Abstract

The along-strike variability in sediment provenance within the Nanaimo basin is important for understanding the tectonic evolution of North America’s Late Cretaceous Pacific margin, providing context for paleogeographic reconstructions. Here, we provide 35 point-counted sandstone samples and 22 new detrital zircon samples from the Nanaimo basin. These new detrital zircon samples compose a portion of a basin-wide data set (N = 49, n = 10,942) that is leveraged to discern spatio-temporal changes in sediment provenance. Provenance data demonstrates that the majority of Nanaimo basin strata were sourced from regions within and east of the Coast Mountains Batholith, while only the southernmost Nanaimo basin, exposed in the San Juan Islands, was supplied sediment from the North Cascade thrust system. Additionally, near-identical age modes and synchronous changes in detrital zircon facies are used to hypothesize a correlation between the Nanaimo Group and the protolith of the Swakane Gneiss. These observations, along with previously identified events in the Cordillera, are used to define two basin-wide events that affected the Nanaimo basin: the first at 84 Ma and the second at 72 Ma. The first event is correlated to the onset of Kula-Farallon spreading, which affected basin subsidence, introduced Proterozoic detrital zircon to the central and southern Nanaimo basin, and uplifted the North Cascade thrust system. The second basin-wide event, which is speculated to have been driven by increased rates of subduction and obliquity, resulted in localized high-flux events in the arc, increased exhumation of the Cascade Crystalline Core, underplating of the Swakane Gneiss, and coarse-grained sedimentation across the basin. The data presented here provides added context for the evolution of the basin and provides insight into the protracted geodynamics of forearc basins undergoing oblique subduction.

INTRODUCTION

Detrital minerals preserved in sedimentary basins reflect tectonic processes that deform, uplift, or bury sediment sources in the hinterland (Pettijohn et al., 1972; Ingersoll, 1978; Dickinson, 1982; Fedo et al., 2003; Whitchurch et al., 2011; Sharman et al., 2015; Orme and Laskowski, 2016; Horton, 2018). As such, the tectonic evolution of orogenic belts is often deciphered from their eroded, detrital components preserved in sedimentary basins (Dickinson, 1982; Fedo et al., 2003; Gehrels et al., 2008). However, the interpretation of tectonic events from basin stratigraphy is complicated by diachronous, along-strike variations that are common in convergent orogens (e.g., Kley et al., 1999; Yin, 2006; Horton, 2018). Along-strike variability can be controlled by several factors, including: changes in magmatic tempo (e.g., Paterson and Ducea, 2015; Cecil et al., 2018), style of deformation and uplift (e.g., Whitchurch et al., 2011; Fitz-Díaz et al., 2018), or arc/basement lithology (e.g., Sundell et al., 2018). Thus, the use of spatially and temporally limited provenance data sets to examine the evolution of hinterland source areas may lead to an incomplete understanding of hinterland evolution or the extent of tectonic mechanisms that drive sedimentation (e.g., Yin, 2006; Cecil et al., 2018; Horton, 2018). Large, basin-scale data sets paired with information regarding hinterland evolution (e.g., exhumation rates, radiometric ages bracketing deformation) reveal intimate details of orogenic processes and their effects on basin sedimentation (e.g., Whitchurch et al., 2011; Saylor et al., 2012; Sundell et al., 2018).

The Nanaimo basin of British Columbia, Canada, is one of many sedimentary basins that sequestered sediments along the Pacific margin of the North American Cordillera throughout Late Cretaceous to Early Paleogene time. These basins preserve a record of the final building stages of the Cordillera (Dickinson, 1976; Surpless et al., 2014; Busby, 2004), including periods of hinterland deformation (Mustard, 1994; Brown, 2012; Sharman et al., 2015), high-flux volcanism (Kimbrough et al., 2001; Englert et al., 2018), changes in subduction angles (Moxon and Graham, 1987; Dickinson et al., 1983; Fitz-Díaz et al., 2018), and differential uplift rates (Saleeby, 2003; Busby, 2004). Therefore, provenance analysis of basin fill deposits, such as the Nanaimo Group, can aid in understanding the protracted diachroneity, spatial extent, and interaction of these processes within the framework of the Late Cretaceous to Paleogene evolution of the Cordillera.

The Late Cretaceous to Paleocene Nanaimo basin is preserved adjacent to many of its sediment source regions (Mustard et al., 1995; Katnick and Mustard, 2003; Brown, 2012; Matthews et al., 2017), including: (1) the Coast mountains batholith, an eroded magmatic arc with along-strike variations in magmatic tempo (Friedman and Armstrong, 1995; Gehrels et al., 2008; Cecil et al., 2018); (2) the Wrangellia terrane, which forms the basement of the basin (Greene et al., 2010); and (3) the North Cascade thrust system, a fold-and-thrust belt composed of lithologically diverse terranes located at the southern margin of the basin (e.g., Brandon et al., 1988; Brown and Gehrels, 2007; Brown et al., 2010; Brown, 2012; Sauer et al., 2017). Although proximal sediment sources are established, interpretations of distal sediment sources are contentious, and have been used to interpret the Late Cretaceous paleolatitude of the basin (Matthews et al., 2017, and references therein). Additionally, the presence, relative sediment contributions, and structural evolution of the different source areas have been used to account for the primary mechanism of basin subsidence. Thus, interpretations of the tectonic setting include forearc (Muller and Jeletzky, 1970; England, 1990; Matthews et al., 2017; Englert et al., 2018; Kent et al., 2019), foreland (Brandon et al., 1988; Mustard, 1994), and strike-slip basin (Pacht, 1984).

Only recently have large-n laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) detrital zircon techniques been used to unravel contentious ideas about the origin of Nanaimo basin strata (Matthews et al., 2017; Englert et al., 2018; Huang et al., 2019; Kent et al., 2019). However, a basin-wide investigation of spatial and temporal changes in provenance has the potential to answer numerous outstanding questions regarding sediment transfer along the Pacific margin of North America. For example, the relative supply from regional sources (e.g., North Cascade thrust system and Coast Mountains Batholith) may help determine basin forming processes. Additionally, the protracted record of the Nanaimo Group (Turonian to Danian) allows for a spatio-temporal analysis of the interplay between subduction, magmatism, and sedimentation in forearc regions along the 160-km-long outcrop belt.

In this study, we use an integrated provenance data set including sandstone petrography and data from 22 new detrital zircon samples (n = 4534) to deduce provenance variability along a 160-km-long depositional-strike-oriented transect of the Nanaimo basin. These new data, along with existing data from Matthews et al. (2017) and Englert et al. (2018), compose a basin-wide detrital zircon database (N = 49, n = 10,942) that provide insight into the along-strike influence of sediment source areas and hinterland/subduction zone processes during basin filling. In doing so, we demonstrate: (1) the sequential emplacement of the terranes that compose the North Cascade Thrust system and their recycling into a limited portion of the Nanaimo basin; (2) the protolith of the Swakane Biotite-Gneiss was derived from source terranes similar to the Nanaimo Group; and (3) a correlation of changes in the Kula plate subduction vectors with changes in both basin sedimentation patterns and hinterland evolution. These results provide strong constraints on the drivers of Nanaimo Group sedimentation and tectonic processes during a contentious period in the geologic history of the northern Cordillera, which can be applied to understand the large-scale geodynamics of forearc systems.

GEOLOGIC SETTING OF THE NANAIMO BASIN

The North American Cordillera is composed of numerous allochthonous fault-bounded terranes that accreted to the western margin of Laurentia as well as volcanic arcs, which, in some cases, stitch terranes together (Monger et al., 1982, and many others). Jurassic to Eocene subduction beneath western North America formed a volcanic arc that spanned from Washington, USA, to Yukon, Canada, the eroded remnant of which is known as the Coast Mountains Batholith (Monger et al., 1982). The Coast Mountains Batholith composes a minor portion of the basement of the Late Cretaceous to Paleogene Nanaimo basin (Mustard, 1994). Plutonism in the Coast Mountains Batholith contains both orogen-wide (ca. 150 Ma) and more localized high-flux events (e.g., Friedman and Armstrong 1995; Gehrels et al., 2009; Beranek et al., 2017; Cecil et al., 2018). Of relevance to the Nanaimo basin are: (1) the southern Coast Mountains Batholith, which ranges in age between 160 and 40 Ma and is associated with high-flux events that occurred between 85–70 Ma and 61–48 Ma (Cecil et al., 2018); (2) the Vancouver Coast Mountains Batholith, which ranges in age between 180 and 60 Ma and is associated with high flux events that occurred between 105–90 Ma and 70–60 Ma (Friedman and Armstrong, 1995) (Fig. 1); and (3) the Cascade Crystalline Core (CCC), which is the most deeply exhumed portion of the Coast Mountains Batholith (Tabor et al., 1987) (Fig. 1), composed of metamorphosed plutonic rocks, volcanic rocks, and forearc sediments that were incorporated into the arc at mid-to-deep-crustal depths (e.g., Swakane Biotite-Gneiss) (Tabor et al., 1987; Matzel et al., 2004; Ducea and Chapman, 2018; Sauer et al., 2017, 2018). Magmatic activity in the CCC spans 96–45 Ma, and high flux events occurred between 96–89 Ma and 78–71 Ma (Miller et al., 2009). Due to changes in subduction and crustal thickening in the Coast Mountains Batholith, magmatism migrated eastward through time and no plutonic activity occurred within 100 km of the modern coastline post–70 Ma (Cecil et al., 2018).

The Insular Superterrane, composed of Alexander and Wrangellia terranes, accreted to the western margin of the Coast Mountains Batholith and already-accreted Intermontane Superterrane during the Late Jurassic to Early Cretaceous. The Wrangellia terrane constitutes the majority of the basement of the Nanaimo basin, composed of >10 km of igneous rocks and lesser sedimentary strata, which ranges in age from Devonian to Jurassic (Yorath et al., 1999; Greene et al., 2010). The timing and latitude of Insular Superterrane accretion to North America is debated. Some suggest accretion occurred during or prior to the Early Jurassic (e.g., Monger et al., 1982; Monger and Gibson, 2019), while others suggest accretion took place during the Early Cretaceous (e.g., Sigloch and Mihalynuk, 2017).

After the collision of the Insular Superterrane, Coast Mountains Batholith magmatism was largely driven by subduction of the Farallon plate. This subduction continued, roughly normal to the margin of North America, until the Farallon plate rifted into the Kula and Farallon plates. Magnetic lineations indicate that seafloor spreading at the Pacific-Kula ridge was ongoing by 79 Ma (Atwater, 1989; Seton et al., 2012), although rifting as early as 85 Ma has been estimated (Woods and Davies, 1982; Engebretson et al., 1985). Rifting reoriented the subduction vector from roughly normal to the North American margin to highly oblique (Woods and Davies, 1982; Engebretson et al., 1985; Doubrovine and Tarduno, 2008). Oblique subduction of the Kula Plate drove the accretion of additional terranes (e.g., Eddy et al., 2017), structural evolution of the Nanaimo basin (e.g., Kent et al., 2019), and changes in stress regime within the Cordillera (e.g., Umhoefer and Miller, 1996).

The southern margin of the Nanaimo basin is defined by the North Cascade thrust system (NCTS). The NCTS is composed of multiple fault-bounded terranes, exposed between the west side of the Fraser–Straight Creek fault zone and Haro Strait (Fig. 1B). Terranes incorporated in the NCTS vary in age from early Paleozoic to Early Cretaceous (Vance 1975; Garver, 1988; Rubin et al., 1990) and are lithologically (Vance, 1975; Brandon et al., 1988; Garver, 1988; Schermer et al., 2018), geochronologically (Brown and Gehrels, 2007; Brown et al., 2010; Brown, 2012; Sauer et al., 2017; Schermer et al., 2018), and thermochronologically diverse (Johnson et al., 1986; Brandon et al., 1988; Brown et al., 2005). Small sedimentary successions within the NCTS (e.g., exposures on Barnes Island and Blakey Island) have been correlated to the Nanaimo Group based on lithostratigraphic descriptions (Vance, 1975) and detrital zircon age distributions (Brown, 2012) (Fig. 1). The thrust nappes within the NCTS include the Chilliwack nappe, Orcas nappe, Shuskan nappe, and Haystack nappe, listed in order of ascending structural level (Brandon et al., 1988) and decreasing relative emplacement age (Brown, 2012). Detrital zircon maximum depositional ages and provenance relationships were used by Brown (2012) to suggest that the terranes exposed in the San Juan Islands were juxtaposed and uplifted between ca. 93 and post–87 Ma, and subsequently recycled into the Nanaimo basin during deposition of the Nanaimo Group.

The Nanaimo basin, along with many of the geologic elements that compose the tectonic framework surrounding the Nanaimo basin (i.e., Coast Mountains Batholith, NCTS, and CCC) are hypothesized to have been located between 1000 and 3000 km south of their current location during the Late Cretaceous (Umhoefer, 1987; Cowan et al., 1997). This hypothesis, known as the Baja-BC hypothesis, was originally established through the use of paleomagnetic methods (e.g., Cowan et al., 1997; Enkin et al., 2001); however, it has since been tested using structural field relationships (e.g., Wyld et al., 2006), biostratigraphic correlations (e.g., Carter and Haggart, 2006), and U-Pb detrital zircon methods (e.g., Matthews et al., 2017), with no clear result. As translation is hypothesized to have occurred during the Late Cretaceous (e.g., Cowan et al., 1997; Enkin et al., 2001), the Nanaimo basin has been a focus of this debate (Matthews et al., 2017, and references therein).

FORMATION OF THE NANAIMO BASIN AND DEPOSITION OF THE NANAIMO GROUP

The Nanaimo basin formed atop the Wrangellia terrane during the early Late Cretaceous (i.e., Turonian) between the Coast Mountains Batholith to the east and the subducting Kula-Farallon Plate to the west (Muller and Jeletzky, 1970; Haggart et al., 2005). Although the Nanaimo basin is in a forearc position, Cretaceous-age thrust sheets within the Coast Mountains Batholith (Rusmore and Woodsworth, 1994; Journeay and Friedman, 1993; Umhoefer and Miller, 1996) and NCTS (Brandon et al., 1988) are commonly thought to have supplied sediment to the basin throughout its evolution (Pacht, 1984; Mustard, 1994; Brown, 2012) (Fig. 1). As such, the Nanaimo basin has been interpreted as a forearc (e.g., Dickinson, 1976; Muller and Jeletzky, 1970; England, 1990; Matthews et al., 2017; Englert et al., 2018; Kent et al., 2019) or a foreland basin (e.g., Brandon et al., 1988; Mustard, 1994; Katnick and Mustard, 2003; Brown, 2012).

The Nanaimo Group is divided into lower and upper intervals based on regional lithostratigraphic correlations and interpreted depositional settings (Clapp, 1914; Muller and Jeletzky, 1970; Mustard, 1994) (Fig. 2A). The lower Nanaimo Group is composed of both non-marine and marginal-marine strata (Mustard, 1994; Jones et al., 2018; Huang et al., 2019; Kent et al., 2019) (Fig. 2A). The upper Nanaimo Group is composed of deep-water depositional system deposits (e.g., submarine canyon or channel deposits) that record sediment routing during the Late Cretaceous to Paleogene (Fig. 2A) (England and Hiscott, 1992; Rowe et al., 2002; Bain and Hubbard, 2016; Englert et al., 2018, 2019).

Provenance of the Nanaimo Group

The sediment provenance of the Nanaimo Group has been interpreted based on sandstone and conglomerate clast petrology (Pacht, 1984; Mustard, 1994; Katnick and Mustard, 2003; Brown, 2012), and U-Pb detrital zircon age distributions (Mustard et al., 1995; Housen and Beck, 1999; Mahoney et al., 1999; Brown, 2012; Matthews et al., 2017; Englert et al., 2018; Huang et al., 2019). The adjacent Coast Mountains Batholith is considered the dominant sediment source for the basin, as evidenced by abundant plutonic detritus, volcanic lithic fragments in the Nanaimo Group sandstones, and Mesozoic detrital zircon ages (Mustard, 1994; Matthews et al., 2017; Fig. 1). The rocks of the NCTS are thought to be a local source of detritus to the central and southern portions of the basin during the deposition of the lower Nanaimo Group providing chert and metamorphic clasts (Pacht, 1984; Brandon et al., 1988), as well as Precambrian zircon (e.g., Mustard et al., 1995; Brown, 2012).

Previous detrital zircon work has identified two detrital zircon facies (cf. LaMaskin, 2012) in the Nanaimo basin, each composed of various age distributions: (1) dominantly Mesozoic ages (150 and 90 Ma age modes); and (2) both Mesozoic (84 and 77 Ma age modes) and Proterozoic ages (1700 and 1380 Ma age modes) (Mustard et al., 1995; Matthews et al., 2017). Mesozoic ages are interpreted to be sourced from the Coast Mountains Batholith (e.g., Mustard et al., 1995; Matthews et al., 2017), whereas source areas spanning from northern Idaho to southern California have been proposed for the Proterozoic zircons, including the Belt-Purcell basin (Mustard et al., 1995; Mahoney et al., 1999; Dumitru et al., 2016), Yavapai-Mazatzal basement (Housen and Beck, 1999), the NCTS (Brown, 2012), and the Mojave-Sonoran region (Matthews et al., 2017). Matthews et al. (2017) dated a transect of detrital zircon samples through the northern Nanaimo basin (N = 16, n = 3964; Fig. 2B) and demonstrated that Proterozoic grains were introduced to the basin during deposition of the Geoffrey Formation, ca. 72–70 Ma. The introduction of these grains has been interpreted as the eastward expansion of river drainages through the Coast Mountains Batholith into sediment source areas east of the arc (Matthews et al., 2017; Englert et al., 2018).

METHODS

Sandstone Petrology and Point Counting Methods

Thirty-five thin sections were cut from sandstones that were used for detrital zircon analysis. Thin sections were stained for potassium feldspar and a compositional and textural analysis was performed using a Nikon Eclipse LV100 petrographic microscope paired with a digital stepping stage. PetrogTM software was used to aid in the tabulation of 400 points per sample. Framework clasts were classified as quartz (chert, monocrystalline, or polycrystalline), feldspar (plagioclase or potassium-feldspar), biotite, muscovite, amphibole, hornblende, clinopyroxene, epidote, and rock fragments (volcanic, plutonic, metamorphic, or sedimentary). Standard Q-F-R (Pettijohn et al., 1972) and Qm-F-Lt (Dickinson et al., 1983) ternary diagrams were used to characterize sandstone composition and broadly interpret tectonic setting. Complete petrologic data is available in GSA Data Repository File DR11.

U-Pb Detrital Zircon Geochronology Methods

Twenty-two new detrital zircon samples from the southern Nanaimo basin were analyzed (Fig. 2). These samples were added to the existing samples of Matthews et al. (2017) and Englert et al. (2018) to compose four transects referred to herein as: (1) the northern Gulf Islands transect, collected in the Comox area (Comox, Extension, and Protection formations), on Denman Island (Cedar District, DeCourcy, and Northumberland formations), and on Hornby Island (DeCourcy, Geoffrey, Spray, and Gabriola formations); (2) the central Gulf Island transect, collected in the Nanaimo area (Extension, Protection, and DeCourcy formations), on Mudge Island (DeCourcy Formation), and Gabriola Island (Northumberland through Gabriola formations) (Figs. 2A and 2B); (3) the southern Gulf Islands transect, collected across Galiano Island (Geoffrey and Gabriola formations), Mayne Island (DeCourcy, Geoffrey, and Gabriola formations), and Saturna Island (DeCourcy, and Geoffrey formations) (Figs. 2A and 2C); and (4) the San Juan Islands transect, collected across Orcas Island (Comox, Haslam, and Extension formations) and Sucia Island (Protection and Cedar District formations) (Figs. 2 and 2D).

For each sample, detrital zircon grains were isolated using standard mineral separation methods (Fedo et al., 2003). Three hundred grains were measured for each sample using LA-ICP-MS to obtain U/Pb ratios and dates following the procedure of Matthews and Guest (2017). The method used to calculate date uncertainties is in alignment with community derived best-practices outlined in Horstwood et al. (2016) but is slightly modified from Matthews and Guest (2017) (see Data Repository for details). In this study 206Pb/238U dates were used for grains yielding dates <1000 Ma whereas 207Pb/206Pb dates were used for grains yielding dates >1000 Ma. Complete isotopic data for the analyzed samples are available in Data Repository File DR2.

The toolset detritalPy (Sharman et al., 2018) was used to create probability density functions (PDFs) for individual samples and groups of samples presented here, as well as for additional published data sets (e.g., Brown, 2012; Sauer et al., 2018). Previously published data were filtered and presented in the same manner as new data reported here to enable comparisons. Reported systematic uncertainties were used to calculate total uncertainties for individual dates (Stotal). When systematic uncertainties were not reported, 1.5% (2σ) was added in quadrature to reported 238U/206Pb ratio uncertainties and 1% (2σ) was added in quadrature to reported 207Pb/206Pb ratio uncertainties.

Calculation of Detrital Zircon Maximum Depositional Ages

Maximum depositional ages of the 22 detrital zircon samples presented here, along with 27 samples presented in Matthews et al. (2017) and Englert et al. (2018), were calculated with Isoplot (v. 4.15; Ludwig, 2012) using the youngest grain cluster that overlaps at 2σ uncertainty method (YGC 2σ) (Dickinson and Gehrels, 2009). The YGC 2σ method is calculated by taking the weighted mean, weighted by ratio uncertainty, of the youngest three or more dates that overlap within uncertainty (Coutts et al., 2019). All dates used in maximum depositional age calculations were obtained from 206Pb/238U ratios with a probability of concordance >1% and uncertainties were propagated in accordance with the methodology outlined in Horstwood et al. (2016).

The YGC 2σ method has been demonstrated to produce conservative ages (i.e., rarely younger than the true depositional age of the deposit), especially when derived from large-n sample sizes or samples with abundant near-depositional ages (Coutts et al., 2019). Weighted average plots for all samples, including those from Matthews et al. (2017) and Englert et al. (2018), are provided in File DR3 (see footnote 1).

Multi-Dimensional Scaling of U-Pb Detrital Zircon Populations

Multi-dimensional scaling plots are a useful tool for visualizing the statistical differences between large numbers of samples (Vermeesch, 2013; Spencer and Kirkland, 2016). In this study, a single multi-dimensional scaling plot was calculated using detritalPy (Sharman et al., 2018) for all large-n detrital zircon samples in the basin, the 22 new samples presented here, and the 27 previously published zircon samples from Matthews et al. (2017) and Englert et al. (2018). Four synthetic age populations (1700, 1380, 150, 90 Ma) were added to the multi-dimensional scaling plot to aid in the visualization of differences between individual samples (Spencer and Kirkland, 2016).

RESULTS

Sandstone Petrology and Point Counting Results

The majority of Nanaimo Group sandstones are arkosic arenites (N = 30) (Table 1) (QFR; Pettijohn et al., 1972). Exceptions include samples DC2, NU3, Or-CO (lithic arenites), Or-HA (sub-litharenite), and Or-EX (sub-arkose) (Table 1) (QFR; Pettijohn et al., 1972). Samples are largely dominated by feldspathic grains (representing between 7% and 82% of framework grains; mean = 35%), and quartz grains (representing between 10% and 50% of counted framework grains; mean = 33%). Chert, micas (biotite and muscovite), and rock fragments (volcanic, plutonic, and metamorphic) are minor components and compose 5% of the framework grains on average (Table 1). Of note, samples of the northern, central, and southern Gulf Islands transects are dominantly composed of feldspathic grains with scarce chert; conversely, samples of the San Juan Islands transect are compositionally distinct and characterized by an increased abundance of chert, and rock fragments, with scarce feldspathic grains (Table 1; Fig. 3).

Nanaimo Group samples populate four fields in the Qm-F-Lt classification scheme of Dickinson et al. (1983): basement uplift, dissected arc, transitional arc, and transitional recycled (Fig. 3; Table 1). In the northern Gulf islands transect, samples of the DeCourcy Formation plot within the transitional arc field, whereas samples of the Extension, Geoffrey, and Gabriola formations plot within the transitional arc and dissected arc fields (Fig. 3A). In the central and southern Gulf Islands transects, the lower Nanaimo Group and DeCourcy Formation plot within in the basement uplift field (Figs. 3B and 3C). Geoffrey Formation samples plot within dissected and transitional arc fields, except for the Geoffrey Formation exposed on Saturna Island (SatGFt), which plots in the basement uplift field (Fig. 3C; Table 1). Samples of the Geoffrey Formation that plot within the dissected and transitional arc fields have an increased proportion (>5%) of volcanic lithic clasts compared to underlying or overlying basement derived compositions (Table 1). Gabriola Formation samples have compositions that plot within the dissected and transitional arc fields, except for samples of the uppermost Gabriola Formation exposed on Galiano and Mayne Islands (GalGBt, MayGBt), which plot within the basement uplift field (Figs. 3B and 3C; Table 1). In the San Juan Island transect, samples from Orcas Island (Comox, Haslam, and Extension formations) plot within the transitional recycled field, and samples from Sucia Island (Protection and Cedar District formations) plot within the transitional arc and dissected arc fields, respectively (Fig. 3D).

U-Pb Detrital Zircon Geochronology Results

Uranium-lead measurements from 22 samples presented here yielded 4534 dates that passed our filtering criteria. The data presented here combined with the data of Matthews et al. (2017) and Englert et al. (2018) form a basin-wide data set of 49 samples and 10,942 dates. A single PDF is displayed for each of the 49 samples considered (Fig. 4). Proterozoic ages compose 16% of the zircons dated and are dominantly bimodal, ranging from 1900 to 800, with age modes at 1700 and 1380 Ma (Table 2). Paleozoic ages compose 1% of the ages dated and range from 550 to 270 Ma with sample-specific age modes at 260, 350, 400, and 450 Ma. Mesozoic ages compose 82% and have sample-specific age modes at 150, 90, 80, and 75 Ma. Archean ages are rare in the Nanaimo basin, representing 0.2% of dated grains.

Maximum depositional ages for samples presented here range from 144.5 ± 2.0 (SuPr) to 67.7 ± 7 Ma (GabSP) and are listed with the sample name in Figure 4. The end of lower Nanaimo Group sedimentation is bracketed by samples PR4 (85.2 ± 1.2 Ma) and GabPRt (84.3 ± 1.2 Ma).

INTERPRETATIONS

Sandstone Petrology and Point Counting Interpretations

We attribute the source and evolution of sediment composition in the Gulf Islands transects (northern, central, and southern) to reflect increased erosion and exhumation of the Coast Mountains Batholith, both through time, and along-strike north-to-south. In the northern Gulf Islands transect, sandstone compositions reflect a shift from transitional arc source regions to dissected arc source regions (Figs. 3A and 5A). We interpret samples from the central and southern Gulf Islands transects to be dominantly derived from uplifted plutonic basement; however, the Geoffrey Formation samples from these transects demonstrate a >5% increase in volcanic lithic fragments and a shift in sandstone composition that reflects erosion of dissected arc source regions. This shift in provenance reflects increased volcanic activity in the Cascade Crystalline core and southern Coastal Mountains Batholith during deposition (e.g., Miller et al., 2009; Cecil et al., 2018) (Table 1; Figs. 3B, 3C, and 5B). Return of sample compositions to basement uplift source terranes by the uppermost Gabriola Formation reflects short-lived hinterland volcanism during the deposition of the Geoffrey Formation (Table 1; Figs. 3B, 3C, and 5B).

We interpret that sandstone compositions of the San Juan Islands transect record the transition from diverse sediment sources in the catchment basin (OrCO, OrHA, OrEX) to more restricted sediment sources (SuPR, SuCD) (Table 1; Fig. 5C). This is documented by a decrease in the variety of framework grains (Table 1; File DR1). The lithic-rich composition, reflected by the transitional- and lithic-recycled sedimentary origin (Fig. 3D), including common chert and exotic framework grains (e.g., amphibole, epidote), suggests derivation from sediment source areas that were not significant for deposits of the Nanaimo Group samples in the Gulf Islands transects yet consistent with sediment sources within the NCTS (e.g., Orcas chert; Pacht, 1984; Brandon et al., 1988) (Table 1; Fig. 5; File DR1).

U-Pb Detrital Zircon Geochronology Interpretations

Detrital Zircon Facies

Five detrital zircon facies are identified based on common age distributions within component zircon spectra, as reflected in a multi-dimensional scaling plot (Fig. 6A), and the presence or absence of key modes in the various samples (Fig. 6B). Zircon facies 1 (F1) is dominated by Mesozoic age grains with modes at 150 and 90 Ma and a minor age mode at 120 Ma. Proterozoic ages are rare (<2%) (Fig. 6; Table 2). Some north-to-south variation is noteworthy in F1 samples, with the northern Gulf Islands transect containing a dominant 90 Ma peak and a subordinate 150 peak, and the southern Gulf Islands transect containing a dominant 150 Ma peak and a subordinate 90 Ma peak (Fig. 4).

Zircon facies 2 (F2) is dominated by Mesozoic age grains that have a primary age mode between 80 and 70 Ma and minor age modes between 250 and 110 Ma. Proterozoic age grains contribute between 15% and 57% of the ages in each sample and have age modes at 1700 and 1380 Ma (Fig. 6; Table 2). This facies varies internally by the prominence of the 90 Ma peak and the age of the youngest grains in each sample (Fig. 4).

Zircon facies 3 (F3) is dominated by Mesozoic age grains that have a bimodal distribution, including age modes at 156 and 115 Ma. Additional minor and discontinuous populations are located between 600 and 300 Ma. Proterozoic age grains constitute ∼1% of the facies with ages between 2550 and 1100 Ma. OrEX is the only sample with this signature (Fig. 6; Table 2).

Zircon facies 4 (F4) is dominated by Mesozoic age grains that have a dominant age mode at 146 Ma and a minor age mode at 190 Ma. Paleozoic age grains compose minor modes at 402, 315, and 265 Ma. Proterozoic age grains constitute ∼30% of the facies and have broad age distributions that range from 1900 to 800 Ma and 700 to 551 Ma. SuPR is the only sample with this signature (Fig. 6; Table 2).

Zircon facies 5 (F5) is composed of a single unimodal zircon population centered at 150 Ma, ranging from 229 to 111 Ma. Remaining Phanerozoic age grains range from 522 to 415 Ma. Proterozoic and older grains range from 2000 to 1700 Ma and 1200 to 900 Ma. Non-Mesozoic ages constitute ∼3% of the facies. SuCD is the only sample with this signature (Fig. 6; Table 2).

Provenance of Crystalline Zircon Ages

Detrital zircon ages of the Nanaimo basin range from 58.7 to 2888.5 Ma and are plausibly derived from crystalline sources that reflect much of the geologic record of North America. However, as zircon is highly resistant to chemical and physical abrasion, it may be recycled multiple times through various geologic processes (e.g., Fedo et al., 2003; Hadlari et al., 2015). Therefore, zircon grains in the Nanaimo basin may have ultimately derived from the crystalline sources discussed below but were subsequently recycled from other sedimentary units prior to being deposited in the basin.

Proterozoic zircon populations in F2 (e.g., 1850–1560 and 1450–1300 Ma) may derive from rocks that make up much of the basement of Laurentia. Zircon with these ages are common in the Yavapai and Mazatzal terranes (1790 and 1610 Ma) (Whitmeyer and Karlstrom, 2007), Granite Rhyolite Province, additional A-type granites (1490 and 1310 Ma) (Goodge and Vervoot, 2006), and the Grenville orogen (1370 and 1000 Ma) (Whitmeyer and Karlstrom, 2007). Proterozoic ages of F4 (e.g., 1700–541 Ma) may derive from similar sources as F2 with the addition of Appalachian and Caledonides orogens (850–541 Ma) (Anfinson et al., 2012).

The Paleozoic zircon populations found in Nanaimo Group samples (i.e., modes at 450, 400, and 350 Ma) are rare in crystalline rocks of North America. Zircons of this age (i.e., 450–350 Ma) are found in oceanic and continental volcanic arcs within the Caledonian and Ellesmerian orogens (Anfinson et al., 2012). Rocks of the Alexander Terrane have age modes that range from 460 to 415 Ma (White et al., 2016). Orthogneiss within the Yellow Aster Complex range in age from 406 to 410 Ma (Schermer et al., 2018). Intrusive dykes within the Yellow Aster Complex and Turtleback terrane range in age from 418 to 375 Ma. (Brown et al., 2010; Schermer et al., 2018). Volcanic strata of the Sicker Group, a dominant portion of the Wrangellia terrane, range in age from 370 to 340 Ma (Sluggett, 2003).

Mesozoic to early Paleogene age zircon found in all Nanaimo Group detrital zircon facies are common to volcanic and plutonic rocks of western North America. The Wrangellia terrane is composed of multiple Late Triassic to Late Jurassic units, including the Karmutsen Formation (ca. 220 Ma), Bonanza Formation (ca. 202–165 Ma), and the Island Intrusions–West Coast Crystalline Complex (ca. 190 and 167 Ma) (DeBari et al., 1999; Greene et al., 2010; Canil et al., 2013). The magmatic history of the Coast Mountains Batholith spans much of the Mesozoic (Cecil et al., 2018). Plutonism in the Coast Mountains Batholith spans 200–60 Ma (Cecil et al., 2018), with previously discussed age modes representative of regional high-flux magmatic events (e.g., Friedman and Armstrong, 1995; Miller et al., 2009; Cecil et al., 2018).

Sediment Source Areas for Nanaimo Group Detrital Zircon Facies

Detrital zircon facies 1 (F1) is composed of Mesozoic zircon grains that we interpret to derive from the Coast Mountains Batholith, as the range of ages and modes coincides with high-flux events within the southern and Vancouver Coast Mountains Batholith, as well as the Cascade Crystalline Core (CCC) (90 and 150 Ma age modes) (Fig. 6) (Friedman and Armstrong, 1995; Miller et al., 2009; Cecil et al., 2018). Zircon age populations that compose the Paleozoic mode in samples OrCO, OrHA, and GabEX (ca. 350 Ma) are interpreted herein to derive from the underlying Sicker Group of the Wrangellia Terrane (Figs. 4A and 4C).

Detrital zircon facies 2 (F2) is composed of Mesozoic grains that we interpret to derive from the Coast Mountains Batholith and Proterozoic ages that derive from sediment sources east of the arc (Figs. 6 and 7A) (Matthews et al., 2017; Englert et al., 2018). We interpret grains yielding latest Cretaceous ages to derive from the CCC and southern Coast Mountains Batholith, where plutonism continued until 71 Ma (Miller et al., 2009) and post–60 Ma (Cecil et al., 2018), respectively. Previously interpreted sources of Proterozoic detrital zircon characteristic of F2 samples (1700 and 1380 Ma age modes) (e.g., Belt-Purcell basin, Lemhi sub-basin, of Idaho and Mojave-Sonoran region of SW North America) are still applicable, and the data presented here do not refine the sediment source interpretation.

Detrital zircon facies 3 (F3) is composed of grains that we interpret to derive from the NCTS with minor contribution from the Coast Mountains Batholith. We interpret zircon grains yielding ages that fall within the 120 Ma and 150 Ma age modes to derive from the Bell Pass mélange and rocks of the Shuskan/Haystack nappes, respectively (Fig. 7). Zircon grains yielding ages ca. 85 Ma are interpreted here to derive from the Coast Mountains Batholith, as Late Cretaceous ages are rare in the NCTS. We interpret Paleozoic age grains to be sourced from the Chilliwack Terrane, Turtleback Complex, or Yellow Aster Complex of the Chilliwack nappe, which all contain prominent ca. 400 Ma age modes (Fig. 7). Proterozoic age zircon grains in F3 are rare (1%); however, we interpret them to derive from the NCTS as they have a large range of ages similar to that of Yellow Aster Complex (2000–900 Ma), and dissimilar age range and age modes to the Proterozoic ages of F2 (Fig. 6). The age modes present in F3 are similar to the age distribution of detrital zircons found in Nanaimo basin strata on Barnes Island reported by Brown (2012) (Fig. 7). Brown (2012) interpreted source areas within the NCTS (i.e., Yellow Aster and Turtleback complexes) for the Barnes Island samples, based on zircon age modes and petrologic analysis of clasts.

We interpret detrital zircon facies 4 (F4) to derive from rocks of the NCTS. A preponderance of grains yielding latest Jurassic to Early Cretaceous ages (150 Ma age mode) and an absence of grains younger than 100 Ma (Figs. 5B and 5C) are attributed to sediment recycling from the Shuskan and Haystack nappes (Fig. 7). Sedimentary rocks with nearly unimodal zircon age modes at ca. 150 Ma (Figs. 5B and 5C) is consistent with this interpretation, rather than derivation from the Coast Mountains Batholith. Paleozoic ages in F4 (400 Ma age mode) may derive from a variety of strata within the Chilliwack nappe (i.e., the Yellow Aster complex, Turtleback complex, and Chilliwack terrane) and Orcas nappe (i.e., Bell Pass mélange) where these ages are common (Fig. 7). Lastly, detrital zircon facies 4 samples contain a small number of Proterozoic ages that are interpreted herein to derive from the Yellow Aster Complex as they are dissimilar to those found in Facies 2 (i.e., larger range of ages, and lack of bimodal distribution) (Figs. 6 and 7).

We interpret detrital zircon facies 5 (F5) to derive from rocks of the NCTS. The facies is composed of a single age mode at ca. 150 Ma and component grains that we interpret to derive entirely from strata of the Shuskan and/or Haystack nappes (Fig. 7). Absence of younger, mid-to-Late Cretaceous ages suggests that derivation from the Coast Mountains Batholith is unlikely.

The Nanaimo Group deposits exposed along the northern transect are characterized by F1 and F2 detrital zircon populations, and the transition between F1 and F2 occurs ca. 71 Ma. The Nanaimo Group deposits exposed in the central and southern transects are characterized by F1 and F2 detrital zircon populations; however, the transition between F1 and F2 occurs ca. 84 Ma. Lastly, the Nanaimo Group deposits exposed in the San Juan Island transect are characterized by F1, F3, F4, and F5 detrital zircon populations. The transition from F1 to F3 occurs ca. 84 Ma, while the transition from F3 to F4 and later F5 occurs post–84 Ma (Fig. 4).

DISCUSSION

Influence of the North Cascade Thrust System on the Nanaimo Basin

Previous models for provenance evolution of the Nanaimo Group have interpreted the NCTS as a major sediment source and driver of sediment transfer to the southern Nanaimo basin (e.g., Pacht, 1984; Mustard, 1994). These interpretations are largely based on the identification of distinct clasts in the Nanaimo Group that are interpreted to derive from units within the NCTS (e.g., Vance, 1975; Pacht, 1984; Brown, 2012). The most widespread of these are radiolarian chert clasts interpreted to derive from the Triassic to Lower Jurassic Orcas Chert of the Orcas nappe (Fig. 7), which have been identified in the lower Nanaimo Group as far north as the Cowichan River area (Fig. 1) (Pacht, 1984). Other exotic clasts that are interpreted to be sourced from the NCTS (e.g., clasts of metamorphic rock in Nanaimo Group equivalents exposed on Sucia, Saturna, and Waldron islands) are much more localized to the southernmost portions of the Nanaimo basin (Pacht, 1984; Brown, 2012).

Our petrographic and detrital zircon data refines these interpretations and suggest that sediment shed from the NCTS (i.e., F3, F4, and F5) was deposited in a spatially restricted part of the Nanaimo basin, immediately adjacent to the NCTS. These data demonstrate that while chert clasts are common within the Comox, Haslam, and Extension formations in the San Juan Island transect, the remainder of the Nanaimo Group is largely devoid of chert (Table 1). Furthermore, detrital zircon facies that we interpret to derive from the NCTS (F3, F4, F5) are limited to the San Juan Island transect (Figs. 6 and 7). Interestingly, samples from the San Juan Islands transect that are most chert-rich (e.g., OrCO, OrHA) exhibit the same F1 detrital zircon signatures as chert-poor samples of the Nanaimo Group from elsewhere in the basin (e.g., CO1) (Figs. 4 and 6).

If the NCTS only contributed sediment to the southernmost part of the Nanaimo basin, an alternative source is needed for the small quantities of chert found within the lower Nanaimo Group exposed far from the NCTS (e.g., GabPRt, EX). It is possible that chert in the lower Nanaimo Group derived from minor chert units within strata of Wrangellia (Pacht, 1984; Brandon et al., 1986), which is thought to have formed basement highs that controlled sedimentation of the lower Nanaimo Group (Pacht, 1984; Mustard, 1994; Kent et al., 2019), as well as acted as a sediment source (Huang et al., 2019). The lack of chert in the upper Nanaimo Group in these regions could signal burial of Wrangellian basement during progressive onlap of the basal unconformity during deposition of lower Nanaimo Group (Johnstone et al., 2006; Kent et al., 2019). Taken together, our data suggest a relatively minor influence of the NCTS as a sediment source and driver of the Nanaimo basin evolution. Although the NCTS was emplaced prior to or during Nanaimo Group deposition, it only supplied sediment to the southernmost portion of the basin with little impact on the basin as a whole. Contraction within the NCTS may have caused enhanced subsidence locally within the southern portion of the basin; however, contraction in the Coast Mountains Batholith ended prior to the deepening of the Nanaimo basin (e.g., Journeay and Friedman, 1993; Rusmore and Woodsworth, 1994). As such, an additional driver for basin-wide subsidence ca. 84 Ma must have been important.

Implications for North Cascade Thrust System Emplacement Timing

There is disagreement regarding the timing and mode of tectonic emplacement of the NCTS, and its role in basin geodynamics. Brandon et al. (1988) suggested that the NCTS were assembled during a single event synchronous with the collision of the Insular Superterrane during the latest Jurassic and Early Cretaceous, then uplifted during the Late Cretaceous ca. 84 Ma. In contrast, Brown (2012) used maximum depositional ages and detrital zircon age distributions of sedimentary bodies within the NCTS (e.g., Barnes Island, Upright Head) to suggest the NCTS was assembled via sequential emplacement of structurally higher nappes onto the southern margin of the basin.

Our detrital zircon and petrographic data provide new constraints on the timing and sequence of nappe emplacement. The lowermost Nanaimo Group deposits of the San Juan Island transect (OrCO and OrHA; 87–86 Ma) contain detrital zircons interpreted to derive from the Coast Mountains Batholith (F1); however, they contain abundant chert that has previously been interpreted to derive from the Orcas nappe, suggesting the Orcas and Chilliwack nappes may have been emplaced by that time (cf. Brown, 2012). If the chert found in these lowermost samples derives from the Orcas nappe, then the underlying Chilliwack nappe was not a significant sediment source at this time as few ca. 400 Ma grains and no Proterozoic grains are present in these samples. The earliest detrital zircon evidence of NCTS uplift is found in the Extension Formation on Orcas Island (F3) ca. 84 Ma, where sediment is dominantly derived from the lower nappes (i.e., Chilliwack and Orcas). During deposition of the Protection and Cedar District Formations on Sucia Island (ca. 83 Ma; Ward et al., 2012) sediment was derived from upper nappes (i.e., Haystack nappes).

The data provided here demonstrates sequential thrusting of NCTS terranes onto the southern margin of the basin, consistent with the model of Brown (2012). The lowermost nappes (Chilliwack and Orcas) were emplaced by 84 Ma, but it is plausible they were emplaced earlier but only the Orcas chert was eroded into the basin. From ca. 84 to ca. 83 Ma the remaining nappes were emplaced (i.e., Shuskan and Haystack nappes), covering the lower nappe, with subsequent sediment derived only from the structurally highest units. In the model of Brandon et al. (1988), where construction of the nappe pile occurred elsewhere along the margin and was thrust into contact with the Nanaimo basin, detrital zircon age populations would either be sourced from the highest structural levels first, or simultaneously from all structural levels, depending on the geometry of the nappe pile.

Stratigraphic Record of Sediment Underplating

The transition from detrital zircon F1 to F2 (i.e., shift to inclusions of significant Proterozoic dates) in the Nanaimo basin occurs diachronously throughout the basin. Previously, Matthews et al. (2017) observed that F2 detrital zircon populations were introduced to the northern Gulf Island transect at the onset of Geoffrey Formation deposition, 71.6 ± 1.9 Ma (Fig. 4C). Englert et al. (2018) demonstrated that F2 detrital zircon populations sourced from east of the Coast Mountains Batholith occurred synchronously with deposition of the Geoffrey Formation (ca. 71 Ma), which is characterized by the basin-wide deposition of large gravel-filled submarine channel-system deposits. The data presented here demonstrates that the introduction of F2 detrital zircon populations occurs 10–15 m.y. earlier in the central and southern Gulf Islands transects (ca. 84 Ma). This observation requires that the mechanism(s) responsible for the transition from F1 to F2 detrital zircon populations in the central and southern portions of the basin (e.g., fluvial drainage expansion) are different from the mechanism that controlled basin-wide gravel dispersal during deposition of the Geoffrey Formation.

Previous authors have discussed the similarity between the detrital zircon age distributions of the Swakane Gneiss and Nanaimo Group, but no definitive correlation between them has been made (e.g., Matzel et al., 2004; Sauer et al., 2018). Correlation is difficult as the Swakane Gneiss is composed of forearc/trench sediments that were incorporated into the Coast Mountains Batholith at depths of ∼40 km, and subsequently metamorphosed (Matzel et al., 2004, Sauer et al., 2018). Sauer et al. (2018) demonstrated that the Swakane Gneiss contains two dominant detrital zircon populations, which are very similar to F1 and F2 described here: (1) rocks yielding maximum depositional ages >84 Ma are characterized by populations with Mesozoic age modes at 90, 110, and 150 Ma (Fig. 8A); and (2) rocks yielding maximum depositional ages <84 Ma are characterized by populations with major age modes at 70 and 80 Ma, minor modes at 90 and 110 Ma, sparse ages between 110–300 Ma, and Proterozoic age populations between 1800 and 1600 and 1440–1360 Ma (Sauer et al., 2018) (Fig. 8A). The similarity between detrital zircon populations in the Swakane Gneiss and those of the Nanaimo Group, as well as the similarity in the timing of the transition from F1 to F2 age distributions within the Swakane Gneiss (ca. 86 and 84 Ma) and the Southern Nanaimo basin (ca. 84 Ma) suggest they share a similar provenance (Fig. 8A). We propose that the protolith of the Swakane Gneiss was either derived from the same sediment sources as the Nanaimo Group and deposited elsewhere along the Pacific margin, or was composed of Nanaimo Group sediments.

The mechanism for burial of the Swakane Gneiss to lower-crustal depths (∼40 km) (Valley et al., 2003; Matzel et al., 2004) is uncertain, and both underplating of trench sediments to the base of the CCC and underthrusting of a forearc region under or into the CCC have been proposed (Matzel et al., 2004; Sauer et al., 2018). Sauer et al. (2019) demonstrated that the Swakane Gneiss had very similar isotopic properties (e.g., U-Pb and Lu-Hf) to well-studied underplated schists exposed in California, suggesting that underplating may have occurred. However, clockwise P-T pathways (Valley et al., 2003; Matzel et al., 2004), and ongoing local magmatism (Miller et al., 2009), neither of which are found in the Californian schists, support an underthrusting mechanism.

We propose a model that details the interaction between CCC deformation and Nanaimo basin sedimentation. The Swakane Gneiss protolith was deposited between 91 and 81 Ma, was sourced from similar sediment source terranes as the Nanaimo basin, and was later incorporated into the arc, reaching maximum metamorphic conditions between 72 and 66 Ma (Sauer et al., 2018). Regardless of the mechanism interpreted to incorporate the Swakane protolith into the CCC, uplift associated with either underthrusting or underplating drove basin sedimentation and culminated in gravel deposition characteristic of Geoffrey Formation, ca. 72 Ma (Fig. 8B). Additionally, the onset of gravel deposition in the Nanaimo basin and peak metamorphic conditions of the Swakane (ca. 72 Ma) are coincident with an increase in exhumation rate in the CCC from 0.6 to 3 mm/yr (Paterson et al., 2004) (Fig. 8B). Identifying the response of sedimentation (see Englert et al., 2018) to changes in lower plate dynamics, identified here, is important, as the response of sediment routing systems to underplating or under thrusting is difficult to quantify in both ancient (Sierra Nevada, California; Ducea and Chapman, 2018) or modern (Raukumara Range, New Zealand; Bassett et al., 2010; Coast Cordillera, Chile; Encinas et al., 2012) settings.

Paleogeography of the Southern Insular Superterrane during the Late Cretaceous

The basin-wide data set presented here demonstrates two basin-scale events recorded in the strata of the Nanaimo Group; the first event occurred ca. 84 Ma and the second ca. 72 Ma (Fig. 9). The first basin-scale event included: (1) the onset of basin-wide deep-water sedimentation (i.e., transition between lower and upper Nanaimo Group), (2) a transition in detrital zircon facies from F1 to F2 across the central and southern Nanaimo basin as well as the protolith of the Swakane Gneiss, and (3) uplift of the NCTS with sediment shed into a restricted portion of the basin (Figs. 9A, 9B, and 9E). We hypothesize that the tectonic driver for this event is the onset of oblique subduction of the Kula plate, which rifted from the Farallon plate and began subducting obliquely beneath the Nanaimo basin as early as 85 Ma (Woods and Davies, 1982; Engebretson et al., 1984). Oblique subduction can increase the rate of subduction erosion causing forearc subsidence (von Huene and Scholl, 1991; Fildani et al., 2008), along with driving forearc deformation and uplift opposite to the direction of subduction (i.e., uplift at the southern end of the Nanaimo basin) (Malatesta et al., 2013, 2016). This process may have driven the transition between F1 to F2 detrital zircon populations in the central and southern Nanaimo basin transects.

The second distinct event occurred at 72 Ma and included: (1) basin-wide introduction of gravel to the basin, as recorded by the Geoffrey Formation (Englert et al., 2018), (2) the change from F1 to F2 detrital zircons facies in the northern Gulf Islands transects, (3) emplacement and metamorphism of the Swakane Gneiss (Sauer et al., 2018), and (4) increased exhumation rate of the CCC (Paterson et al., 2004; Figs. 9C–9E). Detrital zircon samples from this event record a high abundance of near-depositional ages, and sandstone compositions from the central and southern Gulf Islands transects record an increase in volcanic rock fragments. We speculate that these processes may have been caused by an increase in the rate of Kula plate subduction and obliquity starting at 73 Ma, as documented by Doubrovine and Tarduno (2008).

Events in the Nanaimo basin reflect the timing of Kula plate formation, along with changes in convergence and obliquity of subduction. Our results indicate that changes in subduction have a regional impact and are a dominant control on forearc basin evolution; however, they may manifest differently across the basin due to along-strike variability of these processes. Our data suggest that the lower plate had a dominant influence on the evolution of the basin, which is inconsistent with the foreland model of the basin (cf. Mustard, 1994). Due to this interaction and the correlation between changes in Kula Plate spreading with basin and hinterland events, we conclude that the Nanaimo basin should be considered a forearc basin.

The identification of the events at ca. 84 Ma and ca. 72 Ma, as well as their relationships to the history of the Kula Plate is important for understanding the Late Cretaceous paleogeography of the Pacific margin of North America. The formation of the Kula plate has been invoked for the northward translation of the Baja-BC block (e.g., Umhoefer, 1987; Beck, 1991) and the events documented here may record initial excision and or docking of the block. The Kula-Farallon ridge is interpreted to have formed south of the Nanaimo basin (Woods and Davies, 1982; Engebretson et al., 1985; Doubrovine and Tarduno, 2008; Seton et al., 2012). However, if the Nanaimo basin, as a part of the Baja-BC Block, was located at the latitude California during the Late Cretaceous (e.g., Housen and Beck, 1999; Matthews et al., 2017; Sauer et al., 2019), then the Kula-Farallon ridge was located in southern California or northern Mexico at 85 Ma. This position of the ridge is consistent with oceanic plate reconstructions proposed by Engebretson et al. (1985); however, this is inconsistent with paleogeographic models for the southwestern North America during the Late Cretaceous (Jacobson et al., 2011; Sharman et al., 2015; Yonkee and Weil, 2015), which generally favors a higher latitude for both the Baja-BC block and the Kula-Farallon ridge north of California in the latest Cretaceous. A northerly position for the Baja-BC block would allow for the Kula-Farallon ridge to form north of California (e.g., at the latitude of northern Oregon or Washington; Doubrovine and Tarduno, 2008; Seton et al., 2012). Regardless of proposed paleolatitude, the Baja-BC block was located south of its current position in the Late Cretaceous (e.g., Wyld et al., 2006). The events identified here at 84 Ma and 72 Ma, as well as other events (e.g., subduction of the Kula-Farallon Ridge), should be recorded in units outside the Baja-BC block, providing evidence for its paleolatitude.

CONCLUSIONS

Thirty-five point-counted sandstone thin sections and 22 new U-Pb detrital zircon samples complete a basin-wide north-to-south perspective of provenance evolution in the Nanaimo basin of western British Columbia, Canada, and northern Washington, USA. Samples are divided into Gulf Islands transects (northern, central, and southern) and a transect through the San Juan Islands. Petrologic trends in the Gulf Islands transects demonstrate increased exhumation of the Coast Mountains Batholith both through time and from north to south along the margin of the basin. Evolution of sediment composition in the San Juan Islands transects reflects derivation from various thrust nappes of the North Cascade thrust system (NCTS) during their construction early in the basin’s history.

The basin-wide detrital zircon data set incorporates data from previous studies (N = 49, n = 10,942), and is characterized by five detrital zircon facies (F1–F5). These facies are interpreted to represent sediment derivation from the Coast Mountains Batholith, hypothesized sediment sources east of the Coast Mountains Batholith, NCTS, and Wrangellia basement. The Nanaimo basin is largely dominated by sediment derived from the Coast Mountains Batholith (F1) and sources east of the Coast Mountains Batholith (F2). Because the detrital zircon age distributions, as well as the change between F1 and F2 zircon populations observed in the majority of the Nanaimo basin (ca. 84 Ma) occurs synchronously in the protolith of the Swakane Gneiss, we suggest that the protolith of the Swakane Gneiss was derived from similar sources or was composed of Nanaimo Group sediments. Sediment derived from the NCTS (F3, F4, and F5) was deposited in a spatially restricted zone, directly adjacent to the NCTS, rather than distributed across the Nanaimo basin, as suggested by previous models.

Two large, basin-scale tectonic events are resolved in this data set. The first event ca. 84 Ma included deepening of the basin, a change in detrital zircon age distributions, and uplift of the NCTS. This event is hypothesized to be driven by onset of oblique Kula Plate subduction. The second event included basin-wide coarse-grained sedimentation, localized high-flux events, increased exhumation of the CCC, and the underplating and metamorphism of the Swakane Gneiss at ca. 72 Ma. These events are important for future paleographic reconstructions of the Pacific margin, as they may represent key processes involved in northward translation of the Baja-BC block. Additionally, they highlight the interplay between oblique subduction, underplating or underthrusting, and basin sedimentation, which is applicable to other modern and ancient forearc settings.

ACKNOWLEDGMENTS

Funding for this research was provided by a Natural Sciences and Engineering Research Council (NSERC) Discovery Grant (No. RGPIN/341715-2013) to S. Hubbard and a Queen Elizabeth II scholarship from the University of Calgary to D. Coutts. We would like to thank residents of the Gulf Islands and San Juan Islands for access to private property, Washington State Parks for permission to conduct field work in Sucia Island State Marine Park, and Nate Averna for help with transport to Sucia Island. The manuscript evolved thanks to insightful discussions with John Garver and Mark Brandon, as well as constructive reviews from Sarah Roeske, Kathleen Surpless, and Andrew Leier.

1 GSA Data Repository Item 2020133, File DR1: Raw/full point-counting data, File DR2: All isotopic data from Nanaimo basin, and File DR3: Maximum depositional age plots, is available at http://www.geosociety.org/datarepository/2020, or on request from editing@geosociety.org.
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.