Using discordant U-Pb zircon data to re-evaluate the El Paso terrane: Late Paleozoic tectonomagmatic evolution of east-central California (USA) and intense hydrothermal activity in the Jurassic Sierra Nevada arc

The Cordilleran margin of western North America is a tectonic amalgamation of translated parautochthonous and accreted exotic terranes, stitched to the rifted continental margin by a chain of latest Paleozoic to Cenozoic volcanic arcs. Understanding the geometry and constitution of the assembled terranes is required not only to reconstruct the tectonic history of the active margin but also to constrain possible upper-mantle source regions and crustal assimilation histories of subsequently intruded arc magmas. Reconstructing tectonic histories is complicated by thermal disturbances and intra- arc deformation accompanying Mesozoic magmatism, in addition to shuffling by Cenozoic tectonism. Key pieces of tectonic information may be obscured by intrusion of arc magmas along structural weaknesses, and even moderate grades of metamorphic recrystallization can obliterate the fossil record of depositional age and paleoenvironment. Moreover, intense hydrothermal alteration accompanying arc magmatism has the potential to alter zircon geochronologic records, one of the best tools for determining ages and provenance of suspect terranes (e.g., Nemchin and Cawood, 2005; Gehrels, 2014; Vermeesch, 2021).

The Permian to Late Cretaceous Sierra Nevada arc was emplaced into a complex metamorphic framework composed of lithologically distinct terranes. For example, the Kings-Kaweah ophiolite terrane is a compound Ordovician–Pennsylvanian ocean floor–serpentinite complex accreted along the arc’s southwestern edge (Saleeby, 2011); and in the central Sierra Nevada, interpreted protolith compositions identify the Snow Lake block as a parautochthonous terrane correlative with the Death Valley shallow-water facies of southwestern Laurentia’s Neoproterozoic to early Paleozoic passive margin (Schweickert and Lahren, 1990; Memeti et al., 2010). Though the translation histories of these crustal blocks are debated, general agreement ties the Snow Lake block to the local passive margin of southwestern Laurentia and assigns an exotic origin to the Foothills ophiolite belt (e.g., Schweickert, 2006; Chapman et al., 2015; Saleeby and Dunne, 2015).

In contrast, little agreement exists regarding the El Paso terrane, a crustal ribbon of poly-deformed metasedimentary rocks that extends ~200 km from the Kern Plateau of the Sierra Nevada southward into the northcentral Mojave Desert (Fig. 1A). General consensus exists that deepwater marine protoliths of El Paso terrane rocks are displaced relative to shallow-water, passive-margin facies boundaries projected southwestward from the Death Valley region across the Mojave Desert (Stewart, 1972). Various correlations have been proposed between the El Paso terrane and lithostratigraphic elements of the Antler orogeny present in central Nevada. (1) Lower Paleozoic rocks from the central and southern El Paso terrane may correlate to outer-shelf–inner-rise, passive-margin strata of southwestern Laurentia found in the lower plate of the Roberts Mountains thrust (Carr et al., 1984; Stevens et al., 2005). (2) Chert and stratiform barite found in the northern El Paso terrane may correlate to lower to middle Paleozoic, deep marine strata within the exotic Roberts Mountains allochthon that was thrust over the southwestern Laurentian passive margin during the Antler orogeny (Dunne and Suczek, 1991; Poole et al., 1992). (3) Upper Paleozoic clastic strata in the central El Paso terrane may correlate to strata deposited in the Antler foreland basin that formed by crustal flexure inboard of the Roberts Mountains allochthon (Poole, 1974). Detrital zircon data demonstrate that the northern El Paso terrane and Roberts Mountains allochthon strata share a similar provenance (Linde et al., 2016; Attia et al., 2018). However, the apparent lack of “Antler-type” deformation within the Kern Plateau metamorphic pendants has been used to argue against a direct correlation with the Roberts Mountains allochthon (Saleeby and Dunne, 2015).

In this paper, we present new detrital zircon geochronology from two Kern Plateau metamorphic pendants and single-crystal U-Pb-Hf isotope data from associated granitoid plutons deformed within the Kern Plateau shear zone. We model isotopic discordance to lift the Mesozoic metamorphic veil, revealing both an expanded provenance of the metasedimentary protoliths and the earliest age of arc plutonism, enabling reassessment of the ties that bind the El Paso terrane to western North America.

The El Paso terrane is composed of pre-Mesozoic marine sedimentary and metasedimentary rocks cropping out from the Kern Plateau near the eastern Sierra Nevada crest southeastwards toward locations south of the Garlock fault in the northwestern Mojave Desert. The terrane is named for the El Paso Mountains (Fig. 1A), which sit directly north of the Garlock fault. The primary basis for the El Paso terrane displacement origin hypothesis stems from the apparent contrasting depositional environments between these deepwater marine rocks and the neighboring shallow-water continental shelf strata present in eastern California (Burchfiel and Davis, 1981; Carr et al., 1981, 1984, 1997; Dunne and Suczek, 1991). Isotopic compositions of plutons intruded through deep-water strata form a secondary basis for proposing an exotic origin for the El Paso terrane: specifically, the presence of relatively primitive wholerock isotopic values is interpreted as recording Permo-Triassic magma emplacement through accreted oceanic crust on top of which El Paso strata were deposited (Kistler and Ross, 1990; Miller et al., 1995). Accretion scenarios for the El Paso terrane are incorporated into existing models of the late Paleozoic tectonic evolution of the southwestern Laurentian margin, either as docking within a zone of sinistral transpression (Burchfiel and Davis, 1981; Stone and Stevens, 1988; Walker, 1988; Martin and Walker, 1995; Greene et al., 1997; Saleeby and Dunne, 2015) or by thrusting of the El Paso terrane over the Laurentian margin (Stevens et al., 2005).

Detailed sedimentologic, structural, and geochronologic studies make the El Paso Mountains strata the best understood rocks of the entire El Paso terrane (Carr et al., 1984, 1997; Rains et al., 2012; Macdonald, 2016; Cecil et al., 2019). Published maps and structural investigations make the Kennedy Meadows and Bald Mountain metamorphic pendants (Fig. 1B) the best understood parts of the northern El Paso terrane (Dunne and Suczek, 1991; Saleeby and Dunne, 2015), despite relatively poor outcrop, dense vegetative cover, limited accessibility, and complex structural-lithologic relationships. In the following section, we highlight proposed lithologic, geochronologic, and structural correlations between the El Paso terrane and neighboring regions that pertain to the terrane’s potentially exotic origin.

Paleozoic marine strata of the El Paso Mountains can be loosely divided into four packages: (1) Upper Cambrian to Ordovician deep-water clastic sediments with interstratified chert and limestone (e.g., metasedimentary rocks of El Paso Peaks and Colorado Camp; Carr et al., 1984, 1997) correlated by Stevens et al. (2005) to the outer continental rise facies of Laurentia’s passive margin; (2) calcareous sandstones and sandy marble (e.g., metasedimentary rocks of Gerbracht Camp; Carr et al., 1997) possibly correlating to a Middle Devonian submarine-fan complex deposited across the White-Inyo Mountains and Mount Morrison block (e.g., Mount Morrison Sandstone; Stevens and Greene, 1999; Stevens and Pelley, 2006); (3) Mississippian to Pennsylvanian clastic strata (e.g., Benson Well, Apache Mine, and Robbers Mountain formations) having similar age and compositions as mixed clastic-carbonate rocks of the Keeler and Kearsarge Formations in the White-Inyo Mountains, the latter of which may correlate with Antler foreland basin and overlap strata (Poole, 1974; Carr et al., 1984, 1997; Miller and Heller, 1994; Stevens et al., 2001; Stevens and Peters, 2018; Gannon et al., 2021); and (4) dominantly clastic, Permian sediments capped by ca. 260 Ma andesites of the Goler Formation that were deposited in proximity to the nascent Sierra-Mojave arc (e.g., metasedimentary rocks of Holland Camp; Martin and Walker, 1995; Carr et al., 1997; Rains et al., 2012; Macdonald, 2016). Inboard of the nascent arc, locally derived carbonate and siliciclastic detritus was deposited in the Darwin and Lone Pine Basins, which developed in response to transpressional tectonics (Stevens et al., 2015; Lodes et al., 2020). Particularly relevant to this tectonic discussion is the regional juxtaposition of lower Paleozoic, relatively deep-water strata of the central and southern El Paso terrane against Neoproterozoic–lower Paleozoic, shallow-water, passive-margin strata, because the deposition of the Devonian and younger rocks in the El Paso terrane may have postdated terrane emplacement (Poole et al., 1992).

Zeolite- to greenschistfacies El Paso Mountains strata record a complex history of deformation, including tight to isoclinal, gently plunging folds with hinges cut by faults that are sub-parallel to bedding and axial planes. Carr et al. (1984) attribute this deformation to Permian subduction initiation. However, citing a significant coarsening in clastic strata coincident with the Devonian to Mississippian Antler orogeny, the presence of strata correlated to Antler foreland basin and overlap deposits, and numerous unconformities indicating mid-Paleozoic tectonism, these authors conclude: “No structures have been found in the El Paso Mountains that are unequivocally related to Antler-age deformation, but neither can an Antler age be dismissed for some structures in the range.”

The Bald Mountain, Kennedy Meadows, and Indian Wells pendants are the largest metasedimentary bodies of the Kern Plateau, with ~20 km separating the most southeasterly, Indian Wells pendant from the El Paso Mountains (Fig. 1). Based on mapping of the Bald Mountain and Kennedy Meadows pendants, Dunne and Suczek (1991) characterize the protoliths as ~90% argillite and sandy argillite, with lesser basalt, serpentinized peridotite, limestone, barite, conglomerate, and quartz arenite. Some rock types support deposition largely in deep-water, possibly fault-controlled, marine basins. With the exception of low-pressure greenschist strata located near the western margin of the Bald Mountain pendant, protoliths are thoroughly recrystallized under amphibolite-facies conditions, with higher-grade assemblages present within strands of the Kern Plateau shear zone (Fig. 1). Outcrops on the Kern Plateau are below tree line and located south of the maximum extent of the Pleistocene ice sheet contributing to a “middling quality of exposure” (Dunne and Suczek, 1991). Basic lithologic similarities exist between these Kern Plateau pendants and lesser studied pendants located both to the northwest (e.g., Hockett Peak and Rattlesnake pendants) and southeast (e.g., Indian Wells pendant), leading to their inclusion in the El Paso terrane.

Chert, barite, and basalt support a possible correlation between protoliths of the Kern Plateau pendants and deep-water strata of the Roberts Mountains allochthon (Dunne and Suczek, 1991). A potential correlation is also supported by similarity of detrital zircon age populations in three samples from the Bald Mountain and Kennedy Meadows pendants (Attia et al., 2018) with the detrital zircon age populations characterizing Roberts Mountains allochthon strata sourced far to the north near the Peace River Arch (Gehrels, 2000; Linde et al., 2016). However, the apparent lack of Antler-aged deformation features, including a basal thrust fault, is cited as contraindicating the long-proposed correlation of the Kern Plateau pendants to the Roberts Mountains allochthon (Saleeby and Dunne, 2015).

The identity of tilted, relatively undeformed greenschist strata near the western margin of the Bald Mountain pendant is also unclear. Stevens and Pelley (2006) speculate that calcareous sandstone and sandy limestone containing shale chips (Dunne and Suczek, 1991; location 11) may correlate with the Middle Devonian Mount Morrison Sandstone, distal deposits of a submarine-fan complex emanating from the White-Inyo Mountains and deposited across the Mount Morrison block located to the north of the El Paso terrane. Dunne and Suczek (1991) suggest the correlation of conglomeritic lithosomes present in both the Kennedy Meadows and Bald Mountain pendants with Early Permian conglomeritic and volcanic strata of the El Paso Mountains. Despite these suggestions, the depositional age(s) of greenschist-facies “overlap” strata in the Kern Plateau are unknown, and proposed correlations between the Kern Plateau pendants and both the El Paso Mountains and the Mount Morrison block are untested.

Weighted-mean U-Pb zircon ages identify the oldest pluton in the Sierra-Mojave arc as crystallizing at 273.7 ± 4.4 Ma within the El Paso Mountains (Cecil et al., 2019), consistent with an upper-intercept, multi-grain U-Pb zircon age (281 ± 8 Ma) derived from an andesite flow in the western part of the range (Martin and Walker, 1995) and with a 273 ± 2.4 Ma metatuff near the western tip of the Mojave Desert (Chapman et al., 2012). Based on these ages, Cecil et al. (2019) conclude that Sierra-Mojave arc magmatism initiated in the vicinity of the central El Paso terrane. Permian plutons display a gneissic fabric that parallels the axial plane foliation in metamorphic host rocks. Textural development intensifies inwards toward the contacts with metamorphic host rocks, grading into blastomylonite. Deformation of El Paso Mountains granitoids is restricted to 274–256 Ma plutons; younger plutons (258–238 Ma) are unfoliated and interpreted as intruding after a change to more oblique plate convergence within the nascent Mojave-Sierra subduction zone (Cecil et al., 2019).

Based on relatively primitive wholerock strontium-neodymium isotope ratios from Permo-Triassic plutons in the El Paso Mountains and the El Paso terrane south of the Garlock fault (Sr_i ~0.704; εNd_i = +2.3 to +5.4), Miller et al. (1995) propose that the El Paso terrane is rooted in oceanic crust through which Middle Permian to Early Triassic plutons were intruded. These authors propose an ensimatic El Paso terrane was thrust onto the passive margin of the Laurentian continent in the early Mesozoic (between 240 and 179 Ma) such that post-thrust, Middle Jurassic plutons intruded through both oceanic and continental lithosphere. However, single-grain zircon εHf data from plutons exposed in the El Paso Mountains clearly record interaction of Early Permian arc magmas with ancient continental crust (zircon εHf <−9; Cecil et al., 2019). Relatively primitive whole-rock Sr-Nd isotope values coupled with isotopically evolved zircon εHf values might be explained by development of the Permo-Triassic arc on oceanic-attenuated continental crust underlain by depleted mantle and overlain by passive-margin rise facies strata. Cecil et al. (2019) propose that such transitional oceanic-continental crust developed to the north along the rifted Laurentian margin and docked within the sinistral truncation zone prior to initiation of Permian subduction. These authors attribute a secular εHf trend defined by Permian zircons toward more primitive Triassic values to variations in the stress regime of the upper plate due to changing subduction angle: a 275–250 Ma compressional regime produced by orthogonal plate convergence during forced subduction initiation (Rains et al., 2012), which promoted crustal contamination as recorded by variably evolved zircon εHf ratios (−9 to +11), followed by a 245–238 Ma extensional regime produced by oblique subduction in which more primitive-, possibly depleted-mantle–derived magmas traversed the arc with lesser crustal interaction (zircon εHf = +6 to +13).

Previous work failed to identify Permian plutons within the Kern Plateau segment of the northern El Paso terrane. Instead, highly discordant, multi-grain thermal ionization mass spectrometry (TIMS) analyses of deformed granitoids within the Kern Plateau shear zone yield lower-intercept dates that were interpreted as indicating a range of Triassic crystallization ages impacted by significant 1.9 Ga inheritance (239–206 Ma; E3–E7; Saleeby and Dunne, 2015). We note that isotopic discordance has only been observed in plutons intruding strands of the Kern Plateau shear zone surrounding the Bald Mountain and Kennedy Meadows pendants. Plutons surrounding the more southeasterly Indian Wells pendant are concordant, having Early to Middle Triassic (248 Ma and 240 Ma) and Early to Middle Jurassic (182 and 171 Ma) crystallization ages (samples E1, E2, E8, and E9; Saleeby and Dunne, 2015). Multi-grain fractions from a nondeformed quartz diorite pluton that crosscuts deformed granitoids southeast of the Kennedy Meadows pendant define a concordant latest Jurassic crystallization age (148 Ma; sample E12; Saleeby and Dunne, 2015). Relatively undeformed Middle to Late Jurassic intrusions of similar age are found throughout the El Paso terrane (Miller and Glazner, 1995; Bartley et al., 2007; Cecil et al., 2019).

Anastomosing ~N35W-trending strands of the Kern Plateau shear zone surround the largest and best studied metasedimentary rocks of the Kern Plateau (Fig. 1). Most fault strands are marked by mylonitic granitoid rocks; other strands contain lenses of upper-amphibolite–to lower-granulite–facies rocks juxtaposed against lenses of greenschist-facies rocks (Saleeby and Dunne, 2015). Sense-of-shear indicators support a polyphase history of offset (e.g., sinistral, dextral, normal, and reverse offset; Saleeby and Dunne, 2015). Specifically, these authors interpret s-folds in metamorphic rocks as indicating late Paleozoic sinistral shear formed during southeastward translation and ultimate accretion of a parautochthonous El Paso tectonic ribbon; development of a pervasive, normalsense shear affecting both granitoids and metamorphic rocks is attributed to Mesozoic down-to-the-northeast extension.

In the El Paso Mountains, transitional hornblende-hornfels–to garnet-amphibolite–facies metamorphic rocks present within mylonitic zones adjacent to pluton contacts are interpreted as reflecting syntectonic intrusion of Permian plutons (Carr et al., 1984; Cecil et al., 2019). Alternatively, Permian plutons of the El Paso Mountains were emplaced into active shear zones having a prolonged history of deformation similar to that proposed for the Kern Plateau shear zone, with higher-grade metamorphic assemblages potentially reflecting pre-intrusion tectonism and lower-grade assemblages forming during initial arc magmatism.

Kistler (1990) depicts the Kern Plateau shear zone as part of an extensive intra-batholithic break, with the Kern Plateau pendants trapped as slivers along a major lithospheric boundary between North American and Panthalassan plates. This connection derives largely from the observation that the Walker Pass and Long Valley plutons in the Kern Plateau east of the shear zone yield initial strontium isotope ratios more primitive than plutons to the west of the Kern Plateau shear zone (Kistler and Ross, 1990), reversing the eastward trend of increasing continental affinity of plutons recognized across the arc (e.g., DePaolo, 1981; Nadin and Saleeby, 2008). Saleeby (2011) later clarified that the Foothills suture, located west of the western range divide, forms the eastern boundary of accreted Panthalassan lithosphere. Subsequent models portray the region between the Kings-Kaweah ophiolite terrane to the west and the truncated Laurentian passive margin (e.g., the Mount Morrison pendant) to the east as being occupied by a series of crustal ribbons built on Panthalassan lithosphere (e.g., the El Paso terrane, Kernville terrane, and Shoo Fly terrane; Saleeby and Dunne, 2015; Attia et al., 2018). Such models invoke ribbons of oceanic lithosphere being sliced off the southern end of an elongate Golconda–Slide Mountain basin (purportedly located northwest of the Eastern Klamath terrane; Fig. 1A) and subsequently accreted along sinistral transcurrent faults that developed subparallel to the main continental truncation structure. A major implication of these models is that much of the southern Sierra Nevada arc would have intruded through accreted oceanic lithosphere.

In this paper, we revisit the original hypothesis that the deep-water lithologies of the Kern Plateau pendants correlate with the upper, allochthonous plate of the Roberts Mountains thrust. Our sampling strategy was designed to search for hypothesized outcrops of a lower thrust plate, which would be recognized by the well-documented ca. 1.1–1.4–1.75 Ga detrital zircon age “triad” common in southwestern Laurentian passive-margin strata (Memeti et al., 2010; Gehrels and Pecha, 2014; Linde et al., 2014; Mahon et al., 2014; Yonkee et al., 2014; Chapman et al., 2015). Given published detrital zircon data consistent with correlation to the Roberts Mountain allochthon (e.g., a dominant 1.8 Ga population), identifying lower-plate passive-margin strata would provide clear structural evidence for correlation of the northern El Paso terrane with the truncated Roberts Mountains thrust. Moreover, doing so would change the orientation and context of the purported late Paleozoic sinistral transform system, from one transporting slices of oceanic lithosphere southeastward from the southern end of the hypothetical Golconda–Slide Mountain marine basin (Saleeby and Dunne, 2015) to one translating a slice of Laurentian continental lithosphere capped by the Roberts Mountains allochthon southward within the Southwestern Laurentian Borderland (SLaB; Lawton et al., 2017). As such, southward translation of the El Paso terrane would constitute one event occurring within the ~100- m.y.-long Antler-SLaB-Sonoma transpressional-transtensional orogeny (Chen and Clemens-Knott, 2021).

Successful correlation of the Kern Plateau pendants to the Roberts Mountains allochthon would have significant implications for Mesozoic arc magmagenesis. If the Roberts Mountains allochthon plus the structurally overlying Golconda allochthons were thrust over Laurentian passive-margin strata and underlying rifted continental lithosphere, then the mantle source and deep crustal column through which Sierran arc magmas intruded would have been similar to the lithospheric columns traversed by arc magmas, both to the northwest (in displaced passive-margin strata of the Snow Lake block) and to the east-northeast (by magmas intruded through in situ passive-margin strata). Hafnium isotope analyses presented here targeted previously dated plutons with the intent to assess the nature of the underlying lithosphere.

Is the El Paso terrane a slice of exotic oceanic lithosphere, with clasts transported southward from the Peace River Arch region by longshore currents that mixed with detritus from fringing arcs? Or, is the El Paso terrane a parautochthonous sliver of the truncated Roberts Mountains allochthon, together with its underlying Laurentian lithosphere, that was translated southward from central Nevada during Antler-SLaB-Sonoma orogenesis? Answers to these questions bear on the location of the southwestern edge of Laurentia, the late Paleozoic tectonic history of the Californian segment of the Laurentian margin, and the character of mantle lithosphere from which Mesozoic arc magmas of the southeastern Sierra Nevada arc were ultimately derived.

Four detrital zircon samples were collected from the Bald Mountain and Kennedy Meadows pendants, and four plutonic samples were collected from deformed granitoids within strands of the Kern Plateau shear zone that wrap around these metasedimentary pendants (Fig. 1B). Mapping by Dunne and Suczek (1991) guided targeted sampling of quartz-rich clastic metasediments; sampling of deformed granitoids for hafnium isotope analysis was guided by the geochronologic work of Saleeby and Dunne (2015). Standard petrographic analysis provided basic constraints on metamorphic grade of the detrital samples and assured minimal low-temperature alteration of the plutonic samples.

Zircons were separated from bulk-rock samples at California State University, Fullerton using traditional techniques that leverage the high-density and low-magnetic character of zircon. Anticipating low zircon abundance in deep-water protoliths, atypically large (~5 gallon) detrital zircon samples of clastic metasediments were collected (Fig. 1; File 1 in the Supplemental Material¹). Small quantities of tiny detrital zircons were separated from the amphibolite-facies samples (FR-211, FR-246, and FR-249; Fig. 2A), for which zircon size and abundance formed the limiting constraints on the number of analyzed zircons. Similar methods were used to separate the relatively more abundant zircons from the greenschist-facies sample FR-190 (Fig. 2B) and from the granitoid samples. Zircons were mounted, polished, and imaged by cathodoluminescence (igneous samples) and backscatter electron imaging (detrital samples). Images guided ablation pit placement within single growth zones away from inclusions and potentially xenocrystic cores; ablation was achieved using Photon Machines Analyte G2 Excimer lasers. U-Pb isotope data were collected using a Thermo Element2 single-collector inductively coupled plasma mass spectrometer (ICPMS; Gehrels et al., 2006, 2008); Lu-Hf isotopic data were collected on concordant plutonic zircons using a Nu Plasma multi-collector high-resolution ICPMS (Gehrels and Pecha, 2014). Analytical parameters for isotopic data collected at the Arizona Laserchron Center at the University of Arizona are detailed in Files 2 and 3 (see ^{footnote 1}).

New single-grain, laser ablation (LA)–ICPMS data from four plutonic samples collected at previously dated locations are generally concordant and were therefore processed in Isoplot using traditional 20% (5% reverse) relative-age discordance filters (Ludwig, 2008). In contrast, published multi-grain TIMS U-Pb zircon data from deformed plutons of the Kern Plateau document significant isotopic discordance (Saleeby and Dunne, 2015).

Due to significant isotopic discordance, new detrital zircon U-Pb data were first processed with a large (95%) relative-age discordance filter, such that only the most discordant analyses were rejected. Likely zircon crystallization ages were then identified by statistical discordance modeling of the detrital samples using R-based macros developed by Reimink et al. (2016). Zircons yielding U-Pb dates younger than modeled likely zircon crystallization ages were then excluded, unless they were identified as members of a >95% (and <101% reverse) relative-age concordant cluster defined by at least three zircon grains. Finally, the culled data were processed with a tight 5% (1% reverse) concordance distance filter using the IsoplotR routine of Vermeesch (2021) to produce kernel density estimate (KDE) plots. Visual data analysis is supported by plots generated using Isoplot, AgeCalcML, and IsoplotR (Ludwig, 2008; Pullen et al., 2018; Vermeesch, 2021). When referring to U-Pb calculations, we use “date” to refer to a number calculated from measured isotopic ratios and “age” to refer to U-Pb dates that we interpret to be both isotopically concordant and geologically meaningful. In places, we use “true” to emphasize an interpreted primary crystallization age undisturbed by lead loss.

U-Pb isotope data for four foliated quartz diorites to granites are plotted on Wetherill concordia diagrams (Fig. 3 and File 4, see ^{footnote 1}). Weighted-mean averages were calculated using IsoplotR after excluding 23 grains from a total of 125 igneous zircon analyses due to apparent lead loss or anomalous ²⁰⁶Pb/²⁰⁷Pb ratios. Calculated averages provide estimated pluton crystallization ages ranging from 273.8 to 249.4 Ma. When combined with concordant data from two plutons collected near the Indian Wells pendant located directly southeast of the Kennedy Meadows pendant (E1 [248 Ma] and E2 [240 Ma]; Saleeby and Dunne, 2015), the age of plutonism in the Kern Plateau (273.8–240 Ma) matches the range of Permian–Middle Triassic pluton ages from the El Paso Mountains (273.7–239.5 Ma; Cecil et al., 2019). Crystallization ages from throughout the El Paso terrane demonstrate that subduction initiation, as recorded by the footprint of the Permian arc, occurred over an ~200-km-long region extending from the El Paso Mountains northwestward into the Kern Plateau. Plutons younger than 258 Ma from the El Paso Mountains are undeformed; in contrast, Permian deformation of Kern Plateau plutons continued at least to ca. 250 Ma (sample FR-188; File 1, ^{footnote 1}).

Highly discordant, multi-grain zircon fractions of samples previously collected from the same four outcrops in the Kern Plateau were interpreted as Triassic in age, with significant discordance generated by inheritance of a 1.9 Ga xenocrystic fraction (Saleeby and Dunne, 2015; File 1, ^{footnote 1}). Based on extensive discordance within our detrital samples, we propose that the published multi-grain TIMS data from deformed plutons collected near the Bald Mountain and Kennedy Meadows pendants are instead discordant, largely due to lead loss, which shifts isotopic data to the right off Tera-Wasserburg concordia curves to generate artificially younger, Triassic dates (e.g., Saleeby et al., 1987). If correct, the small ablated zircon volume and the use of pre-analysis imaging to guide spot selection effectively excluded most discordant plutonic zircons from LA-ICPMS analysis.

Zircon hafnium isotopic values range from +9.7 to −11.4 (File 5, ^{footnote 1}), spanning a range from near-primitive mantle (εHf ₂₄₀ = ~+15) to highly radiogenic values consistent with incorporation of a significant component of ancient continental crust, which is expected to have strongly negative εHf values. Individual plutons exhibit ranges in εHf values from 5.4 up to 18.3 epsilon units despite relatively close clustering of their ²⁰⁶Pb/²³⁸ U zircon ages (Fig. 3). U-Pb and paired εHf data from the four plutons together define a trend of increasing εHf with time that suggests progressively larger additions of juvenile, mantle-derived material as the Permian arc matured. This trend completely overlaps the zircon U-Pb versus εHf field defined by Permo-Triassic plutons from the El Paso Mountains (Fig. 4), with sample FR-243 extending the secular trend observed by Cecil et al. (2019) to slightly more evolved compositions at arc initiation.

Detrital zircon samples from the Kern Plateau pendants display a wide range of U-Pb isotopic discordance (Fig. 5; Files 6 and 7, ^{footnote 1}). The most concordant sample is FR-190, a relatively undeformed, sub–greenschist-facies, sandy meta-siltstone collected near the western margin of the Bald Mountain pendant (Figs. 1 and 2B). Lead-loss trajectories (solid arrows) project to upper intercepts of ca. 1.1 Ga, 1.4 Ga, and 1.75 Ga (Fig. 5A), an age “triad” that characterizes a detrital zircon chronofacies widely recognized in southwestern Laurentian strata: a 1.2–1.0 Ga population derived from the Grenville province, with 1.48–1.34 and 1.78–1.6 Ga populations derived from Mesoproterozoic midcontinental plutons intruded through the Paleoproterozoic Mojave-Yavapai-Mazatzal provinces of southwestern Laurentia (Gehrels and Pecha, 2014). In contrast, lead-loss trajectories are not easily recognizable in U-Pb data from a quartzite from the Kennedy Meadows pendant (FR-249), but some discordant grains yielding Paleozoic U-Pb dates align on a crude trajectory having a ca. 1.75 Ga upper intercept (red arrow), which projects into a discordant cloud of zircons yielding Mesozoic dates (Fig. 5A inset).

Discrete lead-loss trajectories are also not readily apparent for amphibolite-facies samples FR-246 (a calcsilicate rock from the Bald Mountain pendant) and FR-211 (a quartzite from the Kennedy Meadows pendant). Instead, discordant zircons form scattershot distributions relative to concordia (Fig. 5B). Sixtyeight percent of FR-211 detrital zircons pass a 20% relative-age discordance filter (orange field), and these zircons yield dates spanning a continuous range between 1.9 Ga and 180 Ma. A schematic lead-loss trajectory for 1.75 Ga zircons bounds the majority of FR-246 zircons, 49% of which pass a 20% relative-age discordance filter. Note that relative-age discordance filters are typically not applied to grains younger than ca. 700 Ma by convention due to the limited curvature of this section of concordia and relative imprecision of ²⁰⁷Pb data (Spencer et al., 2016). Subhorizontal lead-loss trajectories inferred from highly discordant Paleozoic and Mesozoic ellipses (e.g., Fig. 5B inset) may reflect analyses with high ²⁰⁷Pb/²³⁵U errors or for which the common lead correction was inappropriate (Nemchin and Cawood, 2005; Spencer et al., 2016).

Previously published detrital zircon data from the Bald Mountain and Kennedy Meadows pendants were processed using a 20% relative-age discordance filter (Attia et al., 2018) limiting our ability to visually identify lead-loss trajec-tories. Hypothetical lead-loss trajectories drawn for 2.75 and 1.8 Ga zircon populations in published sample K-D1-10 (Fig. 5C) suggest these samples may also have experienced lead loss.

Given the signals of moderate (e.g., FR-190) to virtually complete (e.g., FR-249) lead loss displayed by our detrital zircon samples, we first follow the probabilistic discordance procedure of Reimink et al. (2016) to model the likely upper- and lower-intercept ages for lead-loss chords (File 7, ^{footnote 1}). Results are expressed as peaks on a likelihood scale (left column, Fig. 6) corresponding to primary crystallization age (upper intercepts) and time since isotopic disturbance (lower intercepts). We compare modeling results to graphs of the degree of relative concordance between ²⁰⁶Pb/²³⁸U and ²⁰⁷Pb/²⁰⁶Pb dates versus ²⁰⁶Pb/²³⁸U dates (right column, Fig. 6). Such relative-age concordance-versus-date plots are instructive for visualizing the upper intercepts of lead-loss trajectories due to the sensitivity of calculated ²⁰⁷Pb/²⁰⁶Pb dates to lead mobilization. In each concordance-versus-date plot, a red line depicts a lead-loss trajectory for a hypothetical 1.75 Ga zircon population.

Interplay between Figures 5 and 6 is best illustrated by sample FR-190. Clusters of ~100% concordant zircons plot at ca. 1.1, 1.4, and 1.75 Ga (right column, Fig. 6); these southwestern Laurentian “triad” age populations are also marked by clusters of concordant grains on the Wetherill concordia plot (Fig. 5A). Modeled likely upper intercepts (1.11, 1.38, and 1.73 Ga peaks on the blue curve in left column, Fig. 6) neatly match the Laurentian “triad” (light-blue bar).

Red arrows highlight lead-loss trajectories in Figures 5 and 6 (right column). Lead loss shifts peak ages to lower dates as the % concordance decreases: in an approximate sense, the 1.73 Ga peak in FR-190 (Fig. 6A, left column) corresponds to the 1.65 Ga (FR-246) and 1.56 Ga (FR-211) peaks shifted by lead loss. Sample FR-211 best illustrates the necessity for discordance modeling in provenance analysis of metasedimentary rocks because identifying concordant age populations in either Figure 5B or 6C (right column) is challenging. However, a triad of likely upper-intercept peaks reminiscent of the Laurentian “triad” is produced by discordance modeling, albeit shifted to younger ages by lead loss (left column, Fig. 6C).

Of these graphs, concordance-versus-date plots alone clearly reveal the presence of ~100% concordant, ca. 700–500 Ma zircon populations (right column, Fig. 6). On traditional Wetherill concordia plots, such ca. 700–500 Ma zircon populations cannot be distinguished from older zircons that were shifted by extensive lead loss (Figs. 5A and 5B). Even discordance modeling will not identify these young populations as upper intercepts because lead-loss trajectories of older grains are essentially superimposed on the steep, young segment of concordia. Detrital zircons having U-Pb dates between 700 and 500 Ma are indeed present in passive-margin strata northeast of the Sierra Nevada where they are interpreted as crystallizing from magmas related to Neoproterozoic rifting of Laurentia or within the Southern Oklahoma Aulocogen (Lund et al., 2010; Yonkee et al., 2014; Chapman et al., 2015).

Discordance modeling also provides constraints regarding the timing of lead loss, information that is difficult to constrain on such graphs. Three of the four new samples display Mesozoic lower-intercept peaks (orange vertical bar, left column, Fig. 6), with sharp likelihood peaks in the Early Jurassic (195 Ma, FR-211) to Middle Jurassic (177 Ma, FR-246). We interpret the sharp lower intercepts as indicating that a major lead-loss event, likely a period of increased hydrothermal activity, affected the southeastern Sierra Nevada arc during the Early to Middle Jurassic. In contrast, greenschist-facies sample FR-190 yields a broad, Early Cretaceous peak having a maximum at 120 Ma. Plutons of this age are rare in all but the westernmost Sierra Nevada foothills (Clemens-Knott and Saleeby, 1999); therefore, the geologic significance of this lower-intercept date in the Kern Plateau is unclear. One possibility is that this weakly recrystallized sample (FR-190) hosted prolonged hydrothermal circulation from the Middle Jurassic into the early Late Cretaceous (ca. 100–90 Ma) potentially resulting from the emplacement of the nearby Domelands intrusive suite (Fig. 1B; Saleeby et al., 2008), the southernmost of the massive High Sierra granitoid complexes. If correct, the broad 120 Ma lower-intercept peak may represent the continued hydrothermal circulation in such relatively permeable, greenschist-facies “overlap” strata through the Late Jurassic–Early Cretaceous and into the Late Cretaceous. Regardless of the specific geologic significance, we interpret the lower-intercept results as indicating that greenschist-facies FR-190 experienced a different history of hydrothermal alteration than did the amphibolite-facies samples FR-211 and FR-246, despite all three samples sharing a Laurentian provenance. We therefore suspect that FR-190 has a distinctly different origin than FR-211 and FR-246 and will treat it separately in the following tectonic analysis.

The majority of zircons (n = 51 of 61) from quartzite sample FR-249 yield Mesozoic U-Pb dates. Closer scrutiny, however, reveals that only four of these pass a 5% relative-age discordance filter (right column, Fig. 6D; File 6, ^{footnote 1}). These zircons cluster between 185 and 182 Ma, a Middle Jurassic date over-lapping the age of the surrounding Sacatar granodiorite (Bartley et al., 2007; Saleeby and Dunne, 2015; Fig. 1B). In light of abundant evidence for significant lead loss in sample FR-211, we conclude that these seemingly concordant grains are instead zircons for which the U-Pb isotope systematics were completely reset during Jurassic plutonism. If correct, FR-249 serves as a cautionary tale relative to the potential for extreme hydrothermal-driven lead loss to produce deceptively young detrital zircon dates that could support erroneously young maximum depositional age estimates of metasedimentary rocks.

One challenge to modeling lead loss in metamorphic framework rocks of active Cordilleran arcs is to confirm that calculated Mesozoic or Paleozoic lower intercepts are not “anchored” by young zircon populations that (1) were inadvertently included during sampling (e.g., Mesozoic arc veins and small dikes intruded into their metamorphic wallrock), or (2) represent true detrital populations derived from Neoproterozoic–Cambrian rift magmas or by hypothesized mid-Paleozoic fringing arcs. Such determination requires scrutiny of the geochemical data coupled with field observations and consideration of the regional context: (1) No zircons in FR-190 have calculated “best ages” younger than 366 Ma (File 6, ^{footnote 1}); so the modeled ca. 120 Ma lower-intercept peak (Fig. 6A) cannot be anchored by Mesozoic zircons. (2) Inspection of the concordance-date plot for sample FR-246 (Fig. 6B) reveals that dates younger than 500 Ma are likely attributable to lead loss from zircons that crystallized at either ca. 500 Ma or ca. 1.1 Ga; therefore, the sharp 177 Ma lower-intercept peak is not “anchored” by true Jurassic arc zircons.

Due to extreme isotopic scatter resulting from lead loss from Proterozoic zircons and the steepness of the Mesozoic segment of concordia, assessing whether or not any zircons in sample FR-211 might have crystallized from Mesozoic arc magmas is challenging. We note that the near-continuous spread of concordance from 15% to 135% in sample FR-211 zircons having calculated U-Pb dates less than 300 Ma (right column, Fig. 6C) is similar to the discordance range in comparable FR-246 zircons that clearly appear to be ca. 500 Ma zircons reset by lead loss (right column, Fig. 6B). We conclude that <95% concordant zircons yielding dates less than 700 Ma in samples FR-211 and FR-246 are not actually derived from either the Mesozoic arc (i.e., <275 Ma) or from Neoproterozoic–Cambrian rifts (i.e., ca. 700–500 Ma) but instead yield artificially young dates due to lead loss. Therefore, we include these zircons in the discordance modeling but exclude them from the following age-distribution analysis unless they are members of a concordant age cluster (n ≥ 3; Gehrels, 2012; see footnotes in File 6).

Following Vermeesch (2021), we next apply a 5% (1% reverse) concordia-distance filter to generate kernel density estimate (KDE) plots from the <95% discordant data modified by exclusion of zircon interpreted as yielding erroneous U-Pb dates younger than 500 Ma (see Fig. 7 caption). Only one-third of pre-Mesozoic zircon dates in sample FR-246 pass this filter, compared to 47% and 66% for FR-211 and FR-190, respectively. Once filtered, age populations corresponding to the Laurentian detrital zircon “triad” (blue bars) are resolved in all but the most altered sample (FR-249), though lead loss has shifted “triad” age peaks for FR-246 to lower dates (Nemchin and Cawood, 2005). We retained all zircons in FR-249 (Fig. 7D) to illustrate how simply applying a narrow (5%) concordia-distance filter is an insufficient filter for isotopic discordance and may inadvertently generate samples that appear to contain Mesozoic zircons.

Only FR-211 contains a statistically significant (n > 3) Neoproterozoic–Cambrian zircon population (green bar) that passes a 5% concordia-distance filter (Fig. 7C); the effect of partial lead loss on this ca. 550 Ma zircon population is highlighted by a lead-loss trajectory in Figure 6C (right column). Inspection of concordance-date plots for FR-190 and FR-246 suggests that these samples also contained Neoproterozoic–Cambrian zircons (Figs. 6A and 6B), but the small number of grains (FR-190) and extreme isotopic discordance (FR-246) compromise this conclusion.

Cecil et al. (2019) document initiation of the Sierran-Mojave arc in the El Paso Mountains (Fig. 1) at 273.7 ± 4.4 Ma. Our results confirm the timing of local arc initiation (273.8 ± 2.3 Ma; FR-243) and extend the footprint of the Permian arc ~130 km northwestward into the Kern Plateau. Crystallization ages of our remaining samples (263.5 Ma; 257.6 Ma; 249.4 Ma; Fig. 3) combined with published concordant ages from the southeastern Kern Plateau (248 and 240 Ma; Saleeby and Dunne, 2015) span the Permian to Middle Triassic age range defined by published plutonic ages within the central and southern El Paso terrane (Miller et al., 1995; Cecil et al., 2019). The coincidence of Permo-Triassic plutonic ages between the Kern Plateau, the El Paso Mountains, and ranges within the northcentral Mojave Desert strengthens the correlation between the distant outcroppings of variably metamorphosed deep-water strata comprising the El Paso terrane.

Hafnium isotope data from Permian zircons in both the Kern Plateau and El Paso Mountains demonstrate that the earliest arc magmas incorporated significant ancient continental crust (Fig. 4). When combined with these newly interpreted detrital zircon data, this result is consistent with emplacement of Permian arc magmas through Neoproterozoic to Early Cambrian passive-margin strata deposited on rifted crystalline rocks of the Mojave crustal province and sourced in Laurentia, regardless of whether they additionally traversed the Roberts Mountains allochthon. These data support the proposal of Cecil et al. (2019) that the observed trend toward more primitive zircon εHf values (Fig. 4) is likely due to increasing extension with time within the upper plate of the nascent subduction zone, which permitted depleted-mantle–derived magmas to rise and differentiate while experiencing less crustal contamination. Thus, existing isotopic data do not require either an oceanic or transitional oceanic-continental lithosphere underpinning the El Paso terrane to generate relatively primitive isotopic values (Kistler and Ross, 1990; Miller et al., 1995), which we instead interpret as reflecting decreasing amounts of crustal assimilation through time.

Permian plutons intruding the Kern Plateau shear zone and El Paso Mountains strata are variably deformed. In the El Paso Mountains, Permian plutons display a gneissic fabric, which parallels the axial plane foliation in meta-morphic host rocks. Textural development intensifies inwards toward the contacts, grading into blastomylonite (Carr et al., 1984). Cecil et al. (2019) attribute fabrics of the oldest plutons to contractional deformation, possibly occurring during forced subduction initiation. Cecil et al. (2019) proposed early compressional, orthogonal convergence transitioning into extensional, oblique convergence to explain the transition from foliated Permian arc plutons containing isotopically evolved zircons to non-foliated Triassic plutons containing more isotopically primitive zircons. This proposal, however, conflicts with our new zircon εHf data (cf. FR-243, Fig. 4) that extend the secular trend of Cecil et al. (2019) to even more highly evolved compositions at arc initiation, thus implying that upper-plate extension began at arc initiation.

As an alternative model for arc initiation, we propose that the last phase of sinistral translation required to transport the Kern Plateau segment of the Roberts Mountains thrust overlapped the first phase of Permian arc plutonism. The juxtaposition of disparate metamorphic assemblages within the shear zones in both the El Paso Mountains (i.e., associated hornblende-hornfels– and garnet-amphibolite–facies assemblages; Carr et al., 1984) and within the Kern Plateau shear zone (i.e., associated greenschist- and upper-amphibolite–to lower-granulite–facies assemblages; Saleeby and Dunne, 2015) is permissive of a poly-metamorphic history generated by intrusion of Permian plutons into long-lived sinistral shear zones. Given evidence discussed in the Permian Sierra-Mojave Arc section for a long-lived, sinistral transform boundary preceding subduction initiation (e.g., Lawton et al., 2017; Chen and Clemens-Knott, 2021), we propose that the variably deformed plutons intruding the entire length of the El Paso terrane record a transition from a Permian sinistraloblique convergent margin to a Triassic orthogonal convergent margin. Finally, we propose that the steady increase in zircon εHf from Permian arc initiation into the early Late Triassic records an ~50- m.y.-long period of arc extension generated by rollback (Lallemand et al., 2005) of the dense Panthalassan slab.

The Laurentian detrital zircon “triad” chronofacies is clearly visible in the age distribution plots of samples FR-190, FR-211, and FR-246 (Figs. 7A–7C), linking deposition of their sedimentary protoliths to southwestern Laurentia. Samples FR-211 and FR-246 (Fig. 6) may also contain minor ca. 650–500 Ma zircon populations similar to some passive-margin strata of the White-Inyo Mountains, which may be partly sourced in the Southern Oklahoma Aulacogen (Chapman et al., 2015) or from magmas associated with the late rifting stages of the Rodinian supercontinent (e.g., Armin and Mayer, 1983; Lund et al., 2010; Yonkee et al., 2014; Link et al., 2017). A fourth detrital zircon sample (FR-249) likely has a similar Laurentian provenance, but extreme lead loss prohibits detailed interpretation.

Two micaceous quartzite samples (FR-211 and FR-249; Fig. 2A) were collected from isolated quartz-rich horizons in amphibolite-facies sections dominated by shales but also containing rare marble horizons. This lithologic assemblage does not require deposition in particularly deep marine environments, but instead is grossly consistent with environments interpreted both here and in the El Paso Mountains as transitional or outer shelf-slope environments (Carr et al., 1984; Dunne and Suczek, 1991; Stevens et al., 2005).

Stacked age distributions of data filtered using relative-age discordance filters (20% for published data; 5% for new data) form the basis for interpreting the new Kern Plateau detrital zircon data in their regional context (Fig. 8). Blue bars highlight age populations characterizing the Laurentian “triad” that characterizes Laurentian passive-margin strata (Linde et al., 2014) and is present in our new Kern Plateau samples. Note that the 5% relative-age discordance filter (Fig. 8) generates broader, younger age peaks for FR-211 and FR-246 compared to age peaks generated by a 5% concordia-distance filter (Figs. 7B and 7C). In contrast, previously published Paleoproterozoic (ca. 1.8–1.9 and 2.5–2.7 Ga) U-Pb zircon populations in the samples from the Kennedy Meadows (K-D1-10; Fig. 1) and Bald Mountain (B-D1-10) pendants, coupled with the absence of ca. 1.1 Ga and 1.4 Ga populations, correlate these published Kern Plateau samples with the Roberts Mountains allochthon (yellow bars), an interpretation that aligns with earlier correlations based on a presumed deep-water provenance for barite and chert protoliths in the Kern Plateau (Dunne and Suczek, 1991).

Thus, detrital zircon U-Pb data support the presence of Roberts Mountains thrust fragment(s) in the Kern Plateau: specifically, the spatially abrupt juxtaposition of deep-water metasediments having a Roberts Mountains–like detrital zircon chronofacies (e.g., sample K-D1-10) topographically above shallow-water metasediments having a Laurentia-derived detrital zircon chronofacies (e.g., sample FR-211; Fig. 8). Though the provenance of the extensively altered Kennedy Meadows sample (FR-249) is impossible to assess confidently, we assign this sample to the lower plate and propose that these three detrital zircon samples of the eastern Kennedy Meadows pendant tightly constrain the location of the Roberts Mountains thrust (Fig. 1B; structure K1; Fig. 9A): sample FR-249 was collected 0.6 km southeast and 210 m (700 ft) lower in elevation than upper-plate sample K-D1-10; and sample FR-211 is separated from this upper-plate sample by 2.5 km laterally and 410 km (1350 ft) in elevation. In contrast, the proposed thrust location within the Bald Mountain pendant (Fig. 1B; structure B1) is less well constrained, chosen primarily to locate sample FR-246 in the lower plate and to place mapped locations of folded phyllite, quartzite, chert, and deep-water barite in the upper plate (Fig. 9B; Dunne and Suczek, 1991). Detrital zircon sample B-D1-10 correlates with the Roberts Mountains chronofacies, but it is separated from the contiguous Bald Mountain pendant (Fig. 1) so does not help constrain the Roberts Mountains thrust location.

Previous studies have ruled out the possible presence of the Roberts Mountains allochthon in the Kern Plateau based on the lack of deformation described as “Antler-age.” Antler-age deformation presumably refers to the presence of complex folding and multiple, imbricate thrust faults (Evans and Theodore, 1978; Holm-Denoma et al., 2011), comparable to those documented by detailed field mapping and foraminiferal biostratigraphy in the structurally overlying Golconda allochthon (see fig. 1 of Miller et al., 1984). Comparably detailed structural-biostratigraphic studies would be difficult to undertake in the Kern Plateau pendants, where outcrop is poor and the sedimentary protoliths have suffered multiple episodes of metamorphism and deformation. We note, however, that a tightly folded quartz-mica phyllite from the central Bald Mountain pendant (Fig. 9B) strongly resembles shales folded under low strain conditions in the Golconda allochthon (see fig. 6 of Miller et al., 1981). Whereas the presence of Antler-age deformation has not been unequivocally demonstrated in the Kern Plateau pendants, we suggest that Late Paleozoic deformation might be accommodated within the four, undated deformation events identified by Dunne and Suczek (1991), including their mention of a possible thrust fault located somewhere in the northeastern region of the Kennedy Meadows pendant. We conclude that the presence or absence of Antler-type deformation in Sierra Nevada pendants is not a practical metric by which to rule out a possible correlation to the Roberts Mountains allochthon.

Sandstone sample FR-190 is significantly less recrystallized and meta-morphosed (i.e., a friable, greenschist-facies rock; Fig. 2B) than the other three quartzites. Moreover, discordance modeling does not produce a sharp Jurassic lower intercept for this sample (Fig. 6A), unlike FR-246 and FR-211 (Figs. 6B and 6C). Given these differences, we suspect that FR-190 differs in both age and origin from the other detrital zircon samples despite containing similar detrital zircon age populations (Fig. 7). The depicted extent of such sub–greenschist-facies strata in the Bald Mountain pendant (i.e., rocks west of structure B2 in Fig. 1B) is highly speculative, influenced strongly by the appearance of laterally continuous outcrops visible in satellite imagery in the forests north of Sherman Pass Highway. Based on the low metamorphic grade and apparent lack of penetrative deformation, we suspect these tilted strata were deposited on, or were structurally juxtaposed against, the deformed high-grade rocks of the lower and upper Roberts Mountains thrust plates. If correct, the absence of arc-derived detrital zircons in FR-190 constrains the depositional age of these tilted strata to the Devonian to Early Permian. U-Pb age-probability peaks dominated by the southwestern Laurentian detrital zircon “triad” (Fig. 7A) are consistent with derivation of these strata by reworking of older passive-margin strata. These broad constraints are consistent with the proposed correlation of carbonate-bearing quartzites within the western Bald Mountain pendant to the Devonian Mount Morrison Sandstone (Stevens and Pelley, 2006). Based on the dominance of graphite-bearing slates in the section surrounding sample FR-190 (File 1, ^{footnote 1}), we suggest an alternate correlation of these strata to the mixed carbonate-siliciclastic lithologies (including black shale) characterizing Antler foreland basin deposits (e.g., the Kearsarge and Rest Springs Formations in the Inyo Mountains, the Bright Dot Formation in the Mount Morrison block, and Mississippian strata in the El Paso Mountains). Future sedimentological studies of the western Bald Mountain pendant strata, as well as a near-vertical section of strata visible immediately east of the North Fork of the Kern River near the Hockett Peak pendant, are needed to test these possible correlations.

In summary, detrital zircon data suggest that the eastern edge of the Kennedy Meadows pendant contains a segment of the Roberts Mountains thrust (structure K1; Fig. 1). This thrust places allochthonous, mid-Paleozoic deep-water strata (e.g., sample K-D1-10), thought to be originally deposited ~1500–2000 km to the north near the Peace River Arch (Linde et al., 2016, 2017), over early Paleozoic passive-margin strata having sediment sources stretching eastward across Laurentia and including the western Neoproterozoic rift zone (Linde et al., 2014). Dunne and Suczek (1991) note a possible thrust fault somewhere in the northeastern part of this pendant that might align with the proposed K1 thrust (Fig. 1). Sample density in the Bald Mountain pendant does not provide strong guidance regarding the placement of thrust B1 separating lower-plate (sample FR-246) from upper-plate (sample B-D1-10) strata, which we place such that the folded phyllites shown in Figure 9B are in the upper plate. The location of structure B2—representing either an unconformity or a fault separating lower grade, seemingly younger, pre-arc strata (e.g., sample FR-190) of Laurentian provenance from early Paleozoic rocks juxtaposed by the Roberts Mountains thrust—is comparably speculative.

New detrital zircon data require a structure in the Kern Plateau pendants that juxtaposes an exotic Peace River Arch–derived detrital zircon chronofacies above a parautochthonous Laurentian passive-margin detrital zircon chronofacies. We correlate this structure to the Roberts Mountains thrust, while recognizing that this term implies multiple fault strands (Cashman and Sturmer, 2021). At present, the western end of the Roberts Mountains allochthon is generally agreed to crop out on the southern slope of Miller Mountain, located ~60 km east of Mono Lake (Fig. 10; Stevens and Greene, 1999). Identification of a segment of the Roberts Mountains thrust within the Kennedy Meadows pendant requires that the western Roberts Mountains allochthon was truncated somewhere west of Miller Mountain. Correlation between passive-margin strata in the White-Inyo Mountains and the Mount Morrison block (Fig. 10) seemingly requires southward transport of a truncated, western tip of the Roberts Mountains thrust to the Kern Plateau along a sinistral fault located west of the Mount Morrison block (Greene et al., 1997; Stevens and Greene, 1999). This constraint implicates extension of the Roberts Mountains thrust westward into the Sierra Nevada in the vicinity of the Saddlebag Lake and northern Ritter Range pendants.

Whether the Roberts Mountains and Golconda allochthons extend from west-central Nevada into the Sierra Nevada arc framework has long been debated and remains an open question (Stevens and Peters, 2018). Schweickert and Lahren (1987, 1991) mapped possible correlatives to Roberts Mountains allochthon strata in the central Saddlebag Lake pendant, a correlation that was later extended to a chertargillite unit in the adjacent northern Ritter Range pendant (Greene et al., 1997). Possible Golconda allochthon strata have been mapped north of Mono Lake and in the northern Saddlebag Lake pendant (Schweickert and Lahren, 1987; Cao et al., 2015). Detrital zircon studies, however, have failed to identify rocks containing the 1.8 Ga–dominated detrital zircon chronofacies that characterizes both the Roberts Mountains and Golconda allochthons, leading Attia et al. (2018) to conclude that the Roberts Mountains thrust package never extended westward across the Owens Valley. Instead, U-Pb dates from the Twin Lakes assemblage of the northern Saddlebag Lake pendant (N = 6; total zircon n = 369), including a sample previously assigned to the Golconda allochthon (sample S08-15b; Cao et al., 2015), initially appear to span an almost continuous distribution from the Neoarchean through the Paleozoic. A single chertargillite sample (TP-8) from the central Saddlebag Lake pendant lacks the spread of dates below 1 Ga but otherwise has a similar, multimodal distribution of U-Pb dates between 1 and 2 Ga (Ardill et al., 2020; Fig. 11A). These broad, multimodal age distributions are unusual. Attia et al. (2018, 2021) speculate that these enigmatic distributions may record either an exotic provenance sourced in a late Paleozoic fringing arc system or sources in the northern and eastern orogens of Laurentia.

Figure 11A demonstrates that the near-continuous, multimodal age distribution of the Twin Lakes assemblage is remarkably similar to age distributions of samples FR-211 and FR-246 when similarly processed with a 20% relative-age discordance filter (Fig. 11A). This similarity suggests that the broad spread of U-Pb dates within the Twin Lakes assemblage is an artifact of processing discordant zircon data derived from hydrothermally altered passive-margin strata through a 20% relative-age discordance filter without consideration of complications discussed herein arising from extreme lead loss. Likewise, similarity between published data from the Kern Plateau correlated with the Roberts Mountains allochthon (e.g., sample K-D23-10) and the chertargillite unit of the central Saddlebag Lake pendant (sample TP-8; Ardill et al., 2020) suggests that the broad spread of U-Pb dates between ca. 2 Ga and 1 Ga is an artifact of simply using a 20% relative-age discordance filter (Fig. 11A). These observations support an alternate hypothesis that the almost continuous span of measured U-Pb dates on which the existence of the purportedly exotic Twin Lakes assemblage is based is the result of extensive lead loss in samples containing Archean- and Proterozoic-age populations and therefore does not represent an exotic provenance containing Paleozoic and Mesozoic zircon. Indeed, concordia plots of discordance-filtered data for the Twin Lakes assemblage reveal the hallmark isotopic characteristics of lead loss, with error ellipses for dates exceeding 2 Ga projecting toward Phanerozoic dates in a manner similar to concordia plots for samples FR-211 and FR-246 (Fig. 11B).

We use probabilistic discordance analysis (Reimink et al., 2016) to test the hypothesis that many of the previously published U-Pb dates younger than ca. 1 Ga in the Twin Lakes assemblage do not represent true zircon crystallization ages, while recognizing that these particular U-Pb data are not optimal for upper- and lower-intercept modeling due to their small sample size (n = 11 through 169) and previous relative-age discordance filtering (20%). Nevertheless, modeling of sample 16-522 (Cao et al., 2015) from the Twin Lakes region identifies likely upper-intercept ages consistent with a Laurentian provenance (e.g., 1.02 Ga and 1.38 Ga; Fig. 11C; File 7, ^{footnote 1}) and does not identify any likely sub-1 Ga upper intercepts. In contrast, U-Pb discordance modeling of a single chertargillite sample from the central Saddlebag Lake pendant (sample TP-8; discordance filter not stipulated) identifies a most likely upper intercept of ca. 1.82 Ga accompanied by a range of older Paleoproterozoic to Archean upper intercepts (Fig. 11D), which together are consistent with the zircon chronofacies of the exotic Roberts Mountains allochthon (Fig. 8). We concur with Ardill et al. (2020) that a cluster of five zircons in sample TP-8 likely records true early Paleozoic igneous activity, though lead loss potentially decreased the ca. 435 Ma U-Pb dates yielded by these zircons. This Silurian age estimate falls within the depositional age range for Roberts Mountains allochthon strata (Poole et al., 1992), though only ca. 500–490 Ma igneous activity has yet been documented in the allochthon within Ordovician strata (Linde et al., 2016). We note that sample K-D23-10 from the Kennedy Meadows pendant (Attia et al., 2018) contains two concordant grains of similar age (431 Ma and 422 Ma), as well as a multimodal spectrum of U-Pb dates between ca. 2 Ga and 1 Ga (Fig. 11A).

Based on similar age distribution plots of samples K-D23-10 and TP-8, as well as discordance modeling of sample TP-8 (Fig. 11D), we conclude that Schweickert and Lahren (1987) correctly identified either rocks or detritus derived from the chert-rich Roberts Mountains allochthon in the central Saddlebag Lake pendant (e.g., the chertargillite sample TP-8, collected from rocks originally mapped as Palmetto Formation). Additionally, we conclude that rocks from the northern Saddlebag Lake pendant (i.e., the Twin Lakes assemblage) represent hydrothermally altered passive-margin strata similar to our lower-plate samples from the Kern Plateau pendants (e.g., samples FR-211 and FR-246). If correct, the broad, “multimodal” distributions of detrital zircon U-Pb dates used to characterize both the Twin Lakes assemblage and the chertargillite unit result from extensive lead loss and do not identify influx of exotic detritus or the presence of an exotic terrane within the arc’s framework. Due to limitations noted previously regarding our preliminary modeling of the published data, we do not ascribe any specific significance to the broad, lower-intercept estimates.

Sinistral transpression-transtension has long been the dominant paradigm for late Paleozoic tectonism along the western margin of Laurentia (e.g., Davis et al., 1978; Walker, 1988; Martin and Walker, 1995; Dickinson, 2000; Stevens et al., 2005; Colpron and Nelson, 2009; Saleeby, 2011), with deformation in eastern California-Nevada occurring within the ~450-km-wide Southwestern Laurentian Borderland (SLaB; Lawton et al., 2017). General agreement exists regarding the late Paleozoic entrance and exit of certain terranes from the California-Nevada segment of SLaB: for example, the truncation of Laurentian passive-margin strata in the vicinity of the western Mojave Desert and southward translation of the resulting Caborca terrane ~950 km to northern Mexico (e.g., Dickinson, 2000); accretion of the exotic Foothills ophiolite terrane against the truncated Laurentian margin (Saleeby, 2011); thrusting of the exotic Roberts Mountains allochthon over Laurentian passive-margin strata in central Nevada during the Antler orogeny (e.g., Trexler et al., 2004); and later stacking of the Golconda allochthon on top of the Antler highlands (Fig. 1; Miller et al., 1981). Debate exists, however, regarding the locations of the Roberts Mountains and Golconda depocenters, with some models favoring 1500–2000 km of tectonic transport of exotic strata along a long-lived sinistral transpressional plate boundary (Linde et al., 2016; Chen and Clemens-Knott, 2021) and others favoring significant longshore transport of detritus having a northerly provenance followed by relatively proximal marine deposition (Saleeby and Dunne, 2015; Attia et al., 2018). These latter models specifically address how the El Paso terrane and the Roberts Mountains allochthon could have the same exotic detrital zircon provenance but have potentially different postdepositional structural histories, a need dismissed by our identification of the Roberts Mountains thrust in the northern El Paso terrane (Figs. 8 and 9).

We integrate new zircon U-Pb data with published tectonic and depositional models to construct a plausible late Paleozoic tectonic, sedimentologic, and magmatic history of the El Paso terrane and of east-central California as this region evolved within a long-lived sinistral translation regime that ultimately transitioned to a convergent margin (Fig. 12). We focus on the relative positions of three crustal blocks built on Neoproterozoic–early Paleozoic passive-margin strata, each of which is traversed by the Roberts Mountains thrust: the White-Inyo Mountains; the Mount Morrison block, exposed as pendants within the east-central Sierra Nevada batholith; and the El Paso terrane exposed in the Kern Plateau and extending southward into the northern Mojave Desert (Fig. 10). We overlay these crystalline blocks with schematic representations of pre- arc depobasins in order to depict possible late Paleozoic connections between the El Paso terrane and Laurentia, building on existing models (e.g., Stevens et al., 1997; Stevens and Greene, 1999). Stages of this model are as follows:

1. Middle Devonian (Fig. 12A): The submarine-fan complex of the Mount Morrison Sandstone provides a Middle Devonian depositional link between the White-Inyo Mountains and Mount Morrison block (Stevens and Pelley, 2006). This fan system may have extended westward across the El Paso block, potentially evidenced by carbonate-bearing quartzites present along the western margin of the Bald Mountain pendant near sample FR-190.

2. Mississippian (Fig. 12B): The Kearsarge Formation and overlying siltstones of the Rest Spring Formation extend across much of the White-Inyo Mountains (Stevens and Peters, 2018); similar lithologies comprise the Bright Dot Formation of the Mount Morrison block (Stevens and Greene, 1999). These strata are temporally correlative with Antler foreland basin strata of central Nevada that were largely derived by east-southeastward transport of clasts eroded from the uplifted Roberts Mountains allochthon (Poole, 1974). Having documented continuation of the Roberts Mountains allochthon into the El Paso terrane, we depict the Mississippian Kearsarge–Rest Spring–Bright Dot basin as the southward continuation of the Antler foreland basin blanketing all three crustal blocks inboard of the Roberts Mountains thrust. Contrasting with east-southeastward transport directions in the Antler foreland basin, Stevens and Peters (2018) demonstrate that Mississippian clastic and carbonate debris in the White-Inyo Mountains was shed north-northwestward from uplifts generated by normal faulting and not from an Antler highland located to the northwest. We propose a possible correlation between FR-190 and these Mississippian strata based on the abundance of carbonaceous slates cropping out near sample FR-190 in the western Bald Mountain pendant.

3. Late Pennsylvanian to Early Permian (Fig. 12C): During the Pennsylvanian, southward propagation of sinistral strike-slip faults within the SLaB region broke these blocks apart, initiating the southward translation of the El Paso terrane. We speculate that the depocenter of the latest Pennsylvanian to earliest Permian Keeler Canyon Formation was centered on a left-stepping transtensional basin within the latest Paleozoic sinistral shear zone. We depict the hypothesized bounding strike-slip faults as approximating the future locations of the Death Valley fault zone and the Tinemaha fault (Stevens and Stone, 2002) and the step-over occurring along Mississippian normal faults inherited from the Kearsarge basin. To form the Keeler Basin, we hypothesize offset of the Roberts Mountains thrust an arbitrary 20–30 km in a sinistral sense between the White-Inyo Mountains and Mount Morri-son block; the offset differs from the interpreted modern dextral offset of ~65 km (Fig. 10; Stevens and Stone, 2002) predicting ~90 km of post–Late Triassic dextral slip. We propose the Keeler basin extended southward toward the El Paso Mountains to accommodate carbonate debris derived from the Bird Spring shelf, which was located near the eastern margin of the Keeler Basin.

4. Middle Early Permian to Middle Triassic (Fig. 12D): Circa 274 Ma plutonism records arc initiation in the El Paso Mountains and Kern Plateau (Cecil et al., 2019; this study). We propose that sinistral shear continued to be partitioned into the late Paleozoic sinistral fault system during the initial phase of arc magmatism. If correct, synmagmatic deformation of arc plutons intruding into the sinistral fault system continued through ca. 258 Ma, ceasing in the El Paso Mountains soon thereafter and continuing to at least 249 Ma in the Kern Plateau (sample FR-188; Fig. 3). To the northeast across the White-Inyo Mountains and Mount Morrison block, contractional deformation within the Last Chance–Death Valley thrust systems is recorded by thrust faults and folds having axes oriented obliquely to the nascent arc (Fig. 12D; Levy et al., 2020). The Lone Pine and Darwin basins formed in response to this thrusting, which we attribute to interplay between the waning sinistral transpressive regime extending from the north and the waxing subduction zone developing to the south. Siliciclastic turbidites of the Lone Pine Formation may have received siliciclastic debris from the adjacent Conglomerate Mesa Uplift, a SW-NE–trending transpressional antiform formed in response to sinistral-oblique convergence (Stevens et al., 2015; Lodes et al., 2020). Southeast of Conglomerate Mesa, the Darwin Basin also received debris from the Bird Spring carbonate shelf located immediately to the east. Meanwhile volcaniclastic deposits of the nascent arc accumulated in the central El Paso terrane (Carr et al., 1997; Stevens et al., 2015). We speculate that members A and B of the Holland Camp metasedimentary rocks were deposited in a transpressional basin similar to the Darwin and Lone Pine basins, and that the footprint of the Last Chance–Death Valley thrust system extended southwestward into the El Paso block (Rains et al., 2012; Macdonald, 2016).

   The absence of Permo-Triassic arc zircons in the interpreted “overlap” strata of the Kern Plateau (i.e., FR-190) limits possible correlations to the Devonian (Mount Morrison Sandstone) to early Permian (Lone Pine Formation, members A and B) strata; so we cease tracking local sedimentation at this stage. Our reconstruction suggests that any of the late Paleozoic sedimentary unit(s) depicted in Figures 12A–12D may have overlapped the El Paso terrane, making the proposed correlations to overlap strata of the Kern Plateau pendants highly speculative and in need of testing with further sedimentological study.

   As the late Paleozoic transform plate margin evolved into the Mesozoic convergent margin, we propose that the relative orientation of plate convergence vectors rotated counterclockwise from Pennsylvanian sinistral translation (light-blue arrow, Fig. 12D), to late Early Permian sinistral oblique subduction (medium-blue arrow), to orthogonal convergence by the Late Permian–Early Triassic (darkblue arrow). In contrast, Cecil et al. (2019) propose that upper-plate extension (dashed double-headed arrow) responsible for the secular variation in εHf was driven by a change from a Permian orthogonal convergence to increasingly oblique convergence in the Triassic. We agree that the secular εHf variation likely records arc extension but we favor slab steepening as the driving force of extension (Lallemand et al., 2005) instead of plate rotation leading to more oblique convergence and lower overall convergence velocity (Cecil et al., 2019).

5. Late Triassic (Fig. 12E): Early to Middle (?) Triassic deposition of the marine Union Wash Formation (not shown) across east-central California was followed by resumption of arc magmatism to the north in the Mount Morrison block. The modern spatial distribution of Triassic plutons (Fig. 10; Barth et al., 2018) has been modified to accommodate a subsequent ~90 km dextral offset along the cryptic Tinemaha fault. Stevens and Stone (2002) proposed 65 km of Mesozoic dextral offset to re-align passive-margin facies trends as well as to align submarine fan deposits in the Mount Morrison block and the Inyo Mountains (Stevens and Greene, 1999). To that we add ~25 km to reverse a hypothetical 20–30 km of sinistral offset potentially generated in the Late Pennsylvanian to Early Permian during opening of the Keeler Basin (Fig. 12C), resulting in a speculative 90 km of future dextral offset.

   Counterclockwise rotation of relative plate convergence vectors (Fig. 12D) likely continued through the Triassic and Jurassic, picking up speed through the Cretaceous resulting in the formation of dextral shear zones cutting the arc (Cao et al., 2015). One implication of this model is that some Cretaceous and younger dextral faults may be reactivated late Paleozoic sinistral faults.

Jurassic hydrothermal alteration has long been recognized in the Mojave arc: extensive hydrothermal alteration of Jurassic plutons was first identified using whole-rock oxygen isotopes (Solomon and Taylor, 1991). More recently, the coupled U-Pb-O isotopic record of skarn garnets indicates hydrothermal alteration and arc extension from ca. 162 to ca. 148 Ma in the north-central Mojave Desert (Gevedon et al., 2021). Results of the present study extend the footprint of Jurassic arc extension and hydrothermal alteration into the southeastern Sierra Nevada arc. Specifically, U-Pb discordance modeling identifies an Early to Middle Jurassic hydrothermal event (195 and 177 Ma lower-intercept likelihood peaks; Figs. 6B and 6C) that produced significant lead loss in metasedimentary rocks exposed in the Kern Plateau. Moreover, hydrothermal alteration in the southeastern Sierra Nevada arc may have continued into the Late Jurassic: discordant, multi-grain TIMS data from two plutons collected within a strand of the Kern Plateau shear zone are interpreted as having lower lead-loss intercepts of ca. 158 Ma (samples E10 and E11; Saleeby and Dunne, 2015). We propose that normal-sense deformation along the Kern Plateau shear zone (Saleeby and Dunne, 2015) records Jurassic extension that facilitated intense, synmagmatic hydrothermal circulation in the southeastern Sierra Nevada arc.

Curiously, Jurassic lead loss from three new metasedimentary samples (FR-211, FR-246, and FR-249) appears to be much more extreme than what is recorded by published detrital zircon samples from the Kern Plateau (K-D1-10, K-D23-10, and B-D1-10; Attia et al., 2018). We attribute this difference to the fortuitous sampling of the upper plate of the Roberts Mountains thrust by Attia et al. (2018) and our targeted sampling of eastern, relatively low-elevation outcrops in the Bald Mountain and Kennedy Meadows pendants. The defining isotopic characteristics of Roberts Mountains allochthon strata (i.e., a dominant 1.8 Ga U-Pb age peak, coupled with the absence of 1.1 and 1.4 Ga age peaks) are readily apparent in the published U-Pb data that were filtered through a standard 20% relative-age discordance filter (Fig. 8). In contrast, discordance modeling and 5% concordia-distance filtering were necessary to reveal the Laurentian detrital zircon “triad” that supports correlation of our new samples to the lower plate of the Roberts Mountains thrust (Fig. 7).

Why were upper-plate strata largely undisturbed by the Jurassic hydro-thermal event that produced extreme lead loss in lower-plate strata (e.g., FR-249; Figs. 5A and 6D)? We propose that the Roberts Mountains thrust and metamorphic rocks of the overlying Roberts Mountains allochthon formed a low permeability cap to the Jurassic hydrothermal system in the southeastern Sierra Nevada arc, focusing hydrothermal circulation and lead loss in the underlying passive-margin strata while shielding the upper plate from extreme isotopic disturbance. A comparable role has been proposed for the Roberts Mountains thrust in central Nevada, where most Eocene Carlin-type gold deposits are concentrated in deformed metasedimentary rocks located below the Roberts Mountains thrust: specifically, clay-rich, Roberts Mountains allochthon strata are invoked as forming a relatively impermeable, hydrologic cap on the Eocene hydrothermal system from which these exceptionally rich ore deposits precipitated (Hofstra and Cline, 2000; Emsbo et al., 2006; Muntean et al., 2011). Modeling results of discordant zircons from the Kern Plateau thus provide independent confirmation of the economic role played by permeability characteristics across the Roberts Mountains thrust in north-central Nevada.

Discordance modeling results engender a general caution that erosion and reworking of hydrothermally altered arc framework rocks will obscure isotopic connections (e.g., lead-loss chords in sample FR-190, Fig. 5A) between unaltered and altered zircon grains of a given age population. Such decou-pling will complicate determination of true crystallization ages and original provenance of subsequently altered and reworked detrital zircons preserved in intra-arc sediments. For example, the majority of Jurassic intra-arc strata collected throughout the Sierra Nevada arc (e.g., Saleeby, 2011; Attia et al., 2018; Christe et al., 2018) contain multimodal distributions of Precambrian–Paleozoic detrital zircons similar to those characterizing the Twin Lakes assemblage and interpreted as originating in distant sources (Attia et al., 2018). Instead, we interpret the appearance of these enigmatic zircons in Jurassic intra-arc strata as potentially recording the erosional release of highly discordant zircons from the local arc framework, yielding a multimodal distribution of U-Pb dates similar to that found in passive-margin strata of the Kern Plateau. We emphasize that the enigmatic multimodal “age” distributions are not composed of true U-Pb crystallization ages having attendant provenance implications. Instead, these multimodal U-Pb spectra were generated in the Jurassic by lead loss resulting from the extreme hydrothermal alteration of predominantly Precambrian zircons that was focused beneath the Roberts Mountains thrust.

This study documents highly discordant U-Pb isotope systematics of detrital zircons that we conclude were separated from rocks comprising the lower plate of the Roberts Mountains thrust in the southeastern Sierra Nevada arc and explores data processing strategies that reveal both the zircon provenance as well as the timing of hydrothermal lead loss. We demonstrate that the Kern Plateau region of the northern El Paso terrane preserves an offset segment of the Roberts Mountains thrust fault. This structure juxtaposes Cambrian–Devonian deep-water sediments derived from sources far north near the Peace River arch over late Neoproterozoic–early Paleozoic passive-margin sediments deposited relatively locally on North American lithosphere. Truncation of the Roberts Mountains thrust and ~350 km of southward, sinistral translation of the El Paso terrane along the Kern Plateau shear zone occurred within the Southern Laurentian Borderland (SLaB) region during the ~100-m.y.-long Antler-SLaB-Sonoma transtensional-transpressional orogeny (Lawton et al., 2017; Chen and Clemens-Knott, 2021). Probable late Paleozoic strata in the El Paso terrane may correlate with either (1) a Middle Devonian submarine-fan complex that spread across the SLaB region from the modern White-Inyo Mountains (Stevens and Pelley, 2006) or (2) Late Devonian to Mississippian Antler foreland basin strata.

Ages and zircon hafnium isotopic compositions of Middle–Late Permian and possibly earliest Triassic (?) deformed granitoids of the Kern Plateau are virtually identical to zircon from Permo-Triassic plutons of the El Paso Mountains (Cecil et al., 2019). These results extend the footprint of Permian arc initiation ~200 km northwest from the north-central Mojave Desert while diminishing the published record of local Triassic plutonism. Textural and metamorphic relations support intrusion of arc magmas into late Paleozoic sinistral transform faults, such as the Kern Plateau shear zone. Arc magmas with evolved isotopic signatures formed due to variable assimilation of parautochthonous Laurentian lithosphere and not with an allochthonous terrane rooted in exotic oceanic lithosphere. Zircon hafnium isotope data are consistent with a progressive decrease of the amount of crustal interaction experienced by arc magmas through the Permian and into the Middle Triassic, likely reflecting upper-plate extension that enabled primitive-mantle–derived magmas to reach higher crustal levels while experiencing less crustal assimilation.

U-Pb isotopic discordance modeling indicates extensive Early to Middle Jurassic hydrothermal alteration in the Kern Plateau. Hydrothermal alteration was facilitated by early Mesozoic extension within the Sierra-Mojave arc (Busby-Spera, 1988). We propose that Jurassic extension reactivated the Kern Plateau shear zone as a normal fault within the Sierra-Mojave rift graben system and that the local Jurassic hydrothermal system was capped by the relatively impermeable upper plate of the Roberts Mountains thrust. Discordance modeling also permits correlation of the seemingly enigmatic Twin Lakes assemblage located within the east-central Sierra Nevada arc to hydrothermally altered Laurentian passive-margin strata and the associated chert-argillite unit to hydrothermally altered Roberts Mountains allochthon strata. We suggest that coupling upper- and lower-intercept U-Pb isotopic discordance modeling (Reimink et al., 2016) with a tight, 5% concordia-distance filter (Vermeesch, 2021) and scrutiny of discordance trends within individual age populations provides a template for processing discordant detrital zircon data from the metasedimentary framework of long-lived Cordilleran arcs.

Identification of Laurentian passive-margin strata in two Kern Plateau pendants tightly constrains the location of the Roberts Mountains thrust in the Kennedy Meadows pendant. Presence of a translated segment of the Mississippian thrust implies that the Kern Plateau of the eastern Sierra Nevada caps a crustal column built on sub-continental lithospheric mantle endemic to western Laurentia and conflicts with models that propose El Paso terrane strata as originally capping a sliver of oceanic lithosphere (Stevens et al., 2005; Saleeby and Dunne, 2015; Attia et al., 2018). Our results imply that Permian–Cretaceous arc magmas intruded through the El Paso terrane were generated by flux melting of a mantle wedge composed of sub-continental mantle and intruded through a range of continental to marine strata, a lithologic environment grossly similar to that presumably experienced by the vast majority of eastern Sierran arc magmas. Models seeking to explain the regionally anomalous isotopic values of arc magmas intruding the El Paso terrane, therefore, should not invoke source characteristics alone.

Funding for this research was provided by the following sources: the Donors of the American Chemical Society Petroleum Research Fund (56245-UR8) and the U.S. National Science Foundation (NSF) (EAR 1348078), both to Clemens-Knott; the Hamilton Visiting Scholars fund, the generosity of the Hamilton family and Southern Methodist University to Gevedon; NSF-EAR 1649254 and 2050246 to the Arizona Laserchron Center; vehicle and laboratory support from California State University, Fullerton (CSUF), Department of Geological Sciences. Sampling permission over multiple years was granted by supervisors of the Sequoia National Forest, by scientific staff of the Bureau of Land Management, and by the United States Forest Service. DCK thanks Sam and Jeff Knott for assistance collecting the detrital and plutonic samples, respectively; CSUF undergraduates, Erin Boeshart, Priscilla Martinez and Matt Pilker, for laboratory assistance; and Michael Diggles for introducing her to Kern Plateau mapping during a 1984 U.S. Geological Survey–National Association of Geoscience Teachers (USGS-NAGT) summer internship. A decade of lab support and U-Pb-Hf education was provided by Arizona Laserchron Center staff, particularly Mark Pecha, Kurt Sundell, Nicky Geisler, Sarah George, and Dan Alberts. Jesse Reimink provided extensive help using his discordance modeling code and insightful conversations regarding zircon discordance. Jamie Barnes and the late Hugh Taylor have greatly shaped our appreciation of water-rock interaction. We are privileged to have had conversations and shared field trips with many Sierra-Mojave arc specialists, notably including Cathy Busby, Jade Star Lackey, Jason Saleeby, and the late Lee Silver, all of whom helped shape our understanding of Sierra Nevada tectonics and magmatism. This manuscript benefitted from thoughtful reviews by Alan Chapman, Cal Stevens, and Basil Tikoff, along with editorial handling by Wentao Cao.