Abstract

We present the first detailed structural analysis of the Yukon River shear zone (YRSZ), which forms an important structural break within the Yukon-Tanana terrane of the Northern Cordillera in Yukon (Canada). The YRSZ is a NW-SE–striking shear zone that juxtaposes Mississippian orthogneiss hanging-wall rocks (Simpson Range suite) against pre-Late Devonian metasedimentary footwall rocks (Snowcap assemblage). Field and microstructural analyses, including quartz c-axis fabric investigation, indicate that the YRSZ initiated as a top-ESE mid-crustal shear zone active through a temperature range of ≥650–500 °C to ∼540–440 °C. Constraints from the footwall associated with top-ESE shearing on the YRSZ at mid-crustal conditions record a decrease in deformation temperature toward the shear zone, coincident with a transition from coaxial to non-coaxial deformation and an increase in fabric intensity, strain rate, and differential stress estimates. Collectively, these spatial trends represent a classic example of a narrowing shear zone that progressively localizes and intensifies deformation as ambient temperature decreases. U-Pb zircon geochronometry of a deformed Permian orthogneiss from within the YRSZ combined with previously published thermochronometry bracket the timing of top-ESE mid-crustal shearing between 259 ± 2 Ma and 176–168 Ma, either during Late Permian–Middle Triassic metamorphism and lithospheric extension or latest Triassic–Early Jurassic metamorphism and crustal thickening. The YRSZ was subsequently reactivated as a top-WNW upper-crustal thrust fault zone during or after Early to Middle Jurassic cooling and exhumation at 176–168 Ma. This top-WNW thrusting within the YRSZ may be responsible for structural separation of Late Triassic and Early Jurassic plutonic rocks in the hanging wall of the YRSZ from Permian plutonic rocks in its footwall.

1. INTRODUCTION

Understanding the structural evolution of crustal-scale shear zones provides important constraints for models of tectonic deformation (e.g., Sibson, 1977; Handy et al., 2007; Faulkner et al., 2010). Unravelling these constraints from the geological record is difficult due to the overprinting nature of progressive and/or polyphase deformation, which may be recorded within a single shear zone over a variable range of pressures and temperatures (e.g., Price et al., 2016; Williams and Platt, 2017; Hunter et al., 2018; Wallis et al., 2018). Here, we present field and microstructural observations from the Yukon River shear zone (fig. 4 in Tempelman-Kluit and Wanless, 1980; Yukon River Thrust in Ryan et al., 2014; Coleman, 2017; Coleman et al., 2017), which represents a polyphase ductile to brittle, mid- to upper-crustal shear zone in central Yukon within the allochthonous Yukon-Tanana terrane of the Northern Cordilleran accretionary orogen. Our data quantitatively constrain variations of deformation temperature, differential stress and strain rate, strain geometry, and strain intensity during movement within the shear zone. These data outline a well-preserved example of temperature-dependent, progressive strain evolution and localization within a mid- to upper-crustal shear zone (e.g., Law, 1987; Price et al., 2016; Aravadinou and Xypolias, 2017; Hunter et al., 2018).

Shear zones accommodate tectonic processes via localization of deformation within the lithosphere (e.g., Sibson, 1977; Handy et al., 2007; Wallis et al., 2015; Aravadinou and Xypolias, 2017). Within the brittle regime, shear zones are typically governed by a Mohr-Coulomb failure criterion in which cohesion (i.e., rock strength), applied stress, and pore-fluid pressure represent important boundary conditions (e.g., Sibson, 1977; Faulkner et al., 2010). Within the ductile regime, shear zones display a power-law creep behavior facilitated by dynamic recrystallization and diffusive mass transfer processes, resulting in pervasively sheared zones of deformation that commonly develop mineral-shape and/or crystallographic preferred orientations (e.g., Law, 1990; Handy et al., 2007). In these ductile shear zones, temperature and differential stress place important controls on strain rate, strain geometry, and the grain-scale deformation processes that accommodate such strain (e.g., Hirth et al., 2001; Stipp et al., 2002a, 2010). In many cases, exhumed crustal-scale shear zones exposed at the surface record deformation from both the brittle and ductile regimes, which correspond to changes in ambient conditions and material properties as the shear zone evolved (e.g., Parsons et al. 2016; Wallis et al., 2018). The overprinting nature of progressive and/or polyphase ductile-to-brittle (or vice versa) deformation that commonly occurs in such structures must be elucidated in order to understand how crustal-scale shear zones evolve.

There are a variety of microstructural and thermobarometric analytical techniques that, when used correctly and in conjunction with well-constrained field observation, provide a means to determine the thermokinematic evolution of shear zones. Characterization of microstructures in fault rocks via microscopy may be used to distinguish between different temperature- and/or pressure-dependent grain-scale processes (e.g., Sibson, 1977; Law, 1990, 2014). Measurements of crystallographic preferred orientations (CPO) and misorientation angles can provide an indication of temperature-dependent slip-system activity and may be used to constrain kinematics, strain geometry, and strain intensity (e.g., Barth et al., 2010; Ambrose et al., 2018; Hunter et al., 2018; Soucy La Roche et al., 2018). Additionally, measurements of recrystallized grain size may be used to estimate differential stress, which in turn may be used to estimate strain rate if the corresponding deformation temperature is known (e.g., Hirth et al., 2001; Stipp and Tullis, 2003; Stipp et al., 2010).

In our study of the Yukon River shear zone (YRSZ), we utilize techniques outlined above to investigate its structural evolution. Our results provide insight into how deformation is accommodated in the middle to upper crust and how resulting structures evolve and may become reactivated during subsequent tectonic events. Our data indicate that the Yukon River shear zone initiated as a mid-crustal top-ESE shear zone at temperatures of ≥650–500 °C to ∼540–440 °C and subsequently reactivated as part of a top-WNW thrust fault zone at deformation temperatures of <440 °C. New U-Pb zircon geochronometry of orthogneisses yielded igneous crystallization ages of ca. 352–347 Ma in the hanging wall of the YRSZ and ca. 261–259 Ma in the footwall of the YRSZ and constrain a maximum age for top-ESE shearing of 259 ± 2 Ma. 40Ar/39Ar thermochronometry suggests that top-WNW thrusting on the YRSZ occurred during or after the Early to Middle Jurassic (e.g., Joyce et al., 2015). The kinematics and timing constraints derived from the YRSZ are not easily reconciled with current models for the Northern Cordillera (e.g., Berman et al., 2007; Nelson et al., 2013; Staples et al., 2016; Parsons et al., 2018; van Staal et al., 2018) and are considered further in our discussion. The spatial trends in deformation temperature, strain geometry and intensity, grain size, and differential stress displayed by top-ESE deformation on the YRSZ represent an excellent example of progressive deformation and strain localization within an exhuming (i.e., cooling) and narrowing shear zone (e.g., Sibson, 1977; Ramsay, 1980; Law, 1990; Handy et al., 2007; Faulkner et al., 2010). Our results are comparable to both natural and modeled examples of strain localization on crustal-scale shear zones and provide valuable insight into how shear zones evolve through time under variable external conditions (e.g., Platt and Behrmann, 1986; Law, 1987, 1990; Schmid and Casey, 1986; Mancktelow and Pavlis, 1994; Parsons et al., 2016; Aravadinou and Xypolias, 2017).

2. REGIONAL GEOLOGY: THE YUKON-TANANA TERRANE

The Yukon-Tanana terrane (YTT) is located in the central belt of the Northern Cordillera (the Intermontane superterrane) and is traceable from west to south-central Yukon (Fig. 1) and into northern British Columbia and eastern Alaska. It is interpreted as an allochthonous terrane of Paleozoic and Mesozoic volcanic arcs built upon a basement of Laurentian continental crust (Colpron et al., 2006; Nelson et al., 2006, 2013; Piercey and Colpron, 2009). Current popular and widely cited tectonic models for the Northern Cordillera suggest that following initiation of the Devonian Ecstall arc along the west margin of Laurentia, backarc rifting resulted in separation of YTT from Laurentia. Separation occurred sometime during the Late Devonian–Early Mississippian accompanied by formation of the intervening Slide Mountain Ocean (Colpron et al., 2006, 2007; Nelson et al., 2006, 2013; Parsons et al., 2018). Several arc cycles were subsequently built upon YTT between the Late Devonian and Permian during east-dipping subduction of the Panthalassa Ocean beneath the western margin of YTT.

Polyphase amphibolite- to eclogite-facies metamorphism recorded by the YTT continental basement occurred during the mid-Permian to Middle Triassic, latest Triassic–Early Jurassic, and Late Jurassic–Early Cretaceous (e.g., Dusel-Bacon et al., 1995; Creaser et al., 1997; Berman et al., 2007; Staples, 2014; Petrie et al., 2016; Staples et al., 2016; Clark, 2017; Gilotti et al., 2017; Morneau et al., 2017). Each of these metamorphic events was followed by rapid cooling (e.g., Dusel-Bacon et al., 2002; Knight et al., 2013; Staples et al., 2016; Parsons et al., 2018).

Several models have been proposed to explain the accretion of YTT to Laurentia, sometime between the Permian and Jurassic. Beranek and Mortensen (2011) proposed that YTT collided with and accreted back onto Laurentia during the mid to Late Permian “Klondike orogeny,” facilitated by west-dipping subduction of the intervening Slide Mountain ocean beneath YTT. However, recent models have proposed that YTT collided with an intra-oceanic arc during the Late Permian (the Dunite Peak intra-oceanic arc—Parsons et al., 2018; or YTTj—van Staal et al., 2018) and did not accrete to Laurentia until after the Middle Triassic, and probably during the Late Triassic–Early Jurassic (Hansen et al., 1991; Stevens et al., 1996; Plint and Gordon, 1997; Hansen and Dusel-Bacon 1998; Parsons et al., 2018).

2.1. The Yukon River Shear Zone (YRSZ)

In the northern Stevenson Ridge area of west-central Yukon (National Topographic System [NTS] Index 115JK), the occurrence of the YRSZ was inferred by Tempelman-Kluit (1974), Templeman-Kluit and Wanless (1980), and Ryan et al. (2013a, 2013b, 2014) based on constraints from regional mapping and geochronometry (Fig. 2A). These authors proposed that this structure formed an important intra-terrane discontinuity separating crustal “panels” containing different plutonic and volcanic assemblages.

At a regional scale, the trace of the YRSZ strikes approximately NW-SE, along the banks of the Yukon River (red fault trace on Fig. 2A). It separates strongly deformed hornblende-biotite dioritic to tonalitic orthogneiss of the Simpson Range suite (355–345 Ma) in the hanging wall from Snowcap assemblage siliciclastic metasedimentary rocks (pre-Late Devonian) and the Sulphur Creek suite orthogneiss (264–252 Ma) in the footwall (Ryan et al., 2014; Colpron et al., 2016a, 2016b). To the SE, the YRSZ has been mapped as terminating in a region intruded by Late Triassic to Early Jurassic plutonic rocks of the Minto suite (ca. 204–194 Ma; Colpron et al., 2016b) and partly covered by Cretaceous volcanic-dominated Carmacks Group. To the NW, compilation mapping (Colpron et al., 2016a) extrapolated the YRSZ along the Yukon River south of a zone of complex NW-SE–trending strike-slip faulting (dark-gray fault trace on Fig. 2A). However, this extrapolation incorporated rocks of the Permian Sulphur Creek suite in the hanging wall of the YRSZ (Fig. 2A) and contradicts the original hypothesized definition of the Yukon River Thrust (Ryan et al., 2014); thus, this extrapolation may be an error in detail.

Ryan et al. (2014) proposed that the YRSZ (then, the Yukon River Thrust) emplaced a crustal panel that hosts plutonic rocks belonging to the Simpson Range (355–345 Ma), Stikine (216–206 Ma), and Minto (204–194 Ma) suites on top of a crustal panel that hosts plutonic rocks of the Sulphur Creek suite (264–252 Ma). These authors suggested that the YRSZ may represent a fundamental crustal break within the YTT between crustal blocks with distinctly different tectonic histories. The implication of this hypothesis is that in Yukon, the YTT is a composite terrane composed of subterranes with distinct substrates. Based on the mapped distribution of Triassic and Permian rocks, Ryan et al. (2014) also suggested that juxtaposition of footwall and hanging-wall rocks along the YRSZ occurred during post–Late Triassic crustal thickening of the YTT. A similar model suggesting that YTT is a composite terrane composed of distinct subterranes has also been proposed by van Staal et al. (2018).

3. STUDY AREA: YUKON RIVER, NORTHERN STEVENSON RIDGE

The Northern Stevenson Ridge area is defined by kilometer-scale folds with NW-SE–striking and NE-dipping axial planes, which fold the dominant tectonic fabric (Fig. 2A). In areas close to the Yukon River (Fig. 2D), foliation strikes NW-SE to WNW-ESE and dips toward the NE. The trace of the YRSZ as mapped by Ryan et al. (2013a, 2013b, 2014) is defined by the contact between the orthogneiss and metasedimentary rocks and dips moderately toward NE (Fig. 2), parallel to the regional foliation. In the study area (Fig. 2D), this foliation defines the SW limb of a regional-scale synform, which folds the YRSZ and the Simpson Range orthogneiss and Snowcap metasedimentary rocks that lie above and below it (Fig. 2B).

During boat-assisted fieldwork in 2016, we identified a ∼180-m-long exposure of the YRSZ on the north bank of the Yukon River (locality JR-140; Figs. 2 and 3), ∼50 km west of the confluence with the Pelly River. All field and microstructural descriptions come from samples and observations taken from this locality unless specified otherwise. At this outcrop, quartzite interlayered with subordinate semipelitic to pelitic schist and carbonaceous horizons, identified previously as the pre–Late Devonian Snowcap assemblage, are structurally overlain by orthogneiss with K-feldspar augens and a syenogranitic to monzogranitic composition, identified previously as the Late Devonian–Mississippian Simpson Range suite (Figs. 2 and 3) (Ryan et al., 2013a, 2013b; Colpron et al., 2016a, 2016b). A strain gradient recorded by various macrostructural and microstructural kinematic indicators across locality JR-140 (see sections 5 and 6) indicates that the inclined contact between the quartzite and orthogneiss is a fault contact associated with the YRSZ (Fig. 3). This contact, which we refer to henceforth as the “YRSZ fault contact,” is covered by colluvium and vegetation, but can be constrained to a ∼1–1.5-m-wide zone that is subparallel to foliation and is offset by a vertical brittle fault (Fig. 3). U-Pb zircon geochronometry of orthogneiss samples collected from across the study region (Fig. 2A), including locality JR-140, are presented below (section 4), followed by field observations and microstructural analyses of samples collected from locality JR-140 and neighboring outcrops (sections 5 and 6).

4. U-Pb ZIRCON GEOCHRONOMETRY

In order to test the predictions of Ryan et al. (2014), who proposed that the YRSZ places Mississippian metaplutonic rocks atop metasedimentary rocks with Permian plutonic intrusion, we conducted zircon U-Pb geochronometric analyses on five orthogneiss samples collected from across the study region: (A) sample 16RAY-JR140A01, collected from locality JR-140, shown in Figure 3; (B) sample 10RAYMC005A02 from the YRSZ footwall, ∼10 km south of the YRSZ fault trace; (C and D) samples 10RAYMC023A02 and 11RAYAZ297A from the YRSZ hanging wall, ∼1–3 km north of the YRSZ fault trace; and (E) sample 11RAYJR248A, collected from ∼12 km south of the YRSZ fault trace. Sample locations are presented on Figure 2 and listed in Table DR1 in the GSA Data Repository Item1. U-Pb isotopic data were acquired using the sensitive high-resolution ion microprobe (SHRIMP II) at the Geological Survey of Canada Geochronology Laboratory. Concordia plots of SHRIMP II U-Pb data and corresponding mean 207Pb-corrected 206Pb/238U crystallization ages for each sample are presented in Figure 4. Analytical procedures and backscattered electron (BSE) and cathodoluminescence (CL) images of representative zircon crystals with spot analysis locations and ages are presented in File DR1 (see footnote 1).

4.1. Sample 16RAY-JR140A1: Permian Orthogneiss

Sample 16RAY-JR140A1 (Fig. 4A) is a strongly foliated, muscovite-bearing K-feldspar augen syenogranitic to monzogranitic orthogneiss, collected from the study exposure of the YRSZ at locality JR-140 in the immediate hanging wall (Label A in Figs. 2A and 2B). Zircons from this sample are 75–200 µm, euhedral, stubby to semi-elongate prisms. Although most grains have well-preserved facets, ∼40% of the grains have subrounded and/or irregular grain boundaries. In transmitted light, the grains are clear and colorless, with abundant colorless and bubble- and rod-shaped inclusions. In SEM-BSE images, all grains exhibit igneous oscillatory zoning (File DR1a [see footnote 1]). The grains have generally moderate to high U content (116–2910 ppm) and low to moderate Th/U (0.23–0.46 ppm). Five grains showed visible sponge-textured cores. Seventeen spots on 17 grains (one spot per grain) returned a weighted-mean 206Pb/238U age of 259 ± 2 Ma (mean square of weighted deviates [MSWD] = 1.5) (Fig. 4A). Two grains were excluded from the age calculation because of elevated common Pb content, and two were excluded due to suspected inherited components. Based on its age and lithology, we assign this sample to the Sulphur Creek suite (264–252 Ma, e.g., Colpron et al., 2016b).

4.2. Sample 10RAYMC005A02: Permian Orthogneiss

Sample 10RAYMC005A02 (Fig. 4B) is a monzogranitic augen orthogneiss collected from the footwall domain, ∼10 km south of the trace of the YRSZ (Label B in Fig. 2A). This sample yielded large (150–250 µm), euhedral, stubby to semi-elongate prisms. In transmitted light, the grains are clear and pale yellow, with abundant colorless and bubble- and rod-shaped inclusions. In SEM-CL images, all grains exhibit igneous oscillatory zoning (File DR1b [see footnote 1]). The grains have generally high U content (359–1515 ppm) and low to moderate Th/U (0.21–0.55 ppm). Seven grains contained visible cores. Seventeen spots within the oscillatory zoning within 17 grains (one spot per grain) returned a weighted-mean 206Pb/238U age of 261 ± 2 Ma (MSWD = 1.2) (Fig. 4B). One grain, which was slightly reversely discordant, was excluded from the age calculation. Based on its age and lithology, we assign this sample to the Sulphur Creek suite (264–252 Ma; e.g., Colpron et al., 2016b).

4.3. Sample 10RAYMC023A02: Mississippian Orthogneiss

Sample 10RAYMC023A02 (Fig. 4C) is a foliated and folded biotite-hornblende dioritic to granodioritic orthogneiss collected from the YRSZ hanging-wall domain, ∼1 km north of the YRSZ fault trace (Label C in Fig. 2A). Zircon grains from sample 10RAYMC023A02 are large (150–250 µm), euhedral, stubby to subrounded semi-elongate prisms. In transmitted light, most grains are clear and pale yellow, with abundant colorless and bubble- and rod-shaped inclusions. In SEM-CL images, all grains exhibit igneous oscillatory zoning, and many also show sector zoning (File DR1c [see footnote 1]). The grains have generally high U content (359–1515 ppm) and low to moderate Th/U (0.21–0.55 ppm). Twelve grains contain visible cores that are too small to place a beam onto. Twelve spots within the oscillatory zoning within 12 grains (one spot per grain) gave a weighted-mean 206Pb/238U age of 347 ± 3 Ma (MSWD = 1.07) (Fig. 4C). Eight grains were excluded from the age calculation due to suspected postcrystallization Pb loss, high common Pb, or inheritance. Based on its age and lithology, we assign this sample to the Simpson Range suite (355–345 Ma; e.g., Colpron et al., 2016b).

4.4. Sample 11RAYAZ297A: Mississippian Orthogneiss

Sample 11RAYAZ297A (Fig. 4D) is a muscovite-bearing, weakly foliated to non-foliated monzogranitic orthogneiss collected from the YRSZ hanging-wall domain, ∼1 km north of the YRSZ fault trace (Label D in Figs. 2A and 2B). This sample yielded zircon grains that were 100–350 µm, euhedral to slightly rounded, stubby to semi-elongate prisms. Under transmitted light, most grains are clear and colorless, with abundant colorless and bubble- and rod-shaped inclusions. In SEM-CL images, all grains exhibit igneous oscillatory and sector zoning, (File DR1d [see footnote 1]). The grains have generally high U content (144–820 ppm) and moderate to high Th/U (0.37–0.98 ppm). Cores were visible in 10% of the grains. Eighteen spots within the igneous zones of 18 grains (one spot per grain) gave a weighted-mean 206Pb/238U age of 347 ± 5 Ma (MSWD = 1.2) (Fig. 4D). Two grains for which the ion beam is suspected to have inadvertently overlapped inherited cores were excluded from the age calculation. Based on its age and lithology, we assign this sample to the Simpson Range suite (355–345 Ma; e.g., Colpron et al., 2016b).

4.5. Sample 11RAYJR248A: Mississippian Orthogneiss

Sample 11RAYJR248A (Fig. 4E) is a strongly foliated hornblende-biotite-granodioritic orthogneiss, interlayered with amphibolite, collected ∼12 km south of the YRSZ fault trace (Label E in Fig. 2A). This sample yielded zircon grains that were 150–400 µm, euhedral, equant, stubby to semi-elongate prisms. In transmitted light, most grains are clear and colorless, with very abundant colorless and bubble- and rod-shaped inclusions. In SEM-CL images, all grains exhibit igneous oscillatory and sector zoning, (File DR1e [see footnote 1]). The grains have generally moderate U content (110–450 ppm) and moderate Th/U (0.42–0.80 ppm). Cores were only visible in two grains. Eighteen spots within igneous zones of 18 grains (one spot per grain) yielded a weighted-mean 206Pb/238U age of 352 ± 5 Ma (MSWD = 0.81) (Fig. 4E). One grain was excluded from the age calculation due to suspected postcrystallization Pb loss, and one was excluded due to reverse discordance. Based on its age and lithology, we assign this sample to the Simpson Range suite (355–345 Ma; e.g., Colpron et al., 2016b).

4.6. Summary of U-Pb Zircon Geochronometry

Our results indicate that at locality JR-140 (Fig. 3), the orthogneiss that forms the hanging wall to the YRSZ fault contact is Permian (259 ± 2 Ma) and correlates with the Sulphur Creek suite (264–252 Ma; Colpron et al., 2016b). An augen orthogneiss collected from the Snowcap assemblage (in the footwall domain) in the footwall of the YRSZ, ∼10 km SW of the YRSZ fault trace (Fig. 2A), yielded an age of 261 ± 2 Ma, which we also correlate with the Sulphur Creek suite. Elsewhere in the Stevenson Ridge area, two orthogneiss samples collected from the YRSZ hanging wall, positioned ∼1–3 km north of the YRSZ fault trace (Fig. 2A), yielded Mississippian ages (ca. 347 Ma) and are assigned to the Simpson Range suite (355–345 Ma; Colpron et al., 2016b). The ages of these two Mississippian orthogneisses are consistent with the findings of Tempelman-Kluit and Wanless (1980), who determined a U-Pb crystallization age of 351.6 +4.5/−3.2 Ma from an orthogneiss also positioned within the YRSZ hanging wall (Fig. 2A). The fifth orthogneiss sample, collected ∼12 km south of the YRSZ fault trace (Fig. 2A), yielded an age of 352 Ma and is assigned to Simpson Range suite.

The occurrence of Mississippian orthogneiss ∼12 km south of the YRSZ (Fig. 2A) is somewhat unexpected; however, we have no constraints on the nature of the contact between that orthogneiss (sample 11RAYJR248A, Fig. 4E) and the surrounding metasedimentary rocks of the Snowcap assemblage. It is likely that this orthogneiss is an isolated klippe separated from the underlying metasedimentary rocks by the YRSZ.

We cannot determine whether the Permian orthogneiss sample 16RAY-JR140A1 at locality JR-140 is derived from the footwall or the hanging wall. However, we saw no other observations to suggest that Permian orthogneiss may be found in the hanging wall of the YRSZ. We therefore interpret this Permian orthogneiss at JR-140 as a footwall-derived entrained block within the YRSZ, as supported by regional mapping (Ryan et al., 2013a, 2013b) and the work of Ryan et al. (2014). This implies that the YRSZ fault contact at locality JR-140 represents only a portion of the YRSZ, comprising the footwall of the YRSZ (quartzite and schist unit) and a lower segment of the core of the YRSZ (Permian orthogneiss unit). The hanging wall to the YRSZ, which comprises Mississippian orthogneiss (Simpson Range suite), must lie structurally upsection of the Permian orthogneiss at locality JR-140.

5. FIELD-STRUCTURAL OBSERVATIONS

At locality JR-140 and nearby outcrops along the Yukon River, several generations of overprinting deformation fabrics are observed. We describe these fabrics with standard notations for foliations (S1, S2, S3, etc.), lineations (L1, L2, L3, etc.), and folds (F1, F2, F3, etc.), with numbers that correspond to a specific deformation event (i.e., L2 lineation is a lineation that formed during the D2 deformation event). Deformation events are given a standard notation (D1, D2, D3, etc.), numbered in chronological order (i.e., D1 is older than D2).

5.1. D1 and D2 Field-Structural Deformation Fabrics

The earliest deformation that we recorded from outcrops along the Yukon River is a rarely observed foliation (S1) in the orthogneiss that defines the limbs of tight to isoclinal recumbent F2 folds (Fig. 5A). Aplite dikes within the orthogneiss are also deformed by these recumbent folds (Figs. 5A and 5B). Limbs and axial planes of these F2 folds are transposed into parallelism with the S2 foliation. S1 foliation and F2 folded aplite dikes are observed within the orthogneiss at localities JR-138 and JR-139 but are not recorded at locality JR-140 (Fig. 2D).

The S2 foliation is observed ubiquitously across the region. The orthogneiss unit displays a strong S to L-S fabric (D2) with a continuous cleavage defined by macroscopic grain-shape alignment (S2) and an associated L2 feldspar mineral stretching lineation (Figs. 5C and 5D). At locality JR-140, S2 foliation in the orthogneiss unit (Figs. 5C and 5D) has a mean dip/dip direction of 52/351; L2 mineral stretching lineation has a mean plunge/trend of 12/285 (Fig. 6). The degree of mineral stretching associated with these fabrics (D2) increases in intensity downsection toward the YRSZ fault contact with the underlying quartzite and schist unit, accompanying a transition from protomylonite to mylonite (D2). Recrystallized grain size in the orthogneiss visibly decreases toward the YRSZ fault contact, including that of the K-feldspar augens (Fig. 5C), which become progressively flattened and sheared (D2) (Fig. 5D) until they are no longer recognizable against the surrounding matrix grains. Where present, augen asymmetry is mostly consistent with a top-ESE shear sense (D2; Fig. 5C), though subordinate augens with opposing asymmetry are also observed. This D2 top-ESE shearing and associated mylonitic fabrics are the earliest deformation associated with the YRSZ.

S1 foliation is not observed within the quartzite and schist. This unit displays a strong S to L-S fabric (D2) with a penetrative spaced cleavage (S2) marked by mineral segregation and lithologically defined layering with L2 stretching lineation defined by alignment of quartz and biotite (Fig. 5E). These D2 fabrics are subparallel to D2 fabrics within the overlying orthogneiss; S2 foliation has a mean dip/dip direction of 58/038; L2 mineral-stretching lineations have a mean plunge/trend of 11/120 (Fig. 6). S2 foliation in the quartzite and schist unit is mylonitic (D2); quartz and calcite veins are transposed parallel to S2 foliation and commonly boudinaged (D2) (Figs. 5F–5H). Boudinage stretching directions (D2) are parallel to L2 mineral stretching lineations. The degree of mineral segregation in this S2 foliation increases in intensity upsection and forms well-defined mineral segregation bands close to the YRSZ fault contact with the orthogneiss (Fig. 5E). The trace of this fault contact is subparallel to S2 foliation in both the quartzite and orthogneiss units (Fig. 3), indicating that it is a D2 structure associated with top-ESE shearing along the YRSZ.

5.2. D3 Field-Structural Deformation Fabrics

Within the quartzite and schist unit, S2 foliation is deformed by meter-scale asymmetric F3 folds (Figs. 3 and 7A) with steep to subvertical NNE- to SSW-dipping short limbs (Fig. 6), S2 subparallel long limbs, and mean axial plane and fold axis orientations of 36/026 and 00/297, respectively (i.e., fold axes approximately parallel to the strike of the outcrop) (Fig. 6). This asymmetric folding is not observed in the orthogneiss. Axial traces of asymmetric folds in the quartzite and schist unit are subparallel to the axial trace of the regional-scale, NW-SE–trending synform mapped to the north of locality JR-140 (Figs. 2A and 2B). This suggests that asymmetric F3 folds at locality JR-140 are parasitic folds on the southern limb of the regional-scale synform (D3).

5.3. D4 Field-Structural Deformation Fabrics

In the quartzite and schist unit as well as in the orthogneiss unit, S2 foliation and F3 asymmetric folds are cut by subordinate brittle D4 thrust faults with meter-scale offsets and a top-WNW sense of shear (Fig. 3). Within the quartzite and schist unit, these thrust faults increase in occurrence toward the YRSZ fault contact with the overlying orthogneiss (Figs. 7B–7D). Some D4 thrust faults cut across the S2 foliation (Fig. 7C), but most are parallel to the S2 foliation (Fig. 3). These D4 thrust planes slip along S2-parallel carbonaceous pelitic horizons and are typically coincident with the subhorizontal fold-hinge surface axes of the asymmetric F3 folds (Figs. 2C, 3, and 7B). Fault-propagation folds (F4) are observed at many D4 top-WNW thrust fault tips and are commonly associated with decimeter- to meter-scale imbrication of S2 foliation (Fig. 7D). Top-WNW D4 deformation is observed in the northern cliff face of the Yukon River for at least 2–3 km west of JR-140. Approximately 8 km WNW of locality JR-140, at locality JR-142 (Figs. 2A and 2B), a large NW-verging monocline of orthogneiss is observed in the cliff face several hundred meters above the northern bank of the Yukon River. Metasedimentary rocks crop out at the base of this cliff, indicating that the YRSZ is located within this cliff face. The monocline is likely to be a macroscopic fault-propagation fold (F4) recording D4 top-WNW thrusting.

5.4. D5 Field-Structural Deformation Fabrics

A late-stage episode (D5) of steep, brittle normal faulting, striking approximately N-S to NW-SE, is evident across both units (Fig. 3). Small brittle fractures produce decimeter offsets that cut the S2 foliation, F3 asymmetric folds, and D4 thrust faults. Larger subvertical D5 faults are also present with 20–50-cm-wide fault gouge and breccia zones. It is not clear whether these faults accommodated any strike-slip motion, but strike-slip faulting in similar orientations has been previously recorded in the region (Ryan et al., 2013a, 2013b; Colpron et al., 2016a). At locality JR-140, the YRSZ fault contact between the orthogneiss and quartzite units is offset by one of these subvertical D5 faults by an unknown distance, but we note that there is no difference in the appearance and style of deformation immediately on either side of this steep fault (Fig. 3).

6. PETROGRAPHY AND DEFORMATION MICROSTRUCTURES

Five orientated samples from the quartzite and schist unit and three orientated samples from the orthogneiss unit at locality JR-140 were analyzed with polarizing light microscopy in the XZ and XY planes of the kinematic reference frame with respect to the S2-L2 deformation fabric (Figs. 8A–8E) (X, parallel to L2 lineation; Z, perpendicular to S2 foliation; Y parallel to S2 foliation and perpendicular to L2 lineation). Microstructural observations are summarized in Table 1. Crosscutting steep normal faults with unknown offsets prevent calculations of vertical structural distances between samples. Horizontal distances between samples serve as a proxy for relative structural distances between the S2 foliation planes from which the samples were derived. These horizontal distances are given in Table 1, relative to the YRSZ fault contact position.

6.1. Quartzite and Semipelitic Schist Samples (Footwall to the YRSZ Fault Contact)

Quartzite and semipelitic schist contain variable proportions of quartz + biotite + muscovite ± zoisite ± garnet ± kyanite/sillimanite ± staurolite ± chlorite ± sericite ± calcite. Quartz microstructures vary with proximity to the YRSZ fault contact. Mean quartz recrystallized grain size decreases toward the YRSZ fault contact, from 250 µm in sample YR001A1 (Fig. 8A and Table 2) to 115 µm in sample YR005A2 (Fig. 8D and Table 2). Quartz grains are subhedral to anhedral with amoeboid to polygonal grain boundaries (Figs. 8A and 8D). Quartz grain shapes are generally elongate, with long axes aligned parallel to S2 foliation and L2 lineation in all samples except YR005A1 and YR005A2 in which quartz grain long axes are slightly inclined with respect to S2, defining an SPO consistent with a top-ESE shear sense (D2) (Fig. 8D). Grain-boundary pinning of quartz growth by phyllosilicates is common, particularly in semipelitic schist samples YR003A1 and YR008A1. Amoeboid and serrated grain-boundary morphologies are prevalent in all samples and are consistent with microstructures produced by grain-boundary migration (GBM) dynamic recrystallization and subordinate grain-boundary bulging (BLG) dynamic recrystallization of quartz (Figs. 8A and 8D). In addition to GBM microstructures, partially annealed quartz subgrain microstructures are commonly observed closest to the YRSZ fault contact in samples YR005A1, YR005A2 (Fig. 3B), and YR008A1 (Fig. 3), indicative of subgrain rotation (SGR) dynamic recrystallization. Polygonization of some quartz grain boundaries in quartz-rich domains indicates that these samples have been partially annealed during static recrystallization.

Phyllosilicates, zoisite, kyanite, and sillimanite are present as elongate to acicular laths with long axes aligned parallel to S2 foliation and L2 lineation (Figs. 8A−8C). Zoisite grains appear to be boudinaged with an extensional direction in the plane of the S2 foliation, parallel to L2 lineation. Rarely, zoisite appears to have partially replaced feldspar. Some phyllosilicate-rich domains display a weak S-C fabric (D2) suggestive of a top-ESE sense of shear (Fig. 8D). In sample YR001A1, kyanite, zoisite, and calcite ± staurolite display intergrowth microstructures in which kyanite and calcite ± staurolite appear to grow with or replace zoisite (Figs. 8B and 8C). We note that this unusual mineral assemblage has been associated previously with hydrous eclogite-facies metamorphism (e.g., Winkler, 1979; Franz and Selverstone, 1992; Deer et al., 1997) and will be considered further in our discussion. Zoisite contains inclusions of quartz. Some kyanite grains are pseudomorphed by sillimanite, which forms part of a post-peak pressure assemblage. Chlorite is present in most samples as a minor phase along with sericite and calcite, apparently associated with the breakdown of garnet, mica, and feldspar.

Garnet is present in some samples from the quartzite and schist unit. In semipelitic schist sample YR008A1, garnet grains are 250–1000 µm in diameter, subhedral to anhedral, and strongly altered, sometimes entirely replaced by phyllosilicates. In quartzite sample YR005A1, small, rounded, subhedral garnet grains (125–400 µm) are located in S2 foliation-parallel garnet-rich mineral bands (Fig. 8E). These grains have elliptical grain shapes with long axes aligned subparallel to S2 foliation (Fig. 8E), and many of them also have asymmetric biotite mantles (Fig. 8E). Garnet grain shape and orientation and biotite mantle asymmetry form delta-type sigmoids and strain shadows, which are consistent with a top-ESE sense of D2 shear (Fig. 8E).

Beyond locality JR-140, at outcrop exposures farther away from the YRSZ fault contact (Fig. 2D), deformation fabrics in quartzite and semipelitic schist (footwall to the YRSZ fault contact) are less well developed; mineral segregation is not as strong, and grain shapes are not as elongate or as strongly aligned as those observed at JR-140. Quartz deformation is still indicative of GBM dynamic recrystallization, and in one sample (JR-142A1), rare chessboard subgrain microstructures are observed. In quartzite sample JR-143, recrystallized quartz grains have a bimodal mean recrystallized grain size (180 µm and 1475 µm) and are segregated into foliation subparallel domains. Both recrystallized grain-size fractions are dominated by GBM dynamic recrystallization fabrics, and some grains have a weak grain-shape elongation, aligned subparallel to S2 foliation. Garnet grains are anhedral with no preferred alignment.

6.2. Orthogneiss Samples (Hanging Wall to the YRSZ Fault Contact)

The orthogneiss contains variable proportions of K-feldspar + quartz + plagioclase + muscovite + allanite + epidote + sericite + chlorite + biotite + stilpnomelane ± calcite. Feldspar grains are subhedral to anhedral with elongate grain shapes aligned parallel to S2 foliation and L2 lineation. Amoeboid feldspar grain boundaries are indicative of GBM dynamic recrystallization (Fig. 8F). Pericline and albite twinning characteristic of microcline is commonly displayed by K-feldspar. Granophyric intergrowths of plagioclase and quartz are common, and myrmekite is locally observed. Perthitic intergrowths of plagioclase and K-feldspar are also observed. Where present, large K-feldspar augens have been recrystallized to form aggregates of smaller K-feldspar grains with an asymmetric and elongate grain-shape orientation suggestive of a top-ESE shear sense (D2). Sericitization of feldspars is common; growth of calcite aggregates appears to be associated with this sericitization.

Quartz grains are subhedral to euhedral with amoeboid and occasionally serrated grain-boundary morphologies indicative of GBM and subordinate BLG dynamic recrystallization (Fig. 8F). In addition to GBM microstructures, domains of quartz grains in sample YR007A1 commonly display subgrain microstructures indicative of SGR dynamic recrystallization. Quartz grain shapes are generally elongate and are aligned parallel to S2 foliation and L2 lineation. Partial annealing of quartz grains during static recrystallization is also evident.

Muscovite is typically present as isolated, anhedral grains surrounded by quartz and feldspar. Biotite and chlorite are more commonly present in discontinuous S2 foliation subparallel bands of phyllosilicate and allanite-epidote intergrowths that bound quartz-feldspar microlithons. In some cases, these phyllosilicate bands delineate mm-scale elliptical aggregates of K-feldspar, which are likely to be the recrystallized remnants of mylonitized K-feldspar augens. Allanite grains are euhedral with concentrically zoned cores rimmed by epidote-clinozoisite overgrowths. Allanite with epidote overgrowths are elongate and are aligned subparallel to S2 foliation and L2 lineation; where inclined to the foliation, grain-shape orientation and asymmetry, along with chlorite mantle asymmetries, suggest a top-ESE shear sense (D2). Euhedral epidote included within plagioclase is also observed.

6.3. Quartz c-Axis Crystallographic Preferred Orientations

Quartz c-axis fabrics were measured in four quartzite samples and one semipelitic schist sample from locality JR-140 (Fig. 9) using a G60+ automated fabric analyzer (e.g., Larson et al., 2017; Larson, 2018) at a 10 μm optical resolution at the University of British Columbia, Okanagan. Similar instruments have been demonstrated to yield quartz c-axis fabrics indistinguishable from those obtained through other methods such as electron backscatter diffraction (Wilson et al., 2007; Peternell et al., 2010). Quartz c-axis fabrics in samples of orthogneiss were also analyzed; however, these samples yielded random c-axis fabrics and are not presented. Quartz c-axis fabrics from the quartzite and semipelitic schist are summarized in Table 1.

Two sets of pole figures are plotted in the XZ plane of the kinematic reference frame (Fig. 9), with respect to D2 deformation fabrics (X parallel to L2 lineation, Z perpendicular to S2 foliation). In all pole figures, the X direction of the kinematic reference frame equates roughly to the ESE direction of the geographic reference frame, whereas the Z direction of the kinematic reference equates to the upward-vertical direction in the geographic reference frame.

The first set of pole figures (Figs. 9A–9E) are plotted with the same contour scale (multiples of uniform density [m.u.d]) to highlight relative changes in topology and density of c-axis distributions between samples. Fabric strength is measured from this first set of pole figures using the eigenvalue-based intensity parameter, I (Lisle, 1985), 
graphic
where S1, S2, and S1 are the normalized eigenvalues of the c-axis orientation data. I ranges from 0 to 5 for point maxima distributions and 0–3.75 for girdle distributions.

The second set of pole figures (Figs. 9F–9K) are plotted with sample-specific m.u.d. contour scales to better define c-axis crossed-girdle fabrics. Poorly defined crossed-girdle limbs were strengthened with an additional subset of c-axis orientation data (YR005A2, YR005A1, YR003A1, and YR001A1; Figs. 9F–9I). For a given sample, additional data were measured at random from a subset of grains with c-axis azimuths that lay within a specified pole figure quadrant. For example, to strengthen a crossed girdle with a weak upper-left limb, additional c-axis measurements were taken at random from a population of grains with c-axis azimuth orientations of 270°–360°. These strengthened crossed girdles provide more accurate measurements of crossed-girdle opening angles, which are used for deformation thermometry (e.g., Kruhl, 1996; Law, 2014; Faleiros et al., 2016). Additionally, c-axis orientation data from sample JR-143 are split into small and large grain-size fractions that are plotted separately for investigation of c-axis crossed-girdle fabrics (Figs. 9J and 9K).

All samples yielded organized c-axis fabrics (Fig. 9 and Table 1), typical of quartz mylonites (e.g., Barth et al., 2010; Parsons et al., 2016). From the structurally lowest to highest c-axis data sets (i.e., upsection toward the YRSZ fault contact), quartz c-axis fabrics transition from symmetric type I crossed girdles (Figs. 9C–9E and 9H–9K) to point maxima distributed along an asymmetric single girdle with a top-ESE shear sense (Figs. 9A, 9B, 9F, and 9G). Fabric intensity, I, increases upsection from 0.12 (Fig. 9E) to 1.05 (Fig. 9A), whereas c-axis crossed-girdle opening angles decrease upsection from 83° (Fig. 9J) to 64° (Fig. 9F). The c-axis fabric from the semipelitic schist sample YR003A1 is an exception to these trends (Figs. 9C and 9H), with a weaker fabric intensity than the underlying sample and a smaller fabric opening angle than the overlying sample, probably due to grain-boundary pinning of quartz grains by surrounding mica grains. Small and large grain-size fractions from sample JR-143 have fabric opening angles of 75° and 83°, respectively (Figs. 9J and 9K). Based on the quartz c-axis opening angle thermometer of Faleiros et al. (2016), recorded changes in opening angle equate to a reduction in quartz deformation temperature from 621 ± 50 °C to 490 ± 50 °C toward the YRSZ fault contact.

6.4. Microstructure-Derived Deformation Constraints for the Yukon River Shear Zone

6.4.1. Kinematic Constraints

Microstructures observed in samples from the quartzite and schist unit and orthogneiss unit record deformation with a top-ESE shear sense. Parallelism of these microstructural fabrics with S2 foliation and L2 lineation recorded at locality JR-140 and neighboring outcrops indicates that they correspond to top-ESE shearing on the YRSZ, during D2 deformation.

Within the quartzite and schist unit, samples collected close to the YRSZ fault contact display: (1) S-C fabrics defined by phyllosilicates (samples YR008A1, YR003A1, and JR-142); (2) alignment of quartz grain-shape long axes inclined to S2 foliation; (3) asymmetric biotite mantles on garnet cores; and (4) garnet grain-shape anisotropy and alignment (samples YR005A1 and YR005A2). These are all indicative of non-coaxial plane strain with a top-ESE shear sense (Figs. 8D and 8E). Similarly, asymmetric K-feldspar augens in the orthogneiss closest to the YRSZ fault contact (YR007A1) record non-coaxial deformation with a top-ESE sense of shear (Fig. 5C).

Samples farthest away from the YRSZ fault contact in the quartzite and schist unit, display quartz, zoisite, kyanite, and phyllosilicate grain-shape long-axis alignment and boudinage stretching directions parallel to S2 foliation and L2 lineation. These structures are indicative of coaxial plane strain. In the orthogneiss unit, farther away from the YRSZ fault contact, K-feldspar augen asymmetries show a mix of top-ESE and top-WNW shear senses, which is consistent with coaxial deformation.

The spatial transition in quartz c-axis fabrics, from (1) symmetric type I crossed girdles (Figs. 9C–9E) to (2) point maxima distributed along an asymmetric single girdle (Figs. 9A and 9B), is consistent with a transition from coaxial plane strain to non-coaxial plane strain (e.g., Lister et al., 1978; Behrmann and Platt, 1982) with increasing proximity to the YRSZ fault contact (Table 1). Where asymmetries are present, these c-axis fabrics display a top-ESE shear sense (Figs. 9A and 9B). This transition is coincident with an increase in fabric intensity, I, (Table 1) which reflects a positive strain gradient toward the YRSZ fault contact (e.g., Law, 1986; Heilbronner and Tullis, 2006; Morales et al., 2014; Larson et al., 2017; Hunter et al., 2018; Larson, 2018).

6.4.2. Temperature and Pressure Constraints

Petrological and microstructural constraints from locality JR-140 and neighboring outcrops indicate that D2 top-ESE shearing on the YRSZ initiated at or close to eclogite- to amphibolite-facies peak metamorphism. The occurrence of zoisite with kyanite and zoisite with inclusions of quartz, as observed in sample YR001A1, has been associated with hydrous eclogite-facies metamorphism where zoisite + kyanite + quartz form from the breakdown of lawsonite or anorthite at pressures and temperatures exceeding 9 kbar and 620–650 °C (e.g., Winkler, 1979; Franz and Selverstone, 1992; Deer et al., 1997). This is consistent with existing thermobarometric constraints from nearby metasedimentary and metaigneous rocks in the Stewart River area and northern Stevenson Ridge area, which yielded peak metamorphic conditions of 600–665 °C and 7.6–9.8 kbar (Berman et al., 2007; Morneau, 2017; Morneau et al., 2017).

Elongate kyanite-zoisite intergrowths aligned parallel to L2 stretching lineations (Fig. 8A) and delta-type garnet sigmoid grains (Fig. 8E) record D2 top-ESE shearing at or close to peak conditions. Kinematic constraints derived from quartz and feldspar microstructures and quartz c-axis fabrics indicate that these features also record D2 deformation. Feldspar microstructures in the orthogneiss unit display GBM dynamic recrystallization microstructures (Fig. 8F), suggestive of deformation temperatures of >550 °C (Fitz Gerald and Stünitz, 1993; Rosenberg and Stünitz, 2003). Quartz microstructures in these samples and others are dominated by GBM dynamic recrystallization microstructures (Figs. 8A, 8D, and 8F), which are indicative of deformation temperatures of >500 °C (Stipp et al., 2002a). In quartzite sample JR-142 (Fig. 2D), some quartz grains contain chess board subgrains, suggestive of prism-[c] slip at temperatures of >600 °C (Kruhl, 1996; Wallis et al., 2017). These constraints are consistent with quartz c-axis crossed-girdle fabrics (samples YR001A1 and JR-143, Figs. 9D and 9E) that are indicative of mixed <a> slip with a strong component of prism <a> slip (e.g., Barth et al., 2010; Parsons et al., 2016), which is typically suggestive of quartz deformation temperatures of ≥500 °C (Baëta and Ashbee, 1969; Mainprice et al., 1986; Schmid and Casey, 1986). Crossed-girdle opening angle thermometry (Faleiros et al., 2016, their Equation 1) from these samples yields quartz deformation temperatures of 552 ± 50 °C, 566 ± 50 °C, and 621 ± 50 °C (Table 2 and Figs. 9H and 9I). The formation of hydrous metamorphic phases (e.g., zoisite) during D2 deformation suggests that dynamic recrystallization of quartz and feldspar could have occurred under hydrous conditions. If this assertion is correct, then the absolute temperatures of quartz and feldspar D2 deformation may fall toward the lower bounds of our estimated temperature ranges if hydrolytic weakening occurred (e.g., Kruhl, 1996; Kohlstedt, 2006; Faleiros et al., 2016).

In samples closest to the YRSZ fault contact (YR005A1, YR005A2, YR008A1, YR006A1, and YR006A2), quartz SGR microstructures occur in addition to quartz GBM microstructures and probably represent a later stage overprint of the former on the latter, during continued D2 deformation at lower temperatures of ∼400–500 °C (Stipp et al., 2002a). This is consistent with quartz c-axis point maxima and crossed-girdle fabrics (samples YR003A1, YR005A1, and YR005A2; Figs. 9A–9C) that are indicative of mixed <a> slip with a dominance of rhomb <a> slip (e.g., Barth et al. 2010; Parsons et al., 2016), which is typically suggestive of quartz deformation temperatures of ∼400–500 °C (Baëta and Ashbee, 1969; Mainprice et al., 1986; Schmid and Casey, 1986). Crossed-girdle opening angle thermometry from these samples (Faleiros et al., 2016, their Equation 1) yields quartz deformation temperatures of 490 ± 50 °C and 476 ± 50 °C (Table 2 and Figs. 9F and 9G). Collectively, the spatial distribution of these deformation temperature constraints indicates that D2 top-ESE shearing on the YRSZ became progressively localized around the YRSZ fault contact at locality JR-140, as temperature decreased from eclogite-amphibolite–facies peak conditions to ∼540–440 °C.

6.4.3. Flow-Stress and Strain-Rate Estimates (Quartzite and Schist Unit)

Mean recrystallized quartz grain sizes may be used to provide an estimate of differential stress during dynamic recrystallization through application of the quartz recrystallized grain-size piezometer (Stipp and Tullis, 2003). Previous work has demonstrated that this piezometer is robust for small recrystallized grains (typically <100 µm) produced via BLG and SGR dynamic recrystallization, but it underestimates differential stresses from larger grains formed via GBM dynamic recrystallization (e.g., Stipp and Tullis, 2003; Stipp et al., 2010; Cross et al., 2017). In our study of quartzite samples from the YRSZ (YR003A1 omitted due to effects of grain-boundary pinning), mean recrystallized quartz grain size decreases toward the YRSZ fault contact at locality JR-140 (Table 2). Farthest from the YRSZ fault contact, large and small grain-size fractions in sample JR-143 have mean recrystallized grain sizes of 1477 µm and 180 µm, respectively. At locality JR-140, mean recrystallized grain size decreases from 251 µm (YR001A1) to 115 µm (YR005A2) with proximity to the YRSZ fault contact. Using the quartz recrystallized grain-size piezometer (Stipp and Tullis, 2003), these grain sizes yield differential stress estimates of 2.0 MPa and 10.9 MPa (large and small grain-size fractions in JR-143), to 8.3 MPa (YR001A1), to 15.5 MPa (YR005A2) (Table 2). These values are taken as minimum values and are probably underestimates due to the limitations of the grain-size piezometer for large grains formed via SGR and GBM dynamic recrystallization (e.g., Stipp and Tullis, 2003; Stipp et al., 2010). Nonetheless, the spatial variation in recrystallized grain size between these samples can be reasonably interpreted as a relative increase in differential stress with proximity to the footwall–hanging-wall contact.

Using the theoretically derived quartz flow laws of Paterson and Luan (1990) and Hirth et al. (2001) (as recommended by Stipp et al., 2002b), changes in quartz deformation temperature and differential stress estimates correspond to an increase in strain rate with proximity to the YRSZ fault contact from 10−15 s−1 to 10−13 s−1 and 10−18 s−1 to 10−16 s−1, respectively (Table 2). Strain-rate estimates from the flow law of Hirth et al. (2001) are unrealistically slow, probably due to the underestimation of differential stress values. However, the relative trends in differential stress and strain-rate estimates are still significant. The increase in strain-rate estimates by two orders of magnitude (Table 2), together with the increase in quartz c-axis fabric intensity (Table 1), is indicative of a positive strain gradient toward the footwall–hanging-wall contact.

6.5. Summary and Interpretation of Microstructure-Derived Deformation Constraints

Quartz c-axis–derived deformation temperature and strain geometry constraints are consistent with petrography-derived deformation constraints described in the previous section and indicate that D2 top-ESE shearing in the YRSZ initiated at, or close to, amphibolite- to eclogite-facies peak metamorphic conditions during pervasive coaxial plane strain and continued through decreasing ambient temperatures down to temperatures of ∼540–440 °C (Figs. 9F–9K and Tables 1 and 2). During this reduction in temperature, non-coaxial plane strain deformation localized at the YRSZ fault contact. Spatial variations in quartz c-axis fabric intensity and flow-stress and strain-rate estimates from the quartzite and schist unit record a positive strain gradient toward the YRSZ fault contact, consistent with top-ESE shearing in the YRSZ evolving from pervasive low strain-rate coaxial deformation to localized high strain-rate non-coaxial deformation as ambient temperatures decreased. The temperature range of ∼540–440 °C for this lower-temperature deformation is based on the ±50 °C error on the c-axis opening angle thermometer, which, in part, accounts for possible variations in strain rate between samples. As such, our interpretation of increasing relative differential stress and strain rate localized to the YRSZ fault contact during decreasing ambient temperatures appears to be valid, although we are unable to determine absolute values for the strain-rate increase or temperature decrease. Spatial variations in feldspar augen morphology in the orthogneiss unit are also consistent with a positive strain gradient toward the YRSZ fault contact. A subordinate retrograde metamorphic assemblage of subhedral to euhedral sillimanite, sericite, stilpnomelane, and chlorite, observed in samples of quartzite and semipelitic schist, and allanite with epidote overgrowths in samples of orthogneiss probably formed during the waning stages of D2 deformation and may indicate that the YRSZ exhumed on a clockwise pressure-temperature path during this deformation event, with pressure decreasing at a faster rate than temperature.

7. THERMOKINEMATIC EVOLUTION OF THE YUKON RIVER SHEAR ZONE

Field and microstructural observations from outcrops along the Yukon River suggest five phases of deformation (D1 to D5) affected this region. Two of these deformation events correspond to motion along the YRSZ (D2 and D4). Synthesis and interpretation of these data are developed here, described from oldest to youngest.

7.1. D1: S1 Cryptic Foliation (Orthogneiss Unit)

The earliest deformation (D1) is a locally and rarely observed S1 foliation identified within the orthogneiss unit at localities JR-138 and JR-139, but not at JR-140. At these outcrops, the S1 fabric is folded into parallelism with S2 foliation (Fig. 5A). The S1 foliation may have also been the precursor fabric to the S2 spaced cleavage observed within the quartzite and schist unit at locality JR-140. Its regional extent and significance are unknown.

7.2. D2: Mid-Crustal Deformation and Top-ESE Shearing on the YRSZ

D2 deformation is characterized by pervasive ductile shearing (Figs. 5A–5H) and corresponds to the earliest recognized deformation along the YRSZ. This deformation formed a strong, mylonitic S to L-S fabric with a penetrative spaced to continuous cleavage (S2) subparallel to the YRSZ fault contact and an associated L2 mineral stretching lineation (Figs. 5A–5E). At localities JR-138 and JR-139, older fabrics (S1) and structures (e.g., aplite dikes) have been folded (F2) into parallelism with the S2 foliation. These F2 folds are tight to isoclinal, recumbent, and have axial planes orientated parallel to the S2 foliation (Figs. 5A and 5B). Field-structural and microstructural shear-sense indicators associated with D2 deformation record a top-ESE shear sense, corresponding to the earliest motion on the YRSZ (Figs. 5G, 5H, 8D, 8E, 9A, and 9B).

Grain-shape anisotropy and alignment of peak metamorphic assemblages (garnet + kyanite ± zoisite) and quartz deformation temperatures suggest that D2 top-ESE deformation on the YRSZ initiated at, or close to, peak metamorphic amphibolite- to eclogite-facies conditions (Fig. 8E) at temperatures of ≥650–500 °C. Quartz microstructures and c-axis fabrics from samples collected in close proximity to the YRSZ fault contact indicate that D2 top-ESE shearing localized and intensified around the fault contact as temperature decreased to ∼540–440 °C (Table 2 and Fig. 9F–9H). The absence of D2 fabrics with deformation temperatures lower than ∼540–440 °C suggests that D2 terminated at these conditions.

The outlined spatial trend in deformation temperature (i.e., decreasing deformation temperature toward the YRSZ fault contact) correlates with spatial trends in quartz flow-stress and strain-rate estimates (i.e., decreasing recrystallized quartz grain size toward the YRSZ fault contact), quartz c-axis fabric intensity (i.e., increasing strain magnitude toward the YRSZ fault contact), strain geometry (i.e., increased component of non-coaxial deformation toward the YRSZ fault contact), and feldspar augen morphology (Tables 1 and 2). With increasing proximity toward the YRSZ fault contact at locality JR-140, these constraints for D2 deformation record a transition from pervasive, coaxial plane strain at relatively low strain rate and higher temperatures, to localized non-coaxial plane strain at relatively high strain rate and lower temperatures. An increase in the strength of non-coaxial deformation (relative to coaxial deformation) toward the core of a shear zone is observed in other natural shear zones and is a common prediction of models of strain localization across a shear zone (e.g., Platt and Behrmann, 1986; Law, 1987, 1990; Schmid and Casey,1986; Mancktelow and Pavlis, 1994; Parsons et al., 2016; Aravadinou and Xypolias, 2017). This progression of D2 deformation associated with shearing on the YRSZ is typical of a narrowing shear zone that localizes and intensifies deformation as the ambient temperature decreases (e.g., Sibson, 1977; Ramsay, 1980; Handy et al., 2007; Faulkner et al., 2010).

7.3. D3: Regional Folding during NE-SW Compression

In the YRSZ footwall, the S2 foliation is subsequently deformed by asymmetric F3 folds (Figs. 3 and 7A) with NW-SE–striking axial planes (Fig. 6), which control the outcrop- and map-scale orientation of S2 foliation (Figs. 2B, 3, and 7A). F3 fold axes are subparallel to the L2 lineation (Fig. 6). This asymmetric F3 folding is not observed in the orthogneiss; however, regional mapping to the north of locality JR-140 (e.g., Ryan et al., 2013a, 2013b) suggests that the large body of orthogneiss is folded into a regional-scale, NW-SE–trending synform (Figs. 2A and 2B). We therefore interpret these asymmetric F3 folds within the footwall of the YRSZ (quartzite and schist unit) at locality JR-140 as parasitic folds on the southern limb of a regional-scale F3 synform (Figs. 2A and 2B). This F3 fold formed after the D2 top-ESE deformation event on the YRSZ and folded the YRSZ and associated S2 foliation and shear fabrics during ∼NE-SW–directed compression (Fig. 2B). It is thus likely that the YRSZ forms the base of this folded orthogneiss body (Figs. 2A–2C).

7.4. D4+D5: Top-WNW Upper-Crustal Thrust Faulting and Fault-Propagation Folding

In the quartzite and schist unit at locality JR-140 and in the cliff sections for at least 2–3 km NW of JR-140, S2 foliation and asymmetric F3 folds are deformed by WNW-verging D4 thrust faults and associated F4 thrust-fault propagation folds, indicative of non-coaxial contractional deformation (Figs. 3 and 7B–7D). In the quartzite and schist unit, D4 thrust faults commonly propagate along subhorizontal S2 carbonaceous schist horizons and are often coincident with the hinges of asymmetric F3 folds (Figs. 2C, 3, 7B, and 7C). Some D4 thrust faults cut across S2 foliation in the quartzite (Fig. 7C), whereas others form localized meter-scale F4 duplexes that tightly fold S2 foliation (Figs. 2C, 7B, and 7D). Brittle top-WNW thrust faulting (D4) is less commonly observed in the orthogneiss unit. A large NW-verging orthogneiss monocline observed at the top of a cliff section ∼8 km NW of locality JR-140 probably formed during this phase of deformation (F4). We suggest that the minor discordance between footwall and hanging-wall S2 foliation and L2 lineation orientations at locality JR-140 developed during D4 WNW thrust faulting and associated duplex development as a result of translation and rotation of thrust-bounded blocks (Fig. 6). Despite the apparent intensity of D4 deformation, it is not recorded by microstructural deformation fabrics. As such, D4 deformation must have occurred at or close to the brittle-ductile transition at temperatures <440 °C. We suggest that this indicates reactivation of the YRSZ as an upper-crustal top-WNW fault zone. Following D4 thrusting, the YRSZ appears to have remained inactive at upper-crustal levels. D1 to D4 fabrics are deformed by N-S to NW-SE–striking, subvertical, brittle faults and fractures (D5) that are regionally extensive and unrelated to the YRSZ (Fig. 3).

7.5. Timing Constraints and Tectonic Significance of the YRSZ: A Complex Record of Polyphase Deformation during the Structural Evolution of the Yukon-Tanana Terrane

Regionally, the oldest deformation recorded by the Simpson Range suite orthogneiss and Snowcap assemblage metasedimentary rocks occurred during the latest Devonian to Early Mississippian (Berman et al., 2007). This places a maximum age on the S1 fabric that we observed at localities JR-138 and JR-139 (Figs. 5A and 5B).

Within the orthogneiss unit at locality JR-140, D2 top-ESE shear fabrics and kinematic indicators are solid-state deformation features. Orthogneiss sample 16RAY-JR140A01 (Fig. 4A) from the hanging wall of the YRSZ fault contact at locality JR-140 yielded a mean crystallization age of 259 ± 2 Ma, and thus, D2 top-ESE shearing on the YRSZ must have initiated after this date (Fig. 10A). This may explain the occurrence of S1 foliation and F2 folded aplite dikes in orthogneiss at localities JR-138 and JR-139 and their absence in orthogneiss at locality JR-140; the former orthogneiss is Mississippian (Figs. 4C and 4D), and the latter orthogneiss is Permian (Fig. 4A).

D2 deformation initiated at or close to peak metamorphic amphibolite- to eclogite-facies conditions. Previous studies recorded amphibolite- to eclogite-facies metamorphism in YTT during the mid-Permian–Middle Triassic and the latest Triassic–Early Jurassic (Fig. 10A) (e.g., Berman et al., 2007; Petrie et al., 2016; Staples et al., 2016; Clark, 2017; Gilotti et al., 2017; Morneau et al., 2017). Minimum temperature estimates for D2 deformation (∼540–440 °C) are greater than or approximately equal to the closure temperature for Ar loss in muscovite (380–450 °C, Harrison et al., 2009). Local to the YRSZ in the Stevenson Ridge area, 40Ar/39Ar ages from both the hanging wall and the footwall (Joyce et al., 2015) indicate that the YRSZ exhumed through the Ar-closure temperatures for hornblende, muscovite, and biotite between 180 and 160 Ma, 176–168 Ma, and 172–162 Ma, respectively (Fig. 2A and Table DR2 [see footnote 1]). K-Ar ages from the region (Figs. 2A and 2B) are comparable to these 40Ar/39Ar ages (Wanless et al., 1966, 1978; Hunt and Roddick, 1992).

Together, the geochronometric constraints outlined above indicate that D2 deformation and top-ESE shearing on the YRSZ (Fig. 10A) initiated after 259 ± 2 Ma (Late Permian), either during Late Permian–Middle Triassic or latest Triassic–Early Jurassic metamorphism and ceased during or prior to regional cooling at ca. 176–168 Ma (Fig. 10A) (e.g., Joyce et al., 2015). We are unable to determine whether D2 top-ESE occurred during lithospheric extension or contraction, because the original S2 shear fabric has been subsequently folded (F3) during D3 deformation and may have been further reoriented during later deformation events. At ca. 262 Ma, the Buffalo Pitts peridotite was emplaced within metasedimentary rocks of the Snowcap assemblage (Canil et al., 2003; Johnston et al., 2007) ∼40 km SE of locality JR-140 (Fig. 2A). This peridotite is interpreted as an orogenic peridotite, exhumed and emplaced during hyperextension of continental lithosphere in a similar setting to the Red Sea rift (Johnston et al., 2007). As such, emplacement of the Buffalo Pitts orogenic peridotite at ca. 262 Ma suggests that YTT underwent lithospheric extension at that time, synchronous with emplacement of the Sulphur Creek suite (Johnston et al., 2007; Parsons et al., 2018; van Staal et al., 2018). If D2 deformation initiated during Late Permian–Middle Triassic metamorphism (e.g., Creaser et al., 1997; Villeneuve et al., 2003; Staples, 2014; Petrie et al., 2016; Gilotti et al., 2017), then the available constraints suggest that top-ESE shearing on the YRSZ occurred during lithospheric extension (Fig. 10A). Alternatively, if D2 deformation initiated during latest Triassic–Early Jurassic metamorphism, then current interpretations of this metamorphic event suggest that top-ESE shearing on the YRSZ initiated in response to convergence and crustal thickening of the YTT during collision and/or accretion with either the Stikine terrane (e.g., Colpron et al., 2015), the Insular terranes (e.g., Monger, 2014), Laurentia (e.g., Berman et al., 2007; Staples et al., 2016; Parsons et al., 2018), or another component of YTT (e.g., YTTj; van Staal et al., 2018).

The timing of F3 regional folding with NW-SE–trending axial traces during NE-SW–directed compression (D3) cannot be constrained to any greater precision than post-D2 (i.e., after 259 ± 2 Ma) and pre-D4 (Fig. 10B). Our findings demonstrate that D4 top-WNW thrusting along the YRSZ occurred at temperatures of <440 °C. D4 deformation must therefore have taken place during or after regional cooling through the Ar-closure temperature for muscovite between ca. 176–168 Ma (Fig. 10C) (e.g., Joyce et al., 2015).

Previously, Ryan et al. (2014) suggested that the YRSZ (then named the Yukon River Thrust) forms one of several structural breaks within YTT in Yukon that separate distinct crustal blocks. The implication of this hypothesis is that YTT is a composite terrane, built from smaller subterranes. Our findings are consistent with this hypothesis, but they do not validate it. The occurrence of Permian orthogneiss (Sulphur Creek suite) in the hanging wall of the YRSZ fault contact at locality JR-140 was unexpected given the original suggestion made by Ryan et al. (2014) that Permian plutonic rocks were confined to the YRSZ footwall. For this hypothesis to hold true, the Permian orthogneiss at JR-140 must be an entrained block within the YRSZ and another structural contact between Permian and Mississippian orthogneiss units structurally upsection of this locality. This implies that the YRSZ is much wider than the outcrop at locality JR-140 and forms a zone of pervasive ductile deformation with multiple high-strain zones localized along lithological contacts. This is a common feature of mid- to lower-crustal ductile shear zones (e.g., Sibson, 1977; Ramsay, 1980; Handy et al., 2007; Faulkner et al., 2010).

The occurrence of two distinct episodes of shearing on the YRSZ in opposite directions (D2 and D4) makes interpretation of the role of the YRSZ during tectonic evolution of YTT challenging. If the interpretations of Ryan et al. (2014) are correct, then structural separation of different plutonic suites is most easily explained by D4 top-WNW, post–Late Triassic thrust faulting along the YRSZ. Accounting for the structural separation of these plutons by D2 top-ESE shearing on YRSZ requires a complicated initial geometry and large amounts of subsequent tilting in order to ensure that the plutonic suites confined to the hanging wall could not have intruded the footwall. As such, D4 top-WNW contractional deformation on the YRSZ is probably congruent with the original interpretation of the Yukon River Thrust (Ryan et al., 2014) and occurred in response to significant post-Triassic collision and crustal thickening (based on local muscovite Ar/Ar ages and metamorphic constraints). This collisional event occurred either between distinct subterranes of a composite YTT (e.g., Ryan et al., 2014; van Staal et al., 2018), between YTT and the Stikine terrane (e.g., Colpron et al., 2015), between YTT and the Insular terrane(s) (e.g., Monger, 2014; Sigloch and Mihalynuk, 2017), or between YTT and Laurentia (e.g., Berman et al., 2007; Johnston, 2008; Hildebrand, 2009; Staples et al., 2016; Parsons et al., 2018).

Based on the orientation of deformation fabrics associated with D4 deformation and its prevalence along the Yukon River beneath the orthogneiss unit, we prefer to extrapolate the trace of the YRSZ north to northeastward around the F3 synform of the Simpson Range suite orthogneiss (Figs. 2A–2C), rather than continuing northwestward away from the contact between Snowcap assemblage and Simpson Range suite as drawn by Colpron et al. (2016a). Our preferred extrapolated trace of the YRSZ is drawn on Figure 2A in blue.

8. CONCLUSIONS

The Yukon River shear zone (YRSZ) in west-central Yukon cuts across the Yukon-Tanana terrane (YTT), juxtaposing siliciclastic metasedimentary rocks of the Snowcap assemblage and Permian Sulphur Creek orthogneiss in the footwall below Mississippian orthogneiss in the hanging wall. We have identified a portion of the YRSZ at a structural contact between Permian orthogneiss and quartzite at locality JR-140 on the northern bank of the Yukon River.

Petrography of metamorphic mineral assemblages indicates that the Yukon River shear zone initiated at or close to peak metamorphic amphibolite- to eclogite-facies conditions during D2 top-ESE ductile shearing. Microstructural analyses record a transition from coaxial plane strain to non-coaxial plane strain with increasing proximity to the YRSZ fault contact. This is coincident with a decrease in deformation temperatures from ≥650–500 °C to ∼540-440 °C, an increase in flow-stress and strain-rate estimates and quartz c-axis fabric intensity (proxy for strain magnitude), and an increase in the degree of flattening and elongation of feldspar augens. Kinematic indicators associated with D2 deformation consistently show a top-ESE shear sense and are observed more frequently and are better defined with increasing proximity to the YRSZ fault contact. Collectively, these spatial trends represent a classic example of a narrowing shear zone that progressively localizes and intensifies deformation as ambient temperature decreases (e.g., Sibson, 1977; Ramsay, 1980; Handy et al., 2007; Faulkner et al., 2010).

Following an episode of regional-scale folding (F3) during NE-SW–directed compression (D3), the YRSZ was reactivated as a brittle thrust fault zone during D4 top-WNW deformation. The absolute timing of ductile shearing (D2) and brittle reactivation (D4) of the YRSZ is poorly constrained. D2 top-ESE shearing on the YRSZ occurred after crystallization of orthogneiss at locality JR-140 at 259 ± 2 Ma and prior to regional cooling through the closure temperature for Ar loss in muscovite at ca. 176–168 Ma, either during Late Permian–Middle Triassic metamorphism and lithospheric extension or latest Triassic–Early Jurassic metamorphism and crustal thickening. D4 top-WNW thrusting on the YRSZ occurred during or after regional cooling at ca. 176–168 Ma.

A variety of different compressional and extensional events have been hypothesized for the Yukon-Tanana terrane, and it is unclear which of these events may have been responsible for D2 and D4 deformation on the YRSZ. Our findings are consistent with the hypothesis that the YRSZ represents a structural boundary between distinct subterranes within the Yukon-Tanana terrane; however, further investigation is required to validate this theory.

More broadly speaking, we note that the spatial trends in deformation temperature, strain geometry and intensity, grain size, and differential stress displayed by D2 deformation on the YRSZ represent an excellent example of progressive deformation on a narrowing shear zone during decreasing ambient temperatures. Our results are comparable to both natural and modeled examples of strain localization on crustal-scale shear zones and provide a valuable insight into how shear zones evolve through time under variable external conditions.

ACKNOWLEDGMENTS

This research was funded by Natural Resources Canada GEM-II Cordillera project and the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement 639003 “DEEP TIME”). We thank Science Editor Laurent Godin and reviewers Terry Pavlis and Paris Xypolias for thorough reviews that greatly improved the quality and impact of this research. Cees van Staal is thanked for scientific discussion relating to this manuscript and the Northern Cordillera. We gratefully acknowledge the assistance and expert guidance of the Science Laboratory Network staff (GSC-Ottawa) involved with U-Pb zircon geochronometry: Greg Case, Raymond Chung, and Ron Christie for U-Pb sample preparation; Pat Hunt for zircon SEM imaging; Ellie Knight for U-Pb analyses; and Nicole Rayner, Tom Pestaj, and Bill Davis for their expert guidance and advice in the SHRIMP II Laboratory.

1GSA Data Repository Item 2018357, Table DR1: Single grain U-Th-Pb concentrations, isotopic ratios and calculated apparent ages from SHRIMP II analyses of zircon in orthogneiss samples; File DR1: Description of analytical procedures for U-Pb zircon geochronology via Sensitive High Resolution Ion Micro Probe (SHRIMP II) and backscatter electron (BSE) and cathodoluminescence (CL) images of representative zircon grains; Table DR2: Compilation of published geochronology presented in Figure 2, is available at http://www.geosociety.org/datarepository/2018, or on request from editing@geosociety.org.
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.