Structure and internal deformation of thrust sheets in the Sawtooth Range, Montana: insights from anisotropy of magnetic susceptibility

Abstract Geological strain analysis of sedimentary rocks is commonly carried out using clast-based techniques. In the absence of valid strain markers, it can be difficult to identify the presence of an early tectonic fabric development and resulting layer parallel shortening (LPS). In order to identify early LPS, we carried out anisotropy of magnetic susceptibility (AMS) analyses on Mississippian limestones from the Sawtooth Range of Montana. The Sawtooth Range is an arcuate zone of north-trending, closely spaced, west-dipping, imbricate thrust sheets that place Mississippian Madison Group carbonates above Cretaceous shales and sandstones. This structural regime is part of the cordilleran mountain belt of North America, which resulted from accretion of allochthonous terrains to the western edge of the North American continent. Although the region has a general east–west increase in thrust displacement and related brittle deformation, a similar trend in penetrative deformation or the distribution of tectonic fabrics is not observed in the field or in the AMS results. The range of magnetic fabrics identified in each thrust sheet ranges from bedding controlled depositional fabrics to tectonic fabrics at a high angle to bedding.

The initial formation of a penetrative tectonic fabric or cleavage usually develops as a response to coaxial layer parallel shortening (LPS) in fold-and-thrust belts (Mitra et al. 1985;Cooper & Trayner 1986;Mitra 1994;Yonkee & Weil 2010). This process is typically important in accommodating shortening (up to 60%) in fold-and-thrust belts, and develops through a combination of mechanisms, such as pressure solution, grain rotation and grain recrystallization (Ramsay 1967(Ramsay , 1969Engelder & Marshak 1985;Passchier & Trouw 1998). In the absence of valid strain markers, accurately identifying and quantifying LPS can be particularly difficult. In order to counter this, anisotropy of magnetic susceptibility (AMS) analyses have been applied to Mississippian carbonates from the Sawtooth Range of northwestern Montana. AMS data are capable of revealing the susceptibility tensor of all the minerals that contribute to the magnetic fabric and lineation of a sample: the method is, therefore, ideal for determining a rock's petrofabric (Borradaile & Jackson 2004).
The Sawtooth Range is the front range of one of the world's classic fold-and-thrust belts, and is associated with the deformation and development of the North American Cordillera (Fig. 1). The range is composed of numerous allochthonous thrust sheets of Carboniferous-aged limestones and dolomites that were part of the footwall of the regional-scale Lewis, Eldorado, and Hoadley (LEH) Thrust complex (Mudge 1972a, b;Mudge & Earhart 1980;Sears 2001). Despite considerable regional shortening (c. 60%), penetrative strain in the Mississippian carbonates has been largely limited to brittle deformation (Holl & Anastasio 1992), with only limited development of a penetrative tectonic fabric. In order to determine the extent of the development of this penetrative LPS fabric in the Sawtooth Range, AMS data were collected on samples from five thrust sheets, all exposed along the Sun River in the Sawtooth Range (Fig. 2). The Diversion, Sawtooth, French, Norwegian and Beaver thrust sheets are all well exposed by road cuts and natural outcrops along the Sun River (Fig. 3).
west-dipping, imbricate thrust sheets and associated folds composed of Paleozoic and Mesozoic sedimentary rocks (Fig. 3;Holl & Anastasio 1992). These eastward-propagating thrusts typically placed dominantly Mississippian-aged carbonate rocks of the Madison Group above Jurassic and Cretaceous shales and sandstones. Locally Devonian carbonate sequences are also present in the thrust system ( Fig. 3; Mudge et al. 1962;Mudge 1970;DeCelles 2004).
The interbedded limestones and dolomites of the Madison Group are the most prominent lithologies exposed in the Sun River area (Fig. 3). These Carboniferous carbonate rocks rest unconformably on Cambrian and Devonian carbonate rocks, with subsidiary thin siliciclastic units. Precambrian Belt Supergroup strata consist of marine siliciclastic rocks with subordinate carbonate rock units (Fig. 4;Holl & Anastasio 1992).
The Madison Group itself can be divided into the older Allan Mountain Limestone and the younger Castle Reef Dolomite Formations (Mudge 1972a). The Allan Mountain Limestone Formation is characterized by thin beds of dark-grey limestone whereas the Castle Reef Dolomite Formation is mostly thick beds of light-grey dolomite ( Fig. 4; Mudge et al. 1962).
The thrust sheets typically climb from a basal décollement at the top of the Devonian succession that culminates in the Cretaceous, with minor detachments in the Mississippian Allan Mountain Limestone Formation (Mitra 1986). Close spacing of thrust surfaces led to the back-rotation and steepening of individual thrust faults in imbricate arrangements and sigmoidal geometries (Mitra 1986).
The structural regime and deformation in the Sawtooth Range is generally thought to be synchronous with the emplacement of the LEH thrust sheets     The coarse-grained texture, while ideal for strain analysis, is rarely observed. Microstructural deformation observed is mainly grain boundary bulging and type 1 calcite twinning. (Fig. 1;Boyer & Elliott 1982;Boyer 1992;Sears 2001). The crustal-scale LEH thrust package is a large allocthonous sheet composed of siliciclastic Mesoproterozoic to Phanerozoic strata, 70-110 km wide and up to 30 km thick, with an eastward taper (Sears 2001). The total displacement on the thrust sheet varies from 40 to 140 km, with eastward transport initiating at 74 Ma and ceasing by 59 Ma (Sears 2001;Fuentes et al. 2012). These ages are constrained by disruption in the structural and stratigraphic continuity of Campanian-Maastrichtian volcanogenic formations that are capped by 74 Ma tuffs (Sears 2001 and references therein) and undeformed porphyritic dykes with an age of 59 Ma that cross-cut thrusts at the leading edge of the LEH thrust sheet (Sears 2001). These age constraints are conformable with direct dating of authigenic clay formation (68-73 Ma) in fault gouge from the Lewis Thrust in SW Canada (van Der Pluijm et al. 2006).
With the emplacement of the LEH thrust sheet, the strata in the footwall experienced elevated temperature conditions during deformation and . Increases in P j , the degree of anisotropy, implies an increasing strength of the ellipsoid shape. T j represents the shape parameter; positive numbers imply an oblate ellipsoid, whereas negative values imply a prolate ellipsoid; perfectly triaxial ellipsoids are represented by T j values of 0. Representative block diagrams of the progressive development of a tectonic fabric (redrawn from Ramsay & Huber 1983) are matched with the AMS ellipsoid types outline in (b). (b) The evolution of ellipsoid orientations by progressive deformation (LPS) of an originally horizontal bedding fabric (Type 1). As LPS deformation continues the AMS ellipsoid becomes triaxial and starts to resemble Type 2. The first visible stage of deformation is associated with the development of a lineation (Type 3), typically represented by a prolate ellipsoid. As deformation continues this lineation becomes a foliation (Type 4) that is perpendicular to the original bedding plane. Modified from Bakhtari et al. (1998). imbrication. Maximum temperature conditions have been constrained between 100 and 175°C, from illite-bearing mineral assemblages recovered from Cretaceous shale (Hoffman et al. 1976;Gill et al. 2002;O'Brien et al. 2006O'Brien et al. 2007. Both pre-and synfolding chemical remagnetizations have been identified in the Sawtooth Range (O'Brien et al. 2006(O'Brien et al. 2007Nemkin et al. 2016). These remagnetizations are thought to be related to fluid flow rather than a thermal or structural event.
This thermal regime largely concurs with vitrinite reflectance studies that suggest that very localized frictional heating can only be associated with large-scale thrusting (Bustin 1983). These data are further interpreted to indicate that any heating associated with the thrust-related deformation of the Sawtooth Range did not exceed temperatures of 100-175°C. Holl & Anastasio (1992) estimated that the deformation of the strata of the Sawtooth Range accommodated a minimum bulk shortening of 60% based on section balancing. This shortening was primarily enabled by thrusting associated with the forward developing imbricate fan; thrusting, in turn, was facilitated by progressive development of mesoscopic fault arrays that allowed the base of the thrust sheets to deform by cataclastic flow (Holl & Anastasio 1992). Tectonic fabrics, where developed, are consistently at a high angle to bedding, and are limited to stylolitization and spaced cleavage dominated by pressure solution (Fig. 5). This is clearly suggestive of an early (pre-thrusting) localized LPS fabric developed during progressive deformation (Holl & Anastasio 1992).

AMS sampling and methodology
Oriented block samples were collected from the Madison Group Limestone along the Sun River Valley in a transect arranged from east to west and parallel to the direction of thrust transport. Samples orientations were recorded using a compass-clinometer, with strike and dip indicators marked on each sample. Samples were collected from outcrops with well-defined bedding/cleavage relationships. Lithologies with complex sedimentary fabrics, such as synsedimentary deformation, burrowing and cross-bedding were avoided, as these might add further complexities to the relationship between bedding and tectonic fabrics. AMS samples and structural data were obtained from 72 block samples. Block samples were typically 20 × 40 × 20 cm in size. Between 8 and 14 core samples were drilled  (Jelínek 1977), where F should be greater than 4.2 and 3.0 for 95% and 90% confidence intervals respectively. Any specimens that did not satisfy the F-test were removed from interpretations.

AMS analysis
Magnetic susceptibility (k) is the induced magnetization (M ) that is acquired within an externally applied field (H ), k = M/H (Borradaile & Jackson 2004).
The preferred orientation of all magnetic minerals contributes to the observed AMS. Therefore, the total AMS is dependent on the magnetic mineralogy, i.e. the susceptibility and intrinsic anisotropy of minerals and their concentration, as well as their preferred orientation, and in the case of ferromagnetic minerals with a high spontaneous magnetization, their shape and grain size (e.g. Tarling & Hrouda 1993). AMS results are represented by the ellipsoids of magnetic susceptibility, similar to the strain ellipsoid, represented by three mutually orthogonal principal axes K1 ≥ K2 ≥ K3 (Borradaile 1988;Borradaile & Jackson 2010). These axes are the eigenvectors and eigenvalues of the bulk susceptibility tensor or K mean : AMS records the net magnetic contribution of all the minerals in a sample, whether they are diamagnetic, paramagnetic, ferrimagnetic, ferromagnetic (senso stricto) or anti-ferrimagnetic (Tarling & Hrouda 1993). Therefore, AMS is dependent on the magnetic (mineral susceptibility and anisotropy) and physical (shape, size and preferred orientation) properties of these components (Tarling & Hrouda 1993), and can be representative of all fabrics formed at different times and by different mechanisms.
Consequently, AMS represents a composite fabric that can be related to depositional, diagenetic, magmatic and tectonic processes, and as a result, fabric interpretation is not always straightforward (e.g. Borradaile & Jackson 2004). Despite these complications, AMS is typically sensitive to weak tectonic fabrics and their associated slightly preferred orientations of minerals, which contribute to the overall magnetic fabric (Fuller 1963;Borradaile & Tarling 1981;Kligfield et al. 1981;Kissel et al. 1986;Lowrie et al. 1986;Aubourg et al. 1991;Averbuch et al. 1992;Lüneburg et al. 1999;Parés et al. 1999;Borradaile & Jackson 2010). It is also important to note that the magnetic ellipsoid, despite accurately representing the samples petrofabric, cannot be simply correlated with the estimated strain ellipsoid or actual strain. This is due to a number of factors, including but not limited to the following: minerals present in a sample have a fundamental control on the degree of anisotropy and not strain; the predeformation magnetic ellipsoid is not necessarily spherical; and the magnetic ellipsoid may also represent the sum of two competing fabrics, such as primary sedimentary fabrics and cleavage (Hirt et al. 1988(Hirt et al. , 1993. Similar problems with non-isotropic original fabrics have been described in traditional strain markers (Dunnet & Siddans 1971).
A magnetic foliation (the plane perpendicular to K3, defined by K1 and K2) and lineation (parallel to K1) can be obtained from the AMS ellipsoid (Borradaile & Jackson 2004). Additionally, the overall shape of this ellipsoid can be useful for structural interpretations, with three main geometries being oblate (K1 ≅ K2 > K3, with K3 perpendicular to magnetic foliation), prolate (K1 > K2 ≅ K3, with K1 parallel to magnetic lineation) and triaxial (K1 ≠ K2 ≠ K3). In order to quantify and represent these geometries in 2D space the shape and anisotropy parameters of Jelinek (1981) are used. The shape parameter, T j , is defined as: while the degree of anisotropy, P j , is defined as T j and P j can be plotted against each other in Cartesian space (Fig. 6a). T j values range from −1 (prolate) to +1 (oblate), with a T j value of 0 representing a triaxial neutral ellipsoid. P j describes the relative strength of ellipsoid shape anisotropy, with increasing P j values suggesting a stronger fabric or lineation.

Fabric types
There is now a considerable amount of work in compressional settings detailing the development of tectonic fabrics in sedimentary rocks with a primary bedding fabric, as observed by AMS (Graham 1996;Kligfield et al. 1981;Bakhtari et al. 1998;Parés et al. 1999;Robion et al. 1999;Parés 2004;Burmeister et al. 2009). This development can be described using four types of ellipsoid geometries, summarized below and in Figure 6a and b. For a more complete description, see McCarthy et al. (2015). It should be noted that if there is any shear component of the deformation, then these relationships become more complex. Type 1: An initial sedimentary fabric is typically characterized by an oblate ellipsoid, with slight flattening parallel to bedding. In this case, the K1 and K2 axes are scattered in a girdle representing the magnetic foliation and roughly conforming to bedding, while K3 is perpendicular to the magnetic foliation/bedding. Strong magnetic lineations are rarely present, owing to the highly scattered K1. Type 2: The first sign of an incipient tectonic fabric is typically weaker than the primary sedimentary fabric; therefore the AMS ellipsoid may still be weakly oblate and conformable with bedding. In this case, the K1 axes may start clustering in the direction of extension and defining a magnetic lineation parallel to the intersection of an incipient LPS fabric with bedding. Type 3: As deformation continues, the magnetic ellipsoid becomes prolate, the K1 axes become strongly clustered and the K2 axes are roughly equal to the K3 axes. Type 4: The final stage involves a magnetic foliation perpendicular to bedding, with K1 and K2 axes forming a great circle girdle parallel to cleavage. The K1 axes may still be clustered at the intersection of bedding and cleavage, forming a magnetic lineation, or scattered in the plane of cleavage. This stage typically has flattened oblate AMS ellipsoids perpendicular to bedding.

Results
Results from the AMS analyses are presented in Table 1 and summarized in this section. Bulk susceptibility varies from −3.8 × 10 −5 to 1.9 × 10 −4 SI, with the majority of samples yielding a negative (diamagnetic) or extremely weak susceptibility (Fig. 7a). Negative and extremely weak positive susceptibilities are common in very pure limestones that lack a volumetrically significant Fe-Ti oxide component or other magnetic Fe-bearing silicate phases. Calcite and dolomite, which are diamagnetic minerals (Hunt et al. 1995), are the dominant carrier of the AMS fabric in samples with negative bulk susceptibilities. The carrier of the AMS in the specimens with positive susceptibility values up to 19 × 10 −5 may be a mixture of ferromagnetic and diamagnetic minerals. The corrected degree of anisotropy (P j ) varies from 1.01 to c. 2.00, suggesting a range of fabric strengths, which is comparable with deformed limestones elsewhere (Borradaile et al. 2012). It is important to note that P j values become exaggerated as the bulk susceptibility approaches zero (Fig. 7a); this is because the calculation of P j is dependent on division by near-zero in these cases. In order to resolve this some authors have used ΔK (K1 − K3) as an alternative to describe the degree of anisotropy (Almqvist et al. 2010(Almqvist et al. , 2011Hirt & Almqvist 2012). A comparison of P j and ΔK for the sample means is presented in Figure 7c & d; regardless of the anisotropy descriptor used, it should be noted that similar trends are seen. Following the recommendations of Tarling & Hrouda (1993), P j and T j are used for structural comparisons. Additionally, there is no obvious correlation between the shape parameter (T j ) and bulk susceptibility (Fig. 7b), which implies that T j is controlled either by primary or tectonic fabrics, rather than the composition of the limestones. P j and T j values are presented in Figure 8a-e for all specimens in each main thrust sheet. It is evident from these plots that all thrust sheets sampled exhibit a range of AMS ellipsoid geometries from weak oblate through prolate with some samples exhibiting strong oblate geometries.
The contribution of diamagnetic minerals in the sample suite from the Madison Group limestones complicates AMS interpretations. In pure calcite and dolomite, the principal negative susceptibility axis is aligned along the c-axis of the crystal (Borradaile et al. 2012), which is typically perpendicular to schistosity or tectonic cleavage (Flinn 1965). Therefore, the maximum negative susceptibility axis in diamagnetic materials largely coincides with the normal to the dominant foliation (Borradaile et al. 2012). In order to compare the diamagnetic fabrics to paramagnetic fabrics, the orientation of the maximum (most negative) and minimum (least negative) axes are exchanged (Hrouda 2004;Schmidt et al. 2006;Borradaile et al. 2012).
In an attempt to identify regional magnetic fabrics, specimens have been split into two groups, (A) paramagnetic and (B) diamagnetic, and AMS principle axes are plotted on lower hemisphere equal-area projections with bedding and cleavage  (Fig. 9). These plots show a considerable amount of scatter for both paramagnetic and diamagnetic samples, regardless of being corrected for bedding tilt. The K1 axes for paramagnetic samples typically plot at the intersection of cleavage and bedding or along the cleavage plane (Fig. 9b). There is considerable scatter of the paramagnetic K2 axes (Fig. 9c), although a number plot near the bedding plane, which is expected for a low-strain regime. The paramagnetic K3 axes appear to have a bimodal distribution, either plotting as a cluster normal to bedding or as a girdle parallel to bedding, which is indicative of Type 1-2 and Type 3 geometries, respectively.
The diamagnetic axes are presented in Figure 9e & f, and although they have a similar orientation to the paramagnetic axes, there is more scatter of the data. The diamagnetic K1 and K2 axes typically plot as a girdle parallel to bedding, but there is some clustering at the bedding cleavage intersection ( Fig. 9f, g). The diamagnetic K3 axes have a similar distribution to the paramagnetic K3 axes (Fig. 9h).

Interpretation
The AMS fabrics exhibit a range of fabric types that are commonly seen in fold-and-thrust belts (Graham 1996;Kligfield et al. 1981;Bakhtari et al. 1998;Parés 2004;Weil & Yonkee 2009;Yonkee & Weil 2010;McCarthy et al. 2015). These fabric types evolve from bedding controlled to tectonic cleavage through an intermediate stage with intersecting fabrics (Bakhtari et al. 1998;Borradaile et al. 2012). This evolution of fabric type is evident in the P j -T j plots, whereby ellipsoid shapes vary from weakly oblate with flattening parallel to bedding, to prolate with stretching parallel to the extension direction and a final stage of oblate geometries with   flattening parallel to cleavage (Fig. 8;Graham 1996;Parés 2004).
The range of AMS ellipsoid shapes are suggestive of composite fabrics, with bedding largely responsible for the magnetic foliation, while the magnetic lineation typically correlates with the bedding/cleavage intersection. This evolution of a magnetic fabric is discussed further by Housen et al. (1993).
It is interesting to note that each type of magnetic fabric is present across all thrust sheets sampled in this study (Figs 8 & 10). This is further illustrated in Figure 11, whereby a range of ellipsoid types are present across each thrust. This is opposite to what was expected prior to analysis, whereby penetrative deformation was thought to increase towards the hinterland in the steeper Beaver and Norwegian thrusts. These observations suggest that there is no simple east-west increase in penetrative deformation across the five thrusts studied, as is commonly found in fold-and-thrust belts (Yonkee & Weil 2010).
Furthermore, there does not appear to be a regular distribution or localization of bedding-controlled v. cleavage-controlled fabric types within each thrust sheet (Figs 10 & 11).
Where deformation fabrics are observed in the field, they are at a high angle to bedding and largely limited to stylolitization and occasional spaced cleavage/dissolution seams. Despite this weak development of a tectonic fabric, there is a regular correlation with AMS fabrics and the observed tectonic fabric, with the K1 lineation axes regularly plotting along a cleavage plane or at the cleavagebedding intersection lineation (Figs 9-11). This further confirms that AMS is capable of identifying tectonic fabrics in samples with low and negative bulk susceptibilities. Interestingly, owing to the brittle nature of dissolution seams and stylolites, most specimens were prepared from segments of block samples that had no visible tectonic fabric, suggesting that further mineral alignment is occurring in addition to these two strain indicators.

Discussion
The main structures of the Sawtooth Range are characterized by thrust faults that place Madison Limestone over Cretaceous shale (Holl & Anastasio 1992). The emplacement of these thrusts was largely enabled by progressive development of mesoscopic fault arrays that allowed the base of the thrust sheets to deform by cataclastic flow (Holl & Anastasio 1992). Brittle deformation is the most pervasive style of deformation in each thrust sheet, with limited development of a penetrative tectonic fabric.
The poor development of penetrative fabrics in the Madison Limestones may be attributed to the relatively low burial temperature conditions experienced. The temperatures of 100-175°C constrained by illitic mineral assemblages (Hoffman et al. 1976;Gill et al. 2002;O'Brien et al. 2006) are below the temperatures required (200-300°C) for intra-crystalline plastic flow of calcite to become a dominant deformation mechanism (Engelder & Marshak 1985). Analysis of thin sections (Fig. 5) reveals that grain-scale deformation is limited to Type 1 calcite twinning (Ferrill et al. 2004) and grain boundary bulging (Passchier & Trouw 2005). Both of these textures indicate deformation temperatures below 170°C. The presence of a tectonic stylolitic fabric consistently at a high angle to bedding suggests that this fabric developed prior to thrusting. This is further confirmed by the coaxial folding of stylolites with bedding described at the nearby Swift Reservoir (Ward & Sears 2007). Fig. 11. Stereographic projections of principal susceptibility axes for representative block samples across the sampled thrust sheets. Also shown is the inclination of magnetic foliation relative to bedding and tectonic stylolites. Magnetic fabric types are indicated. Inset illustrates evolution of magnetic fabric types assuming horizontal bedding. This is significantly different from the model of tectonic fabric development during thrust emplacement described by Sanderson (1982), whereby if cleavage developed during thrusting, it would be expected to develop at an oblique angle to bedding (Fig. 12). Similarly, Evans & Dunne (1991)   identified four key deformation events associated with thrust sheet evolution: (1) initial layer parallel shortening; (2) bending and folding at a ramp hinge; (3) synthrusting-related simple shear; and (4) post-emplacement flattening. These models suggest that LPS development precedes or is synchronous with thrust sheet emplacement, which is then followed by further deformation. Evans & Dunne (1991) also highlighted that the style of penetrative strain recorded in thrust sheets is dependent on whether the right temperature and pressure conditions are present to accommodate grain-scale deformation, and that these conditions can vary temporally and spatially within a thrust sheet.
The AMS results presented here do not identify any penetrative deformation that can clearly be linked to a synthrusting strain. Furthermore, the only tectonic fabrics identified were consistently at a high angle or perpendicular to bedding and appeared to be of a domainal nature. This is in agreement with the field evidence that LPS occurred prior to thrust sheet emplacement (Holl & Anastasio 1992).
Therefore, a schematic model for LPS evolution in the Sun River area is presented in Figure 13. The first stage of deformation involves the development of widespread LPS or weak tectonic fabric. This is followed thrust fault initiation and related folding, facilitated by brittle deformation in the hanging wall fault boundary as described by Holl & Anastasio (1992). This effectively rotates the early LPS. Further thrusting promotes fracturing in structurally competent units such as the Allan Mountain Limestone and Castle Reef Dolomite Formations, but may also enhance LPS development in the footwall owing to a change in PT conditions. Similar studies in the Wyoming fold-and-thrust belt suggested that LPS developed progressively in individual thrust sheets prior to thrusting and as a consequence of shortening under the influence of the overriding thrust sheet, which would have caused a change in temperature, pressures and fluid conditions (Wiltschko & Dorr 1983).

Conclusion
Analysis of AMS of the carbonate-dominated thrust sheets in the Sun River valley suggests the presence of a weak tectonic fabric that developed prior to thrusting. The magnetic susceptibility of specimens is either diamagnetic, controlled by either calcite and/or dolomite, or paramagnetic, and possibly controlled by minor amounts of phyllosilicates and ferromagnetic minerals. The observed AMS ellipsoid shapes vary from bedding-controlled oblate ellipsoids parallel to bedding through to tectonically modified prolate ellipsoids with the K1 axes parallel to the observed tectonic fabrics. Occassional oblate ellipsoids perpendicular to bedding are also observed.
The formation of the carbonate-dominated thrust sheets in the Sun River area was facilitated by brittle deformation at the base of the thrust sheets, as well as ductile deformation in the Cretaceous strata of the footwalls. The emplacement of these sheets effectively rotated a weak LPS fabric. This weak tectonic fabric is limited to stylolites and dissolution seams in outcrop. The AMS analysis suggests that the tectonic fabrics vary from lineations, parallel to stylolites, to a weak foliation at a high angle to bedding.
These magnetic lineations and foliations are present in all of the sampled thrust sheets, and there appears to be no localization of penetrative deformation represented by AMS ellipsoid types within each sheet. Furthermore, no penetrative fabric associated with thrusting has been detected by the AMS analyses.
The correlation of K1 axes with tectonic fabrics further confirms that AMS is capable of identifying early LPS in both diamagnetic and paramagnetic specimens with low bulk susceptibilities.