A juvenile Paleozoic ocean floor origin for eastern Stikinia, Canadian Cordillera

Detrital zircon geochronology is a powerful tool for evaluating the paleogeographic and tectonic history of a region by determining the ages of rocks being eroded during sedimentation. Pairing U-Pb ages with Hf isotopic compositions of zircon allows insights into the origin of the source rocks, particularly those that are poorly preserved (e.g., Pecha et al., 2016). In the Cordilleran orogen of western Canada and Alaska, such data have proven critical to understanding the evolution of constituent terranes, and most terranes now have robust detrital zircon U/Pb-Hf data sets (e.g., Beranek et al., 2013; Malkowski and Hampton, 2014; Colpron et al., 2015; Pecha et al., 2016; Romero et al., 2020; Alberts et al., 2021).

Stikinia, the largest suspect terrane in the northern Cordilleran orogen (Fig. 1; Colpron et al., 2007) comprises mostly Upper Triassic to Lower Jurassic volcanic, sedimentary, and intrusive rocks (e.g., Logan and Mihalynuk, 2014). George et al. (2021) presented the first detrital zircon U/Pb-Hf data set for Upper Triassic to Lower Jurassic sedimentary rocks in western Stikinia. The basement to these rocks, the Stikine assemblage, comprises Devonian through Permian supracrustal and intrusive rocks but is poorly preserved and exposed (Monger, 1977; Brown et al., 1991, 1996; McClelland, 1992; Logan et al., 2000; Gunning et al., 2006a, 2006b, 2007). Although understanding this basement and its antiquity has a bearing on Cordilleran terrane models, proto–Pacific Ocean (Panthalassa) evolution, and porphyry Cu-Au endowment of Stikinia (Fig. 1; Logan and Mihalynuk, 2014), paired zircon U/Pb-Hf data have hitherto been lacking.

The Asitka Group is the deepest exposed basement in eastern Stikinia. Previously identified fossils indicate deposition during the early Permian (e.g., Ross and Monger, 1978), and mudstone, phyllite, and chert indicate a deepmarine environment. During bedrock mapping, rare rhyolite and sandstone units were identified within the southern-most exposures of the Asitka Group (Ootes et al., 2020a, 2020b). We present the first igneous and detrital zircon U/Pb-Hf data from Stikine basement. We pair these data with zircon trace-element data, the advent of which allows for tectonic discrimination of the zircon source (Grimes et al., 2015). This study includes the first radio larian ages from cherts in the Asitka Group; the biometric and radiometric results refine the timing of Asitka Group deposition. Included are Hazelton Group sandstone (Telkwa Formation; Lower Jurassic) detrital zircon U/Pb-Hf and trace-element data. The zircon εHf(t) and trace-element data are used to evaluate the role of primitive and ancient crust in the evolution of Stikinia and the tectonic environment of parental magma crystallization, enabling comparison to other terranes in the Cordilleran orogen.

The northern Cordilleran orogen of western Canada and Alaska is a collage of parautochthonous and allochthonous terranes. The parautochthonous group of terranes originated as part of the Laurentian margin, whereas the allochthonous group of terranes has paleogeographic affinities from peri-Laurentia to Panthalassa to Baltica (Colpron and Nelson, 2021). Stikinia, part of the classical Intermontane morphological belt, is the largest accreted terrane in the Cordilleran orogen (Monger et al., 1982). It is preserved between Yukon-Tanana and Alexander terranes to the west and Cache Creek and Quesnellia terranes to the east (Fig. 1). Accretion of allochthonous terranes began in the Late Triassic and continued through the Jurassic and involved complex collisional processes between the allochthonous and the parautochthonous terranes (Mihalynuk et al., 1994; Logan and Mihalynuk, 2014; Colpron et al., 2015; Golding, 2020). The relationship between Stikinia and Yukon-Tanana terrane is ambiguous; some consider that Yukon-Tanana terrane is basement to Stikinia (e.g., Jackson et al., 1991; Mihalynuk, 1999), whereas others suggest the terranes remained separate until Mesozoic collision (e.g., Currie and Parrish, 1993; George et al., 2021).

Stikinia comprises a series of unconformity-bounded successions including upper Paleozoic basement rocks of the Stikine assemblage and Asitka Group (Fig. 2; Brown et al., 1991, 1996; McClelland, 1992; Logan et al., 2000; Gunning et al., 2006a, 2006b, 2007), Late Triassic rocks of the Takla, Stuhini, and Lewes River groups (e.g., Monger, 1977; Brown et al., 1996; Logan et al., 2000; Miller et al., 2020; Greig et al., 2021), and Late Triassic to Middle Jurassic rocks of the Hazelton Group (e.g., Tipper and Richards, 1976; Marsden and Thorkelson, 1992; Thorkelson et al., 1995; Gagnon et al., 2012; Barresi et al., 2015; Nelson et al., 2018; Greig et al., 2021). The Late Triassic to Jurassic volcano-sedimentary rocks formed in a suprasubduction zone setting (e.g., Barresi et al., 2015) and are volumetrically the most significant component of Stikinia. All of the supracrustal units have been intruded by plutonic suites, many of which contain porphyry Cu-Au mineralization (e.g., Logan and Mihalynuk, 2014).

The Stikine assemblage includes the oldest stratified rocks in Stikinia (Early Devonian to middle Permian) and forms the basement of western Stikinia. This basement is a mixed carbonate and volcanic succession with minor interbedded siliciclastic rocks (Fig. 2; Monger, 1977; Brown et al., 1991, 1996; McClelland, 1992; Logan et al., 2000; Gunning et al., 2006a, 2006b, 2007). In eastern Stikinia, Paleozoic basement is referred to as the Asitka Group (Figs. 3–5). It comprises felsic-intermediate volcanic flows and tuffs, argillite, mudstone, lime-stone and dolostone, and chert (Lord, 1948; Monger, 1977; Ootes et al., 2020a, 2020b). Although the base of the Asitka Group is not exposed, Lord (1948) suggested a minimum thickness of 2600 m (8500 ft). Fossils, including fusulinids, brachiopods, bryozoa, and sponges indicate it was deposited during the Asselian to Artinskian stages of the Permian (Lord, 1948; Rigby, 1973; Monger, 1977; Ross and Monger, 1978). Paleomagnetic data from Asitka Group volcanic rocks indicate deposition at ~21° to 23°N (Irving and Monger, 1987). Biostratigraphic and sedimentological constraints of the Stikine assemblage (western Stikinia) indicate similar paleolatitudes (Gunning et al., 2006a).

Mafic volcanic and associated sedimentary rocks of the Stuhini Group (western Stikinia) and Takla Group (eastern Stikinia) were deposited disconformably on the Paleozoic basement during the Late Triassic (Lord, 1948; Monger, 1977; Brown et al., 1991, 1996; Logan et al., 2000; Miller et al., 2020; George et al., 2021; Greig et al., 2021). In eastern Stikinia, siliciclastic rocks of the Dewar Formation are the lowest part of the Takla Group. The Dewar Formation is overlain and intercalated with mafic flows and pyroclastic rocks of the Savage Mountain Formation, which is capped by mainly coarsegrained volcaniclastic rocks of the Moosevale Formation (Monger and Church, 1977; Monger, 1977). The Takla Group has not been radiometrically dated, but fossil evidence indicates it is, in part, upper Carnian to Norian and time-equivalent to the Stuhini Group in western Stikinia (Fig. 2; Monger and Church, 1977; Monger, 1977; Logan et al., 2000; George et al., 2021; Greig et al., 2021).

The Takla Group is overlain by the Hazelton Group (Sinemurian to Toarcian; e.g., Tipper and Richards, 1976) in both eastern and western Stikinia (Fig. 2). In eastern Stikinia, the Hazelton Group includes three formations (from base to top: the Telkwa, Nilkitkwa, and Smithers; Tipper and Richards, 1976; Monger, 1977), whereas in western Stikinia, further subdivisions have been detailed (e.g., Nelson et al., 2018). Only the Telkwa Formation was examined in the present study area, where it comprises volcaniclastic rocks, mostly pebble to cobble conglomerates and interbedded mudstone that alternate between maroon and pale-green weathering. These rocks were assigned to the Sikanni facies of the Telkwa Formation by Tipper and Richards (1976). In western Stikinia, the Telkwa Formation is biometrically and radiometrically dated between 205 and 194 Ma (e.g., Barresi et al., 2015). In eastern Stikinia, fossil evidence indicates deposition in the Sinemurian (Fig. 2; Tipper and Richards, 1976; Monger, 1977). Paleolatitudes for the Takla and Hazelton groups indicate deposition between ~23° and 42°N (Monger and Irving, 1980).

This study investigated the most southerly exposures of the Asitka Group where it is overlain by the Dewar Formation (Takla Group) and the Telkwa Formation (Hazelton Group; Figs. 1 and 3–5; Lord, 1948; Monger and Church, 1977; Richards, 1976; Tipper and Richards, 1976; Monger, 1977; Evenchick et al., 2007; Ootes et al., 2020a, 2020b). These units are in fault contact with Cache Creek terrane to the east, and both the Stikinia and Cache Creek rocks are juxtaposed with supracrustal and intrusive rocks of the Quesnel terrane along the dextral strike-slip Pinchi-Ingenika fault (Fig. 3). To the west, the Stikine rocks are covered or are in fault contact with younger Hazelton, Bowser Lake, and Sustut group volcano-sedimentary rocks (Jurassic to Cretaceous; Fig. 3; Lord, 1948; Richards, 1976; Evenchick et al., 2007).

All Stikinian rocks in the study area contain two generations of ductile deformation fabrics. The first generation is most prominent as a moderate to strong foliation (S₁) that is typically bedding (S₀) parallel. The second generation is a locally developed cleavage (S₂) that is axial planar to folds (F₂; Ootes et al., 2020b). Ductile deformation postdated deposition of the Telkwa Formation (i.e., <190 Ma).

In the study area, the Asitka Group was examined at three isolated mountains (Fig. 3). Although the stratigraphy at each location differs in detail, rock types common to all include chert, phyllite, and basaltic andesite.

At location 1, the Asitka Group is on an isolated plateau immediately west of the Ingenika fault and Whistler Basin (Fig. 3). Younging indicators are lacking, and the Asitka-Takla contact is not exposed. The plateau is mostly basaltic andesite, and other rock types (chert, phyllite) are mostly limited to cliff exposures. In the southern part of the plateau, however, a unit of red and white chert is preserved. The chert unit appears to be >10 m thick, although it is folded and structurally thickened. Adjacent to the chert is ~2 m of mudstone, which is in contact with a 30 cm-thick, fine-grained rhyolite unit that can be traced along strike for 5 m (Fig. S1¹). The rhyolite has relict spherulite features observable in thin section (Fig. S6A).

Location 2 is west of the Ingenika fault, 12 km southeast of location 1 (Fig. 3). There, the Asitka Group is exposed on three isolated ridgelines; only the southern two were investigated (Figs. 3 and 5A). Again, younging indicators are lacking, and the Asitka-Takla (Dewar Formation) contact is not exposed; it remains unclear if the contact is stratigraphic or structural. The southernmost ridgeline is mostly basaltic andesite; on its northeast flank are mudstones overlain by white ribbon chert. The central ridgeline consists of fragmental rhyolite on the west and lesser basaltic andesite on the east (Fig. S2, ^{footnote 1}). The fragmental rhyolite alternates from white to red weathering, indicating differing oxidation states. The clasts are all felsic volcanic and range from lapilli to bomb size. Some fragments or pseudo-fragments appear to be welded, but deformation precludes distinguishing between an epiclastic or pyroclastic origin (Figs. S2B and S2C). Above and east of the rhyolite is massive basaltic andesite. The western edge of the outcrop contains a fine-grained gray limestone unit interbedded with minor amounts of basaltic andesite (Fig. S2C).

The most complete stratigraphic section in the study area is in the Sikanni Ranges at Mount Carruthers (location 3), ~10 km southwest of location 2 (Figs. 3–5). The section youngs to the northeast, from Asitka Group to the Dewar Formation to the Telkwa Formation (Figs. 3–5B; Monger, 1977; Ootes et al., 2020a, 2000b), and mapping results were used to build the composite stratigraphic section (Fig. 4). The Asitka Group is thickest in the northwest part of the section and is structurally cut out to the southeast (Fig. 3). The lower part of the section includes carbonate rocks and basaltic andesite (Fig. 4). These rocks pass upsection to gray and beige phyllite and beige and black mudstone (Fig. S3A), locally containing silicified echinoderms (crinoids). The mudstones are locally interlayered with beds of carbonate and sandstone (up to 4 m thick) that contain quartz and lithic pebbles (Fig. S6B). Above the mudstone is green and steel-gray phyllite with carbonate beds and local thin feldspathic sandstone beds. Red to maroon and locally black chert are above the phyllite and contain radiolarians. The top of the section consists of maroon, gray, and beige mudstone, phyllite, and carbonate rock (Figs. 4 and S3).

Along the length of the Sikanni Range (location 3; Fig. 3; Monger, 1977), the Dewar Formation was deposited on different lithological units of the Asitka Group, indicating the contact is an angular unconformity (Fig. 4). The unconformity is nowhere exposed; at the southeast exposure of location 3, the base of the Dewar Formation is an ~1-m-thick black graphitic mudstone (Fig. S4A), and the unconformity is hidden by a talus-covered interval of ~2 m. The remaining lower part of the Dewar Formation comprises black and brown locally pyritic mudstone and siltstone and rare mafic volcanic rocks (Figs. 4 and S4B). The middle part of the section is mostly medium-to coarse-grained sandstone containing plagioclase and pyroxene. Individual beds are up to 2 m thick and are both massive and cross-bedded. The sand-stones commonly have green argillaceous interbeds with well-preserved primary sedimentary structures (e.g., laminations, flame structures; Fig. S4C). Inter-leaved with the sandstones are monomictic and oligomictic clast-supported pebble and cobble conglomerate beds (Figs. 4 and S4D). Clasts range from well-rounded to subangular. The uppermost part of the Dewar Formation is black mudstone and siltstone with thin (<10 cm), fine-grained feldspathic sandstone beds (Fig. 4). From north of location 3 in the Sikanni Range, bivalve fossils (Halobia) in the Dewar Formation are interpreted by Monger (1977) as upper Carnian to lower Norian (ca. 227 Ma).

The Telkwa Formation was only investigated at location 3, in the Sikanni Range (Figs. 3 and 4). The depositional contact with the Dewar Formation has not been directly observed but has been interpreted as an erosional unconformity (Tipper and Richards, 1976; Monger, 1977). In part, this is because the Moose vale and Savage Mountain formations (Takla Group) are missing in the southeast section and interpreted as eroded before Telkwa Formation deposition (Monger, 1977). In the study area, the lower part of the Telkwa Formation includes alternating pale-green and maroon and/or mauve mudstone (Figs. S5A and S5B, ^{footnote 1}). Above are hetero-geneous siliciclastic rocks that include monomictic to polymictic, pebble and cobble conglomerates that are both clast and matrix supported (Fig. S5C). The clast compositions vary but are typically feldspar-rich and of the volcanic porphyry type. The conglomerates have interbeds that include purple sandstone and green and maroon mudstone (Fig. S5).

Tipper and Richards (1976) and Monger (1977) interpreted the volcaniclastic and epiclastic conglomerate and sandstone as non-marine, assigning them to the Sikanni facies of the Telkwa Formation. The Sikanni facies is not fossil bearing, but correlation to other fossil-bearing rocks of the Telkwa Formation to the south indicates a Sinemurian (199.3–190.8 Ma) time of deposition (Tipper and Richards, 1976).

For microfossil separation, 5 kg of red chert (sample 19lo7-3c) were collected from the Asitka Group at location 1 (Figs. 3 and S1, ^{footnote 1}). To maximize the quality of the microfossils, one sample was selected with specific detection techniques applied to pelagic strata from ophiolites and suture zones, using a strong hand lens and investigating specific microfacies such as argillaceous layers within chert beds (Cordey and Krauss, 1990). Samples were processed at the Laboratoire de Géologie de Lyon (Lyon, France), where radiolarians were extracted by numerous and repetitive leaching of samples with low-concentration hydrofluoric acid (HF). A large number of etchings were required (~20 × 24 or 48 h) due to the recrystallization of the chert, a result of diagenesis and later deformation (Figs. 6 and S1). Radiolarians were handpicked and mounted on an aluminum stub for scanning electron microscope (SEM) observation and taxonomical identifications (SEM Zeiss Supra 55 VP, Ecole Normale Supérieure de Lyon).

Two rhyolite samples from the Asitka Group, one from location 1 (A-rhy1) and one from location 2 (A-rhy2; Fig. 3; Table S1, ^{footnote 1}), were collected for igneous zircon U/Pb-Hf and trace-element analysis. Two clastic samples were collected for detrital zircon U/Pb-Hf and trace-element analyses from location 3 (Fig. 3). These include an Asitka Group sandstone (A-San1) and a Hazelton Group sandstone (Telkwa Formation; T-san1). Five samples of medium-to coarse-grained plagioclase and pyroxene-bearing sandstone from the Dewar Formation were collected but did not yield zircon through standard mineral separation.

Each sample weighed ~5 kg and was collected from unaltered and lichen-free bedrock. Cathodoluminescence (CL) imaging and zircon U/Pb-Hf and trace-element analyses were completed at the Pacific Centre for Isotopic and Geochemical Research (PCIGR) at the University of British Columbia (Vancouver, British Columbia, Canada) and the Arctic Resources Laboratory (ARL) at the University of Alberta (Edmonton, Alberta, Canada). Results are in Tables S1–S6 (^{footnote 1}), and all analytical techniques and results of data quality control samples are in appendices 3 and 5 in Ootes et al. (2020c).

Igneous zircons were analyzed for U-Pb and trace elements by laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) at PCIGR (Tables S2 and S3, ^{footnote 1}). A subset of the best quality zircon was then selected and analyzed by chemical abrasion–isotope dilution–thermal ionization mass spectrometry (CA-ID-TIMS; Table S5) by techniques modified from Mundil et al. (2004), Mattinson (2005), and Scoates and Friedman (2008). After CA-TIMS, four separate zircons from sample A-rhy2 were selected for Lu-Hf isotopic analysis (Table S6). Zircons were selected based on size, appropriate trace elements (e.g., low Ca, Fe, P, and La), and single zircon ²⁰⁶Pb/²³⁸U ages were determined by LA-ICP-MS. Zircons were dissolved, and after anion exchange by column chemistry, the Hf isotopic compositions were measured by static multi-collection using a MC-ICP-MS following the methods of Goolaerts et al. (2004) and Weis et al. (2007).

Detrital zircon U-Pb and trace elements were determined by LA-ICP-MS at PCIGR, and Lu-Hf isotopes were determined by LA-MC-ICP-MS at ARL. An additional 35 zircon U/Pb-Hf and trace-element analyses were completed on sample A-san1 at PCIGR (Tables S2–S4).

Igneous and detrital zircon data presented in this study (data in the Supplemental Material, ^{footnote 1}) were screened based on trace-element results. This screening was completed to avoid data that may include mineral inclusions (e.g., apatite) intercepted during zircon laser ablation. Previous studies (e.g., Grimes et al., 2015) have commonly screened results based on absolute trace-element values; for example, zircon with >100 ppm Ca. However, if we had chosen this value for the present study, a large amount of the data would have been screened out because of the relatively high Ca contents of Stikinian zircons and high uncertainty attached to the Ca results (Fig. S7A). Instead, we adopted a modified approach in which the data were plotted and the outliers identified and removed. This approach used Ca, ⁵⁷Fe, P, Ti, La, and Ce (Fig. S7). Some minor outliers remained (e.g., relatively high P, Ti, and La), but in these, the zircon U/Pb-Hf and other trace-element data are internally consistent relative to other results. An example of the screening results from one detrital zircon sample (A-san1) is presented in Figure 7A. The trace-element screening procedure independently identified analyses that have non-typical zircon chondrite-normalized, light rare-earth elements (LREEs). A few analyses with typical zircon chondrite-normalized patterns were also screened out because they contained anomalous concentrations of other trace elements (Fig. 7A); retaining or removing those analyses does not impact the U-Pb or εHf interpretations. The screened-out zircons have comparable ²³⁸U/²⁰⁶Pb ages to those with typical zircon trace-element compositions (Fig. S7).

The radiolarian fauna in the chert of the Asitka Group are poorly preserved due to postdepositional silica recrystallization and deformation (Fig. 6). While this hinders species-level identification, the following radiolarian taxa were identified: Latentibifistula sp., Latentifistula sp., Parafollicucullus sp. cf. scalprata m. rhombothoracata (Ishiga and Imoto), Ruzhencevispongus uralicus Kozur, and Ruzhencevispongus spp. (Fig. 6B). This assemblage is correlative with lower Permian P. scalprata m. rhombothoracata and Ruzhencevispongus uralicus assemblages, ranging from the late Artinskian to the late Kungurian (Aitchison et al., 2017).

The paleoenvironment of chert deposition is not well established. The chert differs from typical deep-water ribbon cherts, which consist of up to 80%–90% radiolarian shells that are typically associated with ophiolites. In the Asitka Group, radiolarians only represent 20%–40% of the chert; other components include oxidized ferruginous clays and possibly volcanic jasper. Therefore, the Asitka cherts may represent a mix of both hydrothermal volcanic-related chemogenic and biogenic processes (e.g., Cordey and Cornée, 2009).

Asitka Group rhyolite A-rhy1 is a fine-grained Asitka Group rhyolite from location 1 (Figs. 3 and S1; Table S1). It is strongly recrystallized, but devitrified spherulite-like features are preserved (Fig. S6A). Sample A-rhy1 yielded ~50–100-μm-long, oscillatory zoned zircon fragments (Ootes et al., 2020c). Fifteen zircons were analyzed by LA-ICP-MS (Tables S2 and S3, ^{footnote 1}). Ten zircon results remained after screening using the trace elements (Fig. 7B). From these zircons, six single grains were analyzed by CA-TIMS (Table S5). These grains yielded overlapping concordant ages with a weighted-mean ²⁰⁶Pb/²³⁸U age of 288.64 ± 0.21 Ma (MSWD = 0.71, probability = 0.62), interpreted as the crystallization age of the rhyolite (Fig. 8A). Zircons from A-rhy1 were considered too small for single-grain, solution-based Lu-Hf analysis.

A-rhy2 is a strongly deformed Asitka Group fragmental rhyolite from location 2 (Figs. 3, 5A, and S2; Table S1). The sample yielded <100–200-μm-long oscillatory zoned zircon fragments (Ootes et al., 2020c). Forty zircons were analyzed by LA-ICP-MS, and 29 remained after trace-element screening (Tables S2 and S3). Five single zircons were selected and analyzed by CA-TIMS (Table S5). Four results overlap on concordia at ca. 294 Ma, while one zircon yielded a younger ²⁰⁶Pb/²³⁸U age of 289.6 ± 0.9 Ma (not included in weighted mean) that results from lead loss. The weighted-mean ²⁰⁶Pb/²³⁸U age from the four zircons (293.89 ± 0.31 Ma; mean square of weighted deviates [MSWD] = 1.27, probability = 0.28) is interpreted as the crystallization age of the rhyolite (Fig. 8B).

Four zircons were selected for Hf isotopic compositions by multi-collector (MC)-ICP-MS (Table S6). The largest zircon (T-1) broke, and the two fragments were analyzed separately. These fragments and one other zircon (2-5) have present-day εHf(0) +6.0 (±0.1; 2SE), +6.3 (±0.2), and +6.4 (±0.2). Two other zircons (3-3 and 3-5) have higher Hf concentrations (Fig. S8) and yield lower εHf(0) of +3.7 (±1.7) and +4.0 (±0.4). The results correspond to time-corrected initial εHf(t) = +9.1 to +11.3 (Figs. S8 and S11; Table S6, ^{footnote 1}).

Asitka Group sandstone sample A-san1 is from the northwest section of location 3 (Fig. 3; Table S1).

The sandstone unit is deformed, ~4 m thick, fine to medium grained, and contains quartz and lithic pebbles (Fig. S6B). The unit is near the middle of the Asitka section, where it immediately overlies black mudstone and is overlain by black, brown, and beige mudstone (Figs. 4 and 5B). This is the only quartz-bearing sandstone unit that has been identified in the Asitka Group. One hundred and fifteen prismatic to rounded detrital zircons were analyzed by LA-ICP-MS. Trace-element data indicate that inclusions were encountered in 19 (16.5%) zircons during laser ablation analysis, and these zircons were screened out. The remaining 96 zircons (Tables S2–S4) have typical chondrite-normalized REE patterns (Fig. 7A).

Ninety-one zircons yield a range of ²⁰⁶Pb/²³⁸U ages between 305.4 ± 7.3 and 354.5 ± 9.4 Ma, and five zircons yield Paleoproterozoic ages with ²⁰⁷Pb/²⁰⁶Pb ages ranging from 1647 ± 29–2177 ± 122.2 Ma (Figs. 9A–9D; Table S2). We consider that the maximum deposition age is close to the time of rhyolite eruption at ca. 295 Ma, which is younger than the youngest single zircon age of 305.4 ± 7.3 Ma, and ~20 m.y. younger than the youngest statistical population at 314.8 ± 1.3 Ma (n = 42; MSWD = 1.02; Figs. 9C and S9, ^{footnote 1}). This interpretation is supported by previously examined fossil assemblages, including fusulinids, brachiopods, bryozoa, and sponges that indicate the Asitka Group was deposited during the Asselian to Artinskian (Lord, 1948; Rigby, 1973; Monger, 1977; Ross and Monger, 1978). Two-component finite unmixing of the detrital zircon data yields calculated ages of 318.3 ± 0.7 (77%) and 335.2 ± 1.4 (23%; Fig. 9D).

The Carboniferous detrital zircons from sample A-san1 have present-day εHf(0) from +3.2 to +9.7 (n = 79) with a median at +5.7 (median uncertainty = ±1.4; 2σ), corresponding to εHf(t) from +10.0 to +16.3. The Paleoproterozoic zircons (n = 4) have present-day εHf(0) from 38.7 to −55.4, corresponding to εHf(t) from 9.2–2.3 (Fig. S11; Table S4, ^{footnote 1}).

Telkwa Formation sandstone sample T-san1 is from the southeast section of location 3 (Figs. 3 and 4; Table S1). It is a purple-white, medium-grained, feldspathic sandstone (Figs. S5C and S6d) inter-bedded with monomictic and polymictic pebble and cobble conglomerate with volcanic clasts, and green and maroon mudstone (Fig. S5). Seventy-eight oscillatory zoned whole zircons and zircon fragments were analyzed by LA-ICP-MS. The trace-element data were used to screen out 11 zircons. The remaining 67 zircons have typical chondrite-normalized REE patterns (Fig. 7C).

The zircons yield a range of ²⁰⁶Pb/²³⁸U ages from 173.9 ± 5.7 to 206.6 ± 9.2 Ma (Table S2; Ootes et al., 2020c). To estimate the maximum deposition age, data were first tested for discordance. The peak of the probability curve is centered at ca. 196 Ma (Figs. 9E and S10A). The youngest single population (YSP) was estimated using the weighted average of the youngest sub-sample that yielded a mean square of weighted deviates (MSWD) of ~1 (Coutts et al., 2019). The best YSP result is from the second and third youngest zircon ages, which yield a ²⁰⁶Pb/²³⁸U weighted mean of 178.5 ± 4.3 Ma (MSWD = 1.05; n = 2; Figs. 9E and S10B, ^{footnote 1}). However, LA-ICP-MS data can yield ages that are younger than the deposition age (e.g., Herriott et al., 2019). Assuming a single zircon population and using the Tuffzirc function (Ludwig, 2003) yields a 196.15 ^+1.15/_1.35 Ma age (Fig. S10C). Unmixing yields two-component calculated ages of 181.3 ± 2.7 Ma (12%) and 197.09 ± 0.96 (88%; Fig. S10D). Using three components (Fig. 9F) yields calculated ages of 180.8 ± 1.3 Ma (2.5%), 195.3 ± 0.8 (66%), and 200.8 ± 1.2 (23%). The older result in the two-component and middle result in the three-component unmixing are indistinguishable from the Tuffzirc result at ca. 196 Ma. These data correspond to correlations (Sinemurian) by Tipper and Richards (1976) and to zircon U-Pb dating of volcanic rocks of the Telkwa Formation in western Stikinia (Barresi et al., 2015; Greig et al., 2021). Whereas Herriott et al. (2019) prefer YSP maximum deposition ages, we interpret the youngest zircons in the Telkwa sandstone to be unreliable, likely from undetected lead loss. We interpret the maximum deposition age as 196 Ma (Fig. 9E), consistent with Sinemurian fossil assem-blages (Tipper and Richards, 1976).

Detrital zircons in T-san1 yield a range of present-day εHf(0) from +0.1 to +7.2 (n = 46) with a median +5.2 (±1.5; 2σ). The results correspond to εHf(t) from +4.3 to +11.4 (Fig. S11; Table S4, ^{footnote 1}). Notably, the lowest εHf(t) result (4.3 ± 2.3) has the highest total Hf concentration (11380 ± 160 ppm) but over-laps, within uncertainty, the next lowest εHf(t) result (6.3 ± 1.6; Fig. S11).

Cherts from the Asitka Group have P. scalprata m. rhombothoracata + Ruzhencevispongus uralicus radiolarian assemblages (Fig. 6) that are Late Artinskian to Kungurian (<286 to >273 Ma) and slightly younger than the rhyolite (A-rhy1), which yielded a U-Pb zircon crystallization age of 288.64 ± 0.21 Ma (Fig. 10). This radiolarian assemblage supports chert deposition above the volcanic rocks (Fig. S1A), possibly a few million years later. The rhyolite at location 2 yielded a U-Pb zircon age of 293.89 ± 0.31 Ma, and both rhyolite ages are consistent with previously reported Asselian to Artinskian fossil assemblages (Lord, 1948; Rigby, 1973; Monger, 1977; Ross and Monger, 1978).

Ninety-five percent of the Asitka Group sand-stone detrital zircons range from 355 to 305 Ma (n = 91; Carboniferous–Middle Mississippian to Middle Pennsylvanian). The results yield a YSP at ca. 315 Ma, and the youngest single grain at ca. 305 Ma (Fig. 9C). However, the maximum deposition age is estimated to correspond with the older ca. 295 Ma rhyolite, and the Asitka Group is interpreted to have been deposited during the Asselian to Kungurian (<298.9 to >273Ma; Fig. 10).

Five Paleoproterozoic zircons (1.65–2.18 Ga) represent ~5% of the detrital zircon data set (Fig. 9A). George et al. (2021) reported poorly defined early Paleozoic to Neoarchean detrital zircons (<3%) from the Stuhini and Hazelton groups in western Stikinia. In contrast, Greig et al. (2021) dated >300 detrital zircons from the Stuhini Group and discovered no zircon older than ca. 225 Ma. George et al. (2021) show that of the pre-Triassic zircons in the Stuhini and Hazelton groups, 90% have a peak probability at 320 Ma (Fig. 10). Based on geological and isotopic arguments, they identified provenance from local Stikine assemblage igneous sources. Western Stikinia contains Early Devonian to Early Missis-sippian volcanic and intrusive rocks (e.g., Logan et al., 2000) and corresponding inherited zircon in younger intrusions (Miller et al., 2020; Fig. 10), but only few (n = 8) zircons of those ages were discovered by George et al. (2021) from a single Stuhini Group sample. Neither the Stuhini Group sample of Greig et al. (2021) nor the Asitka or Hazelton groups in the present study have Devonian to Early Mississippian zircons. Apparently, Devonian basement source rocks did not exist or were not exposed to erosion in eastern Stikinia during the early Permian.

The Asitka Group igneous and detrital zircon εHf(t) data are consistent with derivation from juvenile crust during the Carboniferous to Permian, with no evidence of interaction with or derivation from older continental crust (Fig. 11). In western Stikinia, a minor proportion of detrital zircons from the Stuhini and Hazelton groups are more evolved, indicating minor but detectable influence by older crust (Fig. 11B; George et al., 2021). Nonetheless, most detrital zircons from the Stuhini and Hazelton groups lie on the crustal evolution line derived from the Asitka Group (Fig. 11C). The picrite from the Stuhini Group overlaps this crustal evolution line, but its εHf(t) represents mantle values at the time of its eruption (ca. 222 Ma; Milidragovic et al., 2018). The picrite data, combined with the small spread of Asitka Group εHf(t) data, may indicate both normal and enriched mid-ocean ridge basalt (MORB) mantle was involved in the evolution of Stikinia (Milidragovic et al., 2018).

Comparison of zircon U-Pb data indicates eastern Stikinia has a different evolution than Yukon-Tanana terrane but is similar to Wrangellia (Fig. 12A). Figure 12B compares Stikinian zircon U/Pb-Hf (this study; George et al., 2021) with North American basement (Hottah, Davis et al., 2015), Yukon-Tanana terrane (Pecha et al., 2016; Romero et al., 2020), and Wrangellia (Romero et al., 2020; Alberts et al., 2021). Multivariate statistical evaluation of zircon populations younger than 400 Ma (Fig. 12C) demonstrates that eastern Stikinia likely has no relationship with the Yukon-Tanana terrane. The data show that 95% of zircons from western Stiki nia overlap with those from eastern Stikinia. The more evolved zircons from western Stikinia could indicate imbrication with Yukon-Tanana or some other evolved terrane in the Late Triassic. The similarity of Carboniferous detrital zircons in Stikinia and Wrangellia samples is remarkable (Fig. 12).

The zircon trace elements (Figs. 13, S7, and S11; Table S3, ^{footnote 1}) help identify the nature of the underlying crust, be it ocean arc, ocean island basalt (OIB), or MORB. At a given Nb/Yb ratio, the detrital zircons have U/Yb values above the mantle zircon array, most consistent with zircon formation in a predominantly “arc-like” environment, with a negligible contribution of OIB or MORB (Fig. 13; Grimes et al., 2015). Because the Asitka Group has arc-like zircon chemistry and juvenile εHf(t), which does not permit old crust, the environment of zircon crystallization is best interpreted as incipient ocean arc. The igneous zircons have lower Hf, and higher Yb and Nb relative to the detrital zircons (Fig. S11), and their slightly lower U/Yb values overlap the MORB-OIB–arc fields (Fig. 13). This could indicate an arc to backarc transition from the Carboniferous to the Permian. The chert deposits above the volcanic rocks support a change from active volcanism to relative quiescence and deep-water sedimentation (Cordey and Cornée, 2009).

Belasky and Stevens (2006) modeled Permian fossil assemblages in the Stikine assemblage. They are dissimilar to North American assemblages, and they are not of the Tethyan realm. The assemblages share similarities with Eastern Klamath and Wrangel terranes, and Belasky and Stevens (2006) interpret that Permian Stikinia lies between Wrangellia (north) and East Klamath (south), ~2000–3000 km west of North America. The Carboniferous–Permian connection between Stikinia and Wrangellia is further strengthened by the comparison of detrital zircon results (Fig. 12). George et al. (2021) and Alberts et al. (2021) make similar comparisons, also including the Farewell terrane (Alaska), in this collage. They suggest that although these terranes may have accreted to the Laurentian margin at different times, in the Carboniferous to Permian, they constituted volcanic arcs isolated in the proto–Pacific Ocean (Fig. 14). The collective fossil assemblage modeling and zircon U/Pb-Hf results, with the addition of the trace-element interpretations herein (Figs. 11–13), support a Carboniferous to Permian chain of incipient ocean arcs that included Stikinia and Wrangellia and probably Quesnellia, Farewell, and East Klamath terranes (Fig. 14; Colpron and Nelson, 2011; Malkowski and Hampton, 2014; Alberts et al., 2021; Colpron and Nelson, 2021,George et al., 2021). These terranes consistently contain minor populations of older detrital zircon; for example, ~5% of the detrital zircon in the Asitka Group are Paleoproterozoic. However, juvenile zircon εHf(t) compositions prohibit old continental crust in the early evolution of eastern Stikinia (Figs. 11 and 12). Whatever lithosphere contributed the ancient zircons must have been a piece of continental crust in the oceanic domain. Further discovery and interrogation of early Paleozoic and Precambrian detrital zircons in these terranes will be essential to establish what preexisting continent these peripheral crustal fragments rifted and drifted from, and thereby in what realm they were born (e.g., Baltica versus Panthalassa).

The final contribution from our study is that the Hazelton Group (Telkwa Formation) of eastern Stikinia does not contain evidence of interaction with ancient crust. The Telkwa Formation zircons εHf(t) are most easily explained by derivation from juvenile magmatic rocks in an arc environment formed on the same oceanic lithosphere as the volcano-sedimentary rocks of the Asitka Group and the picrites of the Stuhini Group (Figs. 11–13; Milidragovic et al., 2018). The Hazelton Group in western Stikinia does have a few ca. 200 Ma zircons (<5%) with more evolved εHf(t) (Figs. 11, 12B, and 12C; George et al., 2021). Overlooking potential analytical issues (e.g., mineral inclusions intersected during LA-ICP-MS, which should have been screened out on examination of the time-resolved signal traces), the more evolved εHf(t) indicates interaction with Neo protero zoic or older lithosphere. This lithosphere could be the yet undefined peripheral landmass that contributed the rare Precambrian detrital zircons. A preliminary interpretation by George et al. (2021), based on the age spectra of the rare pre-late Paleozoic detrital zircons, is that this landmass does not share affinity with western North America, and like the Alexander terrane, may have been derived from northern Panthalassa or Baltic realms (see Colpron and Nelson, 2011; Beranek et al., 2013). A second option is that the landmass was the Yukon-Tanana terrane, which became imbricated with Stikinia in the Late Triassic and contaminated Stuhini and Hazelton arc magmas. A final consideration is that although the Precambrian zircons constitute <5% in Stikinia, they represent a range of 1.5 billion years (Fig. 11B). The most plausible source of such a broad age range is one or more older sedimentary successions. For example, the Mackenzie Mountains and Shaler supergroups (Meso-to Neoproterozoic) were deposited in northwestern North America but were supplied by continental-scale river systems that drained the Grenville Province in eastern North America (Rainbird et al., 2017). These zircons were recycled into the rift-related Winder-mere Supergroup (Neoproterozoic) and Paleozoic passive margin sedimentary rocks (Cambrian Type 2 of Hadlari et al., 2012). Each of these stratigraphic units contains a pan-Laurentian zircon record. Such a peripheral landmass could explain the broad range of ages recorded in the Stikinian detrital zircons (Figs. 11D and 12).

The new data provide evidence that eastern Stikinia was born as an incipient arc in the Carboniferous (Fig. 14) and remained as oceanic crust with a peripheral crustal fragment until imbrication with another evolved terrane during the Late Triassic. The crustal fragment, its extent, and its birthplace remain to be determined.

We apply igneous and detrital zircon U/Pb-Hf and trace-element analysis to the Asitka Group and Hazelton Group (Telkwa Formation) in eastern Stikinia and complement this data set with radiolarian ages. The Asitka Group was deposited in a deep marine environment, and extracted radiolarians from cherts are only identified on the genus level but are typical of the assemblages P. scalprata m., rhombothoracata, and Ruzhencevispongus uralicus, which range from the Late Artinskian to the Late Kungurian. This is consistent with, albeit slightly younger than, U-Pb ages of two rhyolites (ca. 294 and 289 Ma). Detrital zircons from a rare sandstone unit, which is stratigraphically below the rhyolites, range from ca. 350–305 Ma with minor Paleoproterozoic zircons (5%). The Telkwa Formation detrital zircon yields a maximum deposition age of ca. 196 Ma.

The combined εHf(t) and trace-element compositions of zircon indicate that the Asitka Group was deposited on a Carboniferous juvenile oceanic arc during the early Permian. The zircon trace elements from rhyolites in this study have lower U/Yb than the detrital zircons, potentially indicating evolution to a backarc-like setting from Carboniferous to Permian. The Telkwa Formation detrital zircon trace elements are also consistent with an arc-like environment. The zircon εHf(t) provide no evidence of old basement in the magmatic sources in eastern Stikinia, and the contribution of the Paleoproterozoic zircon to the Asitka Group sandstone must have come from a peripheral continental landmass. This landmass may have been imbricated with western Stikinia in the Late Triassic, but our study provides no evidence of its existence in eastern Stikinia, which was born and evolved from the ocean floor. Multivariate statistical evaluation of the U/Pb-Hf systematics of detrital zircons from the Asitka Group shows a remarkable overlap with Wrangellia, particularly in Alaska. Although these terranes collided with North America at different times, the similar zircon age and chemistry are consistent with published fossil modeling that pairs Stikinia and Wrangellia as volcanic arcs isolated in the proto–Pacific Ocean during the Carboniferous and Permian.

This study is a contribution to the Hogem mapping project by the British Columbia Geological Survey. Field assistance was provided by K. Shortridge, M. Farhat, and J. Lockie. Clémentine Fellah (Laboratoire de Géologie de Lyon) assisted with the radiolarian SEM session. T. Ockerman and K. Gordon (PCIGR) conducted the igneous zircon Hf isotopic analyses. F. Cordey is supported by the Centre National de la Recherche Scientifique (CNRS), France, Unité Mixte de Recherche UMR 5276. This is Natural Resources Canada contribution #20210338. Discussions with colleagues at the British Columbia Geological Survey improved our understanding of Stikinia, and S. George and an anonymous reviewer helped improve the manuscript.