Abstract

The Whitehorse trough is an Early to Middle Jurassic marine sedimentary basin that overlaps the Intermontane terranes in the northern Cordillera. Detrital zircon dates from eight Laberge Group sandstones from various parts of the trough all display a major Late Triassic–Early Jurassic peak (220–180 Ma) and a minor peak in the mid-Paleozoic (340–330 Ma), corresponding exactly with known igneous ages from areas surrounding the trough. Source regions generally have Early Jurassic (ca. 200–180 Ma) mica cooling dates, and the petrology of metamorphic rocks and Early Jurassic granitoid plutons flanking the trough suggests rapid exhumation during emplacement. These data suggest that subsidence and coarse clastic sedimentation in the trough occurred concurrently with rapid exhumation of the shoulders. Isolated occurrences of sandstone and conglomerate units with similar detrital zircon signatures occur west and east of the trough, as well as overlapping the Cache Creek terrane, indicating that either the trough was once more extensive, or isolated basins tapped similar sources. Development of these sedimentary basins and accompanying rapid exhumation in the northern Cordillera were coeval with the onset of orogenic activity in the hinterland of the southern Canadian Cordillera, and subsidence in the western Canada foreland sedimentary basin. The Whitehorse trough is interpreted as a forearc basin that progressively evolved into a collisional, synorogenic piggyback basin developed atop the nascent Cordilleran orogen. Upper Jurassic–Lower Cretaceous fluvial deposits overlapping the Whitehorse trough have detrital zircons that were mainly derived from recycling of the Laberge Group, but they also contain zircons exotic to the northern Intermontane terranes that are interpreted to reflect windblown detritus from the Late Jurassic–Early Cretaceous magmatic arc that developed either atop the approaching Insular terranes to the west or southern Stikinia.

INTRODUCTION

The Jurassic was a period of fundamental change along the western margin of North America (or Laurentia in the Paleozoic). The series of island-arc terranes that flourished in the peri-Laurentian realm between the Late Devonian and Late Triassic (Stikinia, Quesnellia, and others; also known as the Intermontane terranes in the northern Cordillera) collapsed and accreted to the continental margin in Early to Middle Jurassic time, giving rise to the Cordilleran orogen (Fig. 1, e.g., Monger and Price, 2002; Nelson et al., 2013). This change in the geodynamics of western North America has been related to the early Mesozoic breakup of Pangea and subsequent opening of the Atlantic Ocean (Coney, 1972; Dickinson, 2004). By Early to Middle Jurassic (or earlier), the more exotic Insular terranes (Wrangellia and Alexander) were interacting with western North America (McClelland and Gehrels, 1990; van der Heyden, 1992), and a Cordillera-long continental arc developed along the western fringe of the growing orogenic welt (Gehrels et al., 2009)—a tectonic regime that persisted until the early Cenozoic. The growth of the Cordilleran orogen was accompanied by exhumation and surface uplift (as defined by England and Molnar, 1990), and the development of a series of synorogenic sedimentary basins (Fig. 1), culminating with regional subsidence in the western Canada foreland sedimentary basin by the Late Jurassic (Kimmeridgian; Price, 1994).

In the northern Cordillera, the Whitehorse trough is one of the sedimentary basins that developed in the Early to Middle Jurassic atop the northern Intermontane terranes (Stikinia, Quesnellia, and Cache Creek; Fig. 1). It extends some 650 km from near Dease Lake in northern British Columbia to just north of Carmacks in central Yukon (Fig. 2). The trough was originally interpreted as a forearc basin that developed during convergence of Stikinia with North America following subduction of the Anvil Ocean (now known as the Slide Mountain terrane; Tempelman-Kluit, 1979a). Subsequent mapping showed that mid-Paleozoic to early Mesozoic arc terranes (Yukon-Tanana and Quesnellia) separate the Whitehorse trough from the Slide Mountain terrane (Figs. 1 and 2), which is now known to have closed during the Permian–Triassic (Mortensen, 1992; Nelson et al., 2006; Beranek and Mortensen, 2011). Instead, current tectonic models interpret the Whitehorse trough as a forearc basin that developed as a result of subduction of the Cache Creek ocean (part of Panthalassa) beneath contiguous arc terranes of Stikinia and Quesnellia in the Mesozoic (Mihalynuk et al., 1994; English and Johnston, 2005; Nelson et al., 2013).

We present new detrital zircon U-Pb data from sandstone and conglomerate of the Whitehorse trough in southern Yukon (Fig. 2). We also report detrital zircon data from other isolated occurrences of Jurassic conglomerate in southern Yukon (Fig. 2, inset) and relate development of these basins to Early Jurassic exhumation of the metamorphic infrastructure (Yukon-Tanana terrane) and coeval magmatism in the northern Intermontane terranes. These events in the northern Cordillera coincided with early orogenic activity in central and southern British Columbia, and with the onset of subsidence in the western Canada foreland basin.

GEOLOGICAL SETTING

The northern Intermontane terranes in Yukon and northern British Columbia define a south-facing (concave to the south) concentric pattern around the Cache Creek terrane and Whitehorse trough. Older, mid-Paleozoic rocks of the Yukon-Tanana terrane occur along the outer edges of the Intermontane terranes, and progressively younger rocks of Stikinia and Quesnellia occupy their center, forming part of the depositional basement to the trough (Figs. 1 and 2; Wheeler et al., 1991; Colpron and Nelson, 2011a). Stikinia and its eastern twin Quesnellia are primarily characterized by regionally extensive Upper Triassic augite-phyric basalt, basaltic andesite, and volcaniclastic rocks. In British Columbia, arc sequences of Stikinia and Quesnellia are separated by oceanic rocks of the intervening Cache Creek terrane, an accretionary complex that includes oceanic rocks, fragments of intra-oceanic arc, and exotic elements of Tethyan affinity (Fig. 1; Monger and Ross, 1971; Monger, 1977). In Yukon, the Cache Creek terrane does not extend north of Whitehorse, and similar volcanic and volcaniclastic rocks of Stikinia and Quesnellia are juxtaposed along the Teslin fault (Fig. 2). Stikinia and Quesnellia are interpreted to have formed part of a single, continuous arc system that developed in the peri-Laurentian realm along the western edge of the North American plate in the Late Triassic (Mihalynuk et al., 1994). The current geometry of the Intermontane terranes has been ascribed to either strike-slip duplication of the Stikinia-Quesnellia arc (Wernicke and Klepacki, 1988; Dostal et al., 2009) or oroclinal enfolding of the arc around the Cache Creek subduction complex (Mihalynuk et al., 1994, 2004). The Laberge Group, the dominant sedimentary succession within the Whitehorse trough, was deposited during the Early to Middle Jurassic assembly of the Intermontane terranes along the continental margin.

INTERMONTANE TERRANES

The Yukon-Tanana terrane is the oldest of the Intermontane terranes. It consists of at least three overlapping arc successions of Late Devonian to Middle Permian age that developed atop a pre-Devonian metasedimentary basement of probable western Laurentian affinity (Mortensen and Jilson, 1985; Mortensen, 1992; Colpron et al., 2006, 2007a; Piercey and Colpron, 2009). The Yukon-Tanana terrane is characterized by two main pulses of felsic magmatism at 365–330 Ma and 264–252 Ma (Nelson et al., 2006). Most of the terrane is metamorphosed to greenschist and amphibolite facies, and it experienced penetrative ductile deformation in the early Mississippian, Permian–Triassic, Early Jurassic, and Cretaceous (Berman et al., 2007; Beranek and Mortensen, 2011; Staples et al., 2013, 2014). Late Triassic to Early Jurassic plutons related to development of the Stikinia-Quesnellia arc (see following) intrude the Yukon-Tanana terrane, indicating that it formed part of the basement to the Mesozoic arc terranes (Fig. 2).

Stikinia and Quesnellia make up the bulk of the Intermontane terranes (Fig. 1; Wheeler et al., 1991; Colpron and Nelson, 2011a). They incorporate Upper Devonian to Jurassic volcanic and sedimentary strata, as well as comagmatic plutonic rocks (Monger et al., 1991). Both Stikinia and Quesnellia include sporadically exposed Paleozoic arc “basement” rocks. In Stikinia, these include the Stikine assemblage of northwestern British Columbia, a volcanic-sedimentary sequence of Devonian to Early Permian age (Monger, 1977; Logan et al., 2000; Gunning et al., 2006). Correlative rocks in Yukon are the metamorphosed volcanic, volcaniclastic, and minor carbonate rocks of the Upper Paleozoic Takhini assemblage, exposed west of Whitehorse (Figs. 2 and 3; Hart, 1997). In Quesnellia, Upper Paleozoic arc and backarc successions of the Lay Range assemblage and Harper Ranch Group are found in central and southern British Columbia, respectively (Ferri, 1997; Beatty et al., 2006). In south-central Yukon, the Boswell assemblage is an Upper Paleozoic volcanic-sedimentary sequence assigned to Quesnellia (Colpron et al., 2006). It consists of a lower unit of Upper Devonian to Lower Mississippian basalt (normal and enriched mid-ocean-ridge basalt [MORB]) and limestone, and an upper unit of Upper Mississippian to Lower Permian arc volcanic, volcaniclastic, and sedimentary rocks (including Pennsylvanian–Permian fossiliferous limestone and Pennsylvanian tonalite and rhyolite; Simard, 2003; Simard and Devine, 2003; Colpron, 2011). These rocks are unconformably overlain by Upper Triassic volcanic and volcaniclastic strata of the Semenof formation on Boswell Mountain (Fig. 2; informal stratigraphic nomenclature of Tempelman-Kluit, 1984, 2009; Simard, 2003). Paleozoic rocks of the Takhini and Boswell assemblages are inferred to be the depositional basement to Mesozoic arc rocks of Stikinia-Quesnellia beneath the northern Whitehorse trough. Elsewhere, these rocks are correlated with parts of the Yukon-Tanana terrane, which locally forms the basement to Stikinia (McClelland, 1992) and Quesnellia (Simard et al., 2003; Nelson and Friedman, 2004; Roots et al., 2006).

In southern Yukon, Mesozoic Stikinia is represented by volcanic and sedimentary strata of the Middle Triassic Joe Mountain Formation and the Upper Triassic Lewes River Group (Fig. 3; Hart, 1997; Wheeler, 1961). The Joe Mountain Formation consists of a mafic-ultramafic intrusive complex, basalt and volcaniclastic rocks of Ladinian age, and basalt with MORB to backarc basin geochemical affinity (Hart, 1997; Piercey, 2005). The Upper Triassic Lewes River Group (known as Stuhini Group in British Columbia) includes a lower unit of Carnian augite-phyric basalt, basaltic andesite, and volcaniclastic rocks (Povoas formation; informal nomenclature of Tempelman-Kluit, 1984, 2009) and an upper succession of Carnian to Rhaetian epiclastic, volcanogenic sedimentary rocks and limestone (Aksala formation; Tempelman-Kluit, 1984, 2009; Fig. 3). Volcanic rocks of the Povoas formation have the general character of island-arc tholeiite with minor MORB (Hart, 1997; S.J. Piercey, 2005, personal commun.). Augite-phyric volcanic flows and volcaniclastic strata of the Semenof formation (Quesnellia) northeast of the Teslin fault are similar in age and character to the Povoas formation; they are depicted similarly for simplicity on Figure 2. The Aksala formation includes three mappable members (Fig. 3; Tempelman-Kluit, 1984, 2009): (1) the Casca member, a heterogeneous Carnian–Norian sequence of lithic sandstone, argillite, and conglomerate; (2) the Hancock member, a Norian–Rhaetian reefal limestone; and (3) the Mandanna member, a Rhaetian maroon lithic sandstone, siltstone, mudstone, and minor conglomeratic sequence of fluvial origin (Fig. 4A; Long, 2005). The Mandanna member locally interfingers with algal limestone of the Hancock member (Hart, 1997). Lowey (2008, p. 184) reported a U-Pb zircon date of ca. 202 Ma from a tuff horizon in the upper part of the Mandanna member near Whitehorse. These sedimentary rocks record the waning stage of the Lewes River arc.

In British Columbia, arc volcanism resumed in the latest Triassic to Early Jurassic with deposition of the lower Hazelton Group (Thorkelson et al., 1995; Barresi et al., 2015). This phase of volcanism is represented in Yukon only in minor pyroclastic deposits (Nordenskiöld dacite of Tempelman-Kluit, 1984, 2009) in the Whitehorse trough, but it is more substantially expressed by granitoid plutons of Late Triassic to Early Jurassic age (Fig. 2).

Three main suites of plutons associated with development of the Late Triassic–Early Jurassic arc(s) intrude the Yukon-Tanana terrane and parts of Stikinia and Quesnellia in southern Yukon (Gordey and Makepeace, 2001; Colpron, 2015; Fig. 2). The oldest, a suite of ca. 220–206 Ma granodiorite, diorite, and gabbro, occurs mostly as small plutons south and west of Whitehorse (Fig. 2; Hart and Radloff, 1990; Hart et al., 1995). Granodiorite, quartz monzonite, granite, and minor syenite of the ca. 204–195 Ma suite occur mainly along a northwest-trending belt north of Carmacks, whereas granodiorite and granite of the ca. 190–178 Ma suite are mostly found west of the Whitehorse trough, but also locally along its eastern flank (Fig. 2). These plutonic suites are the northern extension of similar suites in Stikinia and Quesnellia in British Columbia, where they locally host porphyry copper-gold deposits (e.g., Logan and Mihalynuk, 2014). In Yukon, the Minto Mine and Carmacks copper deposits are hosted in granodiorite of the ca. 204–195 Ma suite (Fig. 2).

A Middle Jurassic suite (ca. 174–168 Ma) of granodiorite, quartz monzonite, and granite occurs sporadically in southern Yukon (Fig. 2). It intruded Yukon-Tanana, Stikinia, and Cache Creek terranes after their initial imbrication (Mihalynuk et al., 2004), as well as strata of the Whitehorse trough locally.

WHITEHORSE TROUGH

Whitehorse trough (as defined by Wheeler, 1961) includes the upper sedimentary strata of the Upper Triassic Lewes River Group (Aksala formation) and the clastic sedimentary and volcaniclastic rocks of the Lower to Middle Jurassic Laberge Group (Fig. 3). Wheeler (1961) and Hart (1997) described the contact between Laberge and Lewes River Groups as an unconformity along the west flank of the trough passing laterally into a conformable contact near the center of the basin. However, recent stratigraphic studies indicate that the Laberge Group unconformably overlies Lewes River Group everywhere in the Yukon portion of the trough (Lowey, 2004, 2008; Lowey et al., 2009). In the northern Whitehorse trough, Laberge strata locally overlie different units of the Lewes River Group (Colpron et al., 2007b). There is a sharp contrast in depositional environments, from shallow-water, arc-marginal deposits of the upper Lewes River Group to rapidly subsiding, mainly deltaic to deep-marine strata of the Laberge Group (Bultman, 1979; Dickie and Hein, 1995; Lowey et al., 2009). The subsidence that accommodated strata of the Laberge Group recorded the main phase of development of the Whitehorse trough, and these strata are the focus of this paper (Fig. 3).

The Laberge Group consists of a southern unit of deep-water turbidites (Fig. 4B) and mass-flow fan conglomerates (Fig. 4C), the Sinemurian (or possibly late Hettangian?) to Bajocian Richthofen formation (informal nomenclature of Tempelman-Kluit, 1984, 2009; equivalent to the Inklin Formation of northern British Columbia; Souther, 1971; Johannson et al., 1997; Mihalynuk et al., 1999), and a northern, in part coeval, unit of shallow-marine to fluvial, coal-bearing, interbedded sandstone, mudstone, and conglomerates with minor limestone, the Sinemurian to Bajocian Tanglefoot formation (Figs. 3 and 4D; Tempelman-Kluit, 1984, 2009; Hart, 1997; Lowey, 2004, 2008; Takwahoni Formation in northern British Columbia). A crystal-lithic tuff unit, the ca. 188–186 Ma (Pliensbachian) Nordenskiöld facies (Tempelman-Kluit, 1984, 2009), occurs at multiple stratigraphic levels within both the Richthofen and Tanglefoot formations (Figs. 3 and 4E; Colpron and Friedman, 2008). In British Columbia, the Takwahoni represents a western, more proximal facies of the Laberge, and the Inklin represents a deeper-water, more distal facies generally occurring in a higher thrust sheet (Mihalynuk et al., 1999). Locally, the Inklin Formation unconformably overlies Upper Triassic limestone and Permian–Triassic volcanic rocks assigned to the Cache Creek terrane (Gabrielse, 1998).

Long (1986) interpreted the Tanglefoot formation to have been deposited on a broad coastal zone characterized by tidal marshes and river-dominated deltas. Sandstone beds are locally rich in benthonic fauna (e.g., pelecypods, brachiopods, gastropods), trace fossils, and fossilized wood fragments typical of shoreline to shallow-marine environments (Lowey, 2008). In contrast, the Richthofen formation is interpreted to have been deposited in a series of deep-water slope and coalescing submarine fan systems (Dickie and Hein, 1995; Johannson et al., 1997; Lowey, 2005, 2008). The sandstone-siltstone-mudstone turbiditic rhythmites of the Richthofen formation commonly display graded beds, ripples, load structures, flutes, and contorted slump beds. Conglomerates are poorly sorted and commonly matrix supported, and they represent sediment gravity-flow deposits (Lowey, 2005, 2008). The Richthofen formation contains only sparse planktonic fossils (e.g., ammonites and belemnites) consistent with a deep-water marine environment (Lowey, 2008).

Conglomerates from both the Richthofen and Tanglefoot formations show similar clast types and compositional trends (Dickie and Hein, 1995; Hart et al., 1995; Johannson et al., 1997; Lowey, 2008; Shirmohammad et al., 2011). Sinemurian strata generally have higher proportions of sedimentary and volcanic clasts, with limestone characteristically representing up to 10%–20% of the clast mode (Dickie and Hein, 1995; Johannson et al., 1997; Shirmohammad et al., 2011). By early Pliensbachian, volcanic clasts are dominant, and late Pliensbachian and younger conglomerates contain 60%–80% plutonic clasts (Hart et al., 1995; Dickie and Hein, 1995; Johannson et al., 1997; Shirmohammad et al., 2011). Metamorphic clasts are rare in the Yukon portion of the trough (Hart et al., 1995; Dickie and Hein, 1995), but they are abundant in upper Toarcian strata in British Columbia (Mihalynuk et al., 2004; Shirmohammad et al., 2011). Chert clasts occur only in Bajocian and younger strata of the Laberge Group (Hart, 1997; Clapham et al., 2002; Mihalynuk et al., 2004; Shirmohammad et al., 2011). Sandstone compositions display similar (but more subtle) trends as the conglomerates. Sandstones are generally texturally and compositionally immature, fine to coarse grained, moderately to poorly sorted, and dominated by feldspars and lithic fragments (Wheeler, 1961; Hart, 1997; Johannson et al., 1997; Shirmohammad et al., 2011). The amount of matrix is highly variable, but most typically in the range of 10%–15% (Hart, 1997; Johannson et al., 1997). Volcanic grains are the most common lithic fragments, and quartz rarely represents more than 20% of framework grains. Quartz, K-feldspar, and plutonic lithic grains are notably more common in younger strata of the Laberge Group (Johannson et al., 1997). Hornblende is locally a notable constituent. Detrital garnet of ultrahigh-pressure origins is locally present in Pliensbachian sandstone of the Inklin Formation in northern British Columbia (MacKenzie et al., 2005; Canil et al., 2006).

Overall, the conglomerate and sandstone compositional trends in the Laberge Group reflect the progressive exhumation of the arc terranes (and their plutonic/metamorphic roots) flanking the Whitehorse trough (Dickie and Hein, 1995; Johannson et al., 1997; Shirmohammad et al., 2011). Clast compositions can generally be matched with local sources in older strata surrounding the trough. Paleoflow directions from Laberge Group strata indicate predominantly easterly and northeasterly sediment transport along the western side of the trough, southwesterly transport near its northern apex, and local southwesterly transport along its eastern side (Wheeler, 1961; Bultman, 1979; Monger et al., 1991, p. 304–305; Johannson et al., 1997; Dickie and Hein, 1995; Hart, 1997; Lowey, 2005).

To date, U-Pb geochronology in support of provenance studies is limited to dating of nine granitoid clasts from conglomerates and small populations of detrital zircons from three samples of Laberge Group sandstone (Fig. 2). Hart et al. (1995) dated four clasts from two separate localities near Whitehorse. All clasts yielded dates ranging from 215 to 208 Ma and were most likely locally sourced from the belt of Late Triassic plutons (ca. 220–206 Ma) exposed west and south of Whitehorse (Fig. 2; Hart et al., 1995; Dickie and Hein, 1995). Johannson et al. (1997) reported a U-Pb zircon date of ca. 187 Ma from a single clast of monzogranite in the Inklin Formation near Atlin, British Columbia, and Gordey et al. (1998) dated two clasts (one porphyry and one granite) from two localities along the east side of the trough (Fig. 2). They reported U-Pb zircon dates of ca. 230 Ma for the porphyry and ca. 204 Ma for the granite sample; U-Pb titanite dates from both samples are ca. 204 Ma. Shirmohammad (2011) dated two granitic clasts from Toarcian strata of the Takwahoni Formation in British Columbia at 186.6 ± 0.5 Ma and 221 ± 1 Ma. Detrital zircons from three Pliensbachian–Toarcian sandstones yielded dominant populations at 189.6 ± 1.0 Ma (31 grains) and 184.4 ± 1.0 Ma for two samples (35 and 38 grains; Shirmohammad et al., 2011). Two of the samples also contained a few older grains between 220 and 192 Ma and a single Precambrian grain.

The Laberge Group is unconformably overlain by the Upper Jurassic to Lower Cretaceous Tantalus Formation (Fig. 3; Bostock, 1936; Tempelman-Kluit, 1984, 2009; Long, 2005), a coal-bearing sequence of fluvial chert-pebble conglomerate (Fig. 4F) and sandstone that marks the end of deposition in the Whitehorse trough. Outcrop exposures of Tantalus Formation are sparse, with the most extensive exposures near Carmacks and Division Mountain (Fig. 2). Conglomerates form the bulk of the Tantalus Formation (86%), with sandstone (10%) and mudstone representing the remainder (Long, 2015). The conglomeratic units are dominated by well-sorted to moderately well-sorted, medium and large pebble conglomerates, with angular to well-rounded clasts consisting predominantly of varicolored black, gray, white, and rare red and light-green chert, together with subordinate sandstone, igneous, and metamorphic lithologies (Fig. 4F). The chert clasts are similar to chert in the Cache Creek terrane and are possibly derived from a now-eroded part of the terrane to the north or east, as indicated by southwesterly paleoflow indicators (Long, 2015). The Tantalus Formation is interpreted as deposits of shallow to deep braided gravel-bed rivers that accumulated within confined intermontane valleys that developed atop uplifted Laberge Group strata (Long, 2015). Palynomorph assemblages and stratigraphic relationships suggest an Oxfordian (Upper Jurassic) to possibly as young as Aptian–Albian (Lower Cretaceous) age for the Tantalus Formation (Wheeler, 1961; Lowey, 1984; Tempelman-Kluit, 2009; Long 2015).

OTHER JURASSIC BASINS IN SOUTHERN YUKON

Although the Whitehorse trough contains the largest volume of Jurassic conglomerate and sandstone in the northern Cordillera, similar strata also occur at other locations across southern Yukon (Fig. 2). Most proximal to the trough are the graywackes that make up the youngest strata on the Cache Creek terrane (Fig. 2). More isolated (and distal) occurrences of Jurassic conglomerate are also found near Faro, in central Yukon (Faro Peak formation; Fig. 5A), and south of Beaver Creek, in western Yukon (Macauley Ridge formation; Fig. 5B).

Cache Creek Terrane

The Cache Creek terrane in southern Yukon is composed of four main assemblages (Gordey and Stevens, 1994): (1) Pennsylvanian–Permian carbonate (locally containing fusilinids of Tethyan affinity); (2) uppermost Permian–Middle Triassic intermediate to mafic volcanic rocks (correlated with the Kutcho assemblage; English and Johnston, 2005; Schiarizza, 2012; Bickerton, 2014); (3) variably serpentinized peridotite; and (4) an Upper Triassic to Lower Jurassic chert-clastic succession. This last assemblage has been considered a probable correlative of the Whitehorse trough that overlapped the Cache Creek accretionary complex (Monger et al., 1991). We collected a sample of graywacke from this youngest Cache Creek assemblage to test this correlation (11LB045; Fig. 2; Table 1).

The chert-clastic succession consists of intercalations of two main lithologic associations: (1) a chert-dominated succession of Late Triassic (late Carnian–middle Norian) to Early Jurassic (post-Hettangian, possibly late Sinemurian) age; and (2) a siliceous argillite-graywacke succession of Early Jurassic (Pliensbachian or early Toarcian) age (Cordey et al., 1991; Gordey and Stevens, 1994). The graywacke is fine to coarse grained and contains locally abundant angular lithic pebbles of radiolarian chert and argillite (Cordey et al., 1991; Gordey and Stevens, 1994). It is typically composed of quartz, feldspar, and lithic fragments; hornblende is locally a significant (up to 15%) constituent. Most bedding is planar, with rare rip-up clasts, convolute bedding, planar lamination, and graded beds (Cordey et al., 1991). The chert-graywacke succession of the Cache Creek terrane is interpreted to have been deposited in an accretionary complex setting as trench-fill sediments (Cordey et al., 1991).

Faro Peak Formation

The Faro Peak formation occurs near the eastern inboard edge of the Intermontane terranes, northeast of Tintina fault in central Yukon (Figs. 2 and 5A; Pigage, 2004). It consists of a succession of massive, matrix-supported, polymictic conglomerates interbedded with graywacke, sandstone, siltstone, and mudstone. The conglomerate unconformably overlies metamorphic rocks (mainly quartzite) of the Yukon-Tanana terrane immediately southwest of Vangorda fault, a dextral strike-slip fault that separates the Yukon-Tanana and Slide Mountain terranes in the area (Fig. 5A; Pigage, 2004). Conglomerate also locally overlies basalt and chert that Pigage (2004, p. 25) assigned to the basal member of the Faro Peak formation. He recognized, however, that basalt and chert could also be correlative with the Permian Campbell Range formation of the adjacent Slide Mountain terrane, in which case the Faro Peak formation would represent an overlap assemblage on the Yukon-Tanana and Slide Mountain terranes, and the Vangorda fault would not mark their suture. Clasts in the conglomerate include quartzite, chert, basalt, limestone, and lesser granitic gneiss and serpentinite (Tempelman-Kluit, 1979b; Pigage, 2004). Limestone clasts are typically larger than other clasts in the conglomerate, and conodonts from these clasts generally indicate a Late Triassic age (Carnian–Norian), which was interpreted to date the Faro Peak formation (Tempelman-Kluit, 1979b; Pigage, 2004). However, detrital zircons from the Faro Peak formation (see later herein) clearly indicate a Jurassic (or younger?) age (Beranek and Mortensen, 2007).

Macauley Ridge Formation

Conglomerate and sandstone overlying basalt correlated with the Slide Mountain terrane, as well as Triassic sedimentary rocks and Devonian metavolcanic rocks of probable North American affinity, and straddling the Alaska-Yukon border near Beaver Creek, are assigned to the Macauley Ridge formation (Fig. 5B; Murphy et al., 2007, 2008; informal nomenclature). The conglomerate consists mainly of poorly sorted pebbles to boulders of foliated metasedimentary, metavolcanic, and metaplutonic rocks of local derivation; undeformed plutonic and volcanic clasts are also common (Fig. 4G). Immature lithic sandstone and siltstone also occur with the conglomerate (Tempelman-Kluit, 1974; Murphy et al., 2007). Although not exposed in Yukon, the basal contact of the Macauley Ridge formation is an unconformity on Devonian metamorphic rocks in adjacent Alaska (Richter, 1976). The conglomerate was originally considered to be Cretaceous or Tertiary and closely related to extrusion of Carmacks Group volcanic rocks (Upper Cretaceous; Tempelman-Kluit, 1974); however, the detrital zircon data presented herein suggest correlation with the Laberge Group.

DETRITAL ZIRCON GEOCHRONOLOGY

Fourteen samples were collected over nearly a decade of field work in and around the Whitehorse trough (Fig. 2; Table 1). One sample was collected from the youngest stratigraphic unit of the Lewes River Group (Mandanna member; Norian–Rhaetian), eight are from the Laberge Group (2 Tanglefoot, 6 Richthofen; Sinemurian to Bajocian), one is from Lower Jurassic graywacke overlapping the Cache Creek terrane, and four are from the Upper Jurassic to Lower Cretaceous Tantalus Formation. In addition, we also present data from two isolated occurrences of Jurassic conglomerate near Faro (Faro Peak formation), in central Yukon, and Beaver Creek (Macauley Ridge formation), near the Alaska border (Fig. 2, inset; Fig. 5).

Methods

U-Pb zircon dates were obtained by laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) at four laboratories (Table 1). The majority of samples were analyzed at the Arizona LaserChron Center, University of Arizona, and the Isotope Geology Laboratory at Boise State University, with two samples of Tantalus Formation (TB1 and TB2) that were analyzed in both laboratories. One sample of Laberge Group sandstone was analyzed at Laurentian University, and samples of Faro Peak and Macauley Ridge formations were analyzed at the Pacific Centre for Isotopic and Geochemical Research at the University of British Columbia. In addition, the youngest zircons dated by LA-ICP-MS from two samples of Tantalus Formation (TB1 and TB2) were analyzed more precisely by chemical abrasion–isotope dilution–thermal ionization mass spectrometry (CA-ID-TIMS) at Boise State University. Analytical methods for all laboratories are given in Appendix 1.1 LA-ICP-MS analytical results are given in Appendixes 2 and 3 (see footnote 1) and illustrated on Figures 6–8. CA-ID-TIMS data are presented in Figure 9 and Table 2. Overall, LA-ICP-MS results for similar samples from the various laboratories are in good agreement, and the duplicate analysis of two samples shows excellent interlaboratory agreement. Of the more than 1700 grains analyzed, only 2% yielded Proterozoic–Archean dates, with the majority of zircons being younger than 400 Ma (Appendixes 2 and 3 [see footnote 1]; Figs. 6–8). Therefore, the following discussion focuses on zircons with Phanerozoic ages. It is interesting to note, however, that more than 70% of the Proterozoic–Archean zircons were analyzed in samples from the Faro Peak formation.

Probability density and histogram plots of detrital zircon ages were calculated for each sample using the Isoplot 4.15 add-in for Microsoft Excel (Ludwig, 2008). As 98% of the zircons yielded ages younger than 1 Ga, we used the 206Pb/238U dates and a 10 m.y. bin size for these calculations. Individual probability plots for each sample are presented in Appendixes 2 and 3 (see footnote 1). To facilitate comparison between samples, we present these data as a series of stacked normalized age probability plots in Figures 6–8. These plots were constructed using the Excel macro provided on the Arizona LaserChron Center’s Web site (https://sites.google.com/a/laserchron.org/laserchron/home/), which normalizes the probability curve for each sample by the number of analyses used, such that each curve contains the same area.

In addition to being useful for provenance analysis, the youngest dates in a population of detrital zircons can provide an estimate of the maximum depositional age of strata, particularly those deposited in a convergent margin setting such as Whitehorse trough (e.g., Cawood et al., 2012). Several methods can be used to calculate a maximum depositional age, each presenting advantages and potential complexities (see discussions in Dickinson and Gehrels, 2009; Gehrels, 2014). Particular concerns with LA-ICP-MS analyses are the potential for Pb loss and the inherent uncertainty associated with each analysis, which result in measured dates that are both younger and older than the true age (Gehrels, 2014). For these reasons, we refrain from determining a maximum depositional age using the youngest single zircon date because of the lack of reproducibility, and we favor methods that consider a population of young dates (Dickinson and Gehrels, 2009). Maximum depositional ages from LA-ICP-MS data were estimated using three methods:

  • (1) We calculated the age of the youngest age peak (representing a population of grains) using the TuffZirc Age routine in Isoplot (Table 1; Fig. DR1 [see footnote 1]). This method is more conservative and assumes that dates on the shoulder of a peak result from analytical uncertainty. In their analysis, Dickinson and Gehrels (2009) found that the youngest peak method generally yields ages that are equivalent or older than actual depositional ages of the strata.

  • (2) We calculated the weighted mean age for a cluster of youngest zircon dates that overlapped in age at 1σ (Table 1; Fig. DR2 [see footnote 1]). Our calculations considered populations of at least six grains (maximum of 26; average of 12.8) with low mean square of weighted deviates (MSWD; <1) and high probability of fit (average of 0.99; Fig. DR2 [see footnote 1]). These date clusters generally sample the young shoulder of the youngest peak in age and are thus consistently younger than youngest peak ages by 3–14 m.y. for Upper Triassic to Middle Jurassic strata (average of 8 m.y.), and 1–2 m.y. for the Upper Jurassic to Lower Cretaceous Tantalus Formation (Table 1).

  • (3) We used the Unmix routine in Isoplot (Fig. DR3 [see footnote 1]). Unmix determines the Gaussian distribution that best fits two or more groups of ages for a given population of zircons. Because most of our samples display a single young peak commonly supported by a 20–50 m.y. spread of ages, we calculated Unmix ages assuming that these peaks represent mixing of two populations of young zircons. Samples of Tantalus Formation (TB1, TB2, TB3, and C1) and Laberge Group sample 08MC091 showed more complex patterns, and calculations were done assuming three populations of zircons (Fig. DR3 [see footnote 1]). In general, Unmix ages for the youngest group of zircons lie between the age of the youngest peak and the weighted mean of a cluster of young zircons (Table 1).

For two samples of Tantalus Formation (TB1 and TB2), the maximum depositional age was determined by more precise CA-ID-TIMS dating of the youngest zircons dated by LA-ICP-MS (Table 2). Chronostratigraphic ages for the maximum depositional ages (Table 1) were assigned using the time scale of Cohen et al. (2013; updated January 2015).

Mandanna Member

Arkosic sandstone from the Mandanna member (Aksala formation, Lewes River Group) was collected northwest of Whitehorse (05W170; Fig. 2; Table 1) from a succession of bioturbated red mudstone, siltstone, and sandstone. It is fine to medium grained, planar to cross-stratified, and composed of angular plagioclase, hornblende, volcanic lithic clasts, and quartz. Maroon siltstone rip-up clasts are locally abundant. Detrital zircons have a bimodal distribution with peaks at ca. 338 and 205 Ma (Fig. 6; Appendix 2 [see footnote 1]). Paleozoic grains are 403–303 Ma and constitute 34% of the sample, the most significant proportion of Paleozoic grains analyzed in this study. Mesozoic grains are 248–194 Ma, with 34% of zircons at 205–200 Ma. The ca. 205 Ma peak and the weighted mean age of 201.5 ± 1.5 Ma (for 20 of the youngest grains; Fig. DR2 [see footnote 1]) are consistent with the 201.5 ± 0.6 Ma date from a tuff near the top of the Mandanna member ∼15 km to the south (near location GL04-108b on Figs. 2 and 10; Mortensen inLowey, 2008, p. 184) and suggest a Rhaetian maximum depositional age for the Mandanna member (Table 1).

Laberge Group

Most samples analyzed are sandstones from the Laberge Group collected along the length of the Whitehorse trough in Yukon (Fig. 2; Table 1). Sample 08MC091 (Fig. 2; Table 1) is a coarse-grained calcareous sandstone unconformably overlying limestone of the Hancock member (Aksala formation, Lewes River Group) on Grey Mountain in Whitehorse. A limestone-cobble conglomerate locally marks the unconformity. This calcareous facies represents a shallow-water environment that contrasts with typical deep-water turbidites of the Richthofen formation in the region. Detrital zircons have a dominant population (92%) at 220–180 Ma with a peak at ca. 200 Ma, indicating a Hettangian maximum depositional age (Fig. 6; Table 1; Appendix 2 [see footnote 1]). The weighted mean of six young grains is 188.8 ± 3.2 Ma (Fig. DR2 [see footnote 1]), and the youngest Unmix age of 186.4 ± 2.4 Ma (Fig. DR3 [see footnote 1]) suggests a Pliensbachian maximum depositional age (Table 1). The few Paleozoic grains (5%) are Middle Permian and Devonian.

Sample 04SJP594 is a lithic sandstone from the matrix of a mass-flow, polymictic conglomerate in the Richthofen formation along the Alaska Highway, northwest of Whitehorse (Figs. 2 and 4C; Table 1). This is the same locality for two of the clasts dated at 215–208 Ma by Hart et al. (1995), who estimated a Hettangian?–Pliensbachian depositional age for these strata. Detrital zircons have a broad distribution of Mesozoic grains (96%) between 237 and 173 Ma, with twin peaks at 209 and 189 Ma, and a subsidiary peak at ca. 333 Ma (Fig. 6; Appendix 2 [see footnote 1]). These peaks correspond well with ages of local plutons to the south and northwest (see Appendix 4 [see footnote 1]). The ca. 333 Ma zircons were likely derived from local strata of the Takhini assemblage or Yukon-Tanana terrane to the west. The ca. 189 Ma peak (Fig. DR1 [see footnote 1]) and ca. 187 Ma youngest Unmix age (Fig. DR3 [see footnote 1]) support a Pliensbachian maximum depositional age, but the weighted mean of seven young zircons (180.6 ± 2.2 Ma; Fig. DR2 [see footnote 1]) indicates a younger Toarcian age (Table 1).

Sample GL04-108b is sandstone matrix from a conglomerate unconformably overlying the Mandanna member on Jackson Hill west of Whitehorse (Figs. 2 and 10; Table 1). A tuff in the Mandanna member at this locality yielded a U-Pb zircon date of 201.5 ± 0.6 Ma (Mortensen inLowey, 2008, p. 183–184). Detrital zircons have a broad distribution from 371 to 174 Ma, with peaks at ca. 187 and 343 Ma (Fig. 6; Appendix 2 [see footnote 1]). Paleozoic zircons (31%) have broad bimodal distributions between 370 and 320 Ma and between 280 and 260 Ma (Fig. 6). Jurassic (200–174 Ma) zircons represent 57%, and youngest peak ages calculated with both the TuffZirc and Unmix routines are concordant at ca. 187 Ma (Table 1; Figs. DR1 and DR3 [see footnote 1]). A cluster of nine young zircons yielded a weighted mean age of 180.3 ± 2.3 Ma (Fig. DR2 [see footnote 1]). Pliensbachian to Toarcian maximum depositional ages are estimated (Table 1), suggesting that the sub-Laberge unconformity on Jackson Hill spans at least 15–20 m.y (Fig. 10).

Sample 04SJP603 is lithic sandstone from rhythmites of the Richthofen formation exposed along the western shore of Lake Laberge, north of Whitehorse (Figs. 2 and 4B; Table 1). This area was considered the type area for the Richthofen formation by Lowey (2008). The age of these strata is not known locally, but fossil collections from similar strata in the region (10–15 km to the south) indicate a possible Sinemurian–Pliensbachian depositional age (compiled in Colpron, 2011). Detrital zircons have a bimodal distribution with peaks at ca. 197 (93%) and 327 Ma (Fig. 6; Table 1; Appendix 2 [see footnote 1]). The weighted mean of 17 young grains is 190.5 ± 1.5 Ma (Fig. DR2 [see footnote 1]), and the youngest Unmix age is 195.09 ± 0.77 Ma (Fig. DR3 [see footnote 1]). The calculated maximum depositional ages range from Sinemurian to Pliensbachian (Table 1), in agreement with regional biostratigraphic estimates.

Sample 12LB220 is well-sorted, coarse-grained arkosic sandstone from a graded sandstone-siltstone succession typical of the Richthofen formation near Marsh Lake, southeast of Whitehorse (Fig. 2; Table 1). Here, the sandstone is primarily composed of feldspar (85%), volcanic lithic fragments (10%), and only minor quartz (Bickerton, 2014). Detrital zircons have a prominent peak at ca. 198 Ma (93% of grains) and only minor Paleozoic grains (Fig. 6; Appendix 3 [see footnote 1]). The youngest Unmix age of 194.3 ± 2.7 Ma (Fig. DR3 [see footnote 1]) is in close agreement with the peak age, and the weighted mean of the 17 youngest grains is 190.4 ± 2.8 Ma (Fig. DR2 [see footnote 1]); a Sinemurian–Pliensbachian maximum depositional age is also indicated for this sample (Table 1).

Sample 11MC183 is arkosic sandstone from an outcrop of sandstone-siltstone rhythmite of the Richthofen formation near Carcross (Fig. 2; Table 1). Detrital zircons essentially have a unimodal distribution with a peak at ca. 213 Ma and only four grains yielding Paleozoic dates (380–320 Ma; Fig. 6; Appendix 3 [see footnote 1]). The ca. 213 Ma peak (Fig. DR1 [see footnote 1]) and ca. 209 Ma youngest Unmix age (Fig. DR3 [see footnote 1]) indicate a Norian maximum depositional age, and the weighted mean of 10 young grains (203.6 ± 2.6 Ma) suggests a Rhaetian maximum depositional age (Fig. DR2 [see footnote 1]; Table 1).

Two samples of Tanglefoot formation were collected from road cuts along the Robert Campbell Highway, east of Carmacks near the north end of the trough (Eagles Nest and 04MC002; Fig. 2). The Eagles Nest sample is coarse-grained sandstone collected from a succession of maroon-weathering sandstones and poorly sorted conglomerates immediately east of Eagles Nest bluff, a prominent exposure of Hancock limestone along the Yukon River (Fig. 2; Table 1). This outcrop was tentatively correlated with the Mandanna member by Long (2005) on account of its maroon weathering color and its position immediately above limestone of the Hancock member. Clasts of the conglomerate are dominated by volcanic and subvolcanic rocks but also include lesser limestone, red mudstone, and, rarely, eclogite lithologies. Red siltstone-mudstone interbeds (<10 cm) locally display desiccation cracks. Long (2005) interpreted these strata as deposits of a gravel-bed meandering river system. Detrital zircons have a nearly unimodal distribution with a peak at ca. 201 Ma and very few (6%) Paleozoic grains (Fig. 6; Appendix 3 [see footnote 1]). Its youngest peak suggests a Hettangian maximum depositional age, but the youngest Unmix age (197.74 ± 1.30 Ma; Fig. DR3 [see footnote 1]) and the weighted mean of 19 young grains support a Sinemurian age (193.8 ± 1.8 Ma; Fig. DR2 [see footnote 1]; Table 1). This indicates that these strata are part of the Laberge Group rather than the older Mandanna member. We interpret them as fluvial deposits of the Tanglefoot formation.

Sample 04MC002 is a coarse-grained, brownish-yellow weathering calcareous, arkosic sandstone interbedded with minor black shale. It is composed mainly of poorly sorted, subangular grains of feldspar, quartz, and volcanic lithic fragments, and it is commonly graded. It is from ∼1.5 km west of a tuff horizon in the Tanglefoot formation dated at 187.1 ± 0.7 Ma (Pliensbachian; Colpron and Friedman, 2008) and along strike (4 km) from a bivalve assemblage suggesting a probable early Bajocian age (Colpron et al., 2007b). This biostratigraphic age is consistent with nearby occurrence of isolated chert clasts in the Tanglefoot formation, which elsewhere occur in Bajocian and younger strata (Hart, 1997; Clapham et al., 2002; Shirmohammad et al., 2011). Although separated by a fault, this locality is close to overlying Upper Jurassic strata of the Tantalus Formation (Fig. 2). Detrital zircons have a nearly unimodal population with a peak at ca. 196 Ma and only few (6%) Paleozoic grains (Fig. 6; Appendix 3 [see footnote 1]). The youngest Unmix age is ca. 185 Ma (Fig. DR3 [see footnote 1]), and the weighted mean of six young grains is ca. 182 Ma (Fig. DR2 [see footnote 1]; Table 1). The Sinemurian to Toarcian maximum depositional age suggested by detrital zircons (Table 1) is older than the age of local fossil collections but broadly consistent with the nearby tuff horizon.

In summary, eight samples from Laberge Group sandstones yield similar results, with dominant populations of Late Triassic–Early Jurassic zircons (peaks at 213–186 Ma) and subsidiary populations of Paleozoic grains (predominantly Mississippian; Fig. 6). Mesozoic dates match well with Late Triassic–Middle Jurassic plutonic suites surrounding the Whitehorse trough (compilation of igneous ages shown at top of Fig. 6; see also Appendix 4 [see footnote 1]). Paleozoic dates (mostly Mississippian and mid- to Late Permian) match known sources in the Paleozoic parts of the Stikinia, Quesnellia, and Yukon-Tanana terranes.

Cache Creek Graywacke

A sample of graywacke (11LB045) from the youngest strata of the Cache Creek terrane was collected east of Carcross to test possible links between this succession and the Whitehorse trough, as proposed by Monger et al. (1991). It is a medium-grained graywacke composed of equal amounts of quartz and hornblende, and subordinate plagioclase and clinopyroxene, from a monotonous section of interbedded graywacke and siltstone (Bickerton, 2014). Nearby (∼1 km west) radiolarian chert is Middle to Late Triassic in age (Cordey et al., 1991). Detrital zircons have a similar pattern as those of the Laberge Group, with nearly unimodal distribution of Late Triassic–Early Jurassic dates and a peak at ca. 208 Ma (Fig. 6; Appendix 3 [see footnote 1]). Paleozoic grains only comprise 3% and show similar dates (Mississippian and Permian) as those of the Laberge Group. The youngest peak (ca. 208 Ma; Fig. DR1 [see footnote 1]) and Unmix age (ca. 202 Ma; Fig. DR3 [see footnote 1]) indicate a Rhaetian maximum depositional age, and the weighted mean of the 21 youngest grains supports a Hettangian age (199.5 ± 2.0 Ma; Fig. DR2 [see footnote 1]; Table 1). These data are similar to those from Richthofen sample 11MC183 to the west (Figs. 2 and 6) and strongly suggest correlation of these youngest Cache Creek strata with the Whitehorse trough.

Macauley Ridge Formation

Sandstone matrix from two nearby conglomerate localities of the Macauley Ridge formation was collected south of Beaver Creek in western Yukon (Fig. 5B; Table 1; 06DM174 and 06DM176). They have similar detrital zircon signatures (Appendix 2 [see footnote 1]), shown in a composite probability plot in Figure 7, with most grains at 220–180 Ma (peak at ca. 198 Ma, Sinemurian) and minor Paleozoic grains (8%; ca. 334 and ca. 263 Ma). Calculated maximum depositional ages are Sinemurian to Pliensbachian (Table 1; Figs. DR1, DR2, and DR3 [see footnote 1]). This age distribution and maximum depositional age are similar to that of Laberge Group sandstones (Figs. 6 and 7, bottom), clearly suggesting that the Macauley Ridge formation had similar sources as the Laberge Group. However, the formation is distinct in its abundance of metamorphic clasts, which are rare in the Laberge Group in Yukon. In northern British Columbia, metamorphic clasts are common in Toarcian to Aalenian strata of the Laberge Group (Mihalynuk et al., 1999; Shirmohammad et al., 2011).

Faro Peak Formation

Two samples of matrix-supported, polymictic, pebble to boulder conglomerate of the Faro Peak formation were collected near Faro in central Yukon (Fig. 5A; Table 1; 29LB06 and 34LB06) as part of a detrital zircon study of Triassic clastic successions in Yukon (Beranek, 2009). A Late Triassic depositional age for the Faro Peak was estimated by Tempelman-Kluit (1979b) and Pigage (2004) based on conodont faunas in limestone clasts. The samples are from near the base of the formation, where it overlies micaceous quartzite of the Yukon-Tanana terrane (Fig. 5A; Snowcap assemblage; Colpron et al., 2006). They contain clasts of limestone, chert, micaceous quartzite, volcanic porphyry, mica schist, and thinly laminated argillite. The detrital zircon signatures from these two samples are similar (Appendix 2 [see footnote 1]) and are shown as a composite probability plot on Figure 7. As with the Macauley Ridge samples, the Faro Peak samples have dates similar to those in Laberge Group sandstones (Figs. 6 and 7, bottom). Most zircons (60%) range from 220 to 187 Ma, with a peak at ca. 200 Ma and a minor population (6%) of mostly Mississippian grains (Fig. 7; Appendix 2 [see footnote 1]). A distinguishing feature is the large proportion (26%) of Precambrian grains (ca. 2730–1050 Ma), which could reflect the stratigraphic setting of the Faro Peak conglomerate directly above the Snowcap quartzite (Yukon-Tanana terrane), a potential reservoir for Precambrian zircons of northwestern Laurentian affinity (Piercey and Colpron, 2009). It is interesting to note that both conglomerates from the Faro Peak and Macauley Ridge formations contain significant proportions of metamorphic clasts likely derived from local Yukon-Tanana sources, yet only the Faro Peak samples yielded the Precambrian signal typical of Yukon-Tanana detrital zircons (Gehrels et al., 1991; Nelson and Gehrels, 2007; Piercey and Colpron, 2009). A Hettangian to Sinemurian maximum depositional age is indicated for the Faro Peak formation (Table 1; Figs. DR1, DR2, and DR3 [see footnote 1]).

Tantalus Formation

Four samples of Tantalus Formation were collected, three on Tantalus Butte near Carmacks and one east of Division Mountain (Fig. 2; Table 1). Samples TB1 and TB2 were collected in the open pit at the historical Tantalus coal mine, north of Carmacks. TB1 is from a conglomerate 6 m above the main coal seam, and sample TB2 is from sandstone ∼20 m below the main coal seam. TB3 is a conglomerate collected on the south flank of Tantalus Butte, near the portal to underground workings at the Tantalus mine (now collapsed). Sample C1 is from the upper half of the exposed section at the north end of Corduroy Mountain, east of Division Mountain (Fig. 2). The detrital zircon patterns for these samples (Fig. 8; Appendixes 2 and 3 [see footnote 1]) show overall strong similarities with those of older Laberge Group strata (Figs. 6 and 8, bottom), with a predominance of Late Triassic–Early Jurassic dates (peaks between 197 and 186 Ma) and minor populations of mainly Mississippian and Permian grains. This supports petrological observations (Long, 2015) that the Tantalus conglomerates were derived in part from recycling of older strata and from other source areas common with those of the Laberge Group. The Tantalus samples are distinct from the Laberge Group in their relative abundance of grains younger than 180 Ma (9%–24%) and, for Tantalus Butte samples (TB1, TB2, TB3), subsidiary peaks at ca. 159–147 Ma (Fig. 8). These younger dates are particularly interesting in that they are not represented in known igneous rocks of the Intermontane terranes (compare histogram at top of Fig. 8), but they are more common in the more distant Insular terranes to the west (Alexander and Wrangellia; Fig. 1; Gehrels et al., 2009) and parts of southern Stikinia in British Columbia (Evenchick et al., 2010, and references therein). To further investigate this youngest population of exotic zircons, we selected the two samples from the Tantalus mine open pit for replicate LA-ICP-MS analyses and more precise CA-ID-TIMS analysis of six grains from each sample that yielded young LA-ICP-MS dates (TB1 and TB2; Fig. 8; Appendix 3 [see footnote 1]). TB1 yielded CA-ID-TIMS dates between 170 and 159 Ma (Fig. 9; Table 2). The two youngest dates are equivalent with a weighted mean of 159.20 ± 0.08 Ma, indicating an Oxfordian maximum depositional age (Table 1). TB2 yielded more tightly clustered CA-ID-TIMS dates (Fig. 9; Table 2). The five youngest dates are equivalent with a weighted mean of 148.51 ± 0.06 Ma, suggesting a Tithonian maximum depositional age (Table 1). If the youngest detrital zircon approximates the depositional age of the Tantalus strata, then the section in the open pit of the Tantalus mine may be repeated by a thrust fault at (or near) the main coal seam, as the sample collected above the coal (TB1) is apparently older than sample TB2 collected 20 m below the coal. The other sample from Tantalus Butte (TB3) also has a maximum depositional age of Oxfordian–Kimmeridgian, whereas the Corduroy Mountain sample (C1) was deposited after the Bajocian–Bathonian (Table 1), i.e., somewhat older than previous estimates for the Tantalus Formation.

DISCUSSION

Detrital zircons from nine samples from the Mandanna member and Laberge Group in Whitehorse trough have similar age distributions, with a dominant peak in the Late Triassic to Early Jurassic and subordinate peaks in the Paleozoic (Fig. 6). Mesozoic zircons are mainly 220–180 Ma and sourced locally from Late Triassic to Early Jurassic plutonic suites (and their inferred eroded volcanic covers) that intrude terranes surrounding the trough (Figs. 2 and 6, top; Appendix 4 [see footnote 1]). These sources are consistent with interpretations derived from the few previously dated igneous clasts in conglomerates and limited detrital zircon dates from sandstones of the Laberge Group (Hart et al., 1995; Johannson et al., 1997; Gordey et al., 1998; Shirmohammad et al., 2011).

Paleozoic zircons are ubiquitous (∼12% of analyzed grains) but represent significant populations (>30%) in only two samples (05W170 and GL04-108b) that are from west of Whitehorse (Figs. 2 and 11). Paleozoic grains are typically Mississippian (72% of Paleozoic zircons) or Permian (∼13%) and match well with igneous events documented in Paleozoic successions of Stikinia and Quesnellia, and in the Yukon-Tanana terrane, which surrounds the trough on all sides in Yukon (Figs. 1 and 2). The majority of Mississippian zircons cluster at 340–330 Ma (Fig. 6; Appendixes 2 and 3 [see footnote 1]), and a greater abundance in samples from west of Whitehorse (05W170 and GL04-108b; Figs. 2 and 11A) suggests a possible point source for Mississippian zircons. The Mandanna sample (05W170; Rhaetian) has the greatest proportion of Paleozoic zircons (34%; Fig. 6; Appendix 2 [see footnote 1]), indicating that Paleozoic source rocks were exhumed prior to deep subsidence recorded in the Laberge Group. Similar relative abundances of Paleozoic zircons (31%) in nearby basal strata of the Richthofen formation (GL04-108b; Pliensbachian–Toarcian maximum depositional age; Table 1) could indicate recycling of the Mandanna sediment, which occurs below the sub-Laberge unconformity at this locality (Fig. 10), or continued supply from the same sources as older strata. Other possibilities are that denudation of Paleozoic sources persisted throughout Early Jurassic or that there was rejuvenation of the Paleozoic source region in the Pliensbachian–Toarcian.

Four samples were collected from basal strata of the Laberge Group (08MC091, 04SJP594, GL04-108b, and Eagles Nest; Fig. 2; Table 1). The maximum depositional ages from the two samples collected near the central axis of the Whitehorse trough (Eagles Nest and 08MC091) suggest a Hettangian–Pliensbachian age for the sub-Laberge unconformity, whereas the two samples collected at the western edge of the trough have Pliensbachian to Toarcian maximum depositional ages (04SJP594 and GL04-108b; Table 1; Figs. 2 and 11). Assuming deposition occurred shortly after the youngest detrital zircon was crystallized, which is a characteristic of clastic sediments deposited along convergent plate margins (e.g., Cawood et al., 2012), our limited data set from basal strata suggests westerly transgression of the trough between the Sinemurian and Toarcian, following early uplift of the western flank of the trough and progressive subsidence in the Early Jurassic—an hypothesis easily tested with further sampling from basal strata and more precise determination of maximum depositional ages using the CA-ID-TIMS method. This early uplift is consistent with the greater abundance of Paleozoic grains in older strata of the Mandanna member (Fig. 6) and apparent continued supply from this westerly source in the Early Jurassic.

In general, peak ages of Late Triassic–Early Jurassic detrital zircons in the Mandanna and Laberge samples appear to be in good agreement with available depositional age constraints for these strata. The latest Triassic depositional age for Mandanna sample 05W170 (Table 1) is consistent with a dated tuff horizon in an exposure ∼15 km to the south (Figs. 2, 6, and 10). Similarly, Sinemurian–Pliensbachian fossil collections 10–15 km along strike south of Richthofen sample 04SJP603 are consistent with its ca. 197–190 Ma calculated maximum depositional ages (Fig. 6; Table 1). Tanglefoot sample 04MC002 has maximum depositional ages that range from ca. 196 Ma (or Sinemurian) to ca. 182 Ma (Toarcian; Table 1) and broadly overlap a dated tuff horizon (187.1 ± 0.1 Ma) 1.5 km to the west (Pliensbachian; Colpron and Friedman, 2008), but it is distinctly older than Bajocian bivalves collected 4 km along strike to the south.

The detrital zircon signature for a graywacke from the youngest strata of the Cache Creek terrane (11LB045) is similar to nearby samples of the Laberge Group (11MC183 and 12LB220; Fig. 6), supporting correlation of these strata with the Whitehorse trough. Cordey et al. (1991) interpreted the Lower Jurassic graywacke as accretionary wedge sediments, and these likely represented more distal and deeper(?) facies than turbidites deposited in the submarine fan and delta systems of the Laberge Group to the north and west (Lowey, 2008, p. 189).

Detrital zircons from the Faro Peak and Macauley Ridge formations also display strong similarities with those of the Laberge Group (Fig. 7). Hettangian–Pliensbachian maximum depositional ages (Table 1; Fig. 7) suggest that onset of deposition in these more isolated exposures occurred broadly at the same time as in the Whitehorse trough. However, the Faro Peak and Macauley “basins” are distinct in that they occupy positions near the outer edges of the Intermontane terranes, rather than at their core like the Whitehorse trough (Fig. 1). Restoration of 430–490 km of Eocene dextral displacement along the Tintina fault (Gabrielse et al., 2006) places the Faro Peak and Macauley “basins” to the north (present day) of the trough, overlapping the Yukon-Tanana and Slide Mountain terranes at the inboard edge of the Intermontane terranes in a foreland-ward position. Their different setting is reflected in the higher proportion of metamorphic clasts in conglomerates of the Faro Peak and Macauley Ridge formations, but the similarity in detrital zircon signatures with the Laberge Group suggests a common source area. One possibility is that the trough was originally more extensive and covered a large area of the northern Intermontane terranes in the Sinemurian. Alternatively, there may have been distinct, isolated basins receiving sediments from common sources. The relative abundance of Precambrian zircons in the Faro Peak formation (Appendix 2 [see footnote 1]) suggests mixing of different sources in the Intermontane terranes and the western Laurentian continental margin.

The petrology of metamorphic rocks in the Yukon-Tanana terrane, and of the plutonic rocks that intrude them, suggests deep exhumation (15–20 km) of the basin’s shoulders in the Early Jurassic (e.g., Johnston et al., 1996; McCausland et al., 2002; Tafti, 2005; Berman et al., 2007). Thermobarometric data from amphibolite-facies rocks in the Yukon-Tanana terrane northwest of the Whitehorse trough record dynamic burial of these rocks to 7.5–9.0 kbar in the Early Jurassic (ca. 195–187 Ma; Berman et al., 2007; Johnston and Erdmer, 1995). Although not as well studied, indications are that Yukon-Tanana rocks east of the trough likely experienced a similar history (Hansen, 1992). The latest Triassic to Early Jurassic granitoid plutons that surround the trough (and intrude Yukon-Tanana terrane; Figs. 2 and 12) were apparently emplaced during exhumation of the metamorphic infrastructure, with early phases recording Al-in-hornblende crystallization pressures of 5–7 kbar (consistent with wall rocks), and younger phases characterized by pegmatites and miarolitic cavities (Johnston et al., 1996; McCausland et al., 2002; Tafti, 2005). The predominance of ca. 200–180 Ma 40Ar/39Ar and K-Ar mica cooling dates in the Yukon-Tanana terrane (with peak in cooling ages at ca. 190 Ma; Bennett et al., 2010; Allan et al., 2013) indicates that it was exhumed to upper-crustal levels by the Early Jurassic (Fig. 12). Knight et al. (2013) provided documentation for an Early Jurassic (ca. 190 Ma) crustal-scale extension fault that accommodated, in part, ∼15–20 km of exhumation of the metamorphic infrastructure (Willow Lake fault; Fig. 12; Colpron and Ryan, 2010; Ryan et al., 2010). This suggests that thickening of the Yukon-Tanana crust in the Early Jurassic was rapidly followed by regional extension and exhumation. These events were coeval with rapid subsidence of the Whitehorse trough in the Sinemurian–Pliensbachian and probably led to flank uplift and activation of source areas for sediments of the Laberge Group.

The Yukon-Tanana terrane originated as a peri-Laurentian arc that developed outboard of the contemporaneously opening Slide Mountain backarc ocean between Late Devonian and mid-Permian time (e.g., Mortensen, 1992; Nelson, 1993; Mihalynuk et al., 1999; Nelson et al., 2006; Colpron et al., 2007a). The Slide Mountain Ocean was closed by its subduction beneath the Yukon-Tanana terrane by mid- to Late Permian, and the Yukon-Tanana terrane was juxtaposed with the Laurentian margin by the Permian–Triassic (Beranek and Mortensen, 2011). Subsequent arc development in the Middle to Late Triassic was built (at least in part) upon this reconfigured continental margin, implying that at least the Quesnellia arc was built on attenuated North American crust (e.g., Murphy et al., 1995; Fig. 13A). Stikinia probably extended into an intra-oceanic arc festoon similar to the present-day Aleutian arc (Mihalynuk et al., 1994). The Kutcho arc, a ca. 264–242 Ma intra-oceanic arc assemblage that occupies the lowest structural level in the Cache Creek terrane (Gabrielse, 1998; Schiarizza, 2012), is inferred to have entered the Stikinia subduction zone by the end Norian–Rhaetian, shutting down the Late Triassic (Lewes River–Stuhini) magmatism in Stikinia by ca. 212–206 Ma (Logan and Mihalynuk, 2014; Fig. 13B). Accretion of the Kutcho arc is inferred to have induced deformation of Triassic strata in Stikinia (cf. Brown et al., 1996; Logan and Mihalynuk, 2014; Greig, 2014) and caused profound exhumation of Paleozoic sources, providing detritus to the Mandanna member (Fig. 13B). Counterclockwise rotation of Stikinia, leading eventually to entrapment of the Cache Creek terrane (CC; following the model proposed by Mihalynuk et al., 1994), probably occurred in the latest Triassic. In the late Rhaetian, renewed subduction beneath Stikinia led to deposition of the lower Hazelton Group in British Columbia (e.g., Barresi et al., 2015) and emplacement of ca. 204–195 Ma plutons (Fig. 13C). At that time, subduction is inferred to have occurred on both sides of Stikinia (Marsden and Thorkelson, 1992). At the northern apex of the Intermontane terranes, crustal thickening and subsequent exhumation of Yukon-Tanana terrane heralded the rapid subsidence of the Whitehorse trough in the Sinemurian. Evidence for crustal thickening followed by regional exhumation of the Yukon-Tanana terrane suggests that convergence between Stikinia, Cache Creek, Quesnellia, and the North American margin had begun by the Sinemurian–Pliensbachian, and thus that the northern Whitehorse trough developed in part as a piggyback (or wedge-top) basin during the initial stages of development of the northern Cordilleran orogen (Figs. 13C–13D).

Faunal assemblages and paleomagnetic data suggest that Stikinia was ∼1200 ± 680 km south of its present position relative to the North American craton in the Sinemurian–Pliensbachian, placing the Whitehorse trough at similar latitude as present-day southern Alberta (Smith et al., 2001; Smith, 2006; Kent and Irving, 2010; Fig. 13D). At that time (ca. 195–185 Ma), the Yukon-Tanana terrane was being exhumed all around the northern Whitehorse trough, while the first signs of orogenic activity were being recorded in the hinterland of the Cordillera in British Columbia (Fig. 13D). In central British Columbia, Nixon et al. (1997) reported a U-Pb zircon date of ca. 186 Ma for late pegmatite in the Polaris complex, an Alaskan-type mafic-ultramafic complex related to the development of the Quesnellia arc (Fig. 1). The Polaris complex is a sill-like intrusion with a synkinematic thermal aureole at its base, recording east-directed thrusting of Quesnellia over rocks of the North American margin. Further south, in the Kootenay arc, Murphy et al. (1995) provided a similar, albeit less precise, estimate of ca. 185 Ma based on dates on plutons intruding the Intermontane terranes (including the Cooper stock; Fig. 1) after initial imbrication with North American strata.

The timing of these orogenic events in the hinterland of the northern Cordillera also coincides with onset of subsidence in the Alberta foreland basin (Fernie Formation; Price, 1973, 1994; Cant and Stockmal, 1989). Asgar-Deen (2003) and McCartney (2012) proposed that the basal members (Sinemurian–Toarcian) of the Fernie Formation in the foothills of the Rocky Mountains were deposited in a backbulge depozone (Jacobi, 1981; DeCelles and Giles, 1996) and thus record initial, albeit minor, subsidence related to development of the western Canada foreland basin (Fig. 13D). Eastward migration of the forebulge occurred in the Middle Jurassic and is recorded by the sub–upper Fernie unconformity (Oxfordian–Kimmeridgian; Poulton, 1989). In the Kimmeridgian, the upper Fernie Formation (Passage beds) was deposited in a rapidly subsiding foredeep that received the first influx of westerly derived detritus from the Cordilleran orogen (Poulton, 1989). Subsidence of the foreland basin beginning in the Early Jurassic is an indication that the nascent Cordilleran orogen began imposing its load onto the North American lithosphere by at least the Sinemurian.

The Whitehorse trough has traditionally been interpreted as a forearc basin that developed during convergence of the Intermontane terranes with North America (e.g., Tempelman-Kluit, 1979a). The predominance of arc detritus in the Laberge Group (and rare occurrences of ultrahigh-pressure minerals as detritus in Sinemurian–Pliensbachian strata; MacKenzie et al., 2005; Canil et al., 2006; our Eagles Nest sample) is consistent with a forearc setting. However, the trough developed >40 m.y. after the Early Triassic suture between the Intermontane terranes and the North American margin (Beranek and Mortensen, 2011), facing the collapsing Cache Creek “ocean” (Mihalynuk et al., 1994; Fig. 13). Evidence of Early Jurassic crustal thickening and rapid exhumation of the basin shoulders (to the west, north, and east; ca. 195–187 Ma) in Yukon suggest that the northern Whitehorse trough had developed into a collisional, piggyback (or wedge-top) marine basin atop the nascent northern Cordilleran orogen by late Pliensbachian time (Fig. 13D). This basin was probably continuous to the south into an active forearc basin that persisted until cessation of subduction and imbrication of the Cache Creek terrane in the Middle Jurassic (ca. 174–172 Ma; Mihalynuk et al., 2004; Figs. 13D–13E). By the Bajocian, Laberge Group marine sedimentation ended, and the imbricated Intermontane terranes were stitched by Middle Jurassic plutons (ca. 174–168 Ma; Fig. 2; Mihalynuk et al., 2004).

The end of deposition in the Whitehorse trough in the Middle Jurassic heralded the onset of deposition in the Bowser Basin to the south, the next great piggyback marine basin of the northern Cordillera (Fig. 1; Aalenian–Bajocian–Aptian; Ricketts et al., 1992; Evenchick et al., 2007, 2010). Onset of Bowser Basin development corresponds with influx of chert-pebble conglomerate in the Aalenian–Bajocian (Ricketts et al., 1992; Gagnon et al., 2009; Evenchick et al., 2010). The first chert clasts appeared in the Bajocian in the Whitehorse trough as well, but at the end of deposition. The trough therefore can be viewed as a northern precursor to the Bowser Basin, perhaps indicating the southward propagation of convergence within the Intermontane terranes in the Jurassic (Fig. 13; e.g., Ricketts et al., 1992).

Fluvial deposition of the Tantalus Formation began by the Bathonian or Callovian upon the newly uplifted Intermontane terranes (Fig. 13F). The prominent signal in detrital zircon populations from the Tantalus Formation is best interpreted as recycling of underlying Laberge Group strata (or continued supply from similar sources), but the population of zircons younger than 168 Ma cannot be linked to local sources in the Intermontane terranes (Figs. 8 and 9). The period of ca. 165–125 Ma was a magmatic lull in the northern Intermontane terranes (Breitsprecher and Mortensen, 2004), but arc magmatism was active between 157 and 142 Ma in the western Coast belt (Insular terranes; e.g., Gehrels et al., 2009). At that time, the Insular arc was separated from the Intermontane orogenic welt by deep marine basins of the Gravina-Dezadeash-Nutzotin belt (e.g., Berg et al., 1972; McClelland et al., 1992; Fig. 13F). Therefore, the ca. 159–149 Ma zircons in the Tantalus Formation were likely derived from air-fall tuff contributions from the approaching Insular terranes, rather than through direct fluvial connections. Evenchick et al. (2010) proposed a similar origin for zircons in coeval strata of the Bowser Basin. Late Jurassic global climate models show annual ranges of northeasterly to southeasterly wind directions for the Northern Hemisphere (Moore et al., 1992). These windblown zircons signal the proximity of the Insular terranes by the Oxfordian (Fig. 13F). Structural evidence in the Coast Mountains suggests that interactions between the Insular and Intermontane terranes began even earlier in the Early to Middle Jurassic (McClelland and Gehrels, 1990; van der Heyden, 1992; Fig. 13).

Development of northern Cordilleran synorogenic basins continued with deposition in the Bowser, Gravina-Dezadeash-Nutzotin, and other basins to the west in Alaska, which record the Jurassic–Cretaceous interactions between the Intermontane and Insular terranes (Figs. 1 and 13F; e.g., Berg et al., 1972; McClelland et al., 1992; Trop et al., 2002). The Jurassic development of the Whitehorse trough in the northern North American Cordillera also bears resemblance to development of basins related to the Pliocene–Pleistocene collision of the Luzon arc and progressive southward stalling of the Manilla Trench near Taiwan (e.g., Suppe, 1984; Chiang et al., 2004).

Finally, the relationships and interpretations of Cordilleran tectonic evolution documented in this study, particularly evidence for Early Jurassic interactions between the Intermontane terranes and the western North American margin, are incompatible with models that invoke assembly of the Intermontane terranes as part of a distant ribbon continent that was only accreted to western North America in the Late Cretaceous (e.g., Johnston, 2008; Hildebrand, 2009).

CONCLUSIONS

Detrital zircons from Upper Triassic to Middle Jurassic strata of the Mandanna member and Laberge Group indicate local sediment sources in rapidly exhuming Paleozoic and early Mesozoic arc terranes that flanked the Whitehorse trough (Yukon-Tanana, Stikinia, Quesnellia, and related plutons). Similar zircon populations in isolated conglomeratic occurrences of the Faro Peak and Macauley Ridge formations, and in Lower Jurassic strata of the Cache Creek terrane, support correlations of these strata with the Laberge Group. The petrology and mica cooling dates of metamorphic and plutonic rocks in the Yukon-Tanana terrane suggest that burial to 20–30 km in the Early Jurassic was followed by profound exhumation (15–20 km) within <20 m.y. (ca. 200–180 Ma). These events were coeval with subsidence in the Whitehorse trough (Sinemurian–Pliensbachian), onset of compressional deformation in central and southern British Columbia (ca. 187–185 Ma, or Pliensbachian), and initial subsidence in the western Canada foreland basin in the Sinemurian–Toarcian. They collectively record the early stages in the development of the northern Cordilleran orogen. The Whitehorse trough records deposition in a forearc basin that progressively evolved into a collisional, synorogenic piggyback basin, from north to south, during the Early to Middle Jurassic amalgamation of the Intermontane terranes. Deposition in the trough ended in the Middle Jurassic, and overlying Upper Jurassic fluvial deposits of the Tantalus Formation received airborne ash from the Late Jurassic–Early Cretaceous arc that developed atop the approaching Insular terranes to the west and southern Stikinia.

The data and interpretations presented here were collected over a decade of field work in and around Whitehorse trough by the authors, and involved collaborations and discussions with many, including: Grant Abbott, Steve Gordey, Craig Hart, Grant Lowey, Jim Monger, JoAnne Nelson, Steve Piercey, Jim Ryan, and the late John Wheeler. Balz Kamber and Thomas Ulrich produced the U-Pb data for sample 08MC091 at Laurentian University. Janet Gabites analyzed the Macauley Ridge samples at the Pacific Centre for Isotopic Research of the University of British Columbia. Matt Hutchison, JoAnne Nelson, and Terry Poulton provided comments on an earlier version of this manuscript. Reviews by Mitch Mihalynuk and an anonymous reader helped improve the manuscript. This is Yukon Geological Survey contribution #023.

1GSA Data Repository Item 2015263, analytical methods, LA-ICP-MS results, compilation of igneous ages, and calculation of maximum depositional ages, is available at www.geosociety.org/pubs/ft2015.htm, or on request from editing@geosociety.org, Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301-9140, USA.