We present new U-Pb zircon ages for heterogeneous mafic-felsic rocks of the West Coast complex, a midcrustal plutonic component of exposed Jurassic arc-crustal section on Vancouver Island. We examine the timing of juvenile plutonic crust production, and consanguinity of volcanism and plutonism in this arc. The midcrustal plutons were emplaced between 193 Ma and 174 Ma, contemporaneous with the upper-crustal (<10 km depth) plutonic component of the arc (Island plutonic suite). A 12 km thickness of plutonic arc crust was built as a series of sheets at a rate of ∼0.003 km3 yr−1. Deformation and emplacement were contemporaneous, and there are no correlations among age, differentiation (peridotite to granite), or structural level of plutons in the arc. The age range and a weak eastward younging age polarity of the Jurassic arc section on Vancouver Island match that of the Talkeetna arc-crustal section in Alaska, suggesting that the two arcs are correlative and evolved by either forearc erosion above of an east-dipping slab, or slab rollback during west-dipping subduction.
One of the proposed mechanisms for the growth of continental crust is by the accretion of island arcs (Hamilton, 1994). A conundrum arises in the origin of continental crust by this mechanism, however, in that mantle-derived magma fed to arcs is basaltic, yet accretion of island arc crust has somehow resulted in an andesitic continental crust. One explanation is that differentiation occurs within the arc, and there is delamination or loss of mafic material through its base, where material is returned to the mantle (Kay and Mahlburg Kay, 1993). Dynamical constraints likely confine delamination to only a subset of arcs (Jull and Kelemen, 2001). A more recent idea is that crust can be relaminated, from below, due to subduction erosion or sediment subduction (Hacker et al., 2011).
The mass balance of crustal growth in arcs in part involves knowing the thickness and bulk composition of all parts of the arc crust with its age and evolution. An important interface for within-arc differentiation is the middle crust, situated between the better-exposed upper parts of the arc crust and the lesser-preserved lower crust. In modern arcs, the middle crust is never exposed, but seismic studies define it as having wave speeds of 6–6.5 km/s, with a thickness that varies greatly from ∼4 km to 16 km, both spatially and temporally (Calvert et al., 2008). A seismic distinction may be inaccurate, however, because of poor correlation of seismic velocity and composition (Behn and Kelemen, 2006), and considerable heterogeneity in many exhumed midcrustal sections (Kawate and Arima, 1998; Rioux et al., 2010).
In terms of arc magma genesis, the middle crust may filter magmas rising to feed higher-level plutons, may be cannibalized by upward-rising magmas (Otamendi et al., 2009), or may even act as a “sponge” for water or other elements in differentiating arc magmas by crystallization of amphibole or oxides (Davidson et al., 2007; Larocque and Canil, 2010). It is unclear the extent to which the addition of new material versus recycling of older material is involved in the formation of this intermediate layer in arcs. Magmatic “flare-ups” occur in more long-lived arcs, which may relate to changing tectonic parameters (slab rollback) or to the growth of the arc (Ducea, 2001). Thus, age and volume relationships for the middle crust are necessary to better characterize the mass and tempo of magma production and differentiation within arcs (Ducea, 2001).
In this study, we characterize the thickness and age of a midcrustal section of heterogeneous plutonic rocks of the Jurassic Bonanza arc exposed on Vancouver Island (Fig. 1). Age information for midcrustal rocks of the arc constrain the consanguinity of volcanism and plutonism of an arc, and the longevity of the arc and its magma flux rate. We examine whether any of these observations define a polarity to the arc, or correlate with those of Jurassic arcs exposed farther north in Queen Charlotte Island and Alaska (Talkeetna).
Most of Vancouver Island is underlain by the Wrangellia terrane, which consists of a basement of Devonian Sicker Group volcanic and plutonic rocks, succeeded by Permian Buttle Lake carbonates, and Triassic Karmutsen Plateau basalt–Quatsino limestone, and clastic rocks (Muller, 1977). The Jurassic Bonanza arc intrudes or overlies these basement units of Wrangellia and is unconformably overlain by Late Cretaceous sedimentary rocks of the Nanaimo Group (Fig. 2). Wrangellia has an overall northwesterly structural grain. Major north-trending anticlinoria, including Cowichan and Buttle Lake, are attributed to Late Cretaceous transpression overprinted by high-angle brittle deformation that resulted in homoclinal fault-bounded blocks (Nixon and Orr, 2007b; Yorath et al., 1999).
The Jurassic Bonanza arc forms a 15-km-thick crustal section and is divided from lower to higher structural level into the West Coast complex, the Island plutonic suite, and the Bonanza volcanics (Fig. 2). The West Coast complex has been interpreted as a part of the middle crust (DeBari et al., 1999a; Larocque and Canil, 2008) and is a heterogeneous mixture of mostly diorite and quartz diorite, with lesser granodiorite crosscut by intrusions of leucotonalite. In the diorite, recent studies have recognized decameter- to rarely kilometer-scale bodies of conformable but discontinuous hornblende gabbro, pyroxenite, olivine hornblendite, and peridotite with cumulate textures (Figs. 3A and 3B) (Larocque, 2008; Larocque and Canil, 2007, 2010; Marshall et al., 2007). Intrusive relationships in the West Coast complex are heterogeneous on all scales (Fig. 3C), and levels of strain vary within meters in outcrop (Figs. 3D and 3E; Isachsen, 1987).
The Island plutonic suite occurs as a series of unfoliated quartz diorite to alkali feldspar granite plutons, with rare diorite and gabbro. The contact between rocks previously mapped as the Island plutonic suite and the West Coast complex is not defined, and to some degree the distinction between these two units is obscure as is discussed further herein in light of new age information in this study.
The Bonanza volcanics are pillowed and massive flows of aphanitic basalt, andesite, and dacite with pyroclastic deposits. Based on geological and geobarometric constraints, the structural section of the Bonanza arc consists of 2.5 km of Bonanza volcanics, 10 km of Island plutonic suite rocks, and a less constrained thickness of 10–18 km for the West Coast complex (Canil et al., 2010). The entire section was thus likely 20–25 km thick prior to structural thinning to its current 15 km.
Published U-Pb ages for rocks of Bonanza arc vary between 202 and 168 Ma (Nixon and Orr, 2007a) with a weak eastward younging trend on Vancouver Island (DeBari et al., 1999a). There is no information on the age of mafic and ultramafic rocks in the West Coast complex.
We selected seven samples of the West Coast complex from three regions (Fig. 1; Table 1) to investigate age relationships between samples with and without fabrics, and between the interpreted upper and middle parts of the arc crust.
Rocks of the West Coast complex in the Port Renfrew area were mapped at 1:20,000 scale (Larocque, 2008). Most of the region is underlain by massive quartz diorite, diorite, and hornblende gabbro that in places has layering or a deformation fabric, or is crosscut by leucotonalite intrusives. Hornblende gabbro, pyroxenite, and peridotite occur both as conformable layers in the diorite (Figs. 3A and 3B) and as blocks, schlieren, or xenoliths (Fig. 3C; Larocque and Canil, 2007, 2010). Both massive and heavily strained diorite gneisses are recognized, containing mylonite with a foliation that parallels the general northwest strike and shows top-to-the-north shear sense indicators (sheath folds and lineations; Figs 3E and 3F).
We sampled: massive melanodiorite (JL06–110A), and crosscutting leucotonalite intrusive dikes (JL06–110B) in one outcrop; foliated medium-grained leucodiorite layers (DC0507) adjacent to a mylonite zone showing a mineral lineation defined by hornblende crystals (Fig. 3E); and massive medium-grained hornblende-biotite granodiorite containing mafic enclaves (DC06037).
Rocks mapped as both the Island plutonic suite and the West Coast complex occur along the western and northern edge of Catface Peninsula, north of Tofino (Fig. 1; Isachsen, 1987). We sampled massive medium-grained hornblende-biotite quartz monzonite along the western flank of the peninsula in outcrop (DC1002) and in drill core from the Catface ore deposit (CS10–109) (Smith et al., 2012) (Fig. 1).
Both heterogeneous massive diorite and diorite gneiss with a general northwest-striking fabric in the Victoria region (Fig. 3D) are crosscut in several different areas by north-striking leucotonalite dikes. We sampled massive medium-grained diorite near Mount Douglas (DCMD04) (Fig. 1).
U-Pb zircon dating was performed for six samples at the University of British Columbia using procedures described in Scoates and Friedman (2008) with a modification after Mundil et al. (2004). Two of these samples underwent chemical abrasion pretreatment, and four others underwent physical abrasion (listed in Table 2). U-Pb zircon dating for one other sample was done at the University of Alberta following procedures in Heaman et al. (2002). There are no studies of intercalibration between these two laboratories. All errors are quoted at the 2σ or 95% level of confidence (Table 2). Isotopic dates were calculated using the decay constants 238U = 1.55125 × 10−10 and 235U = 9.8485 × 10−10 yr−1 (Jaffey et al., 1971). The reported ages are based on weighted averages of 206Pb/238U data for a number of fractions in each sample.
Zircons were recovered from all samples as both fragments and euhedral prisms. The Th/U ratios of zircons vary between ∼0.2 and 0.5, consistent with an igneous origin (Table 2). In general, the majority of zircon fractions are concordant or nearly concordant, with little if any inheritance, with the exception of sample DCMD04 from Mount Douglas, described next. There is no correlation of age with composition (i.e., mafic vs. felsic) or fabric (foliated or massive) for any of the samples in our data set.
Two multifragment colorless zircon fractions from sample DCMD-04 (massive quartz diorite) have intermediate U contents (310 and 269 ppm, respectively) and similar Th/U (0.45 and 0.42, respectively) (Table 2). Fraction 1 is concordant and has a 206Pb/238U date of 181.1 ± 0.4 Ma (Fig. 4A). Fraction 2 is discordant and likely reflects the presence of an Archean inherited Pb signal in the cores of this zircon population. A reference line constructed to pass through these two analyses yields a lower intercept date of 181.3 ± 0.4 Ma, interpreted to be the emplacement time of this quartz diorite.
Five fractions from sample CS10–109 (massive quartz monzonite) have intermediate U contents (172–448 ppm) but somewhat variable Th/U (0.29–0.41; Table 2). All five fractions are concordant, with a cluster of four that have significant mutual overlap and one slightly younger grain. A weighted average of the four strongly overlapping 206Pb/238U dates gives an age of 184.46 ± 0.41 Ma (mean square of weighted deviates [MSWD] = 0.28) (Fig. 4B).
All five fractions from this sample (massive quartz monzonite) have high U contents (721–1512 ppm) but broadly similar Th/U (0.28–0.37; Table 2). Three concordant and overlapping analyses (Fig. 4C) give a weighted 206Pb/238U age of 186.20 ± 0.19 Μa (MSWD = 1.5).
Four fractions from this sample of massive granodiorite have intermediate U contents (109–255 ppm) and Th/U of 0.38–0.97 (Table 2). All four fractions are concordant (Fig. 4D) and give a weighted 206Pb/238U age of 192.62 ± 0.38 Ma (MSWD = 0.98).
Four fractions from this sample all have low U (24–68 ppm), reflecting the high color index of this melanodiorite (shown in Fig. 3E), and a narrow range of Th/U of 0.34–0.40 (Table 2). Four concordant analyses produce a weighted 206Pb/238U age (Fig. 4E) of 174.75 ± 0.42 Ma (MSWD = 1.05).
This sample is a leucrocratic dike crosscutting JL06–110A. Four zircon grains have variable U contents of 117−1003 ppm and Th/U ratios of 0.21–0.39 (Table 2). A weighted mean 206Pb/238U age of 174.72 ± 0.38 Μa (MSWD = 0.50) is based on two concordant and overlapping analyses. The two other analyzed grains give discordant results with 206Pb/238U ages that are distinctly older for one and younger for the other.
Four fractions from this leucocratic layer in deformed diorite gneiss (shown in Fig. 3E) have very high U contents (972–3266) and Th/U ratios of 0.42–0.50 (Table 2). All four analyzed fractions are concordant and form two groupings of overlapping pairs. Because these grains were air abraded and not chemically abraded, and have a high U content, we interpret the older concordant and overlapping analyses with a weighted mean 206Pb/238U age of 193.90 ± 0.20 Ma (MSWD = 0.025) as the age if emplacement. The other two fractions are interpreted to have undergone minor Pb loss.
Distinction between Upper and Middle Crust
With the exception of sample DCMD04, we observe little if any inheritance in U-Pb data for samples of the West Coast complex (Table 2; Fig. 4), consistent with juvenile sources as evidenced by ɛNd values of between 5 and 7 for this unit (Andrew et al., 1991; Larocque, 2008). The U-Pb ages from this study, when combined with those from previous studies, show a range in age from ca. 193 Ma to 168 Ma for plutonic rocks in the Bonanza arc (Fig. 5). Ages older than 184 Ma are not recognized in the Island plutonic suite, and those younger than 174 Ma are not recognized in the West Coast complex, but a clear and definitive age distinction between these two units is not obvious.
The barely resolvable age distinction between Island plutonic suite and West Coast complex plutons is paralleled by a subtle distinction in their bulk compositions. The major-element bulk composition of rocks mapped as the West Coast complex broadly overlaps the Island plutonic suite in three different regions of Vancouver Island (Fig. 6). The Island plutonic suite has no mafic/ultramafic rocks, and a notably greater abundance of felsic rocks, enriched in incompatible elements with positive Zr anomalies, whereas the West Coast complex and its mafic/ultramafic cumulates have positive Sr anomalies and trace-element patterns identical to the Bonanza volcanic rocks (Fig. 7).
The only consistent field distinctions are that the Island plutonic suite plutons never show a foliation and, unlike the West Coast complex, commonly exhibit intrusive contacts with the Triassic Karmutsen lavas. The Triassic supracrustal rocks are interpreted as the upper ∼10 km of the crust through which Bonanza arc magmas transited to feed the volcanic arc. These field observations can be used to qualitatively infer that Island plutonic suite plutons occur stratigraphically higher than those of the West Coast complex (DeBari et al., 1999b), an inference corroborated by hornblende barometry, which shows crystallization pressures less than 300 MPa for Island plutonic suite plutons and greater than 300 MPa for the West Coast complex (Canil et al., 2010). A caveat in this interpretation, however, is that few samples of the West Coast complex have the mineralogical criteria amenable to quantitative Al-in-hornblende barometry (An < 35 plagioclase, K-feldspar–bearing; Anderson, 1996; Anderson and Smith, 1995). In intermediate to mafic bulk compositions, the Ti and Al in amphibole are mostly temperature- and pressure-dependent, respectively (Ernst and Liu, 1998), and can be used as semiquantitative thermobarometers for the bulk compositions present in the West Coast complex. At a given Ti content, the Al in amphibole of West Coast complex plutons is distinctly higher than in Island plutonic suite plutons, suggestive of a higher pressure of crystallization than those of Island plutonic suite plutons (Fig. 8). Thus, one reason for unfoliated Island plutonic suite plutons is that temperatures sufficient to enhance synintrusive deformation existed only within the warmer middle crust of the Bonanza arc, comprised of what is now the West Coast complex.
In the Victoria area, the West Coast complex was traditionally divided into separate units as the massive “Wark diorite” and foliated “Colquitz gneiss” (Muller, 1983). Both of these units are broadly diorite and can occur in the same outcrop. Rocks of the West Coast complex having a strong fabric (e.g., DC0507; Figs. 3B and 3F) are the same age as those that are massive with no foliation (DC0637). These observations require that strain during intrusion has been focused into other regions causing a gneissosity, and leaving some parts of the crust with no fabric. The strain partitioning occurred on all scales during intrusion (Fig. 3B; see also Isachsen, 1987).
Leucotonalites crosscut less-evolved diorites that are concordant or comingled with ultramafic rocks at Lens Creek (JL06110; Figs. 3A and 3B). The leucotonalites and diorites here have nearly identical U-Pb ages within the same outcrop (Fig. 4), thus providing a minimum age for the ultramafic rocks of 174 Ma. As both Island plutonic suite and West Coast complex plutons overlap ages inferred for Bonanza arc volcanic rocks, it is evident that all parts of the crust were magmatically active during construction of the Bonanza arc.
Magmatic Flux Rates
Jurassic plutons cover an area of ∼11,000 km2 on Vancouver Island (Fig. 1). Constraints from stratigraphy, metamorphic grade, and Al-in-hornblende barometry show that the West Coast complex and Island plutonic suite comprise ∼12 km in total of plutonic crust emplaced into supracrustal rocks of Wrangellia (Canil et al., 2010). If magmatism occurred over a 28 m.y. interval (Fig. 5), the 1.32 × 105 km3 of arc plutonic magma was emplaced at a rate of 0.004 km3 yr−1, which is within uncertainty of other estimates for large batholiths of the Sierra Nevada and elsewhere (0.001 km3 yr−1; Davis et al., 2011; Miller et al., 2011). This emplacement rate is consistent with incremental emplacement of arc crust by plutons and explains the lack of any obvious contacts between different units of the West Coast complex and Island plutonic suite in the field (Glazner et al., 2004).
Arc Polarity and Relation to the Talkeetna Arc in Alaska
Figure 9A shows the U-Pb ages for West Coast complex and Island plutonic suite samples plotted versus distance eastward from the westernmost terrane-bounding faults in Wrangellia on Vancouver Island (the San Juan and West Coast fault). The latter reference point was used to account for sinistral movement along the San Juan fault, which offsets units of the West Coast complex in southern Vancouver Island (Fig. 1). Plutons younger than 170 Ma only occur in the east. As noted earlier by Debari et al. (1999), an age polarity is weak, but it matches that of the Talkeetna arc in south-central Alaska (Fig. 5). In Alaska, plutons in the Chugach mountains are coeval with the ages found in the West Coast complex plutons. Plutons of the Talkeetna mountains, north of the Chugach, are coeval with the Island plutonic suite intrusions. The eastward (landward)-younging age polarity that characterizes intrusions of the Bonanza arc is weak, but it matches that of the northward (landward)-younging trend exhibited by plutons of the Talkeetna arc over a similar age range. Plutons in the Haida Gwaii (Queen Charlotte) Islands also match the ages of the Island plutonic suite, but there is insufficient sample coverage to recognize any pattern in polarity there (Fig. 9A).
There are some notable differences between the Talkeetna and Bonanza arcs. The Bonanza arc was built upon preexisting Devonian to Triassic supracrustal rocks, but no evidence of a pre-Jurassic basement is recognized in the Talkeetna arc (DeBari et al., 1999). The latter also shows a shift in ɛNd isotopes in intrusive rocks to lower, more-evolved values to the north (landward) (Rioux et al., 2010), whereas no such trend is observed for limited data of the Bonanza arc upper-crustal rocks (Fig. 9B). The ɛNd values in plutons of the West Coast complex show no extensive involvement of preexisting basement, whereas there is a trend of decreasing ɛNd westward for upper-crustal components (Bonanza volcanics, Island plutonic suite) in the Bonanza arc, suggesting some assimilation or contamination in the upper crust to the west.
The similar age range and polarity between the Talkeetna and Bonanza arcs are easiest to explain if these two segments were part of a single, originally continuous magmatic arc. The plate setting may have varied from a simple island arc in one portion (Talkeetna) trending along strike onto an arc built upon a preexisting Devonian–Triassic arc–oceanic plateau (Bonanza).
Correlation of the Bonanza and Talkeetna arcs has been suggested before, but tectonic interpretations of the polarity vary. One model shows the age progression northward in the Talkeetna as due to north-dipping subduction and removal of the forearc to a south-facing arc by tectonic erosion, at a rate observed in modern subduction zones of ∼3 mm yr−1, and consuming between 70–130 km of forearc over the course of the arc’s magmatic history, with a matching northward migration of arc magmatism (Clift et al., 2005). As in the case of Talkeetna arc, tectonic erosion of the forearc to west of an east-dipping, west-facing Bonanza arc could explain eastward migration of the arc plutons eastward from 193 to 165 Ma. An alternative interpretation of the age polarity is that the Bonanza- Talkeetna arc faced east to north (Reed et al., 1983), and that the decreasing ages to the east and north resulted from trench rollback and the related accretion of material to the upper plate at the trench (Fig. 10). Significant post-Jurassic large-scale, margin-parallel translation of parts of Wrangellia and adjacent terranes, and removal of any forearc assemblage along strike make testing of these two disparate polarity models difficult. The stacking of the Bonanza crust along west-dipping, east-verging mylonite zones, however, is consistent with a model of accretionary growth and eastward arc migration within an eastward-facing arc (e.g., Johnston, 2008).
Exhumation of Middle Crust in the Bonanza Arc
In the Victoria and Port Renfrew areas, components of the West Coast complex form tabular sheets bound by high-strain, brittle-ductile mylonite zones characterized by top-to-the-east kinematic indicators (Fig. 3F). The mylonites are overprinted by a later brittle deformation. A similar structural style is recognized to the north in the Nootka Sound region (Fecova et al., 2008), but there the mylonite zones are characterized by subhorizontal lineations and kinematic indicators indicative of strike-slip transport. These features suggest the plutons that make up the West Coast complex may have intruded one another as sheets, with massive textures, as has been proposed for the Island plutonic suite (Canil et al., 2010), but that the sheets have been subsequently stacked along high-strain discontinuities, induced at high temperatures in the middle crust. Postintrusive structural stacking explains the inconspicuous contacts between the different intrusive phases, and juxtaposition of middle- and upper-crustal components of the Bonanza arc.
There are no Late Jurassic sedimentary sequences deposited that show an immediate record of denudation of the Bonanza arc. The oldest sandstones of the Late Cretaceous Nanaimo Group, exposed along eastern Vancouver Island, and in correlative sandstones of the Longarm Formation, found east of and offshore of the Haida Gwaii (Queen Charlotte) Islands, largely originated from the Coast belt in the east, and the Wrangellia terrane, which at the time was an unknown distance to the west (Mustard, 1994). Detrital zircons in the Nanaimo and Longarm sedimentary packages have no populations older than 180 Ma. The latter ages are distinct to the westernmost portions of West Coast complex in the Bonanza arc (Fig. 9A), suggesting that the deeper and older (older than 180 Ma) parts of the middle crust of the arc were not yet uplifted and dissected until post–Late Cretaceous time. Indeed, fission-track and other cooling ages on Vancouver Island (England et al., 1997) suggest that the Bonanza arc was not fully uplifted and dissected until Eocene time (Fig. 5), likely commensurate with the formation of the Cowichan fold-and-thrust belt, itself related to an outboard collision of the Crescent terrane (Johnston and Acton, 2003).
Our results place some constraints on the nature and temporal evolution of the middle crust within an active arc. We show that there is only slight or subtle chemical, petrological, and geochronological distinction between the upper-crustal Island plutonic suite intrusions and the midcrustal West Coast complex in the Jurassic Bonanza arc. The middle crust is the feeding zone, active at the same time, for intrusions and extrusions that grow the upper crust, but by existing at higher temperatures, it is weaker and focuses more strain. Indeed, the distinguishing factor between the West Coast complex and Island Plutonic suite is strain: Middle-crustal rocks experienced significant synmagmatic strain partitioning, whereas upper-crustal intrusions largely lack evidence of synintrusive strain. Synintrusive fabric development in the middle crust may render the middle and lower crust prone to removal, and explain the preferential preservation of massive, unfoliated upper crust.
We sincerely thank G. Pearson for introducing us to the Port Renfrew area, and P. Hetherington for his support. D. Marshall kindly directed us to some unpublished ages. We appreciate reviews by M. Rioux and an anonymous reviewer. Research funding was provided by Emeralds Field Resources, Geoscience BC, and Natural Sciences and Engineering Research Council (NSERC) of Canada (to Canil).