New isotopic ages have been obtained from euhedral, first-cycle zircon grains recovered from Rhaetian strata preserved at the Black Bear Ridge section in northeastern British Columbia. Two statistically significant populations are present: an older population ca. 224 Ma, and a dominant younger population ca. 205 Ma. The younger population includes a group of grains with a weighted average 206Pb/238U age of 205.2 ± 0.9 Ma, which is interpreted to represent the maximum depositional age of the sediment. Potential sources for the two populations are found in the Quesnel terrane in central British Columbia, implying close proximity between this terrane and the autochthonous North American margin during the Late Triassic. This is supported by geochemical and structural evidence. The implication that the Quesnel terrane was close to its present-day latitude during the Late Triassic is in conflict with older estimates of paleolatitude based on paleomagnetic and paleontological evidence. The age of 205.2 ± 0.9 Ma obtained from the youngest population of zircons is also consistent with recently published estimates for the age of the Norian-Rhaetian boundary.
During the Late Triassic, the terranes of the central Canadian Cordillera (the Quesnel, Yukon-Tanana, and Stikine terranes; Fig. 1) are interpreted to have been located off the western margin of the North American continent (Nelson and Colpron, 2007; Colpron and Nelson, 2009). However, neither the exact position of these terranes during the Triassic, nor the timing or location of their final accretion onto the North American continent, is well known. It has been postulated that the Yukon-Tanana terrane had already been accreted onto the North American margin by the Early or Middle Triassic, an interpretation supported by the cessation of arc activity on the terrane, the formation of regional unconformities, the deposition of a middle to upper Triassic overlap assemblage, and the presence of a Permian stitching pluton (Nelson et al., 2006; Nelson and Colpron, 2007; Beranek and Mortensen, 2011). Paleomagnetic evidence suggests that the terrane accreted at a position on the margin similar to that which it occupies today (Symons et al., 2015), although paleontological data suggest that the terrane must have accreted farther south (Orchard, 2006). The Quesnel terrane is thought to have accreted to the margin in the Early Jurassic (186–180 Ma; Nixon et al., 1993), and paleontological evidence places it far to the south of its current position relative to the North American margin during the latest Triassic (Stanley and Nelson, 1996; Stanley and Senowbari-Daryan, 1999). This would require substantial postaccretionary translation of the terrane (Johnston, 2008), inconsistent with the presence of a Late Triassic overlap assemblage on the terrane (Unterschutz et al., 2002). The Stikine terrane was outboard of the margin during the Late Triassic, but retained ties to the Yukon-Tanana and Quesnel terranes (Nelson and Mihalynuk, 1993; Mihalynuk et al., 1994; Nelson and Colpron, 2007). However, paleontological evidence suggests a more southern position for the Stikine terrane during the Triassic (Reid, 1985; Reid and Tempelman-Kluit, 1987), and paleomagnetic data imply that the terrane was far south of the latitude of the Yukon-Tanana terrane until after the Cretaceous (Symons et al., 2015). Conflicting data and uncertainty over the relationships between these three terranes during the Triassic limits our current understanding of the mechanisms and timing of the formation of the Canadian Cordillera.
As part of a larger study on the provenance of Triassic sediments in northeastern British Columbia (Golding, 2014; Golding et al., 2016), samples were collected for detrital zircon analysis from the Rhaetian (uppermost Triassic; Fig. 2) portion of the section at Black Bear Ridge on Williston Lake (Fig. 3), which was located on the autochthonous margin of the North American continent during the Late Triassic (Fig. 1). The samples were found to contain predominantly euhedral zircons that are herein interpreted to be first cycle. Age determinations for these zircons improve our understanding of the position of allochthonous terranes with respect to the North American margin during the Late Triassic, and provide independent confirmation of recently published estimates for the age of the Norian-Rhaetian boundary.
TRIASSIC MAGMATISM IN THE CANADIAN CORDILLERA
Triassic magmatism in the allochthonous terranes of the Canadian Cordillera commenced during the Early Triassic with the formation of volcanic and intrusive rocks of the Kutcho (Childe and Thompson, 1997; Schiarizza, 2012) and Sitlika assemblages in the Cache Creek terrane (Childe and Schiarizza, 1997; Struik et al., 2007). A lull in magmatism occurred in the terranes during the Middle Triassic, before intensifying during the Late Triassic. Renewed magmatism led to the formation of the Takla and Nicola groups in the Quesnel terrane (Jago et al., 2014), and the Takla and Stuhini groups in the Stikine terrane (Mihalynuk, 1999; MacIntyre et al., 2001). The Takla Group is thought to range in age from the Carnian into the Jurassic (e.g., Monger and Church, 1977; Nelson and Bellefontaine, 1996), and both the Stuhini Group (e.g., Monger, 1977) and the Nicola Group (e.g., Preto, 1979; Panteleyev et al., 1996) contain fossils that range from Carnian to Norian. The formation of these groups represented the first period of widespread magmatism in the terranes since the middle Permian (Gunning et al., 2006). A small number of volcanic ash layers have been identified within sedimentary rocks of Triassic age in British Columbia; of these, two have yielded reliable isotopic ages, one from the Triassic-Jurassic boundary in the Queen Charlotte Islands (Wrangell terrane; Pálfy et al., 2000), and one from the Norian of southern British Columbia (Quesnel terrane; Diakow et al., 2011).
BLACK BEAR RIDGE SECTION
The section at Black Bear Ridge is located on the north shore of Peace Reach on Williston Lake (Canadian National Topographic System, NTS, map area 94 B/3), with its base at UTM (Universal Transverse Mercator) 497670E, 6215500N (Fig. 3). The vicinity of this section was first examined by McLearn (1960) prior to the damming of the Peace River. The present outcrop was described and logged in detail by Gibson and Edwards (1990, 1992), Zonneveld et al. (1997, 2010), Orchard et al. (2001a, 2001b), Wignall et al. (2007), Zonneveld (2010), and Greene et al. (2012). The section is rich in ammonoid, bivalve, and conodont fossils, and these faunas have been described thoroughly (McLearn, 1960; Tozer, 1994; Orchard, 1991, 2007a, 2007b, 2010, 2013, 2014; Orchard et al., 2001a, 2001b; Hall and Pitaru, 2003; Pitaru, 2005; Hall, 2006; McRoberts, 2007, 2011). The section consists of 270.0 m of siltstone, shale, and bioclastic limestone belonging to the Ludington, Pardonet, and Fernie formations (Fig. 4; Zonneveld, 2010). The fossil collections indicate that the section encompasses the Carnian-Norian, Norian-Rhaetian, and Rhaetian-Hettangian boundaries, and therefore includes much of the upper Triassic as well as the lowermost Jurassic (Orchard et al., 2001a, 2001b; Wignall et al., 2007; Greene et al., 2012).
The authors collected two detrital zircon samples (MG-10-BBR01 and JP-12-BBR01) from the section at Black Bear Ridge during field work in 2010 and 2012. Both samples were taken from the same horizon, a pebble conglomerate located 242.0 m above the base of the exposed upper Triassic section. This horizon is interpreted to be a lag bed, and occurs within the upper part of the Pardonet Formation (Fig. 4). The lag horizon occurs directly above the last occurrence of the bivalve Monotis (Fig. 4; Zonneveld, 2010), a fossil that has been shown to be restricted to the upper Norian stage in North America (McRoberts et al., 2008; McRoberts, 2010). The ammonoid Lissonites sp. has also been recovered from the same bed as the highest occurrence of Monotis (Fig. 4); this ammonoid species is considered to be indicative of the uppermost Norian Cordilleranus ammonoid zone (Fig. 2; Tozer, 1994; Orchard et al., 2001a, 2001b). A carbonate concretion directly above the lag horizon (Fig. 4) (sampled by Orchard et al., 2001a, 2001b) was found to contain the conodont Epigondolella mosheri. This conodont is found elsewhere on Williston Lake (at the Ne Parle Pas Point section; see Zonneveld, 2010, Fig. 1 therein for location), in association with ammonoids belonging to the Amoenum ammonoid zone (Fig. 2; Orchard et al., 2001a, 2001b), which is early Rhaetian in age (Orchard and Tozer, 1997). The Rhaetian age of this conodont is confirmed at a number of other localities in North America (Carter and Orchard, 2007). Fragmental ramiform conodonts (assigned to Epigondolella by C.M. Henderson inHall and Pitaru, 2004) have been reported from 2 beds 1.5 m and 1.7 m above the top of the Monotis bed (Fig. 4). These fossil occurrences suggest that the lag horizon occurs low within the Rhaetian of the Black Bear Ridge section. The first Jurassic ammonoid to appear in the section, Psiloceras plicatum, occurs 3.0 m above the top of the Monotis-bearing bed (Fig. 4; Hall and Pitaru, 2003, 2004; Pitaru, 2005). In the paper herein this level is taken to represent the base of the Jurassic, following Orchard and Tozer (1997), Pitaru (2005), and Wignall et al. (2007); a number of alternative positions for the Triassic-Jurassic boundary have been suggested (McLearn, 1960; Tozer, 1982; Gibson and Edwards, 1990, 1992; Stott, 1998; Sephton et al., 2002; Hall and Pitaru, 2004; Fig. 4).
The two samples collected from the Black Bear Ridge section were processed at the Pacific Centre for Isotopic and Geochemical Research at the University of British Columbia. The samples were cleaned, crushed, ground, and passed over a Wilfley table before being concentrated using methylene iodide. The heavy fraction of the sample was separated and examined under a light microscope. Grains were selected randomly for analysis. The majority of the grains are euhedral and <100 µm in size (Fig. 5). The 81 zircons selected from the 2 samples were analyzed using the Thermo Finnegan Element2 laser ablation–inductively coupled plasma–mass spectrometer at the University of British Columbia. Cathodoluminescence images of representative grains (Fig. 5) show that they are zoned, with no evidence of resorption. Laser power during the analyses was typically 40% and the spot size was 15 µm. Line analyses were used to reduce the effects of laser-induced elemental fractionation, and the edges of grains were avoided to decrease the likelihood of the analyses being affected by postcrystallization Pb loss. Grains were analyzed along with samples of the Plešovice reference zircon standard (338 ± 1 Ma; Sláma et al., 2008) to correct for mass and elemental fractionation, and samples of an internal 197 Ma KL standard to monitor data quality. Data were reduced using the GLITTER macro for Excel (version 4.4; van Achterbergh et al., 2001). Analyses with discordance between 207Pb/236U and 206Pb/238U ages of >±10% were rejected; this left a total of 72 grains that gave concordant U-Pb analyses (Table 1). The 206Pb/238U age is used as the preferred age for all of the grains younger than 1000 Ma, whereas the 207Pb/206Pb age is used for grains older than 1000 Ma.
Nine of the zircons recovered from the two samples are moderately to well rounded. These grains all give ages older than 500 Ma, consistent with derivation from igneous rocks of northwestern Laurentia or the Canadian Arctic (Beranek et al., 2010). The large majority of the zircons are euhedral and all give ages younger than 275 Ma. A concordia diagram and a plot of the weighted average 206Pb/238U ages for these zircons are shown in Figures 6 and 7. Two major age populations and one minor population are present among the younger zircons. Two older grains give ages of 268–263 Ma, whereas the remaining grains define two statistically significant populations ca. 224 Ma and ca. 205 Ma (Fig. 8; calculated using the Age Pick macro of Gehrels, 2009). The youngest population includes 39 grains with a weighted average 206Pb/238U age of 205.2 ± 0.9 Ma (Fig. 7), which is interpreted to represent the maximum depositional age for the lag bed from which the samples were collected. It has been demonstrated previously (Dickinson and Gehrels, 2009) that the depositional age of a sample is best interpreted on the basis of populations of grains, rather than on the age of the youngest single grains, and this approach is followed herein. Four grains that gave slightly younger ages (to 175.8 Ma) do not form a cohesive population, and are therefore unlikely to represent the depositional age; some of the grains are even younger than the age of the Triassic-Jurassic boundary (201.3 Ma; Schoene et al., 2010). This contradicts the biostratigraphic evidence from the section that the samples are Rhaetian in age, and suggests that at least some of the grains have been affected by postcrystallization lead loss. The youngest ages were all obtained from very small grains for which it was impossible to avoid analyzing material near the grain margins; therefore, it is quite likely that the material analyzed in these grains may have undergone some degree of Pb loss. The euhedral grains in the sample from Black Bear Ridge are interpreted to be first cycle due to their shape and age; it is well documented that reworked detrital zircons from the Triassic of British Columbia are moderately to well rounded, and none have been dated as younger than Permian (Ross et al., 1997; Beranek et al., 2010; Golding, 2014; Golding et al., 2016).
Evidence of magmatism that was contemporaneous with the ages of the zircons from Black Bear Ridge is preserved in the Quesnel and Stikine terranes of central and northern British Columbia. Late Triassic and earliest Jurassic ages have been recorded from intrusions and volcanic rocks in both of these terranes, and a summary of the ages reported in the literature is presented in Supplemental Table 11. The limits of this compilation are 237 Ma (the approximate age of the Ladinian-Carnian boundary) and 200 Ma, within the margin of error on the published ages. There are 57 reported ages within this bracket, ranging from 230.2 Ma to 199.6 Ma (Supplemental Table 1 [see footnote 1]; Fig. 8). The geographic distribution of a selection of these intrusive and volcanic rocks is shown in Figure 1. The majority of these igneous rocks are located within the Quesnel terrane, which would have been located closest to the study area during the Rhaetian, based on the reconstructions presented by Wernicke and Klepacki (1988) and Mihalynuk et al. (1994). The Quesnel terrane therefore represents the most likely source of the Late Triassic zircons found at Black Bear Ridge. The main detrital zircon population at Black Bear Ridge is interpreted to have been sourced from coeval volcanic centers of the Copper Mountain and Hazelton volcanic suites (ca. 205 Ma), whereas the older zircons are probably derived from plutonic and volcanic detritus of the Stikine and Nicola suites (ca. 224 Ma), and the Guichon suite (ca. 212 Ma; Fig. 8; Supplemental Table 1 [see footnote 1]). The two late Permian detrital zircon grains are likely from the middle to late Paleozoic basement rocks that are known to underlie the Quesnel terrane (Ferri, 1997; Nelson and Friedman, 2004).
It has previously been suggested that the Quesnel terrane formed as a pericratonic terrane on the outboard margin of North America, based on the geochemical similarity between Triassic strata that formed on the terrane and on the continental margin (Unterschutz et al., 2002). Although accretion of the terrane onto North American margin has been dated as occurring in the Early Jurassic (Nixon et al., 1993), pre–Permian–Triassic deformation and metamorphism linked to accretion have been interpreted (Klepacki and Wheeler, 1985; Beranek and Mortensen, 2011). Derivation of the euhedral zircon at Black Bear Ridge from the Quesnel terrane would support the suggestion of proximity between the terrane and the continental margin during the Late Triassic. This would also imply that the Quesnel terrane cannot have been as far south as suggested by both the paleomagnetic and paleontological data; this discrepancy may be due to the distribution of fossils being controlled by physical factors other than latitude-dependent temperature variations.
Regional provenance studies of Triassic sediment deposited on the North American margin in Canada have previously recognized that this sediment was derived not just from the North American craton to the east, but also from the Arctic and from the pericratonic Yukon-Tanana terrane to the west (Ross et al., 1997; Ferri, 2009; Beranek et al., 2010; Ferri et al., 2010; Golding et al., 2016). Close ties between the Yukon-Tanana and Quesnel terranes have been postulated (see preceding), consistent with the terranes being situated close to one another during the Late Triassic. Therefore, both the Yukon-Tanana and Quesnel terranes were positioned at the western margin of the Western Canada Sedimentary Basin, implying that the Late Triassic sediment was not deposited on a platform facing an open ocean. This terrane model instead suggests that the sediment was likely deposited in a foreland basin, as previously hypothesized by Unterschutz et al. (2002), Beranek and Mortensen (2011), and Golding (2014) on the basis of structural and geochemical evidence.
The ages of the Rhaetian zircons at Black Bear Ridge also provide independent support for the recent estimates of the age of the Norian-Rhaetian boundary by Wotzlaw et al. (2014) and Maron et al. (2015). These studies place the age of the boundary at 205.3 Ma and 205.7 Ma, respectively, both of which are consistent with the presence of the 205.2 Ma zircons in the Rhaetian at Black Bear Ridge.
Thorough reviews by JoAnne Nelson, Nick Mortimer, an anonymous reviewer, and Associate Editor Jan Lindsay helped to improve this manuscript. Logistical support for this study was provided by the Wicked River Outfitters. Financial support was provided by the Yukon Basins Project of the Geological Survey of Canada GEM (Geo-Mapping for Energy and Minerals) Project, Geoscience BC, and Natural Sciences and Engineering Research Council of Canada Discovery Grant 371662 (to Zonneveld).