We present fault data from a belt of Miocene metamorphic core complexes in western and central Arizona (USA) to determine patterns of brittle strain during and after large-magnitude extension, and to evaluate the magnitude of postextensional dextral shear across the region. In the White Tank Mountains, coeval WNW- to NW-striking dextral, normal, and oblique dextral-normal faults accommodated constrictional strain with extension subparallel to the direction of ductile stretching during core complex development. Northwest-striking oblique dextral-normal faults locally accommodated similar strain in the Harquahala Mountains, whereas in the South Mountains, constriction was primarily partitioned on NE-dipping normal faults and conjugate NW- and north-striking strike-slip faults. We interpret brittle constrictional strain to have developed during the late stages of large-magnitude extension associated with core complex development and folding of detachment fault corrugations. The oblique orientation of the Arizona core complex belt with respect to the extension direction likely resulted in a minor component of dextral transtension, accounting for much of the constrictional strain. In addition, far-field stresses associated with the transtensional Pacific–North America plate boundary may have contributed to constriction, which characterizes most Neogene detachment fault systems in the southwest Cordillera. Following cessation of detachment fault slip across the Arizona core complex belt (ca. 14–12 Ma), distributed NW-striking dextral and oblique dextral–NE-side-up (reverse) faults modified the topographic envelope of corrugations to an orientation clockwise of the core complex extension direction. Based on our analysis of this misalignment, we interpret the postdetachment fault dextral shear strain to increase northwestward from 0.03 across the South Mountains (0.5–0.6 km total slip across 18 km) to >0.03–0.07 across the Harquahala and Harcuvar Mountains (1.2–2.5 km of total slip across ∼35 km) and ∼0.2 across the Buckskin-Rawhide Mountains (7–8 km across 36 km). This along-strike variation in dextral shear is consistent with the regional pattern of distributed strain associated with the Pacific–North America plate boundary, as cumulative dextral offset in the lower Colorado River region increases toward the eastern Mojave Desert region to the northwest.

Metamorphic core complexes provide important insight into how strain is accommodated during large-magnitude crustal extension. Previous studies have demonstrated that many core complexes are characterized by three-dimensional (3-D) strain resulting in coeval vertical and horizontal shortening during extension (e.g., Mancktelow and Pavlis, 1994; Chauvet and Séranne, 1994; Fletcher and Bartley, 1994; Holm et al., 1994; Martínez-Martínez et al., 2002; Singleton, 2013; Erkül et al., 2017). This constrictional strain regime may be largely responsible for development of kilometer-scale detachment fault corrugations—upright antiformal and synformal grooves oriented parallel to the dominant extension direction. Elevated horizontal stress perpendicular to the extension direction may arise due to along-strike heterogeneity in extension magnitude (e.g., Fletcher et al., 1995) or due to detachment faulting within a transtensional setting (e.g., Seiler et al., 2010; Fossen et al., 2013). In the southwest Cordillera (from southern Nevada, USA, to Baja California, Mexico), Neogene detachment fault systems developed either inboard or within the growing Pacific–North America plate boundary (Fig. 1A). The late Neogene detachment systems in this region are directly linked to transtensional fault systems associated with the plate boundary (e.g., Axen and Fletcher, 1998; Oldow et al., 2009), but the role of the plate boundary in the development of the early Neogene detachment systems is not clear. Dextral divergence across the plate boundary may have driven this extension in the early to middle Miocene (e.g., Atwater and Stock, 1998; Glazner et al., 2002) and/or far-field stress associated with the plate boundary may have influenced core complex strain geometry (Singleton, 2013). Detailed characterization of ductile and brittle core complex structures is essential for defining the 3-D strain geometry of large-magnitude extension and for evaluating what role, if any, the Pacific–North America plate boundary played in the evolution of these core complexes.

In western Arizona (USA), NW-striking dextral faults cut an extensional detachment fault system that accommodated tens of kilometers of top-to-the-NE slip (Singleton, 2015). This dextral faulting may have initiated near the end of detachment fault slip in response to eastward migration of Pacific–North America shear in the middle Miocene, resulting in a transition from large-magnitude extension to distributed transtension (Singleton, 2015). Alternatively, dextral faults and conjugate north-striking sinistral faults may have initiated during detachment faulting to accommodate subhorizontal y-axis shortening in a constrictional strain regime. While there is some uncertainty in when and why strike-slip faulting initiated across the western Arizona core complex belt, dextral and oblique dextral slip on NW-striking faults clearly represents the dominant style of faulting following cessation of detachment slip at ca. 14–12 Ma (Richard et al., 1990; Singleton, 2015), accommodating a minor component of distributed Pacific–North America shear. In the Buckskin-Rawhide core complex, postdetachment dextral faults reoriented the range-scale topographic trend of mylonitic footwall corrugations to an orientation clockwise of the detachment fault slip direction (Singleton, 2015). Based on the assumption that the corrugations developed parallel to the core complex extension direction (defined by the detachment fault slickenlines and mylonitic lineations), Singleton (2015) interpreted the ∼15° average misalignment between the Buckskin-Rawhide core complex extension direction and the topographic trend of antiformal corrugations to record 7–8 km of postdetachment dextral shear across the ∼36-km-long mylonitic footwall. Most models for the development of detachment fault corrugations involve either initially grooved or segmented detachment faults giving shape to a corrugated footwall (e.g., Jackson and White, 1989; Spencer, 1999), or folding produced by horizontal shortening perpendicular to the extension direction (e.g., Fletcher and Bartley, 1994; Singleton, 2013). Regardless of origin, it is clear that a corrugated geometry was present during at least the late stages of detachment slip (Prior et al., 2018), and that corrugation axes should have paralleled the dominant detachment slip direction during extension, as significant displacement oblique to the axes would not have been mechanically favorable. If corrugations developed via folding in a transtensional regime, it is possible that a minor misorientation between the trend of the corrugation axes and the finite strain x-axis could have developed if the fold axes rotated as material lines (e.g., Fossen et al., 2013). However, in the Buckskin-Rawhide core complex, the parallelism between mylonitic lineations (mean trend 041°–221°) and the corrugation axis defined by detachment surface measurements (040°–220°; Singleton, 2015) suggests either that no rotation has occurred or that corrugation axes rotated with the x-axis, as observed in physical models of transtensional folding (Venkat-Ramani and Tikoff, 2002). Accordingly, any systematic misalignment between the direction of large-magnitude extension and the topographic trend of corrugations can be used to quantify distributed strike-slip displacement lacking vertical-axis rotation.

In this study, we investigate the geometry and kinematics of brittle faults that cut across the belt of Miocene core complexes in western and central Arizona (Fig. 1) with the goal of: (1) determining the 3-D brittle strain during detachment faulting, and (2) evaluating how deformation changed following the cessation of core complex extension. We also evaluate the orientation of the dominant core complex extension direction and compare it to the topographic trend of the footwall corrugations to quantify the magnitude of postdetachment dextral shear strain across the core complex belt.

Much of this study is based on kinematic data from brittle faults across the crystalline footwalls of core complexes in central Arizona (the South Mountains and White Tank Mountains) and western Arizona (the Harquahala Mountains and Harcuvar Mountains; Fig. 1). The vast majority of faults we measured in these footwalls have centimeter- to meter-scale displacement with little to no breccia or cataclasite development. The entire brittle fault data set presented in this study includes 795 fault plane measurements and 618 sets of slickenlines (Supplemental File 11). We determined slip sense using standard kinematic criteria (e.g., Petit, 1987), with offset markers and Riedel shears serving as the most useful slip indicators. In the field, we determined a slip sense on 58% of faults with slickenlines, and we assumed a slip sense on an additional 27% of faults with unknown slip based on strong geometric similarity to spatially adjacent faults with a clear sense of slip, bringing the total slip-sense determinations to 70% of slickensided faults. Excluding the faults with an assumed slip sense does not change the overall kinematic patterns, and the remaining 30% of faults without an assigned slip sense still include many faults geometrically similar to those with a slip sense. We use incremental shortening and extension axes (P- and T-axes) to identify strain patterns and to assess the strain compatibility between fault populations (e.g., Marrett and Allmendinger, 1990). To characterize kinematically compatible fault populations, we present fault plane solutions from linked Bingham P- and T-axes using FaultKin software (based on algorithms presented in Allmendinger et al. [2012]). We identified kinematically compatible populations as those having P- and T-axes falling within the same dilation and compressional quadrants, respectively. In subsequent sections, we present the fault data organized by both geometry and kinematics.

To evaluate the extension direction during core complex development, we compiled published structural data and present new mylonitic fabric data as well as kinematic data from discrete (<1 cm thick) brittle-ductile shear zones. These shear zones are typically characterized by fractures with mylonitic veneers or “hot slickensides” in which quartz has undergone dynamic recrystallization. We use eigenvector statistics to evaluate mean orientations and the strength of point maxima or girdle distributions (with large maximum eigenvalues or relatively large intermediate eigenvalues, respectively; e.g., Vollmer, 1990).

We compare extension directions to the topographic trend of antiformal corrugations to evaluate postdetachment brittle shear strain under the assumption that the corrugation trend should parallel the overall core complex extension direction during detachment fault slip. In all of the Arizona core complexes, the detachment fault has been erosionally removed from the vast majority of antiformal corrugations, but linear topographic elements in the footwall typically reflect corrugation topography and parallel detachment fault slip directions (Spencer, 2000). Topographic trends were evaluated in a Mercator projection using the GeoMapApp application of the Marine Geoscience Data System (http://www.geomapapp.org; Ryan et al., 2009) and Google Earth. To systematically quantify the azimuth of the corrugations, we mapped geographic centers of the corrugation antiforms, determined the range drainage divides from shaded relief maps, summed kilometer-scale linear ridges and drainages as vectors, and made numerous independent visual estimates of the overall trend. Azimuth errors reflect the range of values obtained using these different methods.

Geologic Background

The South Mountains metamorphic core complex in southern Phoenix, Arizona, consists of variably mylonitic crystalline rocks exposed primarily along one ENE-trending, ∼18-km-long antiformal corrugation (Fig. 2). The eastern half of the corrugation is dominated by early Miocene granitoids (the South Mountains Granodiorite and the Telegraph Pass Granite) with U-Pb zircon ages ranging from ca. 20 to 22 Ma (Reynolds et al., 1986; Clements et al., 2013), whereas the western half consists of the Proterozoic Estrella Gneiss. Mylonitic fabrics are primarily restricted to the eastern 10 km of the range, and fabrics are more strongly developed near the top of the footwall (Reynolds, 1985). The footwall was exhumed during early to middle Miocene top-to-the-ENE–directed slip on a detachment fault that has been erosionally removed across most of the range (Reynolds, 1985). Thermochronometric data indicate that the footwall rapidly cooled below the brittle-plastic transition by ca. 18 Ma (Fitzgerald et al., 1993) and was exhumed to <80 °C by ca. 12 Ma (Prior et al., 2017). Brittle and ductile footwall structures consistently indicate that the Miocene core complex extension direction was ∼060° (Reynolds, 1985). Postdetachment faults have not been mapped at 1:24,000 scale (Reynolds, 1985).

Geometry and Kinematics of Brittle Faults

We measured 479 minor fault planes and 423 sets of slickenlines from several different locations in the South Mountains (Fig. 3). Approximately 70% of these measured faults are from outcrops in early Miocene granitoids and 30% from the Precambrian Estrella Gneiss. Northwest-striking, moderately to steeply dipping faults are dominant, and the mean fault orientation dips ∼74°NE (Fig. 3A). Altogether, 70% of measured faults dip ≥60°, and 86% dip ≥45°. Faults within the early Miocene granitoids are more uniformly oriented than those within the Estrella Gneiss (Figs. 3B, 3C). Many of the faults in the early Miocene granitoids parallel pervasive sets of extension fractures that typically dip steeply northeast (Reynolds, 1985), suggesting that shear reactivation of extension fractures was locally important. These extension fractures are not well developed in the Estrella Gneiss. Slickenlines record a range of slip directions, with strike-slip–dominated (<45° rake) and dip-slip–dominated (>45° rake) kinematics nearly equal (Figs. 3D, 3E). Despite the diversity in slickenline rake, the majority of faults record either subvertical or approximately north-south subhorizontal shortening axes and NE-SW to east-west subhorizontal extension -axes (Figs. 3F–3H).

The vast majority (∼90%) of faults with a known slip sense can be grouped into four kinematically compatible populations: (1) NW- to north-striking normal faults associated with NE-SW extension (∼29%), (2) NW-striking, steeply dipping NE-side-up faults (∼11%), (3) NW-striking dextral faults (∼35%), and (4) approximately north-striking sinistral faults (∼16%) (Fig. 4). Clear cross-cutting relationships between these fault populations are rare, but we did observe several examples of NW-striking normal and dextral faults that cut faults dipping ≤30°. In addition, we interpreted reverse dip-slip motion to overprint dextral motion on one NE-dipping fault with two sets of slickenlines.

Normal faults in the South Mountains record subvertical shortening and a mean subhorizontal extension direction of ∼064°–244° (Figs. 4A–4C), similar to the dominant 060° core complex extension direction recorded by mylonitic lineations, extension fractures, and slickenlines on detachment-related low-angle normal faults (Reynolds, 1985). A mean fault-plane solution based on linked Bingham shortening and extension axes yields a dominant nodal plane oriented 325°, 43°NE with a rake of 102° from NW, indicating a minor overall dextral component of slip associated with normal faulting (Fig. 4C). Chlorite and/or epidote mineralization appears to be most commonly associated with normal faults. Reverse and steeply dipping dip-slip faults typically record NE-side-up motion and are associated with shortening and extension axes that plunge moderately northeast and southwest, respectively (Figs. 4E–4F). As with the normal faults, the mean fault plane solution indicates a minor dextral component to the NE-side-up faults (323°, 84°NE, rake 80° from NW; Fig. 4F). These faults are kinematically compatible with some of the gently dipping (<20°) faults that record top-to-the-NE slip, subparallel to the slip direction of the detachment fault responsible for exhuming the mylonitic footwall (Figs. 4F–4H).

Dextral faults and sinistral faults have consistent orientations, primarily dipping steeply northeast and east, respectively (Figs. 5, 6). Mean fault-plane solutions indicate that purely strike-slip motion is dominant (dextral: 320°, 76°NE, rake 2° from NW; sinistral: 002°, 70°E, rake 0°; Figs. 5C, 5F). In some areas, these strike-slip faults form conjugate wedges (Fig. 6A), and we interpret most of the dextral and sinistral faults as a kinematically compatible population that together accommodated NNW-SSE shortening and ENE-WSW extension. The strikes of dextral and sinistral slip faults overlap at ∼340°–343°, similar to the trend of the acute bisector between the mean dextral and sinistral faults (trend and plunge 339°/01°), suggesting that these faults formed in a stress field where σ1 and σ3 were subhorizontal, trending ∼341° ± 2° and ∼71° ± 3°, respectively. Under this state of stress, the NW-striking extension fractures that are prevalent in the early Miocene granitoids would have been susceptible to reactivation as strike-slip faults. Given an effective σ3 near zero, fractures striking just a few degrees clockwise and counterclockwise of σ1 would be reactivated as sinistral and dextral faults, respectively (Fig. 7). The dominant strike of these extension fractures is ∼330° (Reynolds, 1985), so most would have been oriented for reactivation as dextral faults rather than sinistral faults. The abundance of dextral versus sinistral faults appears to reflect the significant role of extension-fracture reactivation. Dextral and sinistral faults are nearly equal in abundance in the Estrella Gneiss (n = 22 dextral or apparent dextral versus n = 19 sinistral or apparent sinistral), which lacks well-developed extension fracture sets, whereas dextral faults are about three times more abundant in the early Miocene granitoids (n = 86 dextral or apparent dextral versus n = 28 sinistral or apparent sinistral; Figs. 5G–5I). These distributions change little if a sense of slip is assumed on all strike-slip faults with an unknown sense of slip, where sinistral faults are assumed to strike north-south to NE-SW (strike clockwise of 345°), and dextral faults are assumed to strike NW-SE (counterclockwise of 340°): n = 31 dextral and n = 30 sinistral in the Estrella Gneiss; n = 120 dextral and n = 46 sinistral in the early Miocene granitoid.

Total Dextral Slip across the South Mountains

Based on our data set of faults with and without a slip sense, we estimate that NW-striking, steeply dipping dextral faults comprise ∼25% of faults across the South Mountains. In addition, NW-striking dip-slip faults (rake >45°) commonly record a dextral component of slip (Fig. 4). Although the vast majority of these faults are small-scale slip surfaces with centimeter-scale displacement, their abundance across the range suggests that total dextral shear is nontrivial.

To estimate the total component of brittle dextral shear, we compare the early to mid-Miocene core complex extension direction to the well-defined antiformal corrugation axis (Fig. 8). Reynolds (1985) documented the extension direction based on >1500 measurements of brittle and ductile structures, including dikes, mineralized extension fractures, striations in chloritic breccia near the top of the footwall, and mylonitic lineations (Supplemental File 2 [footnote 1]). These data are presented in 21 different categories based on type of structure, lithology, and location. The mean and mode extension directions recorded in nearly all of these categories are between ∼055° and ∼065°, and the total average extension direction is ∼059.8° from all categories of brittle structures and ∼060.6° from all categories of mylonitic lineations (Reynolds, 1985; Supplemental File 2). Although some of the individual measurements were rounded to the nearest value ending in 0 or 5, the consistency of the data set suggests that ∼060° is a reliable extension direction (best estimate from the Reynolds [1985] data = 060.2° ± 0.6°). We independently estimated the brittle extension direction in Google Earth by measuring the trace of 100 well-defined, steeply dipping extension fractures in 50–100-m-long segments, yielding a mean extension direction of 058.3° with a standard deviation of 6.8°. These results are similar the 059° mean dike extension direction determined by Reynolds (1985), based on the compilation of nearly 80 km of dike trends in ∼300 m segments.

The South Mountains form one of the most well-defined detachment fault corrugations in the Basin and Range. The clear ENE trend of the corrugation is delineated by the overall trend of the range and by drainages and ridges (Fig. 8). We estimate the trend of the range by: (1) summing individual linear drainages and ridges that are well defined for >1 km (061.6°), and (2) determining the overall trend of the center of the main corrugation (062.0°; Fig. 8). These estimates strongly suggest an overall trend that is ∼1°–3° clockwise of the extension direction (and likely 1.5°–2°). This subtle misalignment between the extension direction and the corrugation trend is potentially within the margin of error, but it is consistent with distributed dextral shear. If a 1.5°–2° misalignment is due entirely to dextral slip along NW-striking faults, the cumulative dextral shear across the ∼18-km-long exposed corrugation is ∼550 ± 80 m. North-striking sinistral faults would counter this misalignment (moving the topography toward a counterclockwise alignment with the extension direction), so the total dextral slip is likely greater, as reflected by the fault data (Fig. 5). To independently evaluate the magnitude of distributed dextral shear strain, we measured apparent strike-slip offset along steeply dipping (∼50°–90°), NE-striking layers (e.g., pegmatitic or aplitic dikes and distinct layers in the Estrella Gneiss; Fig. 6B). Apparent offsets were noted along all such layers with exposures >2 m long, regardless of whether or not they record strike-slip offset. Altogether we noted apparent offsets on 16 transects totaling 50.1 m in length along a 060° trend. The largest apparent strike-slip offset observed along these transects is 55 cm, and the majority of offsets are <10 cm. The total apparent offset across 50.1 m is 3.51 m of dextral separation along NW-striking faults and 0.33 m of sinistral separation along north-striking faults. The net dextral shear strain along NW-striking faults is 0.063 (3.18 m/50.14 m), which would translate to ∼1.14 km total dextral displacement over 18 km, roughly double the estimate from the extension direction–corrugation trend misalignment. The majority (∼73%) of this transect is within South Mountains Granodiorite, which contains a much higher percent of NW-striking dextral faults than the Estrella Gneiss, so 0.063 is likely an overestimate of the total dextral shear strain across the entire core complex. Additionally, scaling up shear strain from 50 m to 18 km is not likely to be accurate, but these results confirm that hundreds of meters of dextral slip could have been accomplished by centimeter-scale faults.

Geologic Background

The White Tank Mountains core complex is located just west of Phoenix, along strike with the South Mountains NE-dipping detachment system (Figs. 1, 9). Most of the range is composed of Late Cretaceous granitoid and Proterozoic crystalline gneiss that is variably mylonitized only in the northeastern part of the range (Reynolds et al., 2002; Fig. 9). Mylonites have ENE-WSW–trending stretching lineations that most likely record the Miocene extension direction (Reynolds, et al., 2002). As in the South Mountains, the top-to-the-ENE–directed detachment system that was responsible for exhuming these footwall rocks has largely been eroded, and hanging-wall rocks (Miocene sedimentary and volcanic rocks) are exposed only along the western edge of the range (Reynolds et al., 2002). The northeasternmost part of the range comprises a poorly defined corrugation ridge trending ∼063°–067°. Compared to other core complexes in Arizona, relatively little is known about the timing and magnitude of exhumation in the White Tank Mountains. One zircon (U-Th)/He age indicates that the northeastern end of the footwall cooled below ∼190 °C by ca. 16 Ma (Prior et al., 2017). Several NW- to NNW-striking, high-angle faults with unknown kinematics are mapped across the range at 1:24,000 scale (Reynolds et al., 2002).

Geometry and Kinematics of Brittle Faults

We measured 184 minor fault planes and 177 sets of slickenlines from the eastern White Tank Mountains (Fig. 10). Nearly all of these measurements are from road cuts in Precambrian crystalline gneiss and foliated granitoid. Most of these fault planes record centimeter-scale offset and have chlorite ± epidote ± Fe-oxide ± quartz mineralization. As in the South Mountains, most minor faults in the White Tank Mountains dip steeply northeast; 73% of all measured faults dip ≥60°, and 95% dip ≥45° (Fig. 10A). Slickenlines record a range of strike-slip, dip-slip, and oblique-slip motion, with east- to SE-plunging sets dominant (Figs. 10B, 10E). Kinematic axes for the entire data set are very coherent, recording overall NE-SW extension axes and a girdle distribution of shortening axes ranging from NW-SE subhhorizontal to subvertical (Figs. 10D, 10E). Altogether ∼80% of measured faults with a slip sense are compatible with this kinematic regime. The dominant fault populations are NE-dipping normal to oblique-normal faults and steep WNW- to NW-striking dextral to oblique-dextral faults (Figs. 11, 12). Extension axes for the dextral and normal faults are similar, with mean subhorizontal extension directions of 066° and 063°, respectively (Fig. 11). In addition, both populations record an overall oblique dextral-normal component, with the mean fault plane solution of 292°, 78°NE, rake 171° from the NW for the dextral set and 327°, 53°NE, rake 99° from the NW for the normal set (with each set separated by a 45° rake boundary; Fig. 11). Steep NE-striking sinistral and oblique sinistral faults are also present but far less common than dextral and normal faults (Fig. 11F). Relative timing relationships between dextral and normal slip are not clear, but both sets commonly record evidence for synkinematic chlorite ± epidote ± quartz (e.g., asymmetric growth in small releasing steps along the slickenline lineation; Fig. 12A). Some NE-striking sinistral faults also record evidence for synkinematic chlorite ± epidote ± quartz, and we noted inconsistent cross-cutting relationships between NE-striking sinistral faults and oblique normal-dextral faults, suggesting that these faults sets are broadly coeval.

In addition to brittle faults, we also measured 51 brittle-ductile shear surfaces with “hot slickensides” in the Late Cretaceous White Tank granite (Figs. 12C, 13A). These brittle-ductile surfaces are characterized by both mechanical striations and zones of dynamically recrystallized quartz typically <2 mm thick. The majority of these surfaces dip moderately to steeply west to SW and record top-to-the-SW–directed normal-sense slip antithetic to the detachment fault slip direction (Fig. 13A). Extension axes associated with this brittle-ductile slip trend ENE-WSW to NE-SW, with a mean trend and plunge of 252°/07°, which is ∼7° clockwise of the overall extension direction associated with brittle faults. Mylonitic lineations extracted from the Reynolds et al. (2002) geologic map of the White Tank Mountains indicate a mean trend of 060° and with maximum bin between 071° and 075° (Fig. 13B). The relatively large uncertainty of the mean core complex extension direction and footwall corrugation trend obscures analysis of potential misalignment between the two.

Geologic Background

The Harquahala and Harcuvar Mountains in west-central Arizona are continuous with other metamorphic core complexes in the lower Colorado River extensional corridor (Fig. 1). This belt of core complexes parallels the WNW-trending margin of the Maria fold-and-thrust belt, a zone of Cretaceous thick-skinned shortening that likely influenced the geometry and location of Miocene extension (Spencer and Reynolds, 1991; Singleton et al., 2018; Spencer et al., 2018). Located ∼60–100 km west-northwest of the White Tank Mountains, these core complexes were unroofed by ∼45 km of top-to-the-NE displacement across the Eagle Eye–Bullard detachment fault system in the early to middle Miocene (ca. 21–14 Ma; Reynolds and Spencer, 1985; Spencer and Reynolds, 1991; Carter et al., 2004; Prior et al., 2016). Both ranges consist of prominent NE-trending antiformal corrugations dominated by Proterozoic, Jurassic, and Cretaceous crystalline rocks (Richard et al., 1994; Bryant, 1995). The eastern ∼25–30 km of the Harcuvar Mountains have a >1-km-thick zone of mylonitic fabrics with NE-trending lineations, whereas mylonites in the Harquahala Mountains are restricted to a <0.5-km-thick zone beneath the detachment fault in the eastern ∼10–15 km of the range (Richard et al., 1990). Isolated exposures of the detachment fault system are exposed along the flanks of the eastern parts of both ranges, defining an antiformal geometry mimicked by the footwall topography (Richard et al., 1990; Bryant, 1995). The Harcuvar footwall corrugation is topographically continuous for ∼45 km, whereas the ∼30-km-long Harquahala footwall corrugation is disrupted by several NW-striking postdetachment faults that form prominent breaks in the range (Fig. 14). The four largest of these postdetachment faults separate the detachment fault and/or topographic envelope of the range 2.4–2.5 km in a dextral sense and 0.5–1.3 km in a NE-side-up sense (Richard et al., 1990).

Geometry and Kinematics of Brittle Faults

Our brittle fault data from the Harquahala Mountains come primarily from minor fault surfaces near the four largest NW-striking postdetachment faults: the Dushey Canyon, Sunset Canyon, Harquahala Mountain, and Socorro-Hercules fault zones (Fig. 15). Faults within and adjacent to the Dushey Canyon fault damage zone record two distinct episodes of brittle slip. The first episode is characterized by oblique dextral-normal slip on steeply NE-dipping surfaces, which accommodated overall ENE-WSW–directed extension and shortening along a moderately south-plunging axis (Fig. 15B). The dextral-normal faults typically contain chlorite and epidote mineralization, suggesting that they were active during the late stages of detachment slip (Fig. 16A). A younger episode of slip on the Dushey Canyon fault is inferred from faults containing clay gouge ± calcite, which likely record slip at shallower structural levels. Minor faults with gouge in the Dushey Canyon damage zone are characterized by oblique dextral-reverse slip, whereas one exposure of the Dushey Canyon fault principal slip plane defined by calcite mineralization and gouge has downdip slickenlines interpreted to record reverse (NE-side-up) slip (Figs. 16B, 16C). These gouge-bearing faults accommodated north-south to NE-SW shortening and extension along a moderately NW-plunging axis (Fig. 15C). Minor faults located <200 m from the Sunset Canyon, Harquahala Mountain, and Socorro-Hercules faults have variable kinematics, but overall these data suggest that these faults dominantly record oblique dextral-NE-side-up slip (Fig. 15D), which is compatible with their map-scale offsets. Several NNE- to NE-striking faults record oblique sinistral-SE-side-up slip that is compatible with dextral-NE-side-up slip on the NW-striking faults (Fig. 15E).

Fault data from the Harcuvar Mountains were collected from nonmylonitic footwall rocks near Cunningham Pass and across the mylonitic footwall to the northeast (Fig. 1). Unlike in the Harquahala Mountains, our data collection was not focused on specific postdetachment fault zones. The dominant pattern of faulting from these data is dextral or oblique dextral-NE-side-up slip on steep NW-striking faults and normal slip on NE-dipping faults (Fig. 15F). Steep NNW- to north-striking sinistral and oblique sinistral-east-side-up faults are also present. One NW-striking fault zone with oblique dextral-NE-side-up slip cuts the Bullard detachment fault near Burnt Well, but the relative timing of all other faults with respect to the Bullard detachment fault is unclear.

Total Dextral Slip across the Harquahala and Harcuvar Mountains

As in the South Mountains, we compare the Harquahala and Harcuvar core complex extension directions with the topographic trend of the footwall corrugations to estimate the magnitude of postdetachment dextral slip. We estimate the Miocene extension direction in the Harquahala Mountains from: (1) greenschist-facies mylonitic lineations <100 m below the Eagle Eye detachment fault at the northeastern end of the core complex, and (2) extension axes (T-axes) from discrete top-to-the-NE normal-sense shear zones with “hot slickensides” in the Stone Corral area (Figs. 14, 17). The mean orientation of mylonitic lineations and the brittle-ductile T-axes indicate extension directed toward 056° and 051°, respectively, and together these structures indicate an overall extension direction of ∼054° (Fig. 17). Mylonitic lineations measured by Richard et al. (1990) from the Stone Corral area record an overall extension direction of 050°, but most of these lineations record top-to-the-SW sense of shear along a gently SE-dipping shear zone that is likely latest Cretaceous in age, based on U-Pb geochronology of synkinematic titanite, and thus unrelated to Miocene core complex development (Pollard et al., 2013; Pollard, 2014). Richard et al. (1990) also presented measurements of early Miocene mafic dikes across the Harquahala footwall, which record an overall extension direction of 035°–040°. Interestingly, extension directions recorded by early Miocene dikes in the lower Colorado River extensional corridor typically do not closely match the extension direction recorded by detachment slip or mylonitic lineations. For example, the mean extension direction associated with early Miocene dikes in the southwest Whipple Mountains (southeastern California, USA) is ∼30° clockwise of the dominant NE-SW core complex extension direction (Gans and Gentry, 2016), and early Miocene dikes in the Chemehuevi Mountains (southeastern California) record dominantly north-south extension at a high angle to the NE-SW extension direction from mylonitic lineations and detachment fault slickenlines (LaForge et al., 2017). Considering all of these factors, we consider 054° ± 2° to be the best estimate of the Miocene core complex extension direction in the Harquahala Mountains (Fig. 18).

Determining the mean extension direction in the Harcuvar Mountains is complicated by both the presence of older mylonitic fabrics and variation of lineation trends across the core complex. Greenschist-facies mylonitic fabrics in the Harcuvar lower plate are concentrated along a zone within 300 m of the Bullard detachment fault and near the mylonitic front, whereas amphibolite-facies mylonitic fabrics are dominant in leucogranite within the core of the lower plate. Multiple lines of evidence suggest that the amphibolite-facies fabrics are latest Cretaceous to Paleogene in age, whereas the greenschist-facies fabrics are almost certainly early Miocene in age (Wong et al., 2013; Singleton and Wong, 2016). However, all fabrics record top-to-the-NE shear and are geometrically indistinguishable. It is possible that the anisotropy created during latest Cretaceous–Paleogene top-to-the-NE–directed shearing governed the Miocene ductile extension direction, and thus mylonitic lineations (regardless of age) may provide an estimate of the Miocene extension direction. Mylonitic lineations southwest of Smith Peak typically trend ∼050°–055°, whereas lineations northeast of Smith Peak trend 060°–065° (Figs. 1921; Bryant and Wooden, 2008; Spencer et al., 2016). Mean lineation trends from our measurements are ∼3°–5° counterclockwise of the mean trends from Bryant and Wooden (2008). However, our mean values may be overly influenced by some anomalous domains, and when viewed as individual domains, our lineation data are consistent with the domain means from Spencer et al. (2016), who also noted a clockwise rotation of the lineation trend toward the northeast (Fig. 19). The change in lineation is even more pronounced in the adjacent Little Buckskin Mountains (Fig. 19) footwall, where lineations gradually transition from ∼015° from the southwestern end of the 10-km-long corrugation to ∼060° at the northeast end (Fig. 22). Epidote veins across the Little Buckskin Mountains corrugation show the same clockwise rotation trend, matching the extension direction of the lineations, despite the fact that the lineations are associated with amphibolite-facies deformation conditions whereas the epidote veins are completely brittle features (Fig. 22). We suggest that microscale anisotropy of the mylonitic fabrics influenced the vein orientations, and that neither type of structure is a good proxy for the original orientation of the footwall corrugation. The Little Buckskin corrugation is well defined by its antiformal foliation structure, which defines a fold axis oriented 238°/0.5° (Singleton, 2011, 2015). Mylonitic foliation across the Harcuvar lower plate also defines a gentle antiform (though not as well defined as the Little Buckskin antiform), and poles to foliation form an approximate girdle distribution corresponding to a fold axis oriented 227°/06° (Fig. 20B). Basaltic to andesitic dikes across the footwall typically dip steeply to moderately northeast and define a mean extension direction parallel to 048° (Fig. 20C). Near Burnt Well along the NW-dipping flank of the Harcuvar Mountains, slickenlines along lower-plate fault surfaces parallel to the Bullard detachment fault record top-to-the-NE–directed slip with an average slickenline orientation of 223°/09° (Fig. 20D). Taken together, these data do not provide a consistent mean extension direction in the Harcuvar core complex.

We estimate the overall trends of the Harquahala and Harcuvar footwall corrugations based on the orientations of the drainage divide near the crest of each range and the approximate geographic center of the corrugation with respect to the flank margins (Figs. 18, 19). The Harquahala corrugation trend is 056° ± 2°, whereas the Harcuvar corrugation trend is 055° ± 1° southwest of Smith Peak and 068° ± 1° northeast of Smith Peak. Despite uncertainties, it is clear that the topographic corrugation trend is clockwise of the extension direction in both ranges (Figs. 18, 19). The ∼2° misalignment between the Harquahala corrugation trend and extension direction can be accounted for with ∼1.2 km of dextral slip across the ∼35-km-long range. Given that the four largest NW-striking postdetachment faults cumulatively have ∼2.45 km of apparent dextral slip, ∼1.2 km may be an underestimate. If the drainage divide trend is taken as the corrugation trend, the misalignment is ∼4°, corresponding to ∼2.45 km of total dextral slip.

If mylonitic lineations are used as a proxy for the extension direction in the Harcuvar Mountains, the misalignment with corrugation topography is ∼0°–5° southwest of Smith Peak and 3°–8° northeast of Smith Peak (Fig. 19). All other indicators of extension direction suggest >7° misalignment. Given the large uncertainties in the mean extension azimuth, we suggest that the best estimate of total dextral slip across the Harcuvar Mountains may arise from data in the adjacent Little Buckskin Mountains. The mapped dextral offset across NW-striking faults in the Little Buckskin Mountains is ∼0.7 km across ∼9.8 km (Singleton, 2011), and the clear foliation-defined antiform trend and topographic trend of the Little Buckskin corrugation are misaligned by ∼2°–3°. Applying this brittle dextral shear strain and misalignment to the Harcuvar Mountains suggests that the total dextral slip is 1.5–2.5 km across the northeastern ∼35 km of the range. We tentatively interpret the total brittle dextral shear strain across the Harcuvar and Harquahala core complexes as ∼0.03–0.07, corresponding to ∼1.2–2.5 km of total slip across ∼35 km.

Synextensional Brittle Constriction in the Arizona Core Complex Belt

Brittle fault data presented in this study record an evolution of distributed Neogene deformation across the western to central Arizona core complex belt. We interpret lower-plate faults associated with chlorite and/or epidote to have formed during or shortly following top-to-the-NE directed slip on the bounding detachment fault system, which is also strongly associated with chlorite and epidote. These faults yield insight into the strain field during the late stages of core complex development. Faults in the White Tank Mountains, which commonly contain synkinematic chlorite ± epidote ± quartz, provide the clearest example of this strain field. West-northwest–to NW-striking dextral faults and NE-dipping normal faults dominate the White Tank Mountains. The dextral faults commonly have a normal-slip component, whereas normal faults commonly have a dextral-slip component, and each set records an ENE extension direction that is consistent with the extension direction from mylonitic lineations and brittle-ductile shears (Figs. 11, 13). Cumulatively, these minor faults accommodated a constrictional strain characterized by shortening along a subvertical plane approximately perpendicular the extension direction. This girdle distribution of shortening axes and clustered distribution of extension axes matches constrictional strain patterns predicted in transtensional settings (Fossen et al., 2013).

Normal faults in the South Mountains are also locally associated with chlorite ± epidote and are kinematically similar to those in the White Tank Mountains, recording extension subparallel to the core complex extension direction. Many of these normal faults also have a dextral component of slip (Figs. 4A–4C). Strike-slip faults in the South Mountains differ from those in the White Tank Mountains in that they do not record an overall normal-slip component, and dextral faults typically strike NW rather than WNW (Fig. 5). These NW-striking dextral faults make up the dominant fault population, recording overall east-west extension and north-south shortening that is seemingly incompatible with core complex extension (Figs. 5A–5C). However, this pattern of faulting was most likely influenced by shear reactivation of NW-striking extension fractures that are abundant in the early Miocene granitoids (but relatively rare in the pre-Miocene rocks that dominate the White Tank, Harquahala, and Harcuvar Mountains), resulting in an unequal distribution of NW-striking (mostly) dextral faults compared to north-striking sinistral faults. Viewed as conjugate pairs, the dextral and sinistral faults are compatible with a subhorizontal σ3 direction of ∼068°, whereas the geometric boundary between dextral and sinistral slip (overlapping with strikes of ∼340°–343°) suggests that these strike-slip faults formed with a σ3 azimuth of 070°–073° (see section South Mountains Metamorphic Core Complex). The likely influence of extension fracture reactivation on this strike-slip fault population may complicate paleostress determinations from mean orientations, but in one location we observed a conjugate strike-slip fault wedge bisected by extension fractures, suggesting a subhorizontal σ3 direction of ∼065° during Coulomb failure (Fig. 6A). Although most strike-slip faults are characterized by Fe-oxide mineralization rather than chlorite or epidote, the similarity between these inferred σ3 directions and the core complex extension direction (060°) suggest that many of these strike-slip faults could have accommodated a component of syndetachment constriction.

Chlorite- and epidote-bearing faults along the Dushey Canyon fault zone in the eastern Harquahala Mountains primarily record oblique dextral-normal slip and associated ENE-WSW extension (Fig. 15B). This faulting closely resembles the overall pattern in the White Tank Mountains (e.g., the mean fault plane solution in the Dushey Canyon fault zone, not including gouge-bearing faults, is 313°, 69°NE, rake 132° from NW, very similar to the mean fault plane solution from all data in the White Tank Mountains: 307°, 57°NE, rake 132° from NW; Figs. 11E, 15B).

Taken together, fault slip data from the Harquahala–White Tank–South Mountains core complex belt indicate that the late stages of core complex development were characterized by a 3-D constrictional strain regime. Subvertical shortening and subhorizontal shortening perpendicular to the NE extension direction were accommodated by normal faults, oblique normal-dextral faults, and conjugate strike-slip faults. In all three core complexes, the dominance of dextral slip over sinistral slip in faults with a strike-slip component suggests that constriction was transtensional in nature as opposed to purely divergent and coaxial with core complex extension. It is difficult, however, to separate postdetachment from syndetachment dextral faults when evaluating the relative magnitude of dextral versus sinistral slip during brittle constriction. In some areas constriction was dominated by transtensional oblique slip, whereas in other areas slip was partitioned between normal faults and conjugate strike-slip faults that may have accommodated constriction coaxial with the NE-directed extension (Fig. 23A). Given that the trend of the core complex belt is oblique to the extension direction, synextensional constriction likely records a component of dextral, pure shear–dominated transtension. This obliquity is particularly clear between the Whipple Mountains and the Harcuvar Mountains, where the trends of the mylonitic front (southwestern margin of footwall mylonite zones) and the northeasternmost footwall exposures are both ∼115°–120°, which is ∼60°–75° clockwise of the extension direction in these core complexes. This 115°–120° trend continues southeast to the White Tank and South Mountains and appears to define the orientation of the core complex deformation-zone boundary, in which case a component of oblique divergence is required given that the extension direction is not perpendicular to the boundary. A 60°–75° angle of divergence across this boundary would have resulted in a ∼9%–18% component of dextral simple shear parallel to the core complex belt and up to 7°–15° of clockwise rotation at infinite strain (Fossen et al., 2013). One potential problem with this interpretation of minor oblique divergence is the lack of systematic orientation differences between ductile and brittle extension indicators, which would be expected if earlier-formed ductile structures recorded more rotation. Brittle and ductile structures record very similar extension directions in the Buckskin-Rawhide Mountains (Singleton, 2015) and in the South Mountains (Reynolds, 1985; Supplemental File 2 [footnote 1]), and paleomagnetic data from Cretaceous granitoids and Miocene dikes in the footwall of the Harquahala Mountains do not record statistically significant vertical-axis rotation (Livaccari and Geissman, 2001). It is possible that transtensional constriction did not become important until after the footwall cooled to brittle conditions, consistent with the interpretation of Singleton (2013) that constriction in the Buckskin-Rawhide Mountains was primarily restricted to the middle to late stages of core complex development. Clockwise rotations due to a component of dextral divergence may have postdated ductile deformation and/or have been too small to clearly identify with structural or paleomagnetic data. In addition, the 3-D geometry of the oblique deformation zone is unclear, so relationships derived by Fossen et al. (2013) for vertical shear zone boundaries may not fully apply to the Arizona core complex belt. The orientation of this belt may have been guided by the location of a Mesozoic crustal welt (Spencer and Reynolds, 1991), by mid-crustal weak zones inherited from Mesozoic tectonism (e.g., Singleton et al., 2018; Spencer et al., 2018), and/or by basement structures that originated during Proterozoic rifting (e.g., Timmons et al., 2001).

Our interpretation of transtensional constriction in the Harquahala–White Tank–South Mountains core complexes is consistent with evidence for constriction and corrugation folding in the Buckskin-Rawhide core complex (Singleton, 2013). Synextensional constriction results in a component of horizontal shortening perpendicular to the extension direction, which likely either formed corrugations or amplified original detachment fault grooves into well-defined corrugations. Singleton (2015) interpreted transtensional and dextral faulting to have initiated near the cessation of slip on the Buckskin detachment fault, following corrugation development and potentially coinciding with a phase of late-stage ENE- to east-directed slip on the detachment system. We interpret distributed oblique dextral-normal and conjugate strike-slip faulting in the Harquahala–White Tank–South Mountains to have initiated in a constrictional strain regime that was likely important during detachment faulting and corrugation development. Interpreted as folds, corrugations in the Buckskin-Rawhide core complex record only 1%–10% NW-SE shortening (Singleton, 2013)—far less than the component of NE-SW extension during detachment faulting, which in mylonitic early Miocene plutons is likely on the order of 50%–90% via ductile stretching (Singleton and Mosher, 2012) and up to 20%–30% via brittle normal faulting (Singleton, 2015). Modeled as a pure shear–dominated transtensional system with a 60°–75° angle of divergence, 100% NE-SW extension would correspond to a slightly prolate strain geometry with ∼1%–5% NW-SE shortening (Fossen et al., 2013). Pure shear–dominated transtension thus may account for most (if not all) of the strain recorded by corrugation folding.

Influence of the Pacific–North America Plate Boundary on the Arizona Core Complex Belt

Syndetachment dextral and normal faults have been documented in several detachment fault systems associated with the transtensional segments of the Pacific–North America plate boundary (e.g., Axen and Fletcher, 1998; Oldow et al., 2008; Seiler et al., 2010; Luther and Axen, 2013). In addition, syndetachment constriction has also been documented in the Central Mojave and Whipple Mountains metamorphic core complexes in southern California (Fletcher and Bartley, 1994; Yin, 1991; Yin and Dunn, 1992), which both record NE-directed extension coeval with extension in the Arizona core complex belt. These core complexes and the Arizona core complex belt are typically not interpreted to be directly linked to the Pacific–North America plate boundary, though they record similar styles of deformation to the transtensional detachment systems. Constriction and associated synextensional corrugation folding may also characterize core complexes in settings not located near transcurrent systems, potentially driven by uniaxial stress fields that arise during large-magnitude extension (e.g., Fletcher and Bartley, 1994; Singleton, 2013). While some constrictional strain in the southern California and Arizona core complexes may be associated with processes inherent to large-magnitude extension and/or preexisting boundary conditions that gave rise to a component of oblique divergence, the particularly well-defined nature of detachment fault corrugations in this region and the association with postmylonitic (post–early Miocene) dextral faults suggests that constriction was enhanced by far-field stresses associated with the Pacific–North America plate boundary. Even if all syndetachment constrictional strain can be accounted for by dextral transtension during oblique divergence, these far-field stresses could have influenced the divergence direction (to an orientation clockwise of pure divergence) and contributed to an elevated magnitude of horizontal σ2 during extension. Dextral shear directly associated with this diffuse plate boundary became increasingly distributed in the Mojave Desert and Colorado River region toward the end of the middle Miocene (McQuarrie and Wernicke, 2005), potentially overlapping with detachment slip and linking transtensional constriction to postdetachment dextral faulting as a continuum of deformation.

In the Harquahala, Harcuvar, and Buckskin-Rawhide Mountains map-scale NW-striking dextral and oblique dextral faults clearly postdate detachment slip (post–ca. 12 Ma), indicating that a significant component of dextral slip was not coeval with synextensional constriction. Singleton (2015) attributed postdetachment dextral faulting in the Buckskin-Rawhide Mountains to the influence of the diffuse Pacific–North America plate boundary. Some dextral shear associated with this transform jumped into the Mojave Desert and Colorado River region starting ca. 12–10 Ma (McQuarrie and Wernicke, 2005), roughly coinciding with the transition from detachment faulting to distributed dextral shear in the western Arizona core complexes. This transition also appears to have corresponded to a shift to approximately north-south shortening and approximately east-west extension (Singleton, 2015).

In the White Tank and South Mountains, the timing of dextral slip on NW-striking faults with respect to detachment fault slip is uncertain due to the lack of detachment fault exposure. However, given the postdetachment timing of map-scale dextral faults in the Harquahala-Harcuvar-Buckskin-Rawhide Mountains, it is likely that some of the minor dextral faults in the White Tank–South Mountains postdate detachment slip. In the South Mountains, evidence for dextral slip on steep fractures striking as close to north as ∼340°–343° is consistent with a postdetachment clockwise rotation of the horizontal shortening and extension directions, as the boundary between dextral and sinistral slip should otherwise occur at ∼330° (perpendicular to the extension direction) if all strike-slip faulting was syndetachment. While many dextral faults across the core complex belt likely originated during synextensional constriction, we suggest that a rotation of stress directions at the end of detachment slip or following cessation of detachment slip resulted in continued dextral slip and reactivation of syndetachment joints as dextral faults (Fig. 23B).

The youngest phase of faulting in the western to central Arizona core complex belt is associated with oblique dextral-reverse (NE-side-up) slip (Fig. 23C). In the Buckskin-Rawhide Mountains, this faulting regime is recorded by the Lincoln Ranch fault system, which includes the only faults observed to cut poorly consolidated postdetachment deposits (Scott, 2004; Singleton et al., 2014b; Singleton, 2015). Several faults in the Harcuvar Mountains (including the only clear postdetachment fault) record oblique dextral–NE-side-up slip, and the clay gouge–bearing faults in the Dushey Canyon fault zone in the Harquahala Mountains record oblique dextral–NE-side-up slip or NE-side-up reverse slip (Figs. 15C, 15F). We did not observe these faults in the White Tank Mountains, but the NE-side-up reverse faults in the South Mountains (which commonly also have a minor dextral component of slip; Figs. 4D–4F) may belong to this fault population. Viewed as one fault population, these oblique dextral-reverse and NE-side-up faults record transpressional shortening along a gently NNE-plunging axis (Fig. 23C). There is no evidence that these faults cut Quaternary deposits, but the late Neogene timing of this faulting is unclear. If this transpressional slip is associated with the Pacific–North America plate boundary, it likely postdates the plate motion shift from dextral divergence to margin-parallel dextral motion ca. 8 Ma (Atwater and Stock, 1998), and/or postdates the late Miocene to Pliocene shift to dextral transpression in central and southern California (e.g., Page et al., 1998; Ducea et al., 2003).

The systematic misalignment between the core complex extension direction and the topographic trend of the detachment fault corrugations provides insight into the cumulative magnitude of dextral slip across the core complex belt. Regardless of the uncertainties over the extension directions and corrugation trends, it is clear that dextral shear strain associated with NW-striking faults increases northwestward. Our estimates for dextral shear strain are 0.03 across the South Mountains (∼0.55 km total slip across 18 km) and >0.03–0.07 across the Harquahala and Harcuvar Mountains (∼1.2–2.5 km of total slip across ∼35 km). Dextral shear strain appears to increase abruptly in the Buckskin-Rawhide Mountains to ∼0.2 (∼7–8 km across 36 km; Singleton, 2015). This distributed dextral fault system is approximately along strike with the Stateline fault system (∼150 km northwest of the Buckskin-Rawhide Mountains; Fig. 1A), which is located near the northeastern margin of the Eastern California shear zone and is thought to record ∼30 km dextral slip since ca. 13 Ma (Guest et al., 2007). Displacement southeast of the Stateline fault system may have been largely accommodated by post–6 Ma normal faults (Mahan et al., 2009) and/or transferred to dextral faults to the southwest. The along-strike decrease in dextral shear from the California-Nevada state line to Phoenix is consistent with previous studies suggesting that Miocene to recent regional dextral shear strain in the lower Colorado River region is less than that along strike in the eastern Mojave Desert region (Richard, 1993; Bennett et al., 2016). The southeastward decrease in dextral shear across the core complex belt thus reflects a regional pattern that is directly tied to the development of the Pacific–North America plate boundary. The western to central Arizona core complexes are located ∼200–290 km northeast of the San Andreas fault, highlighting the wide influence of the Pacific–North America plate boundary on deformation in the southwestern Cordillera.

Fault data presented here indicate that Miocene metamorphic core complexes in western and central Arizona experienced constrictional strain accommodated by coeval normal faults, conjugate strike-slip faults, and oblique dextral-normal faults. This brittle constriction records extension approximately parallel to the ductile footwall stretching direction and is commonly associated with chlorite mineralization, suggesting that this strain regime was coeval with the late stages of detachment faulting. The oblique orientation between the Arizona core complex belt and the extension direction likely resulted in a minor component of dextral transtension, which may account for much of the constrictional strain and folding of detachment fault corrugations. In addition, regional stresses associated with the diffuse Pacific–North America plate boundary most likely contributed to dextral transtension during the late stages of core complex development and to the subsequent transition to postdetachment dextral shear. Following cessation of detachment fault slip at ca. 14–12 Ma, NW-striking dextral and oblique dextral-reverse (NE-side-up) faults slightly reoriented the topography of the core complex corrugations, which now trend clockwise of the core complex extension direction that likely paralleled their original orientation. Based on our analysis of this misalignment between corrugation topography and the extension direction, we propose that distributed dextral shear strain increases northwestward across the core complex belt from 0.03 across the South Mountains (∼0.55 km total slip across 18 km), >0.03–0.07 across the Harquahala and Harcuvar Mountains (∼1.2–2.5 km of total slip across ∼35 km), and ∼0.2 across the Buckskin-Rawhide Mountains (∼7–8 km across 36 km). This zone of dextral shear is along strike with the Stateline fault system, which accommodated ∼30 km of dextral slip near the northeastern margin of the Eastern California shear zone (∼150 km northwest of the Buckskin-Rawhide Mountains; Guest et al., 2007). The along-strike variation in postdetachment dextral shear reflects regional strain patterns associated with the Eastern California shear zone, where cumulative dextral shear strain in the lower Colorado River region is lower than it is along strike in the eastern Mojave Desert region.

Partial funding for this project was provided by National Science Foundation Tectonics Program award 1557265 to J. Singleton and M. Wong and by Colorado State University startup funds to J. Singleton. We thank Evan Strickland, Mike Wyatt, and Jacqueline Benefield for assistance with some of the data collection. This manuscript was improved by reviews from Sue Beard, an anonymous reviewer, and Associate Editor Bob Miller. Discussion of oblique divergence in the Arizona core complex belt was inspired by insightful comments from Sue Beard.

1Supplemental Files. File 1: Brittle fault data from the South Mountains, White Tank Mountains, and Harquahala Mountains. File 2: Compilation of extension direction indicators in the South Mountains (from Reynolds, 1985). Please visit https://doi.org/10.1130/GES02036.S1 or access the full-text article on www.gsapubs.org to view the Supplemental Files.
Science Editor: Shanaka de Silva
Associate Editor: Robert B. Miller
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.