Reconnaissance geochronology and geochemistry of the Mont-Tremblant gneiss of the Morin terrane, Grenville Province, Québec

The Grenville Province is a classic example of an exhumed orogen, exposing the Mesoproterozoic middle crust of the margin of Laurentia (Fig. 1). Research in the Grenville Province has focused on using it as a natural laboratory to investigate geologic questions such as the petrology of abundant yet enigmatic anorthosite plutons, the relationship between anorthosite emplacement and regional metamorphism, the nature of granulite-facies metamorphism, and testing the mid-crustal implications of plate-tectonic models (e.g., Martignole and Schrijver, 1970; Valley et al., 1990; Carr et al., 2000; Hanmer et al., 2000; Martignole et al., 2000; Peck et al., 2010; McLelland et al., 2010).

The Morin terrane is located in the southwestern Grenville Province and is geologically similar to the Adirondack Highlands to the south, both having been metamorphosed to the granulite facies and hosting numerous plutons of the anorthosite-mangerite-charnockite-granite (AMCG) suite (Fig. 2). Because of its geology, most studies in the Morin terrane have focused on anorthosite-suite magmatism, regional metamorphism, and tectonic history (e.g., Martignole and Schrijver, 1970; Indares and Martignole, 1990; Martignole et al., 2000; Peck and Valley, 2000; Peck et al., 2005). A relatively unstudied feature, especially in recent decades, is the nature of the country rocks to the AMCG suite. These include the Laurentian Fundamental Gneiss of the Mont-Tremblant area, long considered to be the oldest rocks in the region (Logan, 1863). Here, I use the term Mont-Tremblant gneiss to refer to the Morin terrane’s granulite-facies orthogneisses, following Adams (1896) and other workers (Osborne, 1936; Katz, 1969), who used the term Trembling Mountain gneiss. Establishing the age and nature of the Mont-Tremblant gneiss is essential for constraining the early history of the Morin terrane. In addition, knowledge of the age of the Mont-Tremblant gneiss will help to constrain the depositional age of the Grenville Supergroup and its ore deposits (e.g., Eppich and Peck, 2006). This contribution shows that the Mont-Tremblant gneiss is made up of a suite of evolved 1.3–1.4 Ga (Geon 14) arc rocks. This recognition helps to tie the geologic history of Morin terrane to that of adjacent terranes and to regional tectonic reconstructions.

Rocks that predate the 1.15 Ga AMCG suite (Doig, 1991) in the Morin terrane include both metasediments and orthogneiss. Katz (1969) described a package of lithologies in the area of Mont-Tremblant mostly made up of quartzofeldspathic orthogneiss (± orthopyroxene [opx] ± garnet [grt]) with intercalated mafic granulites (+ opx ± clinopyroxene [cpx] ± hornblende [hbl]), amphibolite, pelitic metasediments, and quartzite. These rocks can show retrogression to the amphibolite facies when in contact with members of the AMCG suite, and foliated orthogneiss and metasediments are found as xenoliths within AMCG plutons (Martignole and Schrijver, 1970). West of the anorthosite massif, the Morin terrane is bounded by the vertically dipping Labelle shear zone. East of the anorthosite massif, two distinct packages of metasediments yielded apparent Rb-Sr isochron ages of 1576 ± 16 Ma and 1094 ± 43 Ma, both of which were interpreted as metamorphic ages (Barton and Doig, 1973). The earlier age is much older than any known crystallization or metamorphic ages in the Allochthonous Monocyclic Belt (Rivers et al., 1989), but the younger age may reflect the 1.09–1.02 Ga Ottawan orogeny. Alaskite orthogneiss samples yielded a Rb-Sr isochron age of 1460 ± 58 Ma, which was interpreted as an igneous crystallization age (Barton and Doig, 1973). New Nd mantle extraction ages from basement orthogneisses in this area are similar, 1475 ± 87 Ma (n = 21; Dickin et al., 2010). The only directly dated pre-AMCG rock in the Morin terrane is a 1371 +41/–28 Ma tonalitic orthogneiss from the easternmost part of the terrane (Shawinigan domain; Shaw on Fig. 1B; Corrigan and van Breemen, 1997).

Sampling for geochronology and geochemistry was focused on areas containing orthogneiss units northwest and southwest of the Morin anorthosite massif (informally called the Mont-Tremblant massif and Lac Notre-Dame domain, respectively; Martignole et al., 2006; Fig. 2). Nine samples were chosen for analyses based on lack of alteration, range in composition, and penetrative deformation (Fig. 3). These samples were disaggregated using a hammer and anvil to remove visible alteration and weathering. A split of the pea-crushed sample was fully powdered in a tungsten carbide shatter box for major- and trace-element geochemistry at SGS Mineral Services, Toronto. Major elements were analyzed by X-ray fluorescence on glass disks, and trace elements were analyzed by inductively coupled plasma–mass spectrometry (ICP-MS) after lithium metaborate fusion of rock powders (see http://www.geochem.sgs.com for details).

Sm-Nd analyses were performed at the Isotope Geochemistry and Geochronology Research Centre at Carleton University (Chiarenzelli et al., 2010). Four powdered samples were mixed with a ¹⁴⁸Nd-¹⁴⁹Sm spike and dissolved using HF and HNO₃, and then HCl, before column chemistry using Dowex AG50-X8 and Eichrom LN cation resins (after Richard et al., 1976). Dissolved samples were loaded onto double Re filaments and analyzed in a 9-cup ThermoFinnigan TRITON TI multicollector mass spectrometer. Nd analyses were normalized to a ¹⁴⁶Nd/¹⁴⁴Nd ratio of 0.72190. Long-term laboratory averages for international standards can be found at http://iggrc.carleton.ca/info-clients/analytical-information. Average reproducibility for ɛ_Nd values is approximately ±0.5.

Zircons were separated from two samples for geochronology by partial powdering using a tungsten carbide shatter box and sieving to retrieve a <250 μm fraction. The fine powder fraction was removed by flotation in water, and the heavy mineral fraction was separated using methylene iodide. Zircon was handpicked using a binocular microscope and screened for cracks and alteration. Samples were mounted in epoxy with standards, polished to expose equatorial cross sections, gold coated, and examined using backscattered electron (BSE) imaging and cathodoluminescence (CL) prior to and following ion microprobe analysis (Supplemental File¹). U-Pb geochronology of zircon was conducted at the U.S. Geological Survey–Stanford University Sensitive High-Resolution Ion Microprobe–Reverse Geometry laboratory (SUMAC) using standard procedures (e.g., Premo et al., 2008, and references therein). Analyses focused on sampling igneous core material and overgrowths using an O^2– primary beam that produced ∼20-μm-diameter by 1–2-μm-deep spots. Sample analyses (Table 1) were interspersed with analyses of the 1200 Ma VP-10 standard. Data reduction was performed using the Squid and Isoplot programs (Ludwig, 2001, 2008).

Whole-rock major-element geochemistry presented here (Table 2) is combined with that of Katz (1969, n = 9) from the Mont-Tremblant area for this discussion. Most of the samples have compositions that are granitic (SiO₂ = 71%–78%) or granodioritic (SiO₂ = 61%–67%), with only one sample from Katz plotting as a tonalite (SiO₂ = 54%; Fig. 4). Samples show smooth trends on Harker diagrams. On an AFM diagram (A = Na2O + K2O; M = MgO; F = FeOt) samples are arranged in a linear array and have overall low MgO contents, consistent with a calc-alkaline trend (Fig. 4A). These rocks plot as calc-alkalic with some scatter on the major-element discrimination diagram of Frost et al. (2001) (Fig. 4E). On trace-element diagrams for granitic rocks Mont-Tremblant gneisses plot as volcanic arc granites or across the boundary to within-plate granites (Pearce et al., 1984; Fig 4B). Relatively low Nb (avg. = 11 ppm), Ta (avg. = 1.15 ppm), TiO₂ (avg. = 0.6%), and P₂O₅ (avg. = 0.16%) values are all consistent with a subduction signature (Fig. 4C). These rocks show light rare earth element (LREE) enrichment (La_n/Sm_n avg. = 3.8) and depleted heavy (H) REE (Sm_n/Yb_n avg. = 9.8) patterns, and for the most part have negative Eu anomalies (Fig. 4F). One granitic sample (10SH1B) has lower REE, a large positive Eu anomaly, and a concave-up HREE pattern, consistent with this sample being a cumulate.

Neodymium isotope ratios are enriched relative to the Chondritic Uniform Reservoir (CHUR) (Table 3), and ɛ_Nd at 1.3 Ga ranges from 1.4 to 4.0. Depleted mantle model extraction ages (DePaolo, 1981) for three samples are 1.36–1.49 Ga, which are similar to the 1.48 ± 0.09 Ga average for similar rocks from the eastern Morin terrane (Dickin et al., 2010). One rock has a depleted mantle model age of 1.66 Ga, which is unusually high for orthogneiss from the Allochthonous monocyclic belt. This sample has the highest Sm/Nd value of the four samples analyzed, but it is not unusual otherwise.

Geochronology sample 10SD4 is a coarse-grained, pink, strongly foliated biotite granite gneiss. Zircons from 10SD4 are, for the most part, 200–300-μm-long prisms with aspect ratios of 1.5:1–3.5:1. CL shows oscillatory-zoned interiors with rare mineral inclusions, and occasional cores having mottled zoning. Up to 50-μm-thick rims are dark in CL, occasionally give grains protuberances or embayments suggestive of having been formed around other minerals, and rims occasionally erode and truncate core zoning (Fig. 5A). SHRIMP analyses of cores yield an array of points that when regressed have an upper concordia intersect of 1332 ± 36 Ma (Fig. 6; 2σ errors reported, mean square of weighted deviates [MSWD] = 0.8, n = 9 spots). Excluding one 81% discordant spot, this sample yields an age of 1324 ± 38 Ma (MSWD = 0.59, n = 8). This is interpreted as the preferred igneous crystallization age of 10SD4. It is possible that the spread of near-concordant ages of cores defines an ancient Pb loss event that is essentially parallel to concordia, suggesting that this may be a minimum age. Rims on zircons from 10SD4 define a cord that intersects concordia at 1159 ± 44 Ma (MSWD = 2.9, n = 10). Regressing the data without the two spots having anomalously large errors (one being reversely discordant) improves this age to 1159.4 ± 15.6 Ma (MSWD = 0.37, n = 7), which is taken as the age of rim growth.

Geochronology sample 10SH3 is a coarse-grained, gray, streaky hornblende granite gneiss. Zircons from this sample are 200–400-μm-long prisms with aspect rations of 2.5:1–3.5:1. These zircons commonly contain an innermost core with mottled zoning surrounded by oscillatory zoning visible in CL (Fig. 5B). In some crystals, the outer zone of the grain has lower or dark CL, but this is likely a magmatic feature and yields similar ages to inner parts of grains. Thinner dark rims morphologically similar to those found in 10SD4 are rare and were not dated. Analyses of cores have a somewhat poorly defined upper concordia intercept of 1355 ± 72 Ma (Fig. 7; MSWD = 3.8, n = 22). This rock has likely experienced the same metamorphic event that caused the 1160 Ma rims in 10SD4, which may have caused heterogeneous Pb loss in this sample. Exclusion of five outlying points improves this age estimate to 1333 ± 32 Ma (MSWD = 1.19, n = 17), which is the best estimate of the time of igneous crystallization of the protolith.

The reconnaissance geochemistry and geochronology presented here favor correlation between country rock orthogneiss units (Mont-Tremblant gneiss) across the Morin terrane and correlation with other calc-alkaline rocks in the Allochthonous Monocyclic Belt. Zircon geochronology yields identical ages of 1324 ± 38 Ma and 1333 ± 32 Ma for samples separated by ∼70 km; ∼120 km away in the easternmost Morin terrane, a tonalite has been dated by U-Pb in zircon at 1371 +41/–28 Ma (Corrigan and van Breemen, 1997). The major- and trace-element geochemistry presented here and by Katz (1969) behaves like a coherent igneous suite (Fig. 4). In addition, similar ca. Geon 15 Nd model ages for orthogneisses are shown to span the Morin terrane (Dickin et al., 2010), further pointing toward a related suite of rocks.

When rocks of the Morin terrane were described in The Geology of Canada (Logan, 1863), clearly metasedimentary rocks such as marble and quartzite were grouped with the quartzofeldspathic gneisses described here, and were assigned to the Fundamental Gneiss, which was thought to represent “a series of metamorphic sedimentary strata” (p. 22). Later, Adams (1896) recognized the igneous protolith of the quartzofeldspathic gneisses, but stated that basement or intrusive relationships with Grenville metasediments “cannot be determined” (p. 29). Subsequent work (e.g., Osborne, 1936) favored a syndeformational intrusive relationship to explain map patterns between orthogneiss and metasediments.

Regional mapping by Wynne-Edwards et al. (1966) in what is now recognized as the Central Metasedimentary Belt of Québec and Morin terrane supported the igneous protolith of these gneiss units, and these workers proposed that the Mont-Tremblant gneiss was basement to metasediments of the Grenville Supergroup. This hypothesis was based in part on lack of metasedimentary xenoliths and other intrusive features commonly shown by younger intrusive units. West of the Labelle shear zone, these basement rocks include the oldest rocks now recognized in the Central Metasedimentary Belt of Québec: the 1.37–1.45 Ga arc-related Lacoste intrusive suite (Martignole et al., 2006; Nantel, 2008). Pelitic gneisses dated by Friedman and Martignole (1995) contain zircons with concordant ²⁰⁷Pb/²⁰⁶Pb ages of 1203–1275 Ma. The most elongate (and most unequivocally detrital) grains are 1246–1275 Ma, which is a conservative maximum estimate for deposition and consistent with the Lacoste suite being basement to at least this pelite unit (Davidson et al., 2002). A member of the Lacoste suite is dated at 1365 ± 2 Ma, and is mapped as extending from the Central Metasedimentary Belt to the Morin terrane, and as being cut by the Labelle shear zone (Fig. 2; Nantel, 2008). This may mean that the Mont-Tremblant gneiss in the Morin terrane is correlative with the Lacoste intrusive suite (as proposed by Martignole et al., 2006), which is explored in the following.

Other nearby Grenville terranes also host Geon 14 (to late Geon 15) arc-related metaplutonic rocks: the Adirondack Highlands (McLelland and Chiarenzelli, 1990), the New Jersey Highlands (Volkert et al., 2010), the Hudson Highlands (Walsh et al., 2004), the Green Mountains (Ratcliffe et al., 1991), the Chester Dome (Ratcliffe et al., 1996), the Dysart and Redstone thrust sheets of the Central Metasedimentary Belt boundary thrust zone in Ontario (Lumbers et al., 1990), and the Portneuf–St-Maurice domain in Québec (Corrigan and van Breemen, 1997; Fig. 1). This suite may extend as far south as the French Broad massif of North Carolina (Tollo et al., 2010). The limited numbers of samples analyzed here do not allow a full igneous petrology study of the Mont-Tremblant gneiss, but they do allow some comparisons to published data. Arc rocks in terranes to the south (in Ontario and the United States) for the most part show more tonalitic major-element geochemistry (e.g., Lumbers et al., 1990; McLelland and Chiarenzelli, 1990; Volkert and Drake, 1999). In contrast to the Mont-Tremblant gneiss, these suites are dominated by tonalitic to granodioritic compositions with lower SiO₂ (∼60–70 wt%). They also have higher Na₂O (up to 7 wt%) and lower K₂O (less than 3 wt%) than those of the Mont-Tremblant gneiss, which have Na₂O > 5 wt% and K₂O of 3–6 wt%, in addition to having high FeO*/MgO (Fig. 4).

Geon 14–15 arc rocks in the Central Metasedimentary Belt of Québec are most similar in composition to the Mont-Tremblant gneiss, with evolved members sharing the higher SiO₂ and K₂O and lower Na₂O values (Blein et al., 2003; Martignole et al., 2006; Fig. 4D). The regional mafic to tonalitic gneissic Lacoste intrusive suite is interpreted as representing formation in a continental arc environment (Martignole et al., 2006). In the central part of the Central Metasedimentary Belt of Québec, the Bondy Gneiss Complex has felsic rocks comparable to those in the Morin terrane, and it is interpreted as representing a island arc built on thin continental crust in a backarc environment (Blein et al., 2003). In particular, the Mont-Tremblant gneiss shares the high Zr and low Ti/Zr values of quartzofeldspathic rocks in the Bondy Gneiss Complex, which is characteristic of backarc felsic rocks (see Blein et al., 2003). The Central Metasedimentary Belt of Québec and the Morin terrane are stitched by 1.17–1.16 Ga plutonic rocks (Corriveau and van Breemen, 2000; Peck et al., 2004), and similarities in lithologies and geochemistry of Geon 14 rocks between the Morin terrane and Central Metasedimentary Belt of Québec point to an even earlier linked history.

The isotope geochemistry of the Mont-Tremblant gneiss is broadly similar to Geon 14 arc rocks in nearby terranes. Oxygen isotopes from three country rock gneisses of this suite were reported by Peck et al. (2004) and range from δ¹⁸O = 6.7‰ to 9.9‰ (relative to standard mean ocean water [SMOW]). This is comparable to oxygen isotopes of Geon 14 tonalitic arc rocks from the Adirondack Highlands (Valley et al., 1994) and Bondy Gneiss Complex of the Central Metasedimentary Belt of Québec (Peck et al., 2004), and it is consistent with arc volcanic protoliths contaminated by supracrustal rocks. Regional Nd isotopes of these suites yield positive ɛ_Nd values and Geon 15 depleted mantle extraction ages that are in general 50–150 m.y. older than crystallization ages (McNutt and Dickin, 2011, and references therein). One Morin terrane sample has an unusually old Nd model age of 1.66 Ga, which is similar to an anomalously older model age reported from the Lacoste suite (1.59 Ga; Nantel, 2008). It remains to be seen if this lone analysis represents an older block of crust in the Morin terrane or the natural variation in isotopic compositions of the Mont-Tremblant gneiss.

The tectonic setting for these Geon 14 and 15 arc suites is interpreted to be an Andean-type environment that existed at the margin of Laurentia from 1.5(?) to 1.2 Ga, producing continental and backarc magmatism (Hanmer et al., 2000; Rivers and Corrigan, 2000). This environment may have included rifted continental fragments upon which island arcs were established and subsequently re-accreted (McLelland et al., 2010). The extent to which some of the arc rocks represent juvenile island arcs accreted to Laurentia is controversial (Carr et al., 2000; Hanmer et al., 2000).

The Geon 14 tonalite suite in the Adirondack/Green Mountains is separated across strike of the orogeny by >350 km from those in the Dysart and Redstone thrust sheets at the boundary of the Central Gneiss terrane. McEachern and van Breemen (1993) proposed that these rocks may represent remnants of a continental arc rifted by Geon 13 extension and backarc magmatism. The rifted Dysart–Mt. Holly arc is now a component of most tectonic reconstructions of the southwestern Grenville Province (e.g., Hanmer et al., 2000; Rivers and Corrigan, 2000; Chiarenzelli et al., 2010; McLelland et al., 2010). Regional Nd isotope mapping has shown that Geon 13 rifting in the Elzevir terrane of Ontario dies out to the north in the Central Metasedimentary Belt of Québec, perhaps as a result of rifting oblique to the continental margin (Dickin and McNutt, 2007). In this model, rifting produced juvenile magmatism in Elzevir terrane and was ensimatic, while to the north in the Central Metasedimentary Belt of Québec, rifting was ensialic and lacked magmatism. Dickin and McNutt (2007) also proposed that subduction was oblique to the continental margin as well, and that arcs developed in the Adirondack Highlands were built upon ocean crust, while those to the north were built upon continental crust. Thick continental basement to arc volcanism in the Central Metasedimentary Belt of Québec and Morin terrane might be the reason for the first-order difference observed in chemistry of Geon 14 arc rocks, and may be the explanation for the preponderance of tonalitic arc rocks in the Dysart–Mt. Holly arc versus granitic compositions in the Lacoste, Bondy, and Mont-Tremblant suites in Québec, which have higher SiO₂, K₂O/Na₂O, and FeO*/MgO values.

Field evidence is clear in showing that penetrative deformation of the Mont-Tremblant gneiss predated anorthosite emplacement (Martignole and Schrijver, 1970). The deformation and metamorphism have been ascribed to the 1.19–1.14 Ga Shawinigan orogeny based on geochronology in the Central Metasedimentary Belt of Québec and Adirondack Highlands (Corriveau and van Breeman, 2000; Martignole et al., 2000; Peck et al., 2005), although until now, no metamorphic minerals with high closure temperatures had been dated in the Morin terrane. The Shawinigan orogeny is interpreted to have resulted from western-directed docking of the Allochthonous monocyclic belt (including the Dysart–Mt. Holly arc) with Laurentia (McLelland et al., 2010). Delamination of the base of Shawinigan thickened crust is hypothesized to have caused mantle melting and production of 1.15 Ga AMCG magmatism (McLelland et al., 1996; Corrigan and Hanmer, 1997). Zircon rims in 10SH3 are 1159.4 ± 15.6 Ma, confirming Shawinigan heating in the Morin terrane.

I would like to acknowledge the scientific contributions and friendship of James McLelland, and I am glad to have this opportunity to contribute this paper to a volume in his honor. Myongsun Kong is thanked for assistance in the field. Joe Wooden (U.S. Geological Survey) and the Stanford University Sensitive High-Resolution Ion Microprobe–Reverse Geometry laboratory (SUMAC) staff are thanked for their help in operating the sensitive high-resolution ion microprobe–reverse geometry (SHRIMP-RG) and for assistance with data interpretation. Bruce Selleck is also thanked for assistance on the SHRIMP-RG, and Bruce Selleck and Suzie Nantel are both thanked for advice and encouragement in this project. Brian Cousens and Lizzy Ann Spencer (Carleton University) are thanked for Sm-Nd isotope analyses. Jeff Chiarenzelli, David Corrigan, and Jacques Martignole are thanked for helpful journal reviews, and Graham Baird is thanked for editorial comments. This study was supported in part by the Colgate University Research Council through the 2009–2010 Carter-Wallace Fellowship.