Abstract

Determining the earliest history of a multiply reactivated fault is difficult, particularly when the rock record is incomplete and brittle deformation reorients strain markers. We present a method that incorporates strain analysis of an adjacent fold, inferred deformation temperatures, and published thermochronology to find the initiation age of the north-striking Picuris-Pecos fault of the southern Rocky Mountains (USA). The Picuris-Pecos fault has accommodated at least 37 km of dextral slip. The older east-trending Hondo syncline (New Mexico) is refolded adjacent to the fault and has a map pattern consistent with dextral drag. We analyzed quartz grain shapes, deformation mechanisms, and fractures within the Hondo syncline to determine if the refold is genetically related to the Picuris-Pecos fault. Nine calculated grain-shape ellipsoids predate refolding and are rotated about a vertical axis subparallel to strike of the refold limbs. Healed fractures and semibrittle microstructures are most abundant in the hinge of the refold and adjacent to the fault. We interpret refolding and slip on the Picuris-Pecos fault to have initiated during unroofing through the brittle-plastic transition (∼250–310 °C). Regional thermochronologic studies suggest that the rocks of the Hondo syncline were at this temperature range during the Grenville orogeny. The refold is probably a fault-propagation fold that predated throughgoing faulting, or possibly formed during dextral drag on the Picuris-Pecos fault. Refolding in the Picuris Range and the Truchas Mountains accounts for at least 2.6 km of the total dextral shear on the fault.

INTRODUCTION

Preexisting lithospheric weaknesses control the orientation and location of subsequent deformation in orogenic belts (e.g., see reviews by Marshak, 2004; Butler et al., 2006). The most recent strain field and variation in crustal rheology create a complex system of new and reactivated faults (Butler et al., 2006). Thus, fault initiation and the early history can be difficult to discern. Brittle reactivation of middle crustal structures can cause fracturing and cataclasis that can reorient or destroy early strain markers. Older histories are often limited by erosion or poor exposure of the oldest structures (Sanders et al., 2006). Fault reactivation can lead to a complex thermal history, making thermochronology data difficult to interpret (Sanders et al., 2006).

Prior geologic methods for relative and precise dating of particular fault events require complete stratigraphic sequences with well-known ages, crosscutting relationships, and/or the ability to directly date deformation products (Holdsworth et al., 1997). These methods can be inadequate when the record of the oldest event has been tectonically reworked several times. Here, we present a technique that is effective for finding the early history of structures formed in the middle crust that are now exposed at the surface. It is necessary to characterize strain both adjacent to the fault and outside of the damage zone so that late reorientation of structures can be recognized. We combine strain analysis of an adjacent fold, deformation mechanisms, macrostructures, and previously published thermochronologic data with crosscutting relationships to evaluate the early history of the Picuris-Pecos fault (PPF).

Recognition of reactivated middle crustal structures is significant because fluid migration and ore-bearing mineralization can occur in long-lived lithospheric structures (Holdsworth etal., 1997). In addition, fault reactivation can have other consequences related to seismic hazard assessment and economic exploration (Butler etal., 1997; Holdsworth et al., 1997; Lyon et al., 2007; McBride et al., 2007). In regions where fault zones are buried by thick basin sediments, small earthquakes on reactivated structures can be used to identify deep structures that may pose a seismic risk (McBride et al., 2007). Basement-involved faulting may breach hydrocarbon traps (Lyon et al., 2007) or gas storage stratigraphic units (McBride et al., 2007).

Many argue that Phanerozoic structures in the western U.S. Rocky Mountains have a Proterozoic ancestry (Miller et al., 1963; Weimer, 1980; Karlstrom and Humphreys, 1998; Karlstrom etal., 1999, 2004;,Marshak et al., 2000), but the establishment of the northward trend of this orogenic belt (Fig. 1A) prior to the onset of breakup of the supercontinent Rodinia at 0.8 Ga remains undocumented (Karlstrom et al., 2004). An example of this from the southern Rocky Mountains is a network of north-striking, dextral strike-slip faults with a net separation of as much as 170 that could represent multiple movements of uncertain ages (e.g., Chapin and Cather, 1981; Karlstrom and Daniel, 1993). The age and magnitude of dextral slip remain contended, largely because of a lack of piercing points (Cather etal., 2011). The PPF is the largest of this network of strike-slip faults and has been repeatedly reactivated (Fig. 1; Bauer and Ralser, 1995; Erslev et al., 2004; Cather et al., 2005, 2006). Timing estimates for the initiation of the fault and dominant dextral motion range from Mesoproterozoic to Neoproterozoic (ca. 1.4–0.8 Ga; Montgomery, 1963; Bauer and Ralser, 1995; Wawrzyniec etal., 2007), Paleozoic (Ancestral Rocky Mountains, 325–290 Ma; Woodward et al., 1999), and/or Laramide (ca. 80–40 Ma; Cather, 1999; Cather et al., 2005, 2006).

Semibrittle deformation related to the PPF and an adjacent refolded fold indicate that fault initiation occurred when the rocks passed through the brittle-ductile transition ca. 1.0 Ga (Sanders et al., 2006). We characterize the style and degree of deformation with distance from the fault and with location in the refold to determine if the refold and the PPF have a genetic relationship. We then determine the relative timing of PPF slip and refolding based on crosscutting relationships, strain analyses, the degree of brittle versus ductile deformation, and by comparing our inferred temperatures of deformation to past thermochronological studies.

KINEMATIC HISTORY OF STRIKE-SLIP FAULTING IN THE SOUTHERN ROCKY MOUNTAINS

The PPF of northern New Mexico is a multiply reactivated, ∼80-km-long, steeply dipping, north-striking fault (Fig. 1). Dextral slip of ∼37km on the PPF is based on offset of Proterozoic fold hinges, thrust faults, and distinctive marker horizons (e.g., Montgomery, 1963; Grambling, 1979; Karlstrom and Daniel, 1993; Daniel et al., 1995; Cather, 1999; Cather et al., 2006, 2011). In the Picuris Mountains, the PPF truncates the Proterozoic Hondo syncline (Miller et al., 1963), which is refolded adjacent to the northern PPF and has a map pattern consistent with right-lateral or west-side-down drag (Fig. 1B). Previous workers interpret refolding and attenuation of limbs transposed parallel to the PPF as having formed during Proterozoic ductile right-lateral slip, when the rocks were below the brittle-ductile transition (Miller et al., 1963; Bauer and Ralser, 1995). However, no crystal-plastic deformation textures have been linked to refolding; thus, it could have occurred under brittle-plastic or brittle deformation conditions (e.g., McDonald and Nielsen, 2004). Evidence for brittle slip on the PPF in the Picuris Mountains includes a wide breccia zone adjacent to the fault (Bauer and Ralser, 1995) and abrupt truncation of stratigraphic units and foliation (Karlstrom and Daniel, 1993).

Net slip across the PPF is dominantly strike slip. Offset of hinge zones of the Hondo and correlative Brazos Cabin synclines limits the dip-slip component to a few kilometers, and rocks of both were metamorphosed in similar conditions (∼3.5 kbar, 510 °C) (Grambling, 1981; Daniel et al., 1995). In addition, 40Ar/39Ar (Erslev et al., 2004; Sanders et al., 2006) and apatite fission track (Kelley and Chapin, 1995) cooling ages from rocks of roughly the same elevations on either side of the fault are similar. Horizontal slickenlines on the PPF suggest that recent motion also was strike slip (Bauer, 1987; Bauer, 1993; Bauer and Ralser, 1995).

Direction and magnitude of strike slip on the PPF during the Ancestral Rocky Mountains orogeny (ca. 325–290 Ma) is poorly constrained (Cather et al., 2005). Based on Mesozoic isopach maps, Woodward et al. (1999) hypothesized that major dextral offset on the PPF occurred during the Ancestral Rocky Mountains orogeny. Ye et al. (1996) used drill-hole data and basin subsidence studies to suggest that Ancestral Rocky Mountains strain is a result of northeast-southwest–directed shortening, consistent with dextral slip on north-striking faults. In Cather et al. (2011), source terranes west of the PPF were matched with proximal early to middle Pennsylvanian deposits on the east using clast compositions and monazite ages; it was argued that the PPF accommodated ∼40–50 km of dextral separation since the early Pennsylvanian (during the late Ancestral Rocky Mountains and/or Laramide orogenies), and it was inferred that ∼3–13 km of sinistral separation must have existed on the fault during the Ancestral Rocky Mountains orogeny (early Desmoinesian). The occurrence of a sinistral slip event on the PPF is supported by the brittle anticlockwise rotation of foliation in the damage zone of the PPF south of the study area (Cather et al., 2008).

The Laramide orogeny (ca. 75–35 Ma) involved northeast or east-northeast crustal shortening compatible with dextral strike slip on faults oriented subparallel to the PPF (steep, north striking; Cather et al., 2006). The extent of Laramide strike-slip faulting in New Mexico remains a topic of debate (e.g., Bauer and Ralser, 1995; Erslev, 2005; Cather et al., 2005). Structural studies of regional fault patterns suggest minimal (Baltz, 1967; Woodward et al., 1997; Yin and Ingersoll, 1997; Cather, 1999; Erslev, 2001, 2005) to >100 km of Laramide dextral strike-slip motion (Chapin and Cather, 1981; Karlstrom and Daniel, 1993) within the southern Rockies. Yin and Ingersoll (1997) proposed that Laramide slip was negligible; Woodward et al. (1997) estimated 5–20 km of slip; and in Cather (1999), it was suggested that 33–110 km of dextral motion was accommodated on the system of north-striking steep faults at the eastern edge of the Colorado Plateau. Paleomagnetic data from late Paleozoic rocks from near the southern PPF suggest no significant post-Paleozoic vertical-axis rotations (Wawrzyniec et al., 2007).

PICURIS RANGE GEOLOGY

We studied quartzites of the thick (∼3 km) Paleoproterozoic (ca. 1700–1680 Ma) metasedimentary Hondo Group. The base of the Hondo Group consists of ∼1 km of mature (commonly >95% quartz) crossbedded quartzite of the Ortega Formation (Bauer and Williams, 1989). This is overlain by six interlayered pelitic schist and crossbedded quartzite members of the Rinconada Formation (subdivided by Nielsen, 1972). The limbs and axial surface of the Hondo syncline are subparallel to slightly oblique to the dominant S2 schistose or crenulation cleavage fabric in the region (Bauer, 1987, 1993). Two major Proterozoic thrusts, the Pilar and Plomo faults, bound the F2 folds (Bauer, 1987, 1993). These rocks were multiply ductilely deformed by north-south shortening prior to or ca. 1.4 Ga (Bauer, 1987, 1993; Williams et al., 1999), and the refolding of the Hondo syncline overprints these dominant Proterozoic structures (e.g., Cather et al., 2006).

The overturned Hondo syncline is north vergent with a south-dipping axial surface and an east-trending hinge line. It is refolded within ∼1–1.2 km of the PPF (Fig. 1) in a map pattern consistent with drag on a right-lateral strike-slip fault or a west-side-down dip-slip fault. The refold is gentle to open (interlimb angle ∼125°) and plunges ∼42° toward the south-southwest. Adjacent to the PPF, the north limb of the Hondo syncline strikes approximately north-south, subparallel to the PPF, and dips steeply west, ∼90° different from the regional strike to the west. The southern, overturned limb of the Hondo syncline within the refold is more strongly attenuated than the northern limb (Bauer and Kelson, 1997; Bauer et al., 2003), and is also transposed subparallel to the strike of the PPF.

The timing of Paleoproterozoic to Mesoproterozoic deformation in the Picuris Range is debated; it is bracketed by the S2 foliation that is present in granites ca. 1.65 Ga age, but not in those dated as 1.4 Ga (Fig. 1B) (Bauer and Williams, 1989; Bauer, 1993; Melis, 2001; Daniel and Pyle, 2006). Peak metamorphism in the Picuris Range occurred between 510 and 525 ± 10 °C and 4.1 kbar ± 100 bar (Daniel and Pyle, 2006). Monazite ages that record the timing of metamorphism range from 1434 ± 12 Ma in grain cores to 1390 ± 20 Ma on grain rims (Daniel and Pyle, 2006). The 39Ar/40Ar ages from the Picuris Range indicate slow cooling at midcrustal depths for ∼100–200 m.y. following the ca. 1.4 Ga thermal event (Karlstrom et al., 1997), in agreement with other studies in the region (e.g., Sanders et al., 2006). Following the Paleoproterozoic to Mesoproterozoic shortening events and peak metamorphism, nearby areas south of the Picuris Mountains cooled to 200–350 °C by ca. 1.1 Ga (Sanders et al., 2006; Cather et al., 2006).

METHODS

We measured and described macroscopic deformation, characterized quartz grain microstructures, and measured the axial ratios and long-axis orientations of quartz grains within and west of the refold. We then inferred the strain field and mechanical conditions (brittle, semibrittle, or plastic). We were then able to determine the timing of crystal-plastic, semibrittle, and brittle strain relative to refolding and PPF slip.

We collected all data and quartzite samples along three transects (one from the north limb of the Hondo syncline and two from the south limb) from 4.1 km to within 0.1 km of the PPF measured perpendicular to the fault strike (Fig. 2). Data were mostly collected from quartzite layers to reduce rheologic complications and because schist units are not well exposed. Ortega quartzite samples from the north limb of the Hondo syncline are denoted as XhoN and from the south limb as Xho. We use Xhr for the Rinconada quartzite samples, which were all collected in the south limb of the Hondo syncline.

Quartz Microstructures and Deformation Mechanisms

We describe quartz grain microstructures from several samples along strike of the Hondo and Rinconada quartzites. We correlate quartz grain microstructures to those formed by different dislocation creep processes. Dynamic recrystallization processes produce distinct quartz microstructures as a function of temperature and/or strain rate (Hirth and Tullis, 1992, 1994; Stipp et al., 2002; Tullis, 2002; Stipp and Kunze, 2008). Observed microstructures thus allow us to place constraints on the conditions that prevailed at the time of crystal-plastic deformation. We assess the degree and conditions of crystal-plastic, semibrittle, and brittle deformation in each sample relative to the position in the refold. We then combine these data with a strain analysis (described in the following) to differentiate microstructures formed during folding of the Hondo syncline, refolding, and PPF slip.

Strain Analysis

We measured the axial ratios and long-axis orientations of grains in 36 quartzite samples. First, thin sections were cut in two common orientations: horizontal with north-south and east-west edges, and vertical, east-west–striking planes. These were chosen because the horizontal plane should most strongly reflect any preferred orientation of strain ellipses related to distributed strike-slip shear on the PPF, and the vertical, east-west sections should also record any fault-perpendicular shortening, as is suggested by the apparent attenuation of Hondo syncline limbs as the PPF is approached. All 36 samples have a horizontal section, and 20 of them also have an east-west vertical thin section; subsequently nine samples were cut along north-south vertical planes in order to fully describe the three-dimensional (3-D) strain ellipsoids and to determine how these ellipsoids change shape and orientation with position along the strike of the refold or in relation to the PPF.

We assume that quartz grain axial ratios primarily reflect the relative degree of strain (Elliott, 1970) and that the maximum shortening direction in each sample is perpendicular to the average plane of grain flattening (Elliott, 1970; Shimamoto and Ikeda, 1976; Miller and Oertel, 1979; Milton, 1980; Paterson and Yu, 1994). We measured 59–256 grains in each section. Wemanually traced quartz grains in Adobe Illustrator, and then detected the grains using ImageSXM (S.D. Barrett, 2011, http://www.ImageSXM.org.uk), a free program from NIH Image that measures parameters such as perimeter length, grain area, long and/or short axes of best-fit ellipse, and orientation of the long axis. The magnitudes of axial ratios (R = long axis:short axis) of the best-fit ellipses and the trend and plunge orientations of the long axes are used in this study.

We calculated average 3-D grain-shape ellipsoids (Milton, 1980) for 9 samples using three average 2-D strain ellipses in 3 mutually perpendicular thin sections using the visual basic program ELLIPSOID (Laboratoire de Planétologie et Géodynamique de Nantes, 2002, http://www.sciences.univ-nantes.fr/geol/UMR6112/SPO; Robin, 2002; Launeau and Robin, 2005). We found the size of each 2-D strain ellipse from the mean of the axial ratios from each sample (Elliott, 1970; Shimamoto and Ikeda, 1976; Miller and Oertel, 1979; Paterson and Yu, 1994). We plotted the orientation of the long axis of each grain with R ≥ 1.5 on a rose diagram to find the maximum elongation direction. We only used samples for which ≥70% of the grain long axes are oriented within 10° of the mean orientation. The nine sample locations are along strike of the refold in the southern limb of the Hondo syncline and include the three samples with the highest axial ratios in the horizontal plane from each location along the refold (adjacent to the fault; eastern limb, hinge zone, and western limb of the refold).

Fracture Analysis

We measured and described fractures in the field and in thin section with distance from the PPF. We measured fracture densities perpendicular and parallel to the fault in the field; we also measured the orientation of minor faults and striae. In horizontal thin sections, we measured the strike of fractures, including healed fractures to compare with those described in the field.

RESULTS

Quartz Deformation Microstructures and Mechanisms

No strong mylonite fabric exists, but statistical analyses of grain shapes indicate that quartz grains are commonly elongate subparallel to bedding. Quartz microstructures within each sample are similar in each orientation (horizontal, east-west vertical, and north-south vertical). Undulatory or patchy extinction and annealed fractures are present in most samples (Figs. 3A–3D). Most quartz microstructures from the upright and overturned limb are distinct from one another. Samples from the overturned (south) limb (Fig. 3A) show undulatory extinction, small recrystallized grains (≤10 μm), larger subgrains, rare intragranular fractures, and sutured grain boundaries. Most recrystallized grains are concentrated around grain boundaries or within healed cracks (Fig. 3B). Other samples from the overturned limb have sutured grain boundaries, weak undulatory extinction, rare recrystallized grains, and rare subgrains (Fig. 3C). Many samples with similar structures have abundant intragranular fractures. Hondo quartzite samples from the upright limb of the syncline (Fig. 3D) have large quartz grains with amoeboid shapes and checkerboard extinction. “Island grains” (old grains overgrown by new recrystallized grains) enclosed within quartz are also abundant. Quartz grain boundaries also have small-scale sutures.

The microstructures observed in the south limb quartzites are consistent with those produced by bulging recrystallization processes (BLG; ∼280–400 °C; Stipp et al., 2002). Samples that have a higher proportion of subgrains and recrystallized grains are similar to the higher temperature–lower strain rate end (BLG II), where dislocation climb and subgrain rotation become more important (Hirth and Tullis, 1992; Stipp et al., 2002). Samples with more intragranular fractures and rare recrystallized grains fit better with the lower temperature–higher strain rate end (BLG I), where grain boundary migration is slow and activated by strain. North limb samples have checkerboard extinction diagnostic of deformation at temperatures >630°C (Kruhl, 1996), and microstructures similar to those formed by rapid grain boundary migration processes along with dislocation glide and climb (Hirth and Tullis, 1992; Stipp et al., 2002; Tullis, 2002; Stipp and Kunze, 2008). We plotted the inferred deformation mechanism of each sample to determine if there were systematic changes along strike of the refold and with distance from the PPF (Fig. 4).

Brittle-ductile structures overprint the microstructures indicative of plastic deformation. In the hinge of the refold and adjacent to the PPF, grain-size reduction occurred mainly by brittle fracturing, but limited crystal plasticity is indicated by undulatory and patchy extinction, annealed fractures, and recrystallized grains along fractures (Fig. 5). These microstructures are consistent with those that form at low temperatures by semibrittle flow (Hirth and Tullis, 1994; Stipp et al., 2002), during which brittle fractures grade into zones of recrystallized grain. Cracks can contain recrystallized grains (Fig. 5A), representing slow grain boundary migration that is confined to fractures or crack tips. Figures 5B and 5C are examples of several annealed fractures of a brittle-ductile nature, and Figure 5D shows an annealed fracture subparallel to a zone of cataclasis. The semibrittle fracture density is highest in the hinge of the refold and adjacent to the PPF. No semibrittle deformation textures were observed in background samples (provided by P. Bauer) from ∼2–6 km away from the PPF and the refold.

Grain-Shape Analysis

For the 3-D strain analysis, we determined 9 grain-shape ellipsoids from various locations along the refold. We chose the south (overturned, shallower) limb samples for this analysis because they contain the most elongate grains. Also, north-limb grain shapes were formed by high rates of grain boundary migration as opposed to strain. To compare the shape of the ellipsoids, we normalized the magnitude of each principal axis using e3 equal to 1 (where e1, e2, and e3 are the minimum, intermediate, and maximum shortening directions, respectively). The ellipsoids are prolate with e1> e2≈ e3. Rranges from 1.7 to 9.5 (Fig. 6A) and the ratio e2:e3 ranges from 1 to 1.8 (Fig. 6B). Figures 6A and 6B illustrate that there is no systematic change in the shape of the ellipsoid along strike of the refold. Figures 6C and 6D show the trend and plunge of the maximum shortening direction (e3) versus distance from the PPF. In most samples, the plunge of e3 is moderate (∼40°–60°) and does not vary significantly or systematically along the refolded fold. Generally, the trend of e3 near the fault (east of the refold hinge) is toward the east or the west. West of the refold hinge, e3 trends toward the north or south.

The grain-shape ellipsoids are generally elongate parallel to bedding (bedding is subparallel to e1 and e2). Stereonets of the trend and plunge of each axis of the 3-D grain-shape ellipsoid and an approximate bedding measurement from each sample are shown in Figure 7. Locally bedding planes do not contain e1 and/or e2, and, in some cases, bedding contains e3. Furthermore, in some locations, particularly areas where Xho samples are adjacent to the fault, the quartzite units are massive, so bedding orientation was obtained from a nearby outcrop. This could lead to errors if samples were collected from a local heterogeneity, such as a small, parasitic fold.

The shapes of the 2-D grain-shape ellipses do not vary systematically with distance from the PPF (Fig. 8) and are consistent with the 3-D data set. In the horizontal plane, the highest axial ratios are from the samples closest to the PPF. Table 1 contains a summary of the average values from oriented thin-section planes for rock units on the north and south limbs of the Hondo syncline. Quartz grains viewed in the horizontal plane typically have slightly higher R values when compared with the two vertical planes. Because the plunge of the maximum elongation direction is moderate, our thin sections are oblique to the plane containing the maximum elongation.

We infer the 2-D shortening directions for each thin section from the average orientation of e1 of individual grains. In the horizontal sections (Fig. 9), the directions of e1 axes vary along the refold from dominantly northwest-southeast in the far field to east-west directly adjacent to the fault. In the hinge of the refold, e1 orientations range from northwest-southeast, northeast-southwest, and east-west. In the east-west vertical plane, e1 orientations in quartz grains from the far field (>2 km from the fault) plunge from ∼0° to 30° (Fig. 10). Vertical 2-D samples collected closer to the fault (<1.7 km from the fault) have shortening directions plunging ∼45°–75° from horizontal (Fig. 10). We cannot infer much from the 2-D vertical section set, but find that the 2-D horizontal data set is nearly identical to the results from the 3-D ellipsoid data.

FRACTURE ANALYSIS

PPF-related damage consists of brittle, brittle-ductile fractures, and variable degrees of cataclastic deformation and extends to 1.4 km west of the fault. Two north-striking fault-tip zones intersect the south limb, ∼1 and 2 km west of the PPF. As a result, fracture orientations are very scattered; most fractures are subparallel to bedding (Fig. 11A) and only 3% of the fracture data set are right-lateral faults subparallel to the PPF. To avoid this complexity, we only assess the fractures in the north limb of the Hondo syncline. Most fractures are east-west–striking, commonly subparallel to bedding (Figs. 11A, 11B). A subsidiary set of north-south–striking steep fractures strike subparallel to the PPF and are commonly striated (Fig. 11C). Striae range from east-trending dip-slip to dextral, north-trending strike-slip orientations. East-west–striking fractures are most dense in the hinge of the refold (Fig. 12). The north-south–striking fractures decrease in density with distance from the PPF (Fig. 12).

Figure 13 shows the strikes of macrofractures measured in the field compared with strikes of microfractures measured in horizontal thin sections. Most macrofractures and microfractures from the east and west limbs of the refold have similar strikes. Macrofractures in the hinge of the refold are extremely variable.

DISCUSSION

We find that high-temperature quartz microstructures and grain shapes are related to Hondo syncline formation and predate refolding and PPF slip. This interpretation is based on several of our results. (1) Apparent deformation mechanisms vary significantly between the north and south limbs of the Hondo syncline (Fig. 4), but no systematic change occurs along strike of the refold or with distance from the PPF. (2) The 3-D strain ellipsoids and horizontal 2-D strain ellipses are reoriented within the refold. (3)The quartz grain shape maximum elongation direction is commonly parallel to bedding that has been transposed parallel to the limbs of the Hondo syncline.

Quartz microstructures suggest that the Hondo syncline formed at conditions consistent with BLG (Stipp et al., 2002). Picuris samples display microstructures indicative of stark differences in grain boundary mobility, which is controlled by temperature. Sutured grain boundaries, small recrystallized grains, and intragranular fractures form by bulging recrystallization at low temperatures or high strain rates (Hirth and Tullis, 1994; Stipp et al., 2002).

However, microstructures such as checkerboard extinction and amoeboid shapes with island grains form by fast grain-boundary migration, a process that is only active at high temperature (Hirth and Tullis, 1992; Tullis, 2002). Relatively small differences in the interpreted temperature of plastic deformation (<∼150°) along strike of each limb may result from differences in original grain size or local strain rate or total strain differences.

We propose that a thermal pulse heated the rocks of the north limb of the Hondo syncline, causing high rates of grain boundary migration. No remnants of plutonic rocks that may have intruded during this episode are exposed, but this limb is bounded by the Embudo fault, which may have downfaulted any such plutons beneath the fill of the Rio Grande rift. The lack of a metamorphic grade change from north to south suggests that the north limb is unlikely to have been exhumed from much greater depths than the south limb. North limb samples were overprinted by lower temperature and/or higher strain-rate microstructures (small-scale sutures), consistent with the south limb samples.

Refolding of the Hondo syncline and at least some localized PPF slip occurred at or near the brittle-ductile transition, based on the prevalence of semibrittle microstructures that overprint high-temperature microstructures. Annealed fractures and limited crystal plasticity are most abundant in the hinge of the refold and adjacent to the PPF. Macroscopic fractures are also most abundant in the hinge of the fold and, because they are subparallel to microfractures, likely have a semibrittle origin. We attribute north-south–striking fractures to PPF slip because of their orientation, the presence of right-lateral slip on some of them, and their decreasing density westward away from the fault.

High-temperature grain-shape ellipsoids and ellipses are reoriented as the PPF is approached. The trend of the maximum shortening direction rotates from north-south outside the refold, consistent with Mesoproterozoic shortening (e.g., Bauer, 1987), to east-west adjacent to the fault, indicating right-lateral rotation. The plunges of the long axes of ellipsoids are generally shallow and do not change systematically as the fault is approached. Dip-slip motion on the PPF is unlikely during refolding because vertical-plane axial ratios are the same along strike (Figs. 8B, 8C), indicating that modification of high-temperature grain shapes was minimal. However, the highest axial ratios are in horizontal planes and adjacent to the PPF. Thus, localized dextral strain related to PPF slip probably overprints the higher temperature deformation.

We interpret refolding to have been related to brittle-ductile, dextral drag or fault-propagation folding along the PPF. The lack of a mylonite zone on the PPF suggests that the refold may have formed as a fault-propagation fold before the PPF accommodated a large offset as a throughgoing structure. However, the refold may be a drag fold if strain rates within the PPF fault zone were too high for deformation to be accommodated by plasticity, which would explain the lack of a mylonite zone. Because our data suggest that the refold formed during initiation of the PPF, we prefer the interpretation that it formed by fault propagation as the PPF grew toward the surface. In addition, drag folds are generally smaller, within a few tens of meters of the causative fault zone, so this refold may be too large for formation by this mechanism (e.g., Reches and Eidelman, 1995; Groshong, 2006).

We believe that brittle-ductile deformation and modified axial ratios indicate that refolding and initial PPF slip occurred during the Grenville orogeny because (1) the Hondo syncline was unroofed through the brittle-ductile transition ca. 1.1 Ga during the Grenville orogeny (Cather et al., 2006; Sanders et al., 2006) and (2) Sanders et al. (2006) described east-west shortening ca. 1.0 Ga in the southern Sangre de Cristo Mountains.

The PPF likely has a late Mesoproterozoic (Grenville) origin, but because it has been reactivated, our data do not constrain the age of the major dextral slip on the system. Some cataclastic deformation and fractures clearly related to PPF dextral slip overprint deformation related to refolding. Major breccia zones are present along the PPF in the study area and elsewhere; these imply a significant component of brittle slip along the fault. In Cather et al. (2011), it was argued that 3–13 km of sinistral separation existed on the fault in the early Pennsylvanian. If this is correct, it implies that the Grenville-age (ca. 1100 Ma) dextral slip event documented here was followed by a significant sinistral slip event prior to the middle Pennsylvanian, probably in the Neoproterozoic (Cather et al., 2006) or during the early part of the Ancestral Rocky Mountains deformation.

CONCLUSIONS

We conclude that the refolding of the Hondo syncline was related to initial dextral shear on the PPF at temperatures of ∼250–300 °C. The refold is probably a fault-propagation fold, but a drag-fold origin cannot be ruled out. Older grain-shape ellipsoids rotate around the hinge of the refold, consistent with dextral slip. Little to no change of the plunge of the maximum elongation direction across the refold suggests that dip-slip motion is unlikely. This semibrittle shear probably initiated during Grenville orogenesis, based on comparison of temperatures inferred from quartz microfabrics with regional thermochronometry. The PPF has been reactivated brittlely several times since then, and most PPF slip postdates the refolding event. Our results are consistent with the hypothesis that the north-trending structural grain in the Rocky Mountains was established during the Proterozoic (prior to ca. 0.8 Ga) and that subsequent deformation has been localized on these old structures (e.g., Karlstrom et al., 2004; Sanders et al., 2006).

We integrated petrography and microstructures to determine the strain field and temperature ranges of the earliest deformation on a multiply reactivated fault. These data, tied with field observations and map-scale geometry, were compared to published thermochronology to find the age of refolding adjacent to the PPF. Our results suggest that it may be difficult to learn much about the early (semibrittle or plastic) history of a fault like this from studying only rocks near the fault that are strongly overprinted by late brittle fabrics that reorient early-formed strain markers. We recommend analyzing the deformation with distance from the fault to quantify the rotation of strain markers and to characterize overprinting strain events.

We thank Jane Selverstone for insightful discussions and helpful reviews of this paper. Paul Bauer graciously provided background samples from outside our study area. We also thank Jeff Amato, Carol Frost, and Whitney Behr for thorough reviews. Funding was provided by National Science Foundation grants EAR-0809638 to Gary J. Axen and EAR-0809220 to Jane Selverstone. The New Mexico Bureau of Geology and Mineral Resources provided support to Amy Luther for this research as part of her doctoral thesis at the New Mexico Institute of Mining and Technology.