## Abstract

Anisotropy of magnetic susceptibility (AMS) and structural studies of red beds in the Wyoming salient were completed to evaluate relations of magnetic fabrics to layer-parallel shortening and vertical-axis rotation in curved fold-thrust systems. The red beds display cleavage, fractures, veins, minor folds, and minor faults that accommodated widespread early layer-parallel shortening and minor strike-parallel extension. Magnetic susceptibility is carried mostly by paramagnetic phyllosilicates and ferromagnetic hematite that have composite fabrics related to sedimentary deposition, diagenesis, and tectonic processes. Anisotropy of magnetic susceptibility fabrics range from distinctly oblate ellipsoids parallel to bedding that reflect dominant sedimentary fabrics (type 1), to moderately oblate ellipsoids with weak magnetic lineations roughly parallel to the intersection of weak layer-parallel shortening fabrics and bedding (type 2), to triaxial and prolate ellipsoids with distinct magnetic lineations parallel to the intersection of moderate layer-parallel shortening fabrics and bedding (type 3). Type 1 sites occur mostly in the central, frontal part of the salient where layer-parallel shortening is <5%, whereas type 3 sites are found mostly in more interior thrust systems and toward the salient ends where layer-parallel shortening is >15%. Magnetic lineations are subparallel to structural trend and exhibit a tangential pattern around curved fold-thrust systems. Regional patterns of anisotropy of magnetic susceptibility are broadly similar to patterns of finite strain estimated from reduction spots. Combined with paleomagnetic data, anisotropy of magnetic susceptibility data indicate that early layer-parallel shortening fabrics started with minor primary curvature and then underwent significant vertical-axis rotation during large-scale thrusting. Correlations with finite strain, structural, and paleomagnetic data sets indicate that analysis of anisotropy of magnetic susceptibility in weakly deformed red beds is useful for evaluating kinematic evolution of thrust systems.

## INTRODUCTION

Determining the spatial distribution and temporal evolution of strain and vertical-axis rotation in curved fold-thrust belts is critical for characterizing the complete kinematic history and ultimately the mechanical development of mountain systems. Quantifying the nature of strain, however, is difficult in weakly deformed sedimentary rocks within external parts of mountain belts where tectonic fabrics are subtle. One potential tool for acquiring fabric information in such weakly deformed rocks is anisotropy of low-field magnetic susceptibility (AMS). AMS analysis provides a rapid and sensitive technique for measuring preferred orientations of magnetic grains averaged over relatively large volumes of rock (≥10 cm^{3}). Thus, AMS provides a complementary tool that can be combined with typical structural analysis to develop improved kinematic models of fold-thrust belts.

AMS relates the directional variability of sample magnetization (*M*) in response to an applied field (*B*), and is defined by a symmetric second-rank tensor, [*K*], given by *M* = [*K*]*B*. The eigenvectors and eigenvalues of [*K*], *K*_{max} ≥ *K*_{int} ≥ *K*_{min}, give the orientations and lengths of the three principal axes of the corresponding AMS ellipsoid. AMS defines an oblate ellipsoid if *K*_{max} ≅ *K*_{int} > *K*_{min} with *K*_{min} perpendicular to magnetic foliation, a prolate ellipsoid if *K*_{max} > *K*_{int}≅ *K*_{min} with *K*_{max} parallel to magnetic lineation, and a triaxial ellipsoid if *K*_{max} ≠ *K*_{int} ≠ *K*_{min}. Considerable research has shown links between the AMS ellipsoid, preferred orientations of mineral grains, and strain (see reviews by Uyeda et al., 1963; Borradaile and Tarling, 1981; Hrouda, 1982; Rochette et al., 1992; Borradaile and Henry, 1997; Borradaile, 2001).

The AMS ellipsoid depends on both the magnetic (mineral susceptibility and anisotropy) and physical (shape, size, and preferred orientation) properties of grains integrated over a volume of rock (Owens, 1974; Rochette, 1987; Hrouda and Schulmann, 1990; Aubourg et al., 1995; Debacker et al., 2009). An advantage of using AMS is that contributions from all magnetic components are combined to provide a description of the overall magnetic fabric (Borradaile, 1988). This can also be a disadvantage, as AMS may record contributions from multiple ferromagnetic (*sensu lato*), paramagnetic, and diamagnetic minerals that grew at different times and deformed by different mechanisms. Consequently, AMS often represents a composite fabric related to multiple depositional, diagenetic, and tectonic processes, which complicates fabric interpretation (Housen et al., 1993; Aubourg et al., 1995; Robion et al., 1999; Evans et al., 2003; Martin-Hernandez and Hirt, 2004).

Layer-parallel shortening (LPS) is a widespread component of internal strain that accumulates early in the deformational history of fold-thrust belts (Engelder and Geiser, 1979; Mitra and Yonkee, 1985; Geiser, 1988; Ong et al., 2007). LPS results in preferred alignment of mineral grains by multiple processes, including pressure solution, grain rotation, and neocrystallization, which may be recorded by AMS. Graham (1966) first established a link between AMS and LPS fabrics, and subsequent studies have utilized AMS in variably deformed sedimentary rocks for kinematic analysis of fold-thrust belts (e.g., Kligfield et al., 1981; Kissel et al., 1986; Lowrie and Hirt, 1987; Bakhtari et al., 1998; Parés et al., 1999; Cifelli et al., 2004; Smith et al., 2005).

Numerous studies in active and ancient fold-thrust belts have shown that AMS is useful for interpreting deformation fabrics, but the tectonic significance of magnetic foliations and lineations varies with location and lithology. Magnetic foliation (the plane perpendicular to *K*_{min}) is typically parallel to bedding in weakly deformed rocks but becomes parallel to cleavage in strongly deformed areas (e.g., Housen and van der Pluijm, 1990, 1991; Hirt et al., 2004; Averbuch et al., 1992). Magnetic lineation (parallel to *K*_{max}) typically tracks the intersection between LPS and bedding fabrics and is subparallel to regional structural trend (e.g., Borradaile and Tarling, 1981; Bakhtari et al., 1998), or tracks the maximum extension direction and is subperpendicular to structural trend (e.g., Kligfield et al., 1981; Hirt et al., 2000). Mixed patterns also occur (Aubourg et al., 1991), and oblique fabrics locally develop in complex structures (Saint-Bezar et al., 2002). Thus, tectonic interpretation of AMS fabrics must be done judiciously and with care to determine carriers of magnetic susceptibility, nature of primary fabrics, deformation processes, and relations to other structures (Parés and van der Pluijm, 2002).

In general, an idealized suite of AMS ellipsoids develops in fold-thrust belts as primary sedimentary fabrics are progressively overprinted by tectonic fabrics (Borradaile and Tarling, 1981; Parés and Dinarès-Turell, 1993; Sagnotti and Speranza, 1993; Parés and van der Pluijm, 2003). Primary sedimentary fabrics related to deposition and diagenesis (compaction and cementation) are characterized by distinctly oblate AMS ellipsoids with *K*_{min} perpendicular to bedding and *K*_{max} ≈ *K*_{int} lying within the bedding plane (stage A in Fig. 1). With progressive deformation, a range of composite fabrics forms. For minor LPS, the tectonic fabric is weaker than the primary fabric, *K*_{min} remains perpendicular to bedding, *K*_{max} roughly clusters parallel to the intersection of an incipient LPS fabric with bedding, and the AMS ellipsoid becomes triaxial (stage B in Fig. 1). For moderate LPS, the strengths of tectonic and primary fabrics become similar, *K*_{min} begins to scatter away from the bedding pole, *K*_{max} strongly clusters parallel to the intersection of LPS and bedding fabrics, and the AMS ellipsoid becomes prolate (stage C in Fig. 1). For high strain, the tectonic fabric becomes dominant, *K*_{min} clusters perpendicular to cleavage, and *K*_{max} may cluster either parallel to structural trend or down the dip of cleavage, depending partly on the nature of three-dimensional (3-D) strain (stage D in Fig. 1). In detail, the evolution of composite AMS fabrics is complex and depends not only on strain but also on the intensity of primary fabrics, mineralogy, and deformation mechanisms.

To further evaluate relations of AMS to internal strain in weakly to moderately deformed sedimentary rocks, and to test the use of AMS for kinematic analysis of curved fold-thrust belts, integrated AMS, structural, and paleomagnetic studies were completed for the Wyoming salient of the Sevier thrust belt (Fig. 2). The Wyoming salient is an excellent location for such integrated studies because red beds suitable for AMS and paleomagnetic analysis are widely exposed, tectonic structures (cleavage, fractures, and veins) are well developed, finite strain markers (reduction spots) are available for direct comparison with AMS fabrics, and regional relations are well constrained. In this paper we address the following questions:

What are the relations of AMS fabrics, LPS directions, and structural trend on a regional scale around the Wyoming salient?

How are AMS ellipsoid shapes related to strain magnitude and lithology in weakly to moderately deformed red beds?

How can AMS studies best be combined with detailed structural and paleomagnetic studies to determine kinematic evolution of curved fold-thrust belts?

## GEOLOGIC SETTING

The Sevier thrust belt lies in the North American Cordillera and is characterized by folds and thrust faults that shortened and translated mostly sedimentary rocks overall eastward during the Early Cretaceous to early Paleogene (Armstrong and Oriel, 1965; Armstrong, 1968; DeCelles, 2004). The belt is geometrically divided into a number of salients (Lawton et al., 1994), including the Wyoming salient, which is bounded on the north and south by the basement-cored Gros Ventre and Uinta foreland uplifts (Fig. 2). Major thrust systems in the Wyoming salient, from west to east, are the Willard (comprising the Willard, Paris, and Meade thrusts), Crawford (and associated folds), Absaroka (and associated imbricates), and Hogsback system (comprising the Hogsback, Darby, Prospect, and Granite Creek thrusts). Major thrust and fold traces display >90° of regional curvature in map view, from NW trends in the northern part of the salient to NE trends in the southern part, with additional curvature near oblique thrust ramps and transfer zones (Fig. 2). Synorogenic deposits and thermochronologic data record an overall foreland propagating (west to east) sequence of thrusts, from the Early Cretaceous Willard system, to the mid-Cretaceous Crawford system, to the mostly Late Cretaceous Absaroka system, to the early Paleogene Hogsback system (Wiltschko and Dorr, 1983; Burtner and Nigrini, 1994; DeCelles, 2004). Emplacement of the Hogsback system overlapped temporally with development of the bounding foreland uplifts (Dorr et al., 1977; Bradley and Bruhn, 1988).

Cross sections from the Wyoming salient show typical fold-thrust structures, along with a decrease in fold-thrust shortening and change in thrust slip directions toward the salient ends (e.g., Royse et al., 1975; Dixon, 1982; Woodward, 1986; Coogan, 1992; Royse, 1993). Accommodating such along-strike changes in displacement requires additional deformation components, including vertical-axis rotation and tangential (strike-parallel) extension, which, along with early LPS, produced systematic 3-D strain patterns (Yonkee and Weil, 2009a).

## INTEGRATED AMS AND STRUCTURAL STUDIES

Our studies have focused on red beds of the Triassic Ankareh Formation, which display consistent patterns of tectonic structures, contain reduction spots that record finite strain, and carry interpretable paleomagnetic components. Integrated structural, paleomagnetic, and AMS data were collected for 154 sites during the summers of 2003 through 2007 (Fig. 2). Site locations were chosen to optimize distribution along strike and within individual thrust systems of the Wyoming salient, with several detailed sampling arrays designed to check for local variations associated with folds and oblique ramps. Additional data were collected for eight subsites to evaluate effects of lithology on AMS.

Oriented samples were collected in the field using both a portable gas-powered drill and a magnetic compass for collecting cores, or a hammer and chisel for collecting hand samples that were later drilled in the laboratory. Typically 6–12 standard 2.54-cm-diameter cores were drilled from multiple beds at each site. Paleomagnetic analysis was performed on cores to determine vertical-axis rotations, with results given in Weil et al. (2009). Detailed structural data (including orientations of bedding, cleavage, fractures, and veins) were measured at each site, and 3-D strain was estimated for sites containing reduction spots, with results given in Yonkee and Weil (2009a). Thus, AMS data can be directly related to strain and vertical-axis rotation patterns on a regional scale around curved thrust systems of the Wyoming salient.

### Lithology and Structural Style of the Ankareh Formation

Field descriptions of outcrop characteristics, petrographic analysis of thin sections, and X-ray diffraction studies were used to determine the lithology and nature of primary and tectonic structures that influence AMS. The Ankareh Formation consists mostly of hematite-stained, quartzose to arkosic, variably calcareous mudstone to sandstone (Fig. 3A) (Kummel, 1954; High and Picard, 1969; Brandley and Rigby, 1988). The red beds contain abundant quartz (typically 50–90 vol%), feldspar (5–20 vol%), carbonate (10–40 vol%), phyllosilicates (1–10 vol%), and minor hematite (~1–2 vol%). Calcareous mudstone grades into thin layers of silty limestone. Detrital quartz grains vary from silt to sand size and display thin quartz overgrowths and minor suturing related to sedimentary compaction (Fig. 3B). Detrital feldspar grains display minor alteration to sericite. Carbonate (mostly calcite) occurs as widespread cement and as detrital grains and rip-up clasts in some samples. Phyllosilicates include larger detrital mica (biotite, muscovite, and chlorite) and clays (illite and kaolinite) that partly grew during diagenesis. Detrital micas vary from being dispersed with weak preferred orientations to concentrated with strong preferred orientations along thin beds. Hematite occurs mostly as small (~1 μm) grains in pore fillings and along rims of detrital quartz grains that are further enclosed by quartz overgrowths (Fig. 3C). Smaller grains appear to have crystallized during early diagenesis. Larger (~10–50 μm) detrital hematite and ilmenite grains are rare and partly altered to fine-grained hematite along their margins. Some red beds also contain trace amounts of fine-grained magnetite that formed during synthrusting (Cretaceous) alteration and chemical remagnetization (Weil et al., 2009).

Primary sedimentary structures in red beds include bedding defined by changes in grain size and composition, mudcracks, rip-up clasts, root traces, and paleosols with carbonate nodules. Bedding varies from well defined to disrupted in areas with loading, dehydration, and paleosol structures. Red beds were deposited mostly in supratidal, meandering stream, and floodplain environments (High and Picard, 1969; Brandley and Rigby, 1988).

The Ankareh Formation displays regionally consistent tectonic structures, including cleavage, fractures, veins, minor folds, and minor faults (see Yonkee and Weil, 2009a, for details). Cleavage is defined by anastomosing partings that are typically at high angles to bedding (Fig. 3A). Cleavage is perpendicular to the shortening direction defined by shapes of reduction spots and is strongly fanned about large-scale folds, indicating cleavage formed during early LPS prior to large-scale folding in each thrust system. Cleavage is better developed in calcareous mudstone than in coarser-grained quartzose sandstone, and within more western (interior) thrust systems. Fractures are widely developed in sandstone, with most sites displaying two dominant, regionally consistent sets: a high-angle set perpendicular to bedding and subparallel to local structural trend, and a cross-strike set subperpendicular to structural trend. The high-angle set is interpreted to partly reflect fracturing along preexisting weak LPS fabrics during unloading, although some fractures may have formed during large-scale folding and thrusting. Cross-strike fractures and associated veins record a protracted history of minor extension subparallel to structural trend. Additional fracture sets oblique to structural trend are locally developed, especially near oblique ramps and transfer zones. Tear faults formed over a protracted history and accommodated minor wrench shear. Contraction faults at low angles to bedding and minor folds accommodated early LPS in well-bedded intervals.

Microscopically, cleavage is defined by discontinuous, anastomosing seams that are enriched in hematite and clay, and formed mostly by dissolution of calcite and quartz (Fig. 3D). Clay and hematite are partly aligned within seams, reflecting rotation during collapse as material was removed by dissolution, along with neocrystallization of some clays. Material dissolved along seams was partly precipitated in associated thin veins, though samples appear to have undergone net volume loss. Detrital mica grains between seams are locally kinked from minor shortening (Fig. 3E). Crystal plastic deformation is limited, with deformation twins developed in larger calcite grains and undulatory extinction in some quartz grains.

In summary, red beds of the Ankareh Formation display a range of primary sedimentary and secondary tectonic structures (Fig. 4). Bedding varies from well developed to disrupted, with detrital mica weakly to strongly aligned along bedding. Minor hematite occurs mostly as small grains that grew during early diagenesis. Clay and hematite are locally concentrated and partly aligned within anastomosing cleavage seams. Cleavage (best developed in calcareous mudstone), high-angle fracture sets (best developed in quartzose sandstone), contraction faults, and minor folds record widespread LPS early in the deformation history. Cross-strike fractures and veins reflect minor extension subparallel to structural trend over a protracted history. Tear faults accommodated local wrench shear and block rotation.

### Rock Magnetic Experiments and Sources of Magnetic Susceptibility

Rock magnetic experiments were performed on selected samples to determine the nature of ferromagnetic (*sensu lato*) and paramagnetic contributions to total AMS, including temperature dependence of susceptibility, acquisition and alternating-field (AF) demagnetization of isothermal remanent magnetization (IRM), thermal demagnetization of three-component IRM, and measurement of hysteresis properties. Magnetic experiments were carried out in the Paleomagnetic Laboratory at Bryn Mawr College, with hysteresis properties measured at the University of Minnesota Institute for Rock Magnetism.

Low-temperature susceptibility measurements were performed on representative samples to determine the importance of a paramagnetic component to AMS (Richter and van der Pluijm, 1994; Parés and van der Pluijm, 2002) (Fig. 5). Paramagnetic susceptibility is strongly temperature dependent according to the Curie-Weiss law, whereas ferromagnetic susceptibilities are nearly temperature independent except in association with magnetic isotropic points, magnetic ordering temperatures, structural phase transitions, and superparamagnetism (Senanayake and McElhinny, 1982; Radhakrishnamurty and Likhite, 1993; Moskowitz et al., 1998; Richter and van der Pluijm, 1994; Martin-Hernandez and Ferre, 2007). For most analyzed samples, low-field susceptibility decreases between 120 and 290 K (Fig. 5A, B), recording a significant paramagnetic contribution to bulk susceptibility. However, the average ratio of susceptibility measured at 77 K and 290 K is 2.8, which is less than the expected ratio of 3.8 for a pure paramagnetic component (Richter and van der Pluijm, 1994), indicating the presence of a ferromagnetic component. This is supported by the observed Morin (hematite) and Verwey (magnetite) structural phase transitions in some samples (Fig. 5B, C), as well as a decrease in susceptibility between 70 and 100 K that likely represents Curie-Weiss behavior of an antiferromagnetic mineral above its ordering temperature (Moskowitz et al., 1998).

IRM experiments for all analyzed samples show magnetization acquisition up to 2.0 T with concave upward curves, indicating that high-coercivity hematite is the main ferromagnetic component in Ankareh red beds (for detailed results of Ankareh IRM experiments, see Weil et al., 2009). A few samples have a higher proportion of magnetization acquisition below 200 mT, consistent with a minor contribution from magnetite (Dunlop, 1986). AF demagnetization of IRM results in <20% remanence loss for most samples, with up to 40% remanence loss for samples that have a minor contribution from magnetite.

Thermal demagnetization of three-axis IRM (Lowrie, 1990) shows decay in intensity of all three components up to 680 °C, confirming that hematite is the main ferromagnetic component. The majority of samples carry most of their remanence along the hard axis (imparted with a 2.0 T field). The intermediate and soft axes (imparted at 0.5 and 0.15 T) show similar decay patterns as those of the hard axis. A few samples display minor contributions from a lower coercivity mineral, interpreted to be magnetite on the basis of a decrease in remanence near 560 °C (Weil et al., 2009).

Hysteresis loops for Ankareh samples show consistent positive slopes at higher applied fields, indicating a significant paramagnetic component in addition to a ferromagnetic component (Fig. 6). Samples have magnetization and coercivity ratios (*M _{r}*/

*M*and

_{s}*H*/

_{cr}*H*) that lie mainly in the single-domain and pseudo-single-domain mixing fields of a Day diagram (Day et al., 1977). None of the measured samples were saturated in the available fields owing to the presence of hematite, and therefore highfield AMS measurements could not be used to quantitatively isolate paramagnetic components.

_{c}In summary, rock magnetic experiments, combined with petrographic and X-ray-diffraction studies, indicate that AMS has a significant paramagnetic component carried by phyllosilicates and a ferromagnetic (*sensu lato*) component carried mostly by hematite, plus minor magnetite in some samples. Phyllosilicates and hematite have oblate magnetocrystalline anisotropy with *K*_{min} subparallel to the *c*-axis and *K*_{max} and *K*_{int} subparallel to the basal plane (Turner et al., 1985; Borradaile, 1988; Rochette et al., 1992; Borradaile and Henry, 1997). Thus, crystallographic and related shape preferred orientations of phyllosilicates and hematite produce magnetic anisotropy. Ankareh samples have variable bedding fabrics partly defined by preferred orientation of detrital mica, and variable LPS fabrics partly defined by rotation of hematite and clays in cleavage seams and kinking of mica, resulting in a range of AMS ellipsoid shapes (Fig. 4). Diamagnetic minerals, including quartz, feldspar, and calcite, do not significantly contribute to AMS in Ankareh red beds.

### AMS Analysis

AMS was measured for individual cores with an AGICO Kappabridge KLY-3 susceptibility bridge, which operates at a frequency of 875 Hz and has a typical sensitivity of ~2.0 × 10^{−8} SI. Site mean AMS eigenvalues and eigenvectors were estimated from multiple cores using tensor methods, with uncertainties evaluated from distributions of bootstrapped values, following the methods of Constable and Tauxe (1990) and Tauxe (1998). Bootstrapped site eigenvalues typically have quasi-normal distributions, with standard deviations of ~0.001 for normalized (for eigenvalues normalized to *K*_{max} + *K*_{int}+ *K*_{min} = 1.0) *K*_{max} and *K*_{int} and ~0.002 for *K*_{min}. These variations give 1σ uncertainties of about ±0.003–0.006 for lineation (*L* = *K*_{max}/*K*_{int}) and ±0.005–0.010 for foliation (*F* = *K*_{int}/*K*_{min}). Trends of bootstrapped *K*_{max} eigenvectors typically have quasi-normal distributions, with 1σ uncertainties of about ±3–10° for triaxial to moderately oblate ellipsoids. AMS parameters defined by Jelinek (1981) and Tarling and Hrouda (1993), including corrected anisotropy degree, *P*′, and shape parameter, *T*, which varies from −1.0 to 1.0 for prolate to oblate ellipsoids, were calculated for each site. Site AMS ellipsoid data and parameters are listed in Table 1^{102}.

Samples from the Ankareh Formation have low mean susceptibility (*K*_{m} = [*K*_{max}+ *K*_{int}+ *K*_{min}]/3), with most core and site *K*_{m} values between 40 × 10^{−6} SI and 140 × 10^{−6} SI, and a median value of 70 × 10^{−6} SI (Fig. 7A). These low values of *K*_{m} are consistent with susceptibility contributions from paramagnetic minerals (mainly phyllosilicates) and small amounts of ferromagnetic minerals (mainly hematite). Some sites include a minor contribution from trace amounts of magnetite that formed during Cretaceous alteration and remagnetization (Weil et al., 2009), but these sites have only slightly higher bulk susceptibilities (median of 100 × 10^{−6} SI) and do not display obvious differences in AMS ellipsoid shape. The corrected degree of anisotropy, *P*′, is low (<1.1 for most sites) and uncorrelated with *K*_{m} (Fig. 7B), consistent with anisotropy being more closely related to primary and tectonic fabrics rather than mineralogy.

The AMS ellipsoids have foliation values (*F* = *K*_{int}/*K*_{min}) mostly <1.10, and lineation values (*L* = *K*_{max}/*K*_{int}) mostly <1.04, and they display a range of shapes from distinctly oblate to triaxial to prolate (Fig. 8). These shapes correspond with idealized fabric stages A to C in Figure 1, in which a sedimentary fabric becomes overprinted by a weak to moderate LPS fabric. Sites with distinctly oblate AMS ellipsoids related to a dominant sedimentary fabric have *K*_{min} perpendicular to bedding, display large variations in individual core and bootstrap *K*_{max} directions, and show significant overlap in *K*_{max} and *K*_{int} bootstrap values (Fig. 9A, B). Sites with moderately oblate ellipsoids related to a weak LPS fabric display roughly clustered core and bootstrap *K*_{max} directions that define a weak magnetic lineation but still show minor overlap in *K*_{max} and *K*_{int} bootstrap values (Fig. 9C, D). Sites with triaxial ellipsoids have well-clustered core and bootstrap *K*_{max} directions that define a strong magnetic lineation and show no significant overlap between *K*_{max}, *K*_{int}, and *K*_{min} values (Fig. 9E). Sites with prolate ellipsoids also have well-clustered *K*_{max} directions, but *K*_{int} and *K*_{min} bootstrap values overlap, and *K*_{min} directions become variable as the LPS fabric becomes stronger (Fig. 9F). Although the site AMS ellipsoid provides a measure of average magnetic fabric, individual cores display dispersion related to variations in sedimentary fabric and mineralogy. For example, AN121 has a moderately oblate site mean ellipsoid, but individual cores range from distinctly oblate to triaxial (Fig. 9C).

The AMS ellipsoid shapes reflect composite fabrics of varying intensity, with foliation mostly controlled by primary bedding fabric, and lineation mostly controlled by a secondary LPS fabric. Numerical and experimental simulations of composite AMS fabrics indicate that a magnetic lineation develops as a primary foliation is progressively overprinted by a secondary foliation (Housen et al., 1993). Thus, we use lineation intensity, *L*′ = ln(*K*_{max}/*K*_{int}) as a proxy for deformation intensity, and divide Ankareh sites into three gradational fabric types: type 1, *L*′ <0.003, or a *K*_{max} eigenvector half-angle uncertainty >20°; type 2, 0.003 ≤*L*′ ≤0.010; and type 3, *L*′ > 0.010. Type 1 fabrics lack a definable magnetic lineation and correspond mostly to distinctly oblate AMS ellipsoids, reflecting a dominant sedimentary fabric (Fig. 9A, B). Type 2 fabrics have a weak magnetic lineation roughly parallel (generally <20°) to structural trend and correspond mostly with moderately oblate ellipsoids, reflecting intersection of a weak LPS fabric with bedding (Fig. 9C, D). Type 3 fabrics have a distinct lineation subparallel (generally <15°) to structural trend and correspond mostly with triaxial and prolate ellipsoids, reflecting the intersection of a moderate LPS fabric with bedding (Fig. 9E, F). Of the 140 sites with interpretable AMS data, 34 have a dominant sedimentary fabric (type 1), 52 have a weak magnetic lineation (type 2), 46 have a strong lineation (type 3), and the remaining 8 sites have oblique patterns.

#### Spatial Patterns of AMS Fabrics within the Wyoming Salient

Structural relations indicate that widespread LPS occurred early in the deformation history of each thrust system, prior to large-scale folding (Yonkee and Weil, 2009a). Thus, to compare AMS fabrics with other structural data from the Wyoming salient, AMS site eigenvectors were restored to prefolding orientations using successive rotations based on structural relations for each site (Table 2^{202}). Most sites were in limbs of subhorizontal folds or dip panels above frontal thrust ramps, and a single rotation about bed strike was used for restoration. For sites along plunging folds, plunge was first removed, followed by rotation about the partly restored bedding strike. For sites over oblique ramps, a single rotation about bed strike was used, although rotation paths may have been more complex. Some sites near the northern and southern ends of the salient had additional rotation components associated with tilting along margins of foreland uplifts and flexure in footwalls of Neogene normal faults.

In situ *K*_{min} site axes display large dispersion, whereas restoration of bedding to horizontal results in strongly clustered, subvertical *K*_{min} axes, confirming that magnetic foliation is overall parallel to bedding around folds, and formed during sedimentary deposition and diagenesis (Fig. 10). Restored *K*_{max} axes for sites with a definable magnetic lineation range from mostly NW-SE in the northern part of the salient, to N-S in the central part, to NE-SW in the southern part. Because *K*_{max} axes are typically at low angles to local fold axes, restoration of bedding to horizontal results in relatively minor changes in orientation.

Restored *K*_{max} site mean directions display systematic variations with respect to regional structural trend and define a tangential pattern around the salient (Fig. 11A). *K*_{max} directions are interpreted to mostly reflect intersection of bedding with LPS fabrics that formed prior to large-scale folding and thrusting, although magnetic lineations may also partly reflect fracturing and minor tangential extension during thrusting. Corresponding shortening directions (orthogonal to *K*_{max} and subparallel to *K*_{int}) define a radial pattern. A similar radial pattern of shortening directions was interpreted from orientations of cleavage and high-angle fracture sets by Yonkee and Weil (2009a). In detail, restored *K*_{max} directions display minor (±15°) scatter for sites in similar structural domains, related to measurement uncertainties and local structural noise. Overturned fold limbs and transfer zones have more complex deformational histories and display outlier *K*_{max} directions.

Intensity of magnetic lineation, *L*′ = ln(*K*_{max}/*K*_{int}) also displays spatial variations across the salient (Fig. 11B). Most sites with type 1 fabrics (*L*′ < 0.003, dominant sedimentary fabric) are found in central parts of the frontal Hogsback and Absaroka systems. In contrast, most sites with type 3 fabrics (*L*′ > 0.010, moderate tectonic fabric) are found in the more interior Crawford system and toward the salient ends. Cleavage intensity in widely exposed micritic limestone of the Jurassic Twin Creek Formation displays similar patterns, with cleavage weak to absent in central parts of the Hogsback and Absaroka systems, and moderately to strongly developed in much of the Crawford system and toward the salient ends.

AMS, however, also depends on sampled lithology, as illustrated by nearby locations having different fabric types. Differences in AMS fabrics between adjacent exposures of red sandstone and carbonate-rich mudstone to fine-grained sandstone were evaluated for four sites and associated subsites (AN 35B, 112B, 141B, and 145B). Magnetic lineation is consistently weaker for carbonate-rich lithologies (mean *L*′ = 0.003 compared with 0.008 for red sandstone). Bed-parallel magnetic foliation is also consistently weaker for carbonate-rich lithologies. This pattern is also seen regionally for sites in similar structural domains. For example, sites in the Crawford system that contain either carbonate-rich lithologies (AN 74, 131) or quartz-rich sandstone with few phyllosilicates (AN 76, 118) have relatively weak lineations. Carbonate-rich lithologies tend to have fewer detrital micas with more random initial distributions (related to paleosol development and disrupted bedding), which results in less kinking of mica during LPS, and thus lower intensity of magnetic lineation.

### Correlation of AMS Lineation and StructuralTrend

Correlations between changes in *K*_{max} directions and curvature in regional structural trend (relative to a reference trend of 360° for the Wyoming salient) are evaluated using a weighted-least-squares method that incorporates uncertainties in AMS directions (see Yonkee and Weil, 2009b, for detailed description of method). Structural trend is estimated from geologic map data (including fold traces, formation contacts, and bedding strikes corrected for fold plunge) and has small uncertainties (<5°). Because sites from overturned fold limbs and transfer zones display outlier AMS directions related to local complications, analysis is done on a filtered data set of sites in relatively simple structural settings to evaluate systematic regional correlations.

The Ankareh Formation has 85 sites with definable *K*_{max} lineations in relatively simple structural settings. Analysis of these sites yields a best-fit slope of 0.94, with a 95% confidence interval of ±0.09 (Fig. 12), confirming that *K*_{max} directions are on average subparallel to structural trend around the salient. Observed scatter in residuals (standard deviation σ_{r} = 15°) reflects a combination of measurement uncertainty in *K*_{max} directions (σ_{m} ~3–10°), structural noise (σ_{n} ~8°), and additional AMS fabric variability (σ_{f} ~10°, for σ_{r}^{2} = σ_{m}^{2} + σ_{n}^{2} + σ_{f}^{2}). Measurement uncertainty is estimated from the dispersion of bootstrapped AMS directions. Structural noise reflects local strain refraction and small-scale block rotations, and is estimated from spatial variations in cleavage orientations (Yonkee and Weil, 2009b). Fabric variability reflects the composite nature of AMS related to varying primary sedimentary fabrics and multiple deformation increments, and is estimated from correlating AMS and finite strain directions, as described in the next section.

### Comparison of AMS to Finite Strain

Finite strain in the Ankareh Formation was estimated from reduction spots in calcareous mudstone to fine-grained sandstone, with details given in Yonkee and Weil (2009a). In the least deformed samples, reduction spots define distinctly oblate ellipsoids, with circular sections along bedding, and slightly elliptical sections on surfaces perpendicular to bedding. These relations record ~10–15% sedimentary compaction, with *Z* initially perpendicular to bedding and *X* ≈ *Y* ≈1.0 parallel to bedding (where *X* ≥ *Y* ≥ *Z* are principal axes of the strain ellipsoid). This primary fabric was modified by progressive deformation, including widespread LPS and minor tangential extension (TE) parallel to structural trend. Section ellipses along bedding record the ratio of LPS to TE, with axial ratios of *R* ≈ (1 + TE)/(1 − LPS). Section ellipses on surfaces perpendicular to structural trend display more complex patterns, first becoming more circular as LPS approaches 10–15%, and then becoming more elliptical perpendicular to bedding as LPS increases and the tectonic fabric grows stronger than the primary fabric.

The combination of a primary oblate fabric followed by LPS produced systematic changes in reduction spot axial ratios and principal directions (Fig. 13A). Reduction spot fabrics are divided into three gradational categories: (1) very low strain samples having a dominant sedimentary fabric with distinctly oblate ellipsoids (*R _{XY}* ~1.0 and

*R*~1.1–1.2) and

_{YZ}*Z*perpendicular to bedding; (2) low strain samples having composite sedimentary and tectonic fabrics with triaxial to prolate ellipsoids (

*R*~1.1–1.25 and

_{XY}*R*

_{YZ}decreasing to ~1.0) and

*X*subparallel to structural trend; and (3) medium to high strain samples having stronger tectonic fabrics with

*R*increasing as LPS increases,

_{YZ}*X*subparallel to structural trend, and

*Z*switched parallel to bedding. A decrease in

*R*for some higher strain sites records minor extension down the dip of cleavage. This pattern is partly similar to AMS ellipsoids (compare Fig. 13A and B), but AMS fabrics appear to display a smaller tectonic component. Very low strain samples have type 1 AMS fabrics, low strain samples mostly have type 2 AMS fabrics, and medium to high strain samples mostly have type 3 AMS fabrics, although carbonate-rich samples have weaker magnetic lineations.

_{XY}Bed-parallel strain ellipses, which record the ratio of LPS to tangential extension, have long axes that define a tangential pattern around the salient, subparallel to *K*_{max} directions (Fig. 11C). Strain ratios increase overall westward and toward the salient ends, from <1.1 in the central frontal Hogsback and Absaroka systems to 1.3–1.6 in the more interior Crawford system and toward the salient ends. Spatial variations in AMS lineation intensity and strain intensity are broadly similar, although AMS fabric types display additional lithologic control. Strain ratios, along with textural relations, record <5% LPS in the central Hogsback to 15–30% LPS in the Crawford system, ~5% TE in all thrust systems, and minor extension down the dip of cleavage at higher strain sites. Note, because of the nature of 3-D strain (LPS with volume loss and minor strike-parallel extension), *K*_{max} directions are subparallel to both the stretching (*X*) direction of strain and to the intersection of LPS fabrics with bedding.

Relations between AMS ellipsoids and corresponding strain ellipsoids are quantified using a weighted least-squares approach that incorporates uncertainties in both AMS and strain values. Correlation of AMS *K*_{max} and finite strain *X* directions for sites where data are available yields a best-fit slope, m, of 1.07 with a 95% confidence interval of ±0.28 (Fig. 14A), indicating that on average *K*_{max} and *X* directions are subparallel to each other. Residuals have a standard deviation of σ_{r} = 12°, which is greater than expected for measurement uncertainty in *K*_{max} directions (σ_{mK} ~3–10°) and *X* directions (*m**σ_{mX} ~3–5°). Residuals thus indicate additional AMS fabric variability, with σ_{f} ~10° (determined from σ_{r}^{2} = σ_{mK}^{2} + (*m**σ_{mX})^{2} + σ_{f}^{2}). This variability is likely related to the composite nature of AMS that reflects varying primary sedimentary fabrics and multiple deformation increments, producing additional angular dispersion about finite strain axes.

Correlation between *K*_{max} directions and orientations of cleavage and high-angle fractures related to LPS yields a similar result, with a slope of 0.98 and a 95% confidence interval of ±0.12 (Fig. 14B); the smaller confidence interval reflects a larger number of sites from which mesoscopic data are available. Thus magnetic lineation is on average subparallel to the intersection of cleavage and high-angle fracture sets with bedding. Residuals between *K*_{max} directions and cleavage (σ_{r} = 11°) are similar to that for strain *X* directions, but residuals are higher for fractures (σ_{r} = 16°). This is interpreted to reflect additional fracture fabric variability.

Comparison of AMS and strain magnitudes is complicated by the composite nature of fabrics, varying lithology, and multiple deformation mechanisms (e.g., Owens, 1974; Borradaile and Tarling, 1984; Borradaile, 1988). Strain intensity is evaluated using the axial ratio of the bed-parallel strain ellipse, *R*′ = ln(*R*) ≈ ln[(1 + TE)/(1 − LPS)], and AMS intensity is evaluated using magnetic lineation, *L*′ = ln(*K*_{max}/*K*_{int}). Correlation of *L*′ and *R*′ yields a best-fit slope, m, of 0.056 with a 95% confidence interval of ± 0.018 (Fig. 15A), confirming a crude relation between increasing magnetic lineation and strain intensity. Observed scatter of residuals (σ_{r} = 0.064), however, is greater than expected for measurement uncertainties in AMS lineation (σ_{mL} ~0.003–0.006) and strain ratios (*m**σ_{mR} ~.003), indicating additional AMS fabric variability of σ_{f} ~0.004 (determined from σ_{r}^{2} = σ_{mL}^{2} + (*m**σ_{mR})^{2} + σ_{f}^{2}). This additional variability in AMS lineation strength is interpreted to reflect variations in lithology (mineralogy and strength of primary fabrics) between sites. Comparison of the AMS shape parameters *T* and *P*′ with finite strain intensity yields poor correlations (Fig. 15B). These parameters directly incorporate bedding foliation, which varies with the nature of primary sedimentary fabrics, obscuring tectonic contributions.

## DISCUSSION

The AMS fabrics within red beds of the Ankareh Formation reflect interplay of depositional, diagenetic, and tectonic processes that produced preferred orientations of mineral grains (Fig. 16). Rock magnetic experiments indicate that paramagnetic phyllosilicates (detrital mica and clay) and ferromagnetic hematite (that mostly grew during early diagenesis) are the main contributors to AMS. Biotite, chlorite, and phengitic muscoviteillite have mean susceptibilities of ~300 × 10^{−6} to 3000 × 10^{−6} SI (depending on Fe content), and moderate magnetocrystalline anisotropy (*K*_{max}/*K*_{min} = 1.2–1.3) with *K*_{min} subparallel to the c axis and *K*_{max} ≈ *K*_{int} lying close to the basal plane (Borradaile et al., 1987; Martin-Hernandez and Hirt, 2003). Hematite has a mean susceptibility of ~2500 × 10^{−6} SI, and strong anisotropy (*K*_{max}/*K*_{min} >100) with *K*_{min} subparallel to the c axis and *K*_{max} ≈ *K*_{int} in the basal plane (Tarling and Hrouda, 1993). Phyllosilicates and hematite grains have sheetlike shapes parallel to basal planes, such that shape and magnetocrystalline orientations are closely related (Hrouda and Schulmann, 1990; Hrouda and Kahan, 1991). Low mean susceptibility (*K*_{m} ~40–140 × 10^{−6} SI) for most red bed samples is consistent with observed small amounts of phyllosilicates (1–10%) and hematite (~1–2%). Some samples contained trace amounts of magnetite, but these samples had only slightly higher mean susceptibilities. Magnetic foliation values (*K*_{max}/*K*_{int} <1.10) are consistent with observed weak to moderate preferred orientation of detrital mica grains along bedding, and very weak to near random orientations of small hematite and clay grains along quartz grain rims (Fig. 16A). Magnetic lineation varies with deformation intensity and is interpreted to reflect a zone-axis of phyllosilicate and hematite grains parallel to the intersection of LPS fabrics with bedding (Rochette and Vialon, 1984; Rochette, 1987). LPS fabrics progressively developed by rotation and neocrystallization of hematite and clay along cleavage seams and kinking of mica in the rock matrix, resulting in increased lineation intensity (Fig. 16B). Although orientations of individual grain rotation axes vary depending on grain orientation, the average rotation axis is parallel to the intersection of LPS and bedding fabrics.

The AMS ellipsoids in Ankareh red beds define three gradational fabric types that range from a dominant sedimentary fabric lacking a magnetic lineation (type 1), to a weak LPS–sedimentary fabric with a weak magnetic lineation (type 2), to a composite fabric with a distinct lineation (type 3) (Fig. 9). Directions of magnetic lineation define a regional tangential pattern around the Wyoming salient, with local complications in overturned folds and transfer zones (Fig. 11A). Type 1 fabrics are found mostly in the central, frontal part of the salient, where cleavage is absent and LPS is <5%, whereas type 3 fabrics occur mostly in more interior thrust systems and toward the salient ends, where cleavage is better developed and LPS is >15% (Fig. 11B). Although no single measure describes AMS fabrics, ellipsoid shape and lineation intensity provide crude measures of deformation intensity for red beds in the Wyoming salient.

AMS patterns from the Wyoming salient are broadly similar to patterns in weakly to moderately deformed sedimentary rocks from other fold-thrust belts, including the Zagros-Makran belt in Iran (Aubourg et al., 2004), Arc of Fars in Iran (Bakhtari et al., 1998), High Atlas of Morocco (Saint-Bezar et al., 2002), Siwaliks of Nepal (Gautam and Rösler, 1999), Taiwan (Kissel et al., 1986), Pyrenees of Spain (Parés et al., 1999), and Appalachians (Parés and van der Pluijm, 2003). These patterns reflect a primary fabric overprinted by a tectonic fabric of varying intensity. However, details of AMS patterns and relations to regional structures vary within and between different orogenic belts. The question then is how can AMS fabrics best be utilized to evaluate strain patterns and evolution of orogenic curvature?

### Use of AMS to Understand Strain Patterns

Many comparisons have been made between AMS fabrics and strain markers (e.g., Owens, 1974; Kneen, 1976; Hrouda, 1970; Rathore, 1979; Kligfield et al., 1981; Rathore and Henry, 1982; Siddans et al., 1984; Hrouda, 1987; Cogne, 1987, 1988; Borradaile, 1988; Hirt et al., 1988). Although relationships between AMS and strain have been documented in some areas with consistent lithology (Aubourg et al., 1991; Lee et al., 1990; Averbuch et al., 1992), no universal correlation exists, as AMS fabrics result from multiple minerals that have different magnetocrystalline properties and deform by different mechanisms (Hrouda, 1982, 1987; Rochette, 1987; Borradaile and Henry, 1997; Luneburg et al., 1999).

In the Wyoming salient, AMS *K*_{max} directions are on average subparallel to both the intersection of LPS fabrics (cleavage and high-angle fracture sets) with bedding and to the finite strain X direction. This pattern reflects the nature of 3-D strain, which includes LPS with volume loss and minor strike-parallel extension related to development of curvature in the Wyoming salient. Quantitative correlation of AMS and strain indicates that *K*_{max} directions can be used to estimate orientations of incipient LPS fabrics in red beds, even in areas lacking strain markers and obvious cleavage. However, care is needed to evaluate local structural noise, oblique fabrics (especially in complex areas such as oblique thrust ramps), and AMS “fabric variability” related to multiple deformation increments and varying primary sedimentary fabrics.

Relations of AMS and strain magnitude are more complex. Magnetic lineation intensity, *L*′ = ln(*K*_{max}/*K*_{int}), along with cleavage intensity and strain ratios, increases overall toward more interior thrust systems and the ends of the Wyoming salient. In detail, however, nearby sites display variations in lineation intensity related to differences in lithology. Sandstones with more abundant phyllosilicates and stronger initial bedding fabrics tend to have higher L′ compared with carbonate-richer lithologies, and thus are better for estimating general strain intensity, especially in weakly deformed areas. Correlation of AMS *L*′ and bed-parallel strain ratios yields a slope of 0.056 ± 0.018 (Fig. 15A), excluding carbonate-rich samples. The correlation established here may be applicable to other weakly deformed red beds, but additional data from other areas are needed to test this correlation. Note that residuals between observed and expected AMS *L*′ are greater than can be explained by measurement uncertainty alone, reflecting additional lithologic control from varying amounts and subfabrics of detrital mica, clays, hematite, and minor magnetite. Understanding relations between AMS and strain magnitude can be improved by quantifying paramagnetic, hematite, and magnetite contributions to AMS and analyzing corresponding mineral grain fabrics and microtextures (Martin-Hernandez and Hirt, 2004; Martin-Hernandez and Ferre, 2007).

### Use of AMS to Understand Orogenic Curvature

Commonly cited end-member kinematic models for evolution of curved mountain belts include: primary arcs that start with initial curvature and have no secondary rotation, and may form by uniform thrust slip or radial slip; progressive arcs that develop increasing secondary curvature during contraction, and may form by divergent thrust emplacement with curved slip or differential shortening with parallel fault slip; and secondary oroclines that are rotated into curved belts during a subsequent phase of wrenching or bending (Weil and Sussman, 2004). Each of these models predicts distinct patterns of LPS fabrics and vertical-axis rotations along and across orogenic strike that can be compared with AMS and paleomagnetic data. AMS analysis indicates that LPS fabrics in the Wyoming salient formed subparallel to structural trend (Fig. 12), and paleomagnetic analysis indicates that ~75% of present-day curvature in structural trend was acquired by secondary rotation during large-scale thrusting (Weil et al., 2009). Consequently, kinematic models that require primary curvature with no secondary rotation can be eliminated, including uniform and radial slip of a primary arc (models A and B, Fig. 17). Models that require 100% secondary rotation can also be eliminated, including oroclines with superimposed wrenching and bending (models E and F, Fig. 17). Kinematic models consistent with observed rotation and AMS patterns include progressive arcs that form by a combination of divergent thrust emplacement and differential shortening (models C and D, Fig. 17). In the curved slip model, LPS fabrics develop subperpendicular to the initial slip direction (and about but not precisely parallel to trend) and are then rotated by an amount equal to the change in thrust slip direction. In the differential shortening model, LPS directions undergo varying rotation related to a combination of block rotation, wrench shear, and slip along tear faults.

When analyzed together, AMS, strain, mesoscopic structural, and paleomagnetic data indicate that the Wyoming salient is a progressive arc that started with 25% primary curvature and subsequently acquired ~75% secondary curvature. Slight primary curvature was likely related to geometry and lithology of the initial sedimentary prism (Coogan and Yonkee, 1985; Royse, 1993; DeCelles, 1994). Subsequent vertical-axis rotations were mostly related to curved slip paths during divergent thrust emplacement, with rotation concentrated along the leading edge of the growing wedge (Weil et al., 2009). Smaller amounts of rotation occurred during “passive” thrust sheet transport in the wedge interior as active deformation progressively shifted toward frontal thrust systems. Late-stage localized rotation and tilting were superimposed on the frontal thrust system near the bounding foreland uplifts (Weil et al., 2009).

## CONCLUSIONS

Analysis of AMS in red beds from the Wyoming salient reveals patterns of early LPS and constrains 3-D kinematic models of orogenic wedge development. The magnetic susceptibility of samples originates from paramagnetic phyllosilicates and ferromagnetic hematite. AMS ellipsoid shapes range from distinctly oblate parallel to bedding to prolate with *K*_{max} parallel to local trend. Regionally, AMS *K*_{max} directions define a tangential pattern sub parallel to arcuate fold-thrust trends around the salient (Fig. 11A) and correspond to both intersection of LPS fabrics (cleavage and high-angle fractures) with bedding and to stretching (*X*) directions of finite strain ellipsoids (Fig. 11C). General parallelism of AMS foliation with bedding planes and parallelism of magnetic lineation with structural trend reflect composite fabrics between primary sedimentary and tectonic LPS fabrics that result from rotation, kinking, and neocrystallization of phyllosilicates and hematite (Fig. 16). Correlation of *L*′ with strain intensity measured from reduction spots suggests that AMS can be used to crudely approximate strain intensity in red beds (Fig. 15A), but lithologic effects are also important. Correlation of *K*_{max} directions with LPS directions indicates that the AMS in weakly deformed rocks provides a useful proxy of early LPS directions.

Regional patterns of magnetic lineations in the Wyoming salient provide a complementary data set to existing paleomagnetic (Weil et al., 2009), mesoscopic structure, and strain data (Yonkee and Weil, 2009a), which can be integrated to constrain viable kinematic models. Together these data indicate that the Wyoming salient is a progressive arc that underwent significant vertical-axis rotation within individual thrust sheets subsequent to early LPS. This kinematic model of curvature is best explained by a combination of primary thickness variations in the sedimentary wedge, variations in wedge interior and fault zone strengths, differential shortening to maintain wedge taper, and interaction with Laramide foreland uplifts.

This work would not have been possible without the financial support of U.S. National Science Foundation grants EAR-0409103 and EAR-0408653. Many students from Bryn Mawr College, Haverford College, and Weber State University participated in field and laboratory work, which greatly enhanced the quality of this study. These students include Tyler Cluff, Andrea Cutruzzula, Steve Fellows, Melissa Lindholm, Anna Mazzariello, Sarah McCullough, Evan Pugh, Zoe Ruge, Cameron Thompson, Matt Tomich, Kira Tushman, and Virginia Walker. Comments by the editors of *Lithosphere* (especially James Evans) and reviews by Mike Petronis and an anonymous reviewer are greatly appreciated. Some of this research was made possible by a Visiting Fellowship at the Institute for Rock Magnetism (IRM) at the University of Minnesota. The IRM is funded by the Instruments and Facilities Program, Earth Sciences Division, National Science Foundation, and the W.M. Keck Foundation.