U-Pb ages of detrital zircons from metasedimentary rocks of the Sawtooth metamorphic complex (SMC) are key for identifying the stratigraphic age of metamorphosed strata within the Idaho batholith region of the Cordilleran orogen. The SMC is an exposure of medium- to high-grade metasedimentary rocks surrounded by the Idaho and Sawtooth batholiths. U-Pb ages of detrital zircons from SMC quartzites and quartzofeldspathic gneisses yield two distinctive age spectra consisting of primary 2900–2510 Ma and 1990–1760 Ma zircons for one group and 1870–1670 Ma, 1490–1330 Ma, and 1220–1020 Ma zircons for the other group. The suite of samples also yields a small number of zircons with concordant Cambrian and Neoproterozoic ages. Statistical and visual comparisons of age spectra with detrital-zircon data from Proterozoic and Paleozoic strata deposited on the western passive margin of Laurentia suggest that the SMC rocks were deposited in the Cambrian and Middle Ordovician. The identification of lower Paleozoic shelf strata in the SMC, along with similar sections recently identified in the Stibnite and Gospel Peaks inliers of the Idaho batholith, suggest that the Cordilleran passive margin sequence was continuous along the western margin of Laurentia, including the area occupied by the Idaho batholith.

The original distribution and preservation of Neoproterozoic to Paleozoic shelf strata along the Cordilleran passive margin in the region now occupied by the Idaho batholith (Fig. 1A) remain major questions with important implications for the Neoproterozoic rift geometries of western Laurentia and Cordilleran margin segmentation, as well as the role of basement highs in early Paleozoic depositional systems (e.g., Lund, 2008; Dickinson, 2009; Yonkee et al., 2014). It had been proposed that metasedimentary rocks within the Idaho batholith were metamorphosed equivalents of the Proterozoic Belt Supergroup (e.g., Bond et al., 1978; Fisher et al., 2001). This led to a variety of interpretations for why lower Paleozoic shelf strata had not been deposited or present in the Idaho batholith region (e.g., Yates, 1968; Burchfiel et al., 1992; Pope and Sears, 1997; Dickinson, 2004, 2006, 2009). More recent mapping and detrital-zircon analyses of isolated exposures within and surrounding the Idaho batholith have shown that some metasedimentary rocks thought to be Mesoproterozoic or older are actually Neoproterozoic to possibly early Paleozoic (e.g., Lund et al., 2003; Lewis et al., 2012, 2014), suggesting that Paleozoic shelf deposits might be continuous across the region.

The metasedimentary rocks of the Sawtooth metamorphic complex (SMC), Idaho, are upper amphibolite to granulite-facies rocks exposed at high elevations within the Atlanta lobe of the Idaho batholith (Fig. 1B; e.g., Reid, 1963; Dutrow et al., 1995). Recent U-Pb analyses of detrital zircon within the SMC metapsammites revealed Grenville-aged zircons (Bergeron, 2012), opening the possibility that the protoliths were Neoproterozoic or younger. In this study, we present U-Pb data from detrital zircons that help define the stratigraphic age and provenance of metapsammites in the SMC, which suggest the presence of lower Paleozoic shelf strata in the Idaho batholith region.

The country rocks to the Idaho batholith were initially assumed to be pre-Belt Supergroup (>1500 Ma) basement (Armstrong, 1975) or Mesoproterozoic (Belt) metasedimentary rocks (e.g., Bond et al., 1978; Fisher et al., 2001). More recent studies document a northwest-trending belt of Neoproterozoic Windermere Supergroup rocks exposed as roof pendants in Cretaceous and younger granitoids of central Idaho (Lund et al., 2003; Lund, 2004; Lewis et al., 2012). Lewis et al. (2010) also showed that quartzites (<2% feldspar) of the Syringa metamorphic succession in north-central Idaho are likely Neoproterozoic on the basis of detrital-zircon ages. These metamorphosed Neoproterozoic rocks in (north-) central Idaho are generally composed of a wide range of lithologies, including garnet- and kyanite-bearing, micaceous, quartzofeldspathic gneiss; amphibolite; schist; calc-silicate gneiss; marble; quartzite; diamictite; and bimodal volcanic strata (e.g., Lund et al., 2003; Lund et al., 2008). The quartzites are typically meta-quartz arenites intercalated with quartzite-conglomerate (Lund et al., 2003; Lewis et al., 2010). In contrast, documented occurrences of Cambrian–Ordovician (meta)sedimentary rocks in the Idaho batholith region are rare; the nearest ones to the study area are those in northern and east-central Idaho (Fig. 2) (Ross, 1947; Bush, 1989). Lund et al. (2003), however, suggested that the gray marble and calc-silicate gneisses of the Gospel Peaks and Stibnite sections (Fig. 2) in the Idaho batholith are Cambrian, based on their stratigraphic position above Neoproterozoic rocks. This study provides geochronological evidence supporting the presence of Cambrian–Ordovician strata in the Idaho batholith region.

The Sawtooth metamorphic complex (SMC) is located in the northern portion of the Sawtooth Range, Idaho (Fig. 1B). Although glacial moraines cover the eastern margin, it is presumably bounded by the Sawtooth normal fault to the east (e.g., Reid, 1963; Thackray et al., 2013), the Cretaceous Idaho batholith to the south and west, and the Eocene Sawtooth batholith to the north. The SMC section was first mapped by Reid (1963) as undifferentiated Precambrian “Thompson Peak metamorphic rocks.” Lewis et al. (2012) assigned it to the Neoproterozoic–Cambrian Windermere Supergroup on the basis of lithologic similarity to the Gospel Peaks section (Fig. 2). The SMC lithologies present include aluminous, quartzofeldspathic, calc-silicate, mafic, and migmatitic gneisses, along with mica schist, marble, and metapsammite (Fig. 3; Dutrow et al., 1995; Metz, 2010; Bergeron, 2012; Fukai, 2013; Ma, 2015). Mineral assemblages in aluminous (para)gneisses indicate metamorphic conditions consistent with transitional granulite facies (e.g., Dutrow et al., 1995; Metz, 2010; Dutrow et al., 2013). Most SMC paragneisses have been intensely and pervasively deformed by folding, thrusting, and dextral shearing and intruded by dikes and pods of Cretaceous and Eocene igneous rocks (Metz, 2010; Ma, 2015). The structures within the SMC include north-south–striking domains of isoclinal and closed folds alternating with shear zones typical of wrench-dominated transpression (Ma, 2015).

Outcrops that include metapsammite in the SMC are characterized by north-south–striking compositional layers with predominantly steep dips. Photographs of typical outcrops are shown in Figure 4. Overall, the SMC metapsammites contain extensively recrystallized quartz and feldspar (Metz, 2010; Bergeron, 2012) with three lithotypes distinguished in this study. Type I is the most abundant and is a quartzofeldspathic gneiss consisting of quartz (75%–85%), feldspar (10%–15%), and micas (5%–10% biotite, with minor muscovite) (e.g., Fig. 5A). Mylonitic fabrics occur locally in meter-scale shear zones within the type I metapsammites (e.g., Fig. 5B). Type II is a white, foliated quartzite consisting of 98% quartz with <2% biotite and feldspar (e.g., Fig. 5C; Metz, 2010). This type crops out as 3–10-m-thick layers within calc-silicate and biotite-rich quartzofeldspathic gneisses. Type III is a brown, strongly recrystallized quartzite with <1% biotite and occurs as large boudins (tens of meters scale) within biotite-rich quartzofeldspathic gneisses (e.g., Fig. 5D). In this study, 14 samples of the SMC metapsammites (Supplemental Table 1A)1, including ten samples of the type I quartzofeldspathic gneiss (two of which are mylonitic), three of the type II white quartzite, and one of the type III brown quartzite, were analyzed to determine the detrital-zircon U-Pb ages.

Samples were trimmed free of veins and alteration, crushed, milled, and sieved prior to zircon separation using traditional magnetic and density-based methods, followed by handpicking under a binocular microscope. A careful cleaning of crushing and separation facilities with high-pressure air and/or alcohol was performed prior to processing each sample to preclude cross contamination. The zircon grains were mounted in epoxy, polished, and imaged using cathodoluminescence (CL) and backscattered electrons (BSE) to identify internal structures (e.g., cores and rims) and compositional variations. U-Pb isotopic analyses were performed on a Nu-Plasma multi-collector inductively coupled plasma mass spectrometer integrated with a New Wave 213 nm ultraviolet Nd:YAG laser ablation system at the University of Florida following methods of Mueller et al. (2008). FC-1 zircon from the Duluth Complex with an age of 1099.0 ± 0.6 Ma (Paces and Miller, 1993; Black et al., 2003; Mattinson, 2010) was used as the primary age reference for data calibration and drift correction. R33 zircons from the Braintree Complex with an age of 419.26 ± 0.39 Ma (Black et al., 2004) were used as the secondary standard (Supplemental Table 1B; see footnote 1). Two FC-1 zircons were analyzed after every ten unknowns. Both standards and unknowns were ablated for 30 seconds (300 laser shots) with identical laser and mass spectrometer parameters. Drift corrections were made by linear bracketing with the FC-1 standards. Data were reduced using an in-house Excel spreadsheet and plotted using Isoplot (Ludwig, 2012). Age distributions are displayed as kernel density estimations (KDEs) utilizing DensityPlotter (Vermeesch, 2012). Tabulations of sample locations and the complete U-Pb isotopic data are in the Supplemental Tables (see footnote 1).

Quantitative comparisons of detrital-zircon age spectra can be done in several ways, but the success of these tests that are commonly used for comparing detrital zircons (e.g., Gehrels, 2012; Saylor and Sundell, 2016) is dependent on the sizes of the respective data sets in a nonlinear way. Saylor and Sundell (2016) suggested that the reliability of such comparisons is difficult to constrain when applied to sample sets with less than 300 entries, regardless of the method used. In this study, none of the SMC samples yielded >300 analyses with ages <5% (206Pb/238U ages) or <10% (207Pb/206Pb ages) discordance. With this caveat, we provide results of three methods of quantitative comparison. The Kolmogorov-Smirnov (K-S) test provides a test of the possibility that two sample sets were derived from the same population based on the total spectrum of ages and errors from each sample by comparing cumulative probability curves (e.g., Press et al., 1986; DeGraaff-Surpless et al., 2003; Sircombe and Hazelton, 2004). The primary output of the K-S test is the probability, P value, which indicates the likelihood that two suites of detrital zircons were drawn randomly from the same parent population, interpreted here as provenance, at a specific confidence level. A P value >0.05 suggests that the observed age spectra of two groups of zircons could have been derived from random sampling of the same parent population (provenance) at the 95% confidence level. Results of the K-S test for the SMC samples are presented in Table 1, and those for comparisons of the SMC sample age spectra to selected detrital-zircon data from non-SMC localities in the Cordillera are shown in Supplemental Table 1C (see footnote 1). It is important to note, however, that the magnitude of the P value is not diagnostic of relative probability, i.e., a P value of 0.07 does not indicate a higher probability of the test being successful than a value of 0.06 because the K-S test is a form of null hypothesis test. As a consequence, the absolute value of P cannot be used to indicate that any proposed relationship is more or less likely to be correct than the probability assigned to the test initially. We also utilized the overlap and similarity functions described by Gehrels (2000). Although not null-hypothesis–based tests, the values generated (from 0 to 1, no match and perfect match, respectively) are also not surrogates for the relative probability of the samples being from the same parent population or provenance, i.e., 0.6 is not 20% more likely to indicate similarity than a value of 0.5. Results of the overlap and similarity tests are presented in Table 1 for the SMC samples and Supplemental Table 1C (see footnote 1) for comparisons of the SMC and those non-SMC samples in the Cordillera.

For the presumed metamorphic overgrowths on detrital zircons analyzed in this study, the TuffZirc algorithm of Isoplot (Ludwig, 2012) was utilized to calculate the growth ages. The TuffZirc algorithm calculates a median age of the largest coherent (statistically within analytical error, 2σ) group of zircons and interprets it as the true age of the population with an asymmetric uncertainty derived from the 95% confidence errors of the median 206Pb/238U ages (Ludwig, 2012).

Detrital Zircons

Detrital zircons display both euhedral and anhedral external morphologies. Most zircon interiors are rounded and show faint oscillatory zoning (e.g., Fig. 5E). Interiors of 1601 zircons were analyzed, resulting in 552 analyses that are less than 10% or 5% discordant based on 206Pb/238U versus 207Pb/206Pb ages for grains older than 1000 Ma and 206Pb/238U versus 207Pb/235U ages for grains younger than 1000 Ma, respectively. Analyses that show greater than 10% or 5% discordance are omitted from the following discussion and are not shown in the plots. Ten out of the 552 least discordant analyses give ages younger than 1000 Ma (Supplemental Table 1D [see footnote 1]). Six of the 14 metapsammitic samples analyzed contain at least one grain younger than 1000 Ma (Fig. 6).

Samples CA13ST-5 (type I lithology), ST11-08 (type II), MC13ST-6 (type III), and MC14ST-15 (mylonitic type I) all contain a major concentration of ages at ca. 1800 Ma and a secondary concentration at ca. 2700 Ma (Fig. 6A). K-S evaluations of age spectra of detrital zircons from all SMC samples show that these four samples are similar to each other but distinct from the rest, as shown by the cumulative age-probability curves (Fig. 7). Results of the K-S test suggest that the sources of these four samples are indistinguishable at the 95% confidence level, because the P values are all larger than 0.05 (0.148–0.889; Table 1). These four spectra (Fig. 6A), therefore, are classified as Group A. Results of the overlap and similarity tests (Table 1) also support this classification. The combined results from Group A samples yield age concentrations of 2900–2510 Ma (cumulative 29%) and 1990–1760 Ma (cumulative 56%), with a subordinate group at 2120–2040 Ma (cumulative 6%) (Fig. 8A).

Zircons from eight samples, MB11-40, ST11-02, MB11-61, ST09-01, MC14ST-11, MC14ST-16 (type I lithology), along with MC13ST-19 (mylonitic type I) and CA13ST-1 (type II), are all dominated by ages clustered near ca. 1800 Ma, ca. 1400 Ma, and ca. 1100 Ma (Fig. 6B). These age spectra produce similar, but not identical, cumulative age-probability curves (Fig. 7) and therefore are classified as Group B. The slight variations in the age-probability curves are likely caused by the sampling statistics of different numbers of grains available for individual samples due to discordance. Results of the overlap and similarity tests (Table 1) support this classification as well. The Group B samples are the only ones to have Paleozoic or Neoproterozoic zircons. Two samples contain Cambrian zircons (496 ± 10 Ma, 2σ, 0.8% discordant in sample MB11-40 and 529 ± 18 Ma, 2σ, 1.8% discordant in sample ST11-02). Two samples contain Ediacaran zircons (542 ± 12 Ma and 605 ± 14 Ma, 2σ, –1.4% and 4.5% discordant, respectively, in sample MC13ST-19, and 632 ± 9 Ma, 2σ, 1.5% discordant in sample MC14ST-16). Two samples also contain Tonian zircons (742 ± 21 Ma, 2σ, 0.1% discordant in sample ST11-02 and 892 ± 13 Ma, 933 ± 13 Ma, 999 ± 66 Ma, 2σ, 3.2%, 2.0%, –1.9% discordant, respectively, in sample MC14ST-11). The combined zircons from Group B samples show major age concentrations of 1870–1670 Ma (cumulative 27%), 1490–1330 Ma (cumulative 25%), and 1220–1020 Ma (cumulative 37%), along with nine Neoproterozoic–Cambrian zircons (Fig. 8B).

Overall, the results show that Group A comprises dominantly Meso-Neoarchean (ca. 2900–2510 Ma) and Paleoproterozoic (ca. 1990–1760 Ma) ages (Fig. 8A), and Group B is dominated by Paleoproterozoic (ca. 1870–1670 Ma), early Mesoproterozoic (ca. 1490–1330 Ma), and late Mesoproterozoic (ca. 1220–1020 Ma) ages (Fig. 8B). Two samples, CA13ST-4 (type II lithology) and CA13ST-8 (type I), yield high proportions of discordant grains so are not assigned to groups or plotted. Only two (578 ± 11 Ma, 3002 ± 14 Ma, 2σ) out of 100 and three (1863 ± 8 Ma, 2698 ± 6 Ma, 2768 ± 7 Ma, 2σ) out of 137 zircons from those two samples, respectively, meet the concordance criteria (<10% for ages >1000 Ma or <5% for ages <1000 Ma). There is not a significant relationship between the two groups of age spectra and lithology. Quartzofeldspathic gneisses (type I lithology) and white quartzites (type II lithology) are represented in both groups.

Detrital-Zircon Overgrowths

Cathodoluminescence and BSE images show subrounded cores with metamorphic overgrowths up to ∼100 μm wide for many detrital zircons (e.g., Fig. 5E). For example, >30% of the zircons analyzed from samples MC13ST-6 and MC13ST-19 contain overgrowths wider than 30 μm. In CL images, the overgrowths are commonly zoned and generally darker than their corresponding cores (Figs. 9A and 9B). Fifty-five zircon overgrowths from samples MC13ST-6 and MC13ST-19 yield 206Pb/238U ages with less than 5% discordance based on 206Pb/238U versus 207Pb/235U ages.

Sample MC13ST-19 is from a strongly sheared metapsammite that contains winged feldspar porphyroclasts and S-C fabrics indicative of dextral strike-slip shearing. Quartz, feldspar, and biotite are highly recrystallized and/or sheared. Twenty-six overgrowths yield 206Pb/238U ages of ca. 85.3–117.0 Ma, and 13 of those give a TuffZirc age of 89.2 +1.5/–0.8 Ma and a concordia intercept age of 89.5 ± 0.7 Ma (mean square of weighted deviates [MSWD] = 0.52) (Fig. 9A). Sample MC13ST-6 is from a tens of meters–scale boudin of metapsammite within biotite-rich quartzofeldspathic gneiss. Quartz in this sample is strongly recrystallized. Twenty-nine overgrowths give 206Pb/238U ages of ca. 79.7–98.4 Ma, and 15 of those yield a TuffZirc age of 91.8 +1.1/–0.4 Ma and a concordia intercept age of 91.8 ± 0.5 Ma (MSWD = 1.3) (Fig. 9B). As shown in the concordia plots, minimal inheritance of older material and minor Pb loss might have occurred for some of the zircon overgrowths in samples MC13ST-6 and MC13ST-19 respectively. Therefore, the TuffZirc ages that have excluded some grains likely involved inheritance or Pb loss are better estimations of the true ages.

Discordance of the U-Pb Ages

More than half of the detrital zircons analyzed from the SMC samples are discordant at the 5% (<1000 Ma) or 10% (>1000 Ma) levels, respectively. U-Pb data from the overgrowths on detrital zircons from two samples give Cretaceous ages. The overgrowths are generally dark in CL images suggesting that they are relatively high in uranium (Rubatto and Gebauer, 2000). Oscillatory zoning and euhedral forms observed in CL images for many of the wide (>30 μm) overgrowths (e.g., Figs. 9A and 9B) indicate a fluid-rich environment (Corfu et al., 2003). Such environment could facilitate growth of euhedral crystals, and the fluid(s) can have variable uranium contents, which may lead to development of oscillatory zoning at the crystal-fluid interface (Rubatto and Gebauer, 2000). This suggests that the overgrowths grew during a hydrothermal or metamorphic event(s); such growth may well have resulted in Pb loss for some detrital zircons. The overgrowths with ages between 91.8 +1.1/–0.4 Ma and 89.2 +1.5/–0.8 Ma suggest that the metamorphic and/or hydrothermal event(s) were associated with intrusion of the early metaluminous plutons (ca. 100–85 Ma) before emplacement of the dominant peraluminous plutons (ca. 80–67 Ma) of the Atlanta lobe of the Idaho batholith (Gaschnig et al., 2010).

Depositional Ages of SMC Paragneisses

The following discussion is based only on the 552 (34%) detrital zircons that yield data meeting the concordance criteria stated above.

The ages of the youngest detrital zircons constrain the depositional age of a sedimentary stratum, the quality of which depends upon the provenance of the detritus and presence of zircon derived from igneous rocks formed and exposed near the time of deposition (e.g., Fedo et al., 2003; Andersen, 2005; Dickinson and Gehrels, 2009; Cawood et al., 2012). Four of the SMC samples analyzed contain zircons with concordant Cambrian or latest Neoproterozoic ages (ca. 496 Ma, ca. 529 Ma, ca. 542 Ma, and ca. 578 Ma) (Supplemental Table 1D [see footnote 1]), indicating that they were deposited no earlier than the Cambrian or the Neoproterozoic–Cambrian boundary. Six of the samples have maximum depositional ages less than 1100 Ma, while results from three other samples indicate deposition after 1800 Ma.

Lower Cambrian and Middle Ordovician quartzites from the passive margin of western Laurentia in Nevada, Utah, and California commonly show age spectra dominated (>99%) by Mesoproterozoic and older detrital zircons (Gehrels and Dickinson, 1995; Stewart et al., 2001; Baar, 2009; Lawton et al., 2010; Workman, 2012; Gehrels and Pecha, 2014; Yonkee et al., 2014; Chapman et al., 2015). The youngest detrital zircons in some of these Cordilleran Cambrian–Ordovician strata are hundreds of millions of years older than the inferred depositional age (e.g., Figs. 10B and 10C), so that the lack of Cambrian zircons in some of the SMC samples does not preclude them being part of the same depositional succession as those samples that do have Cambrian zircons.

Another approach to constraining depositional age is to compare the age spectra of detrital zircons from the unknown samples to those from other, better constrained, stratigraphic successions (e.g., Fedo et al., 2003; Gehrels, 2012, 2014). The strata must have an established detrital-zircon age spectrum that is referred to as a detrital-zircon fingerprint (Ross and Parrish, 1991), reference (Gehrels et al., 1995), barcode (Sircombe, 2000; Link et al., 2005), or chronofacies (Lawton et al., 2010). This approach of correlating detrital-zircon age spectra is commonly used in supercontinent reconstructions (e.g., Ireland et al., 1998; Rainbird et al., 1998; Berry et al., 2001).

Detrital-zircon age spectra from the following Proterozoic–Paleozoic sedimentary successions in the North American Cordillera are displayed in Figure 10 for comparison with those of the SMC samples: (1) The Missoula and Lemhi Groups of the Mesoproterozoic Belt Supergroup (the lower Belt is not included because it contains 1490–1610 Ma detrital zircons that are absent in the SMC samples) (Ross and Villeneuve, 2003; Stewart et al., 2010); (2) the Neoproterozoic Windermere Supergroup (Ross and Parrish, 1991; Gehrels and Ross, 1998); (3) the Syringa metasedimentary rocks (Lewis et al., 2007, 2010); (4) the Neoproterozoic to Ordovician Cordilleran passive margin strata (quartzites of Caddy Canyon, Mutual Formation, Prospect Mountain, Eureka, and their equivalents in California, Nevada, Utah, and Idaho; Gehrels and Dickinson, 1995; Stewart et al., 2001; Baar, 2009; Lawton et al., 2010; Workman, 2012; Gehrels and Pecha, 2014; Yonkee et al., 2014; Chapman et al., 2015); and (5) the Middle Pennsylvanian–Lower Permian strata (sandstones of the Wood River Formation) in south-central Idaho (Link et al., 2014). The sample locations for these data are shown in Figure 1A.

Although similarities exist between the SMC age spectra and those of the lower Paleozoic sections to the south and east based on a visual comparison of age spectra, a more quantitative comparison of these spectra utilizing the K-S test may provide a more robust statistical test for similarities of sources among samples, largely dependent on sample sizes (e.g., Berry et al., 2001; Saylor and Sundell, 2016). The combined detrital-zircon age data (Fig. 8A) of the Group A samples (Fig. 6A) are characterized by an age concentration between 1990 and 1760 Ma (56%) and a subordinate concentration between 2900 and 2510 Ma (29%), accompanied by a smaller group at 2120–2040 Ma (6%). Each of the first two age concentrations is present in all the Group A samples, while the last one, 2120–2040 Ma zircons, has not been detected in sample MC14ST-15 due to the small number of concordant analyses (n = 22/125, Fig. 6A). The aggregate age spectrum (Fig. 8A) remarkably matches those of the Middle Ordovician Eureka Quartzite (Fig. 10B) from southern and east-central Nevada (Gehrels and Dickinson, 1995; Workman, 2012; Gehrels and Pecha, 2014) and the Kinnikinic Quartzite from east-central Idaho (Baar, 2009). A K-S test of the combined SMC Group A detrital zircons (Fig. 8A) versus published data from the Middle Ordovician quartzites (Fig. 10B) gives a P value of 0.089 (Supplemental Table 1C [see footnote 1]), indicating that there is a better than 95% chance they had a common provenance, i.e., that the zircons in these samples were derived from the same population. Comparisons between the age spectra of the aggregate Group A detrital zircons and those from other Paleozoic–Proterozoic strata in the U.S. Cordillera available for comparison (Fig. 10), however, did not pass the K-S test at the 95% confidence level (P values <0.003; Supplemental Table 1C [see footnote 1]). Therefore, the SMC Group A metapsammites are possible equivalents to some of the Middle Ordovician quartzites from the Cordilleran passive margin successions, which is also supported by the results of the similarity test (Supplemental Table 1C [see footnote 1]). Spectra of some other Ordovician quartzites, perhaps those that are older than the Middle Ordovician in the Great Basin, do not match the SMC Group A spectrum, as shown by the detrital-zircon data from the lower Vinini Formation that is upper Lower to lower Middle Ordovician in age (Gehrels and Pecha, 2014; Linde et al., 2016).

It is noted that the available detrital-zircon age spectrum of the Neoproterozoic Windermere Supergroup contains a large number of ages that exist in the spectrum of the SMC Group A samples (Ross and Parrish, 1991; Gehrels and Ross, 1998). The ca. 2700 Ma peak and the 2120–2040 Ma ages, however, are missing in the Windermere spectrum (Fig. 10G). The detrital-zircon age spectrum of the Cambrian Gold Creek Quartzite from Lakeview of northern Idaho (Fig. 2) also contains abundant ages that occur in the SMC Group A spectrum (Lewis et al., 2010), but it has a much smaller Neoarchean peak, and the 2120–2040 Ma ages are largely missing in the Gold Creek Quartzite. Comparisons of the age spectra of the Windermere Supergroup and Gold Creek Quartzite with the SMC Group A spectrum failed the K-S test at the 95% confidence level (Supplemental Table 1C [see footnote 1]).

The combined age spectrum from Group B samples shows concentrations at: 1870–1670 Ma (27%), 1490–1330 Ma (25%), and 1220–1020 Ma (37%) (Fig. 8B). Age spectra of zircons extracted from Cambrian and Neoproterozoic quartzites from the Laurentian passive margin also show primary peaks at ca. 1800 Ma, 1400 Ma, and 1100 Ma, and most also have a subordinate group of ca. 2700 Ma zircons (Figs. 10C–10E). K-S comparisons between the combined Group B spectrum (Fig. 8B) and those of the Cambrian–Neoproterozoic strata of the Laurentian passive margin (Figs. 10C–10E) failed at the 95% confidence level, presumably because of the much smaller 2700 Ma peak and more pronounced 1100 Ma peak in the SMC spectrum (Fig. 10C), as well as the additional 1490–1610 Ma ages and much smaller 1800 Ma peaks in the spectra of the Neoproterozoic Mutual Formation (Fig. 10D; Gehrels and Pecha, 2014; Yonkee et al., 2014) and Caddy Canyon Quartzite (Fig. 10E; Lawton et al., 2010; Gehrels and Pecha, 2014; Yonkee et al., 2014). The lack of abundant Archean zircons in the SMC rocks may be related to the extensive degree of discordance, which would generally affect the older, more metamict zircons.

The aggregate age spectrum of the Group B samples (Fig. 8B) is most similar to the spectra of some Cambrian quartzites (Fig. 10C) from the western margin of Laurentia, including the Wood Canyon Formation (Stewart et al., 2001; Wooden et al., 2013) and Campito Formation (Chapman et al., 2015) from southern California, the Prospect Mountain Quartzite from central Utah (Lawton et al., 2010; Yonkee et al., 2014), the Geertsen Canyon Quartzite from northern Utah (Gehrels and Pecha, 2014; Yonkee et al., 2014), and the Camelback Mountain Quartzite from southeastern Idaho (Yonkee et al., 2014). The large component of Grenville-age zircons in the SMC samples is observed in the age spectrum of the Cambrian Campito Formation (Stewart et al., 2001; Chapman et al., 2015) and that of the Osgood Mountain Quartzites formed from latest Neoproterozoic through earliest Cambrian time (Linde et al., 2014). This correlation of the SMC Group B metapsammites with those Cordilleran Cambrian quartzites is supported by the results of the similarity test (Supplemental Table 1C [see footnote 1]) and is consistent with the presence of three Cambrian (496 ± 10 Ma, 529 ± 18 Ma, 542 ± 12 Ma, 2σ) and several Neoproterozoic zircons in these samples.

The Neoproterozoic Syringa metasedimentary rocks give a somewhat similar age spectrum to the SMC Group B samples (Fig. 10F; Lewis et al., 2007, 2010), but the 1400 Ma peak is largely missing in the Syringa spectrum, which is probably why it failed the K-S test at the 95% confidence level (Supplemental Table 1C [see footnote 1]). The detrital-zircon age distribution in the Pennsylvanian–Permian sandstones (Fig. 10A; Link et al., 2014), Neoproterozoic Windermere Supergroup (Fig. 10G; Ross and Parrish, 1991; Gehrels and Ross, 1998), and the Mesoproterozoic Belt Supergroup (Fig. 10H; Ross and Villeneuve, 2003; Stewart et al., 2010) also failed the K-S test at the 95% confidence level when compared to the SMC Group B spectrum (Supplemental Table 1C [see footnote 1]). Visual inspection of these data shows that their age spectra either lack one or more age populations (e.g., the Windermere and Belt Supergroups) or have additional populations (e.g., the Pennsylvanian–Permian sandstones) compared to that of the SMC Group B samples.

The depositional ages for the SMC Group A samples, which include types I, II, and III lithologies, are interpreted to be Middle Ordovician on the basis of comparisons of detrital-zircon age spectra with a Middle Ordovician reference derived from the literature (e.g., Gehrels et al., 1995). The Group B samples, including types I and II lithologies, are most similar to some Cambrian quartzites from the western Laurentian margin, which is consistent with the ages of the youngest detrital zircons analyzed from three samples of this group. In addition, the mix of protoliths for the SMC paragneisses may include quartz-rich sandstone, feldspathic sandstone, carbonaceous siltstone, mudstone, and carbonate (Metz, 2010; Fukai, 2013); this mix is similar to the unmetamorphosed Cambrian–Ordovician succession in southeastern Idaho that contains quartzite, argillite, carbonate, shale, and arkosic quartzite (e.g., Armstrong and Oriel, 1965), further supporting correlation of the SMC metapsammites with the Cambrian–Ordovician strata. The identification of highly metamorphosed lower Paleozoic sedimentary rocks in the SMC adds to the area in which metamorphosed Neoproterozoic rocks of the Laurentian margin have been identified in the Idaho batholith region (e.g., Lund et al., 2003; Lewis et al., 2007, 2010) and expands the age range of these strata to Middle Ordovician.

Implications for the SMC Lithostratigraphy

Strong deformation and high-grade metamorphism have obfuscated the primary sedimentary structures of the SMC metasedimentary rocks. The only possible way to reveal the stratigraphic relationship for the lithologic units is placing the inferred Cambrian and Middle Ordovician rocks into the structural context of the SMC, making an assumption about the younging direction in synforms and antiforms that repeat the units across strike. The amount of transposition within the units, however, makes any reconstruction of a stratigraphic column depicting the relative ages of the quartzites and quartzofeldspathic gneisses with the adjacent strata, including calc-silicate gneisses, mica schists, and quartzofeldspathic gneisses, extremely poorly constrained.

Provenance of the SMC Metapsammites

Both of the detrital-zircon age groups identified in the SMC rocks (Groups A and B) are consistent with known sources in Laurentia, either from basement exposed in the early Paleozoic or recycled from Proterozoic sedimentary rocks (Fig. 11). The 1200–1000 Ma zircons were likely derived from the Grenville orogen of eastern or southern Laurentia and coeval igneous rocks such as the Pikes Peak batholith of Colorado, as suggested for other early Paleozoic and Neoproterozoic sequences in western Laurentia (e.g., Rainbird et al., 1992; Stewart et al., 2001; Mueller et al., 2007; Howard et al., 2015). The 1490–1330 Ma grains were likely derived from the Granite-Rhyolite province and/or a variety of possible granitoids of western Laurentia including those within the Belt basin (e.g., Goodge and Vervoort, 2006; Bickford et al., 2015). The 1990–1670 Ma zircons are consistent with sources from the Yavapai-Mazatzal-Mojave provinces, Trans-Hudson orogen, Great Falls tectonic zone, and/or recycled from parts of the upper Belt Supergroup or Neoproterozoic sequences (e.g., Hoffman, 1988; Mueller et al., 2002, 2007; Ross and Villeneuve, 2003; Ansdell, 2005; Foster et al., 2006; Amato et al., 2008; Bickford et al., 2008; Wooden et al., 2013; Gifford et al., 2014; Yonkee et al., 2014). The Archean and early Paleoproterozoic zircons (3000–2000 Ma) have potential basement sources in the Wyoming and Mojave provinces as well as the Grouse Creek block and the Farmington zone, and/or recycled from the Proterozoic sedimentary rocks in western Laurentia (e.g., Mueller and Frost, 2006; Mueller et al., 2007; Mueller et al., 2011; Wooden et al., 2013; Yonkee et al., 2014).

The Neoproterozoic–Cambrian detrital zircons (ca. 742 Ma, ca. 632 Ma, ca. 605 Ma, ca. 578 Ma, ca. 542 Ma, ca. 529 Ma, and ca. 496 Ma) may have originated from local rift-related igneous rocks on the western margin of Laurentia (Lund et al., 2003 and references therein). Reports of rift-related magmatism from east-central Idaho (Fig. 2) include at least two discrete magmatic pulses represented by syenite-diorite suites and tuffaceous diamictite at ca. 665–650 Ma and ca.500–485 Ma (Lund et al., 2010).

Regional Implications

The presence of early Paleozoic shelf strata in the Idaho batholith region (SMC) combined with their likely presence farther north in the Stibnite and Gospel Peaks inliers (Fig. 2) (Lund et al., 2003) suggests that the early Paleozoic Cordilleran passive margin shelf succession was once continuous from the Great Basin through northern Idaho into Canada. This also indicates that the Cambrian–Ordovician shelf strata probably extend from east-central Idaho westward into the Idaho batholith. The results of this study suggest that there may be other not identified Cambrian–Ordovician shelf strata in the Idaho batholith region such as those in the SMC, which could be the metamorphosed and erosional remnants of the original shelf succession in this region. This assumes that the SMC rocks are not an allochthonous slice of the Cordilleran margin translated by Mesozoic dextral strike-slip faults from the Mojave or Great Basin, which seems unlikely because of their position well inboard of the edge of western Laurentia represented by the location of the western Idaho shear zone (Fig. 1A).

U-Pb geochronology of detrital zircons from quartzites and quartzofeldspathic gneisses in the SMC reveals two distinctive groups of age spectra. Group A exhibits primary concentrations of ages at 2900–2510 Ma and 1990–1760 Ma. Group B is characterized by 1870–1670 Ma, 1490–1330 Ma, and 1220–1020 Ma grains. Group B also contains three detrital zircons dated at 542 ± 12 Ma, 529 ± 18 Ma, and 496 ± 10 Ma along with several Neoproterozoic grains. The ages of these detrital zircons are consistent with Laurentian sources. The detrital-zircon age spectra from the SMC Groups A and B metapsammites are similar to those from the Middle Ordovician and Cambrian shelf strata of western Laurentian margin, respectively.

The presence of early Paleozoic shelf strata in the Idaho batholith region represented by the SMC metapsammites and their likely presence in north-central Idaho (Lund et al., 2003) suggest that the Cordilleran passive margin shelf successions were continuous from the Great Basin through the Idaho batholith region into northern Idaho and western Washington.

Reviewer Karen Lund, one anonymous reviewer, the Guest Associate Editor Joshua Schwartz, and the Science Editor Shanaka de Silva are thanked for their helpful revisions, comments, and suggestions (on the earlier version of the manuscript), which greatly improved the presentation and discussion. We thank the Colorado Scientific Society and the Geological Society of America for financial support of Ma’s fieldwork and the Belt Association, the New Orleans Geological Society, Marathon Oil, and the Department of Geology and Geophysics at Louisiana State University for support of Bergeron’s study. The field assistance of Sutie Xu, Andy Whitty, and Denis Norton made collecting and backpacking out samples possible. We are grateful to David Fluetsch and Lieze Dean of the Sawtooth National Recreation Area (National Forest Service) for facilitating the permit for collecting samples. Dr. Ann Heatherington and Dr. George Kamenov are acknowledged for assistance with the zircon analyses, and we thank Celina Will and Xiaogang Xie for CL/BSE imaging assistance. This research was supported by National Science Foundation grants EAR-1145073 to Dutrow and EAR-1145212 to Mueller and Foster.

1Supplemental Tables. Tabulations of sample locations and the complete U-Pb isotopic data. Please visit http://dx.doi.org/10.1130/GES01201.S1 or the full-text article on www.gsapubs.org to view the Supplemental Tables.