica rhyolite) and 19.21 ± 0.15 Ma (BJ13Ch-5; undeformed biotite rhyolite). We note no significant age difference between undeformed and mylonitic dikes from this data set; undeformed dikes range from 20.94 ± 0.13 to 19.21 ± 0.15 Ma, whereas those hosting a well-developed mylonitic lineation parallel to the regional extension direction range in age from 21.45 ± 0.19 to 19.59 ± 0.71 Ma. Although their ages overlap within error, differences in whole-rock composition and concentration of inherited and/or xenocrystic zircon (absent in BJ13Ch-5, CG14Ch-123 hosts <10% xenocrystic zircon with ²⁰⁴Pb-corrected ²⁰⁷Pb/²⁰⁶Pb ages ca. 1.7 Ga; in contrast, BJ14Ch-4 hosts >70% inherited zircon between 1700 and 600 Ma) indicate at least 2 phases of these intermediate to felsic dikes. Ti-in-zircon temperatures span a wider range from 704 ± 26 to 788 ± 26 °C, suggesting multiple phases (and possible sources) of felsic dikes (separated by hundreds of thousands of years or less).

Dikes of the Chemehuevi dike swarm form sets that strike dominantly west-northwest, south-southwest, and east-northeast with a systematic range in dip. Dikes in domains 1 and 2 (Fig. 6) form two sets striking west-northwest and south-southwest with moderate to steep dips. Maximum eigenvector (E1) of poles to dikes in domain 1 for the west-northwest set is 189°/42° and E1 for the south-southwest is 106°/16°; in domain 2, the west-northwest set E1 is 199°/35° and the south-southwest set E1 is 112°/21°. We also note that the west-northwest and south-southwest dike sets of domains 1 and 2 are roughly orthogonal to each other and when combined average an east-west strike (maximum E1 = 178°/44° and 182°/39°, respectively). In domain 3, subvertical dikes strike east-northeast (pole to dikes E1 = 351°/10°). In contrast, dikes in domains 4 and 5 dip gently to the southeast while also striking east-northeast (pole to dikes E1 = 356°/56° and 338°/72°, respectively).

The orientations of many dikes within the Chemehuevi Mountains parallel preexisting anisotropies within the country rock. Dikes in domain 1 commonly parallel quartz-sericite joints associated with late-stage cooling of the Chemehuevi Mountains Plutonic Suite (John and Foster, 1993; John and Mukasa, 1990). In domains 4 and 5, dikes locally intrude moderately south-dipping brittle to mylonitic shear zones. There is no evidence that significant rotation of dikes took place outside of localized ductile deformation, as presented in this study, and minor isostatic doming related to denudation concentrated on the flanks of the core complex (John and Foster, 1993).

Dikes hosted in the footwall to the Chemehuevi detachment fault, proximal to the MWF, are centimeter- to multimeter-thick planar intrusions that extend for meters to kilometers along strike (Fig. 7). This study provides an upper bound on dike abundance with intermediate to felsic Miocene dikes composing as much as 25% of the footwall, observed in domains 4 and 5. Previous studies estimated dikes to compose <1%–2% of the CDF footwall overall and >10% in the center of the range (John and Foster, 1993). The relative abundance of dikes of different compositions also varies spatially across the Chemehuevi Mountains, from southwest to northeast; mafic to intermediate dikes are the most abundant in domains 1–3, whereas intermediate-felsic to felsic dikes are relatively more abundant in domains 4 and 5 (Fig. 3). The results of this study agree with previous studies in that mafic dikes are the most volumetrically significant, especially in the center of the range (domain 3; John and Foster, 1993).

The geometry of dikes also varies across the range. In domains 1–3, dikes are planar and commonly traceable for kilometers. In these domains, dikes with mafic to intermediate compositions truncate intermediate-felsic and felsic dikes, indicating that mafic to intermediate dikes are relatively younger (Fig. 7A). Commonly, the MWF truncates dikes at high angles; in some instances, dikes terminate within meters of the fault zone, and/or are deflected in the extension direction (Fig. 7B). Dikes in domains 4 and 5 can be planar in outcrop (Fig. 7C), but locally define open to isoclinal folds (Fig. 7D) and have undulatory geometries (Fig. 7E). Dikes in these eastern domains are commonly hosted in centimeter-thick, southeast-dipping shear zones that accommodated top-to-the-southeast displacement (normal-sense deflection) of metamorphic foliation (Fig. 7F).

Mafic to intermediate dikes are common within the damage zone of the MWF, and show extreme variability in their shape, thickness, and orientation. These dikes vary in thickness from centimeters to meters, and vary in shape from subparallel walled to irregular (Fig. 8). Like mafic to intermediate dikes outside the damage zone, the age of these dikes is unconstrained due to extreme greenschist facies alteration, but are coeval with MWF slip based on a number of crosscutting relations, including (1) mafic dikes intrude open fractures (Fig. 8A) and cataclasites (Fig. 8B) within the damage zone with locally preserved chilled margins (Fig. 8C); (2) mafic dikes host clasts of cataclasite (Fig. 8B), yet show repeated fracture themselves (Figs. 8A, 8C); and (3) dikes interfinger with quartz-epidote veins within and adjacent to the MWF damage zone (Fig. 8D). These relations are common in domains 1–3 but rare to absent in domains 4 and 5, and indicate episodic and repeated fracturing and magmatism at structurally shallow depths within the MWF zone during slip.

Magmatic and solid-state fabrics are preserved in dikes of the Chemehuevi dike swarm, and vary by position in the detachment footwall downdip. Dikes exposed in the western domains (1–3) preserve fine-grained or porphyritic magmatic textures (Fig. 9A). These textures are characterized by an aphanitic groundmass hosting randomly oriented euhedral to subhedral phenocrysts of quartz, feldspar, biotite, and hornblende in felsic dikes, and feldspar, hornblende, pyroxene, and oxides in mafic to intermediate dikes. In contrast, all intermediate to felsic dikes in eastern domains (4 and 5) are L (linear) > S (foliation) tectonites with fine-grained or porphyroclastic textures (including porphyroclasts of feldspar, amphibole, and mica). Aligned phyllosilicates and quartz and feldspar ribbons define the tectonic fabric (Figs. 9B–9D). Lineations hosted in mylonitic dikes trend northeast-southwest, with the maximum eigenvector oriented 231°/9°, parallel to the regional extension direction (Fig. 6; John, 1987a). The present-day gentle southwest plunge of this mylonitic lineation is likely a result of minor denudation-related footwall uplift and rotation. Mylonitic foliations hosted by the dikes parallel dike margins, and are dominantly subhorizontal to moderately south dipping (Figs. 9C, 9D).

The Chemehuevi dike swarm shows a systematic increase in mylonitic fabric intensity northeastward (downdip) in intermediate to felsic composition dikes. This progression in macroscopic fabric development is mimicked by the country rock, but the transition between country rock hosting undeformed versus mylonitic fabrics occurs at the deepest exposed structural levels in the northeast (John, 1987b; LaForge, 2016). None of the dikes in domains 1–3 host mylonitic fabrics. However, equivalent dikes in domain 4 host a penetrative mylonitic foliation and lineation; the country rock only rarely hosts localized zones of centimeter-thick mylonitic deformation in quartz (Fig. 9C). In domain 5, both dikes and their country rock host mylonitic fabrics of similar orientation; recrystallized mineral phases in county rock are limited to meter-thick zones of quartz and phyllosilicates (Fig. 9D; John and Mukasa, 1990).

This study highlights distinctions in orientation, geometry, and internal fabric between dikes of similar composition and age at different structural levels of the low-angle normal fault system. These observations indicate that syntectonic intermediate to felsic dikes deformed with an elevated brittle-plastic transition (BPT) relative to the surrounding country rock. The BPT in quartz and feldspar associated with these dikes is expressed geographically between domains 3 and 4 (∼18 km downdip from the westernmost exposure of the MWF, at ∼9 km minimum paleodepth), based on penetrative mylonitic fabrics hosted in intermediate to felsic dikes only in domains 4 and 5, and their absence in domains 1–3. The BPT of quartz hosted in country rock is structurally deeper, between domains 4 and 5 (John, 1987b; LaForge, 2016). This evidence substantiates that the MWF was active across the upper limit of the BPT (John and Foster, 1993), with a shallowed BPT in synextensional intermediate to felsic dikes.

We suggest that dikes deformed within an elevated BPT resulting from a combination of characteristics unique to dikes in this system, including the fine-grained polymineralic texture, enhanced fluid activity, and elevated temperatures relative to the footwall. Studies of naturally deformed mylonites (Stünitz and Fitzgerald, 1993; Kilian et al., 2011; Sullivan et al., 2013; Oliot et al., 2014), and experimental formed mylonites (Bos and Spiers, 2001; Holyoke and Tullis, 2006a, 2006b) indicate that fine-grained, polyphase aggregates are weaker than monomineralic aggregates. Alternatively, enhanced fluid activity in dikes may aid in mylonitic fabric development (Selverstone et al., 2012). We suggest that residual heat from dike emplacement during extension may have facilitated deformation in the solid state as dikes cooled. We also propose that ambient heat of the footwall also played a role in deformation. The lack of deformation in intermediate to felsic dikes at shallow structural levels (domains 1–3) suggests that higher ambient temperature in the footwall downdip allowed dikes to deform plastically. This deformation could have been accomplished by (1) dikes cooling slower at deeper structural levels while undergoing differential stress forming a mylonitic fabric or (2) hotter ambient temperatures in deep structural levels after cooling to background temperatures, allowing the rheologically distinct dikes to form a mylonitic fabric. While the exact mechanisms that facilitated the elevated BPT in intermediate to felsic dikes are unclear, the dikes represent a distinct component of the system that localized plastic strain during the early evolution of this low-angle normal fault system.

The orientations of dikes in the Chemehuevi dike swarm are not as expected for the regional northeast-directed extension, but are similar to swarms of dikes exposed elsewhere in the CREC north of the Chemehuevi Mountains. Assuming no preexisting anisotropies, dikes are predicted to intrude roughly orthogonal to the extension direction (Anderson, 1951), as they do in the Newberry Mountains (Spencer, 1985), Whipple Mountains (Gans and Gentry, 2016), Harquahala Mountains (Richard, et al., 1990), and Buckskin Mountains (Singleton, 2015). Large populations of Miocene dikes in the Chemehuevi Mountains strike east-west or northeast-southwest, parallel or oblique to the extension direction. Although the west-northwest dike sets in domains 1 and 2 are close to extensional-orthogonal (E1 of poles to dikes trend within 26° and 36° of being orthogonal extension direction, respectively, in domain 1 and 2), they are still oblique to the extension direction of 45°–55° northeast. Oblique or parallel dikes to extension the direction are more comparable to dikes in the Homer Mountains, Piute Range, and northwestern Sacramento Mountains, which host east-west–trending dikes (Spencer, 1985).

East-west–oriented dikes in the Homer Mountains area occur in structural domains of synformal warping near the detachment breakaway (Spencer, 1985). Spencer (1985) argued that synformal down-warping creates compressive stresses that overwhelm regional extensional stress during denudation, shifting the local extension direction to north-south, accounting for east-west–oriented dike emplacement. However, the Chemehuevi Mountains are hosted in an antiformal uplift geographic-structural domain in the CREC (Spencer, 1985), where dikes are predicted to be north-south or northwest-southeast, and therefore do not clearly fit this explanation. We propose that dikes in the Chemehuevi Mountains were likely controlled by preexisting weaknesses and minor localized ductile rotation rather than the contemporaneous stress state. In domains 1–3 Cretaceous joints (noted in John, 1987a, 1987b; John and Foster, 1993) likely controlled the original orientations of the two sets of dikes hosted in the Chemehuevi Mountains Plutonic Suite. Alternatively, if we consider west-northwest dikes to represent an extension-orthogonal, originally vertical set, they would have rotated to the southwest around a horizontal axis by >40°. This rotation varies significantly from other data constraining CDF footwall rotation to be minor, and this scenario is therefore less likely (John and Foster, 1993). In domains 4 and 5, dikes intruded moderately dipping shear zones hosted in Proterozoic gneiss, and subsequently rotated into parallelism with the overriding low-angle normal faults, an effect that is also displayed in the transitions between domains 3, 4, and 5. With increasing structural depth in the footwall to the fault system, dikes are closer to horizontal; in domains 3, 4, and 5 the maximum eigenvectors of poles to dikes plunge 10°, 56°, and 72°, respectively.

The Miocene Chemehuevi dike swarm correlates both temporally and chemically with local and regional extrusive magmatism. The earliest pulse of the Chemehuevi dike swarm (Table 3) was emplaced during early extension between 21.45 ± 0.19 and 19.21 ± 0.15 Ma, and shortly thereafter magmatism overlaps in age with the ca. 22–18.8 Ma lower volcanic section of Miller and John (1999), exposed in the CDF hanging wall. Compositionally, mafic to intermediate dikes associated with this pulse are similar to the lower volcanic section (Table 3; Miller and John, 1999); felsic dikes are not. These felsic intrusive rocks are similar in composition to the PST (Fig. 2), a unit they predate by as little as 400 k.y. (Table 3). These high-silica dikes are exposed only in the northeast Chemehuevi Mountains (Fig. 3), closest to the recently recognized Peach Spring Tuff caldera, in the central Black Mountains (Ferguson et al., 2013). The second pulse of dike emplacement includes minor 11.6 ± 1.2–11.1 Ma mafic dikes (John, 1987b) similar in composition to post-PST mafic lavas erupted in the CDF hanging wall between 15.5 and 13.9 Ma (Miller and John, 1999). No intrusive rhyolitic dikes postdating deposition of the PST are observed. We interpret the first pulse of the Miocene Chemehuevi dike swarm emplacement to represent part of the magmatic system that fed and rejuvenated a magma system related to the coeval volcanic deposits hosted in the hanging wall to the CDF, and possibly the PST. The second pulse of dike emplacement is not coeval with extrusive volcanism, despite some similarities in composition.

Major and trace element geochemistry highlight similarities between the Chemehuevi dike swarm and local coeval magmatism across the CREC. The Chemehuevi dike swarm is similar to dikes in the Mohave and Whipple Mountains, the Swansea Plutonic Suite in the Buckskin Mountains (Pease, 1991; Bryant and Wooden, 2008; Gentry, 2015), and pre- and post-PST volcanism in the southern Black Mountains (Figs. 6 and 8; Fig. S2 [see ^{footnote 1}]). However, PST-related magmatism differs from the Chemehuevi and CREC magmatism in major and trace element geochemistry (Pamukcu et al., 2013; McDowell et al., 2014; Fig. 4). We suggest that the magmatism in the Chemehuevi Mountains is closely related to other magmatism in the CREC, whereas its relation to the PST is unclear.

The Miocene history of extension in the Chemehuevi Mountains can be described by early brittle extension with dike emplacement and extrusive magmatism followed by rapid brittle faulting accompanied by minor volcanism. A summary of inferred timing and rates of magmatism and faulting associated with the CDF system as constrained from this and other studies is presented in Table 3 and Figure 10.

The syntectonic portion of the Chemehuevi dike swarm was emplaced from ca. 21.5 to 19.2 Ma shortly after (∼1.5 ± 1 to 3.8 ± 1 m.y.) the initiation of extension at 23 (±1) Ma. The first pulse of dike intrusion is distributed across the range and divided into two distinct episodes based on composition, timing, and emplacement and/or deformation style. The earliest dikes were intermediate to felsic in composition, including dikes that developed a mylonitic fabric in deep structural levels of the fault system (domains 4 and 5). The timing of fabric formation is poorly constrained, but occurred in the solid state following emplacement of individual dikes before significant denudation-related cooling of the footwall. Mylonitization of individual felsic dikes at deep structural levels is therefore bracketed as shortly after emplacement from ca. 21.5–19.2 Ma to the inception of rapid slip at or shortly after 19 Ma. The second episode of dike emplacement dominated by mafic to intermediate composition dikes observed cut felsic dikes at moderate to shallow structural depths (domains 1–3; Fig. 7A). While we do not present absolute ages of mafic to intermediate dikes, we suggest that ages cluster ca. 19 Ma because they postdate 19.2 Ma felsic dikes, and are compositionally similar to the ca. 20–19 Ma dikes in the Mohave Mountains and local volcanic rocks as young at 18.8 Ma (Pease, 1991; Miller and John, 1999). We propose here that slip on the MWF ceased near the end of emplacement of the syntectonic dike swarm ca. 19 Ma, based on mutually crosscutting relations with fractures within the MWF fault zone that suggest that fault slip and dike emplacement occurred and stopped at approximately the same time. We also note that extrusive magmatism correlative with the dikes spans a broader age range (ca. 22 to 18.8 Ma), suggesting that dike emplacement may overlap with the onset of extension (Fig. 10; Miller and John, 1999; Ferguson et al., 2013). Thermochronologic and geochronologic estimates of peak slip rates (7–8 mm/yr) along the CDF postdate voluminous magmatism between 19 and 15 Ma. These same data highlight possible accelerated slip at 15 ± 1 mm/yr (Miller and John, 1999; Carter et al., 2006), with the majority of slip accommodated prior to 13.9 Ma (Fig. 10; Table 3; Miller and John, 1999). Brittle extension in the upper crust continued until 11.6 ± 1.2 Ma, evidenced by a minor second pulse of mafic diking that intrudes into and above the CDF (Table 3; John, 1987b; John and Foster, 1993).

The temporal evolution of normal faulting accompanied by magmatism followed by rapid-slip faulting devoid of magmatism is similar to other highly extended terranes throughout the CREC (Gans and Bohrson, 1998; Faulds et al., 1999) and the Basin and Range Province (Gans et al., 1989) suggestive of an active rifting mode of crustal extension.

Intrusive magmatism that predates rapid brittle extension represents an increasingly recognized component of crustal stretching in CREC core complexes. In the Whipple Mountains, the ca. 20.5–19.0 Ma Chambers Well dike swarm intruded prior to rapid tectonic extension beginning at 19.0–18.5 Ma (Gans and Gentry, 2016). In the adjacent Buckskin-Rawhide core complex, the ca. 22–21 Ma Swansea Plutonic Suite predates the inception of extension ca. 21–20 Ma (Singleton et al., 2014). We show in the Chemehuevi Mountains, the 21.5–19.2 Ma dikes predate rapid slip ca. 19 Ma. In the past, studies have emphasized volcanism that predates rapid slip (Gans et al., 1989; Gans and Bohrson, 1998; Faulds et al., 1999), but as these studies show, intrusive magmatism that predates rapid slip may be as widespread. In addition, these studies indicate that many volcanic deposits have intrusive equivalents that intrude adjacent or directly into fault zones. Magmatism is also clearly overprinted by deformation subsequent to intrusion, forming syntectonic L > S tectonites in many metamorphic core complexes (Fletcher and Bartley, 1994; Singleton and Mosher, 2012; Singleton, 2013). These recent studies document magmatism intruding into and adjacent to evolving fault systems prior to rapid slip, and highlight the long-standing question of the role intrusive magmatism plays in metamorphic core complex evolution.

This study presents data related to the synextensional mafic to felsic Miocene Chemehuevi dike swarm intricately involved in the early history of the CDF system, and contributes five main ideas. (1) The Chemehuevi dike swarm intruded into the footwall, hanging wall, and fault zone of a low-angle normal fault as it was active in the upper middle crust deforming with an elevated BPT. (2) Although previous studies noted that the Chemehuevi dike swarm accommodated only a minor component of the regional extension (<2% of the CDF footwall by volume), this study suggests that they are locally more volumetrically significant adjacent to a low-angle normal fault zone, composing as much as 25% of the footwall near the MWF in the deepest exposures of the footwall and fault system (domains 4 and 5). (3) Dike orientations are somewhat anomalous in the Chemehuevi Mountains with large populations striking east-west or southwest-northeast, oblique and parallel to the extension direction. These were likely controlled by preexisting fractures and minor postemplacement rotation. (4) The Chemehuevi dike swarm represents part of the feeder system to local syntectonic volcanism, and is geochemically similar to regional magmatism. (5) The Miocene evolution of the CDF system fits a two-phase tectono-magmatic evolution of other highly extended areas in the CREC and Basin and Range Province with an initial period of faulting accompanied by magmatism ∼1.5 ± 1 to 3.8 ± 1 m.y. after the onset of extension followed by rapid slip of rapid brittle faulting without significant magmatism.

We acknowledge the support from National Science Foundation grants (EAR-1145183 awarded to John and Grimes) as primary funding for this project. LaForge also acknowledges additional financial support for field work from the American Association of Petroleum Geologists Weimer Family Named Grant. We thank R. Heilbronner, H. Stunitz, R. Kilian, C. MacDonald, and J. Brown for useful discussions in the field and laboratory. Rose Pettiette and Connor Marr helped in the field. We thank associate editor Allen J. McGrew and reviewers John S. Singleton and Walter A. Sullivan for constructive reviews of this manuscript. We thank the Bureau of Land Management Needles Field Office and Ramona Daniels for logistical support and extended access to the Chemehuevi Mountains Wilderness Area.