Abstract

Late Paleozoic and Mesozoic intrusive rocks from the Wallowa and Olds Ferry arc terranes of the Blue Mountains Province, Oregon-Idaho, provide constraints on the paleogeographic and tectonic setting of magmatism preserved in both arcs. Sr, Nd, and Pb isotopic data show that the Wallowa terrane represents an isotopically depleted, juvenile intra-oceanic island arc. By contrast, isotopic data for intrusive rocks of the Olds Ferry arc are more isotopically enriched, and thereby establish a clear distinction between the two arcs. This distinction strengthens paleogeographic interpretation of the Olds Ferry terrane as a fringing continental arc, and it provides a basis for correlation to other inboard Cordilleran arc terranes including Quesnellia and Stikinia. The Wallowa terrane is by contrast more similar geologically and isotopically to the outboard Insular terranes.

These isotopic data also constrain interpretations of regional lithospheric architecture. Isotopic profiles generated orthogonal to the inferred Wallowa–Olds Ferry terrane boundary and the western Idaho shear zone show abrupt increases in initial 87Sr/86Sr that mark the transitions between three geochemically distinct lithospheric columns. West-to-east spatial variability in the isotopic compositions of Neogene volcanic rocks is explained by the partial melting of these three geochemically distinct mantle reservoirs coupled to their respective crustal columns since the early Mesozoic, rather than alternative models of lithosphere-scale décollement offset during Sevier shortening. The inherited arc-related mantle of the Olds Ferry arc may also have played a primary role in the petrogenesis of distinctive Neogene low-K, high-alumina olivine tholeiites of the High Lava Plains.

INTRODUCTION

Our understanding of the late Paleozoic to Mesozoic tectonic evolution of the Cordilleran orogen of western North America depends on the ability to reconstruct a comprehensive geologic history for individual accreted terranes. Establishing a thorough geologic framework involves characterizing essential elements including the lithostratigraphy, structure, petrology, and geochemistry of the constituent rocks within targeted field areas, and placing these data within a precise spatial and temporal context. Our understanding of the tectonic evolution of the accreted terranes of the Blue Mountains Province of northeastern Oregon and western Idaho (Fig. 1) has been hampered by a lack of many of these key data types. As with many of the Paleozoic to Mesozoic lithotectonic assemblages exposed along the western Cordillera, constituent terranes of the Blue Mountains Province are largely obscured by extensive Cenozoic cover, such as the thick volcanic sequences of the Columbia River Basalt Group (Fitzgerald, 1982; Swanson et al., 1981). Regional uplift, faulting, and associated erosion of the Snake River and its tributaries have produced only limited exposures of the Paleozoic to Mesozoic “basement” (Brooks et al., 1976; Mitchell and Bennett, 1979). Exposures such as these provide rare opportunities to study the detailed history of the late Paleozoic to Mesozoic assembly and accretion of the Blue Mountains Province.

The rocks of the Blue Mountains Province comprise an assemblage of volcanic arc terranes, sedimentary basins, and mélange complexes that record a protracted history of Permian to Cretaceous subduction, magmatism, and accretion of oceanic crustal fragments to western North America (Brooks et al., 1976; Vallier et al., 1977; Brooks and Vallier, 1978; Dickinson, 1979; Silberling et al., 1984; Vallier, 1995; Gray and Oldow, 2005; Dorsey and LaMaskin, 2007; Schwartz et al., 2010; LaMaskin et al., 2015). The Wallowa arc terrane (WA) and the Olds Ferry arc terrane (OF) represent two of the four major lithotectonic units that underlie the area (Brooks and Vallier, 1978; Dickinson, 1979; Silberling et al., 1984; Vallier, 1995). Proposed tectonic models for the Blue Mountains Province rely heavily on the character of magmatism preserved in these arcs, as well as their paleogeographic constraints. Exposed basement rocks from both of these arc terranes offer critical information pertaining to their magmatic histories and paleogeography. Kurz et al. (2012) provided a detailed geologic framework of the magmatic and deformational history of the Cougar Creek complex (CCC), one of several basement complexes that manifest the midcrustal levels of the WA (Fig. 1). However, questions pertaining to the precise geologic setting of the WA still remain. In addition, an in-depth account of the timing, duration, and geochemistry of the OF has yet to be determined.

In this study, we present new trace-element and Sr, Nd, and Pb isotopic data for Permian and Triassic intrusive rocks of the WA and OF arc terranes. These geochemical and isotopic data for magmatic rocks of the WA and OF terranes are also placed within temporal context using new high-precision U-Pb zircon crystallization ages. These new data particularly clarify the age and duration of igneous activity in both arcs by distinguishing distinct unconformity-bounded cycles of magmatism. These new data from both arc terranes help to distinguish their paleogeography and tectonic affinity and allow, for the first time, a thorough comparison between arc terranes from the Blue Mountains Province and those located in the Canadian Cordillera.

Isotopic characterization of late Paleozoic and early Mesozoic magmatic rocks of the WA and OF arc terranes also provides valuable insights into regional lithospheric structure and the composition of mantle reservoirs that may have impacted subsequent Cenozoic petrogenetic processes. Isotopic analyses for the older magmatic rocks (i.e., prior to Late Jurassic suturing) from the WA and OF arc terranes are sparse. The isotopic characteristics of these two arcs have been assumed to be the same based on the initial 87Sr/86Sr ratios of a few postaccretion plutons from the WA terrane alone (Armstrong et al., 1977). Therefore, previous work has defined chemically distinct lithospheric provinces based on the isotopic character of the intra-oceanic WA terrane only and has not specifically addressed the separate role of the OF fringing arc within this context. The potential influence of a more diverse OF lithosphere on the petrogenesis of subsequent magmatic episodes in the region has also not been investigated (e.g., basaltic and rhyolitic volcanism in the western Snake River Plain and Owyhee Plateau). In this study, isotopic data for late Paleozoic and early Mesozoic intrusive rocks from the OF and WA arc terranes are combined with ∼1600 igneous rock analyses gathered from the literature and the North American Volcanic and Intrusive Rock Database to test hypotheses of regional lithospheric architecture (cf. Leeman et al., 1992) and explore the role of inherited arc-related mantle in the petrogenesis of younger basalt magmatism in the region.

GEOLOGIC FRAMEWORK

The Blue Mountains Province of northeastern Oregon and western Idaho consists of four pre-Cenozoic tectonostratigraphic assemblages (Fig. 1): the Wallowa (WA) and Olds Ferry (OF) arc terranes, the Baker accretionary terrane (which incorporates the Grindstone terrane of Blome and Nestell, 1991), and the Izee Basin terrane (Brooks et al., 1976; Vallier et al., 1977; Brooks and Vallier, 1978; Dickinson, 1979; Silberling et al., 1984; Vallier, 1995; Dorsey and LaMaskin, 2007; Schwartz et al., 2010; Kurz et al., 2012). Regional uplift and subsequent incision through a thick carapace of Cenozoic cover (primarily voluminous successions of the Columbia River Basalt Group) have exposed a number of intrusive basement complexes throughout the Blue Mountains Province (Fig. 1). In the following sections, we provide a brief overview for the WA and OF arc terranes, and descriptions for specific exposures of arc basement rocks that we have used as our primary field laboratories. The scope of this study does not include direct comparisons of these intrusive rocks with rocks of the Baker or the Izee terranes, and thus a detailed description of these terranes is not provided. Schwartz et al. (2010) and LaMaskin et al. (2011) provided recent and thorough descriptions for both the Baker and Izee terranes, respectively.

Wallowa Arc Terrane

The WA terrane (Silberling et al., 1984), also known as the “Wallowa Mountains–Seven Devils Mountains volcanic arc terrane” (Brooks and Vallier, 1978), or the “Seven Devils terrane” (Dickinson and Thayer, 1978), ranges in age from Middle Permian to Early Cretaceous (Vallier, 1995). Stratified felsic volcanic and volcaniclastic sequences of Permian (Guadalupian) age (i.e., Hunsaker Creek Formation and portions of the Clover Creek Greenstone in the Wallowa Mountains area) comprise the oldest known supracrustal rocks in the WA (Vallier, 1967, 1977, 1995). Permian supracrustal rocks are unconformably overlain by thick Middle to Upper Triassic (Ladinian and Carnian) volcanic and volcaniclastic rocks of mostly mafic to intermediate composition (i.e., Wild Sheep Creek and Doyle Creek Formations; Vallier, 1967, 1977, 1995). Late Triassic (late Carnian and early Norian) and Early Jurassic massive carbonate deposits and sandstone-mudstone flysch sequences unconformably overlie the Permian and Triassic volcanic assemblages (Vallier, 1977, 1995). The Pittsburg Landing area in Hells Canyon and Hammer Creek area in the Salmon River canyon host Early Jurassic volcanic and volcaniclastic rocks (informally named Hammer Creek assemblage; Vallier et al., 2016), and an overlapping assemblage of Late Jurassic age (Coon Hollow Formation) overlies the Permian, Triassic, and Early Jurassic rock units unconformably in northern Hells Canyon (LaMaskin et al., 2011, 2015).

Exposures of WA midcrustal rocks occur at several locations throughout the Blue Mountains Province, especially within the Snake and Salmon River canyons (Fig. 1). These include, but are not limited to, the Sparta complex located west of Richland, Oregon (Prostka and Bateman, 1962; Almy, 1977; Phelps, 1978, 1979; Phelps and Avé Lallemant, 1980; Vallier, 1995), the Oxbow complex near Oxbow, Oregon (Balcer, 1980; Schmidt, 1980; Avé Lallemant et al., 1980, 1985; Vallier, 1995), the Cougar Creek complex located just south of Pittsburg Landing, Idaho (Balcer, 1980; Vallier, 1995; Kurz, 2001; Kurz and Northrup, 2008; Kohn and Northrup, 2009; Kurz et al., 2012), and the Wolf Creek–Deep Creek and Imnaha plutonic complexes near the confluence of the Snake and Imnaha Rivers (Morrison, 1963; Chen, 1985; Vallier, 1995).

Cougar Creek Complex

The Cougar Creek complex (CCC) is exposed along 10 km of the Snake River in Hells Canyon from Temperance Creek to Pittsburg Landing, Idaho (Fig. 1). Similar basement exposures are described in the Salmon River canyon, near White Bird, Slate Creek, and Lucile, Idaho, and also to the northeast along the South Fork of the Clearwater River (Walker, 1986; Vallier, 1995; Kauffman et al., 2014). The CCC represents a regionally significant feature, assuming these easterly exposures are correlative (Vallier, 1995). The CCC contains numerous dikes and small plutons that record Middle Permian to Early Triassic felsic calc-alkaline magmatism and Late Triassic mafic to intermediate tholeiitic igneous activity (Walker, 1986, 1995; Kurz et al., 2012).

Metamorphism and deformation in the CCC occurred under lower to upper greenschist-facies conditions (Avé Lallemant, 1995; Vallier, 1995; Kurz, 2001; Kurz and Northrup, 2008; Kohn and Northrup, 2009). The CCC and adjacent rocks of the Permian Hunsaker Creek Formation of the Seven Devils Group record two distinct episodes of deformation (Kurz et al., 2012). Older D1 deformation is documented in the Hunsaker Creek Formation and is characterized by prominent foliation, lineation, folding, and other mesoscopic and microscopic structures that indicate a dominantly right-lateral kinematic regime with minor left-lateral structures (Kurz, 2001; Kurz et al., 2012). D1 deformation is broadly constrained by the Middle Permian (Wordian or Capitanian) fossil age of the Hunsaker Creek Formation (Vallier, 1967, 1977) and the Late Triassic (ca. 229.43 ± 0.08 Ma) crosscutting Suicide Point pluton (Kurz et al., 2012). Younger D2 deformation involved Late Triassic (ca. 229.43 ± 0.08 Ma to 229.13 ± 0.45 Ma) synmagmatic left-lateral, strike-slip mylonitic shearing associated with sinistral-oblique subduction (Avé Lallemant et al., 1985; Avé Lallemant, 1995; Kurz and Northrup, 2008; Kurz et al., 2012). Middle to Late Triassic 40Ar/39Ar cooling ages have been interpreted as the minimum age for peak metamorphism and deformation (Balcer, 1980; Walker, 1986; Avé Lallemant, 1995; Snee et al., 1995; Vallier, 1995; Gray and Oldow, 2005). However, subsequent interpretations of combined quartz recrystallization textures (Kurz and Northrup, 2008), titanium-in-quartz thermometry (Kohn and Northrup, 2009), and high-precision U-Pb ages for magmatic rocks show that hornblende cooling ages likely correspond to magmatic cooling as opposed to metamorphism (Kurz et al., 2012).

Salmon River Canyon

Plutonic basement rocks are exposed at several locations along the Salmon River canyon near White Bird, Slate Creek, and Lucile, Idaho, and likely represent continuations of WA arc basement present in the CCC (Fig. 1; Walker, 1986; Vallier, 1995; Schmidt et al., 2009; Kurz et al., 2012; Kauffman et al., 2014). Previous geochronologic data for intrusive rocks in the Salmon River canyon include multigrain zircon fractions from a mylonitic tonalite unit and one trondhjemite body, which yielded U-Pb ages of 258 Ma and 260 Ma, respectively (Walker, 1986). In this study, three samples were collected from the “metatonalite” map unit described in detail by Schmidt et al. (2009): the metatonalite (SAL09-01), a basalt dike that clearly crosscuts the metatonalite (SAL09-02), and a late-stage hornblende-rich comagmatic phase of the metatonalite unit (SAL09-03). The crosscutting basalt dike may also coincide with lithologies that define the “fine-grained greenstone” map unit of Schmidt et al. (2009). Crosscutting relationships between older and felsic, coarse-grained plutonic rocks and later fine- to coarse-grained mafic to intermediate intrusive units are commonly observed to the west in the CCC (Kurz, 2001; Kurz et al., 2012).

Oxbow Complex

The Oxbow complex (OXC) is exposed along the Snake River near Oxbow, Oregon (in the type locality), and extends to the northeast toward Cuprum, Idaho, along Indian Creek (Figs. 1 and 2). The OXC is lithologically and structurally similar to the CCC; major rock types include basalt, rhyolite, diabase, gabbro, diorite, quartz diorite, tonalite, trondhjemite, and their metamorphosed and deformed equivalents (Vallier, 1967; Phelps, 1978; Balcer, 1980; Schmidt, 1980; Avé Lallemant et al., 1985; Avé Lallemant, 1995; Vallier, 1995). Like the CCC, the OXC is also inferred to be the basement to the WA (Avé Lallemant, 1995; Vallier, 1995). Most rocks in the OXC were deformed and recrystallized under greenschist- to amphibolite-facies conditions (Vallier, 1967, 1995). Many lithologic contacts have been either altered through strong deformation or are gradational, which obscures the original nature of some crosscutting relationships.

Tectonic fabrics in the OXC include well-developed NE-SW–trending protomylonitic to ultramylonitic foliation that dips moderately to steeply to the NW or SE, and NE-SW–trending, horizontal to subhorizontal stretching lineations defined by deformed and/or aligned mineral grains (Schmidt, 1980; Avé Lallemant, 1995; Vallier, 1995). Mesoscopic and microscopic analysis of mylonitic rocks indicated left-lateral transport analogous to the CCC (Schmidt, 1980; Avé Lallemant et al., 1985; Avé Lallemant, 1995).

Four hornblende 40Ar/39Ar plateau ages range from 214 Ma to 228 Ma (Phelps, 1978; Avé Lallemant et al., 1980; Balcer, 1980) and are consistent with Late Triassic cooling. Multigrain fractions of zircon derived from a felsic mylonitic dike yielded a concordant earliest Triassic U-Pb age of 249 Ma for one component of the composite magmatic complex, as well as a maximum constraint for post–Early Triassic mylonitic deformation (Walker, 1986). Two multigrain fractions from a deformed tonalitic pluton yielded discordant (minimum) U-Pb zircon ages of 225–222 Ma.

Unpublished bedrock geology mapping by Vallier provided the base for field investigation and sample collection in this study (Fig. 2). Volcanic and volcaniclastic sequences of the Permian Windy Ridge and Hunsaker Creek Formations of the Seven Devils Group (Vallier, 1977) border intrusive rocks of the OXC to the northwest. The Windy Ridge Formation consists of metamorphosed volcaniclastic and felsic (rhyolitic) pyroclastic lithologies including rhyolite tuff, tuff breccia, and very rare flows and dikes that are tectonically mixed with intrusive rocks of the OXC in some places (Vallier, 1977). No age data (radioisotopic or paleontological) have been obtained from the Windy Ridge Formation. The Hunsaker Creek Formation is Permian (Capitanian or Wordian) in age based on fossil occurrences, and it consists of rhyolite flows and dikes; pyroclastic deposits consisting of rhyolitic block-and-ash flows, tuff, and lapilli tuff; and epiclastic deposits of breccia, conglomerate, sandstone, and siltstone (Vallier, 1977, 1995; Vallier et al., 2016). The Hunsaker Creek Formation is locally separated from the Windy Ridge Formation by a high-angle fault, but elsewhere it overlies grayish-green volcaniclastic rocks of the upper Windy Ridge Formation. More thorough descriptions of the Windy Ridge and Hunsaker Creek Formations were published in Vallier et al. (1977).

Intrusive map units of the OXC include dike complexes composed of variable proportions of felsic (mostly tonalite and trondhjemite) and mafic (gabbro, diabase, and basalt) intrusions (Fig. 2). A sheared rhyolite-amphibolite unit (Pora) is possibly the oldest unit in the OXC, and it contains plagiorhyolite (low amounts of K2O and CaO, and high Na2O contents) and mafic dikes that probably fed the overlying and adjacent Windy Ridge and Hunsaker Creek Formations. Plagiorhyolites have not been identified regionally in Triassic stratified and intrusive rocks, and their occurrence within a WA sequence of rocks implies a Permian age. A sheared tonalite-amphibolite unit (PTRota) consists of mylonitic tonalite, gabbro, and basalt dikes. Field observations indicate older and more intensely deformed tonalitic bodies are crosscut by later mafic bodies that often exhibit lesser degrees of deformation. A sheared amphibolite unit (PTRoa) is composed entirely of deformed gabbro, diabase, and basalt intrusions, locally exposed at the mouth of Indian Creek and across the Snake River on the Oxbow (Fig. 2). The rhyolite-amphibolite, tonalite-amphibolite, and amphibolite dike units described here may represent a continuous unit containing variable proportions of felsic versus mafic material.

A gabbro-diabase unit (PTRogd) is characterized by medium- to fine-grained dikes and tabular intrusive units that exhibit relatively less deformation compared to adjacent rocks within the sheared tonalite-amphibolite unit. Rocks from the gabbro-diabase unit may represent the less-recrystallized/deformed equivalents of “amphibolite” lithologies comprising the sheared dike units described in the previous paragraphs. A comparatively larger medium-grained quartz diorite pluton (Poqd) crops out northwest of Indian Creek and records variable amounts of crystal-plastic recrystallization expressed by poorly to well-developed mylonitic foliation that parallels the dominant NE-SW structural grain of the OXC. Mylonitic fabrics are weaker in the interior portion of the pluton and increase in intensity along its southeastern margin adjacent to the sheared tonalite-amphibolite and amphibolite units. Sample OX08-02 was collected from this pluton for U-Pb geochronology. A small gabbro intrusive body (PTRog) borders the quartz diorite (PTRoqd) along its northeastern margin and is interpreted to be genetically related to the quartz diorite. A small body of dark-gray to black, very fine-grained foliated rock interpreted as a possible roof pendant records foliation parallel to the NE-SW structural grain of the OXC.

In the southwestern corner of the Oxbow 7.5 min quadrangle, just north of Brownlee Dam along the Brownlee-Oxbow Highway, another window of Permian and Triassic rocks is exposed (Fig. 2B). A coarse-grained quartz diorite (Pwqd) comprises the only plutonic lithology in the area. This pluton records prominent thrust-sense faulting that may be related to the nearby Wildhorse shear zone, which juxtaposes rocks of the WA and the Baker terranes within the Wildhorse River canyon located approximately 2 mi (3.2 km) to the east (Mann and Vallier, 2007). Sample OX08-08 was collected from this quartz diorite for U-Pb geochronology. Volcanogenic and marine sequences of the informal Wildhorse formation (Mann, 1988) are locally juxtaposed with this quartz diorite (along the Snake River; Fig. 2B), but they are also described to unconformably (?) overlie similar intrusive rocks in the Wildhorse River canyon (Mann and Vallier, 2007). In the type section within the Wildhorse River canyon, the Wildhorse formation (map units TRwvs and TRsl) includes a buff-colored basal volcanic sandstone with limestone interbeds gradationally succeeded by 300 m of massive, bluish-colored recrystallized limestone and carbonate flysch. Carbonate rocks are depositionally overlain by ∼65 m of poorly sorted, maroon-colored volcanic breccia and conglomerate. Fossil data from bioclastic limestones indicate a Late Triassic (Norian) age for the Wildhorse formation (Mann and Vallier, 2007).

Olds Ferry Arc Terrane

With respect to the North American continent, the OF is the most inboard tectonostratigraphic unit of the Blue Mountains Province, and is considerably less exposed relative to the WA (Fig. 1). The OF has been referred to as part of the “volcanic arc terrane” (Vallier et al., 1977), as the “Juniper Mountain–Cuddy Mountain arc terrane” (Brooks and Vallier, 1978), the “Huntington volcanic arc terrane” (Brooks, 1979), the “Huntington arc” (Mullen and Sarewitz, 1983), and the “Huntington arc terrane” (Dickinson, 1979), and it was eventually mapped as the “Olds Ferry terrane” (Silberling et al., 1984). Several paleogeographic interpretations for the OF have been proposed in previous studies. One interpretation suggests that the OF evolved within an oceanic setting as a younger arc component built upon the older WA, forming a “two-stage” composite island-arc system (Vallier, 1995). Other work has speculated that the OF represents a separate arc system (independent of the WA) that fringed the North American craton (Miller, 1987; Saleeby et al., 1992; Dorsey and LaMaskin, 2007; Schwartz et al., 2010).

Volcanic, volcaniclastic, carbonate, and clastic sedimentary rocks of the lower and upper members of the Huntington Formation comprise the bulk of the OF supracrustal rock record (Brooks et al., 1976; Brooks, 1979; Dorsey and LaMaskin, 2007; Tumpane, 2010; Northrup et al., 2011). The lower member of the Huntington Formation consists primarily of mafic to intermediate massive lava flows, volcanic breccias, and subordinate volcaniclastic, conglomerate, and carbonate lithologies (Juras, 1973; Dorsey and LaMaskin, 2007; Tumpane, 2010). The upper member of the Huntington Formation is distinguished by abundant volcaniclastic and sedimentary lithologies such as volcanic sandstone, turbidites, conglomerates with cobble- to boulder-sized clasts, and laminated shale. An additional contrast is the apparent increase in felsic volcanic rocks within the upper Huntington Formation (Dorsey and LaMaskin, 2007; Tumpane, 2010; Northrup et al., 2011). The Lower to Middle Jurassic Weatherby Formation (Brooks, 1979) of the Izee basin unconformably overlies the Huntington Formation. This relationship is well documented at the former Bay Horse mine located along the Oregon side of Brownlee Reservoir north of Huntington, where a distinctive rhyolite tuff of the upper Huntington Formation is overlain by the McCord Butte conglomerate, which contains clasts of the upper Huntington Formation (Brooks, 1967, 1979; Juras, 1973; Hendricksen, 1975; Tumpane, 2010; Northrup et al., 2011).

Paleontologic data from previous studies of the Huntington Formation place these dominantly volcanogenic rocks in the Late Triassic (late Carnian to Norian; Brooks and Vallier, 1978; Brooks, 1979; Vallier, 1995). Recent high-precision isotope dilution–thermal ionization mass spectrometry (ID-TIMS) U-Pb geochronologic data (Tumpane, 2010) for volcanic lithologies from the lower and upper members of the Huntington Formation provide the first radiometric ages for these supracrustal rocks. Two volcanic samples from the lower member of the Huntington Formation have yielded U-Pb zircon ages of ca. 221–222 Ma (Tumpane, 2010), placing the lower member of the Huntington Formation in the Norian Stage of the Late Triassic. Two samples from the top of the upper Huntington section have yielded ages of ca. 187–188 Ma, indicating an Early Jurassic (Pliensbachian) age for the top of sequence (Tumpane, 2010). A precise age for the bottom of the upper Huntington section has not been determined.

Exposures of intrusive basement rocks of the OF are limited, requiring the use of multiple field areas to obtain a representative suite of samples. For this study, we conducted field investigations near the town of Huntington, Oregon, along Brownlee Reservoir, in the Dennett Creek area near the historic mining town of Mineral, Idaho, and within Rush Creek canyon, located northwest of Cambridge, Idaho. Next, we provide a brief description for key relationships documented in each of the field areas used in this study.

Plutonic Basement Rocks near Huntington, Oregon

A coarse-grained trondhjemite (TRb; HT04-04) represents the primary intrusive body exposed in this area, located ∼2.5 mi (4 km) north of the confluence between the Burnt River and the Snake River (i.e., Brownlee Reservoir) along Snake River road in Oregon (Fig. 3). This trondhjemite is also exposed to the east across Brownlee Reservoir in Idaho, and it is referred to as the Brownlee trondhjemite in this study. Another small body of tonalite (TRb; HT07-02) is exposed on the Idaho side of Brownlee Reservoir about 1 mi (1.6 km) south of the main Brownlee pluton. This small unit is more melanocratic compared to the Brownlee trondhjemite, containing a larger proportion of altered hornblende, but it is interpreted as a portion of the same pluton.

The Brownlee trondhjemite has been variously interpreted as being intrusive into the Huntington Formation (Juras, 1973; Brooks et al., 1976), or as basement upon which the Huntington Formation was deposited (Vallier, 1995; Tumpane, 2010). Field relationships recently documented by Tumpane (2010) show that the Brownlee trondhjemite represents the depositional basement for both the lower and upper members of the Huntington Formation. This stratigraphic relationship also agrees with new and existing geochronologic and paleontologic age constraints for the tonalite and the Huntington Formation (Brooks, 1979; Walker, 1986; LaMaskin, 2008; Tumpane, 2010).

Dennett Creek near Mineral, Idaho

Intrusive rocks exposed in the Dennett Creek area were most recently mapped by Payne and Northrup (2003) as the composite Iron Mountain granodiorite, consisting of equigranular granite, monzonite, hornblende quartz diorite, and smaller outcrops of hornblende ± plagioclase porphyry within the surrounding country rocks. At least one component from the composite Iron Mountain granodiorite has been dated at 200 ± 4 Ma by K-Ar in biotite, indicating Early Jurassic cooling (Hendricksen et al., 1972).

Further geologic mapping conducted during this study identified three distinct intrusive units in the Dennett Creek area. A hornblende-biotite granodiorite (i.e., Iron Mountain granodiorite [TRhgd]; DC07-06) represents the primary intrusive unit within the Dennett Creek drainage (Fig. 4). A biotite granodiorite (TRbgd; DC08-07) was distinguished in the eastern portion of the map area based on the abundance of biotite as a primary mafic mineral and higher quartz contents relative to the granodiorite within the Dennett Creek drainage to the west. Small exposures of hornblende-plagioclase porphyritic hypabyssal andesite (Tha) located throughout the field area are usually observed intruding the Big Hill shale (Jbhs), which is correlative to the Lower to Middle Jurassic Weatherby Formation (Brooks, 1979; Tumpane, 2010). An excellent example of this intrusive relationship is documented in a small drainage southwest of Mineral (location of sample DC08-01; Fig. 4). Intrusive rocks exposed to the west near Lookout Mountain in Oregon exhibit chilled margins described as “hornblende andesite” similar to the Eocene lithologies observed near Dennett Creek (Prostka, 1967).

Rush Creek Canyon

Plutonic basement rocks of the OF are exposed in the footwall block of the Rush Peak fault in the Cuddy Mountains of western Idaho (Figs. 1 and 5). Intrusive lithologies include gabbro (RC07-06; TRgb), quartz diorite (RC07-03; TRqd), porphyritic granodiorite (RC08-04; TRpg), and biotite granite (RC07-05; TRgr). The gabbro unit is locally foliated and exhibits outcrop-scale magma-mixing textures with the quartz diorite map unit (TRqd). Additional map unit descriptions were compiled from previous work in Smith and Wood (2001) and Smith (2003) and references therein. Previous K-Ar ages for these intrusive rocks range from 216 Ma to 190 Ma (Hendricksen et al., 1972).

ANALYTICAL METHODS

Whole-rock trace-element abundances were analyzed on a Thermo Scientific X-Series 2 quadrupole inductively coupled plasma–mass spectrometer (ICP-MS) at Boise State University. Accuracies of ICP-MS analyses are estimated at <5% for trace elements. Isotopic compositions and parent-daughter ratios for Rb-Sr, Sm-Nd, and common Pb were measured by multicollector ID-TIMS, using methods described in the Data Repository item1. Sample preparation and analytical methods for U-Pb geochronology by ID-TIMS are also described in the Data Repository item, as well as in Kurz et al. (2012).

Concordant U-Pb dates (considering decay constant errors) were obtained from 60 individually analyzed zircon grains from the eight dated samples (Data Repository materials and Tables S1–S2) and are illustrated on conventional concordia diagrams in Figures 6 and 7. Each sample yielded a majority cluster of equivalent single-zircon 206Pb/238U dates that we interpret as the igneous crystallization age of the zircons, which approximates the solidification age of the pluton. We discarded from age calculations the minority of grains with dates that were resolvable from the majority cluster at the 95% confidence interval. Relatively more common outliers with older dates were interpreted as antecrysts from an earlier magmatic episode or composite grains with xenocrystic cores, while rare outliers with younger dates were interpreted to have severe Pb loss not completely mitigated by chemical abrasion. Nonsystematic errors on the sample weighted mean ages are reported in the text and Table 1 as internal 2σ for those samples with probability of fit of >0.05 on the weighted mean date. For one sample with probability of fit <0.05, errors are at the 95% confidence interval, which is the internal error expanded by the square root of the mean squared weighted deviation (MSWD) and the Student’s t multiplier for n – 1 degrees of freedom (Ludwig, 2003). These error estimates should be considered when comparing our 206Pb/238U dates with those from other laboratories that used the EARTHTIME tracer solution or one that was calibrated using EARTHTIME gravimetric standards. When comparing our dates with those derived from other decay schemes (e.g., 40Ar/39Ar, 187Re-187Os), the systematic uncertainties in tracer calibration (0.03%) and the 238U decay constant (0.106%) should be added to the internal error in quadrature.

GEOCHRONOLOGY

The temporal resolution of U-Pb zircon ages determined through ID-TIMS has greatly increased due to: (1) significant reduction of common Pb blank contributions; (2) analysis of single zircon and/or crystal fragments; and (3) implementation of aggressive annealing and chemical abrasion procedures to reduce or eliminate the effects of Pb-loss (Bowring and Schmitz, 2003; Mundil et al., 2004; Mattinson, 2005; Matzel et al., 2006; Schoene et al., 2006). The resulting ability to parse time at a much higher resolution has challenged the previous assumption that a zircon date can be simply interpreted as the time a pluton had fully crystallized. In this paper, we utilize terminology from Miller et al. (2007, and references therein) to help describe characteristics in zircon age systematics observed in samples from the CCC and provide a common foundation for the interpretation of our data. We use the terms “autocryst,” “antecryst,” “xenocryst,” and “inherited grain” or “inheritance” to describe temporal interpretations of individual zircon grains and/or clusters of analyses. To review, an autocryst is a zircon crystal for which crystallization or growth is exclusively associated with a distinct pulse or increment of emplaced magma; for our data, this refers to the group of grains used to characterize the final solidification of the magma. The term antecryst pertains to zircon grains that grew in an earlier pulse of magma or within a discrete reservoir and were subsequently entrained in a later pulse. A xenocryst is described as a zircon crystal that is incorporated from the enclosing country rock during magma emplacement. Xenocrysts are in general adequately older (several million years) relative to the intruding pulse of magma so as to be readily interpreted as unrelated to the magma system. Inherited zircon is synonymous with xenocryst, referring to the presence of zircon that has survived partial melting of the source and/or assimilation processes of wall rock during intrusion.

Wallowa Arc Terrane: Oxbow Complex

Previous geochronologic results for intrusive rocks from the OXC (Fig. 2) include U-Pb ages for two deformed tonalite bodies (Walker, 1986). Two multigrain zircon analyses from a mylonitic tonalite dike (CO80-1) yielded concordant 206Pb/238U ages of 250 Ma and 249 Ma, while two multigrain zircon analyses from a deformed tonalite pluton (CO79-1) gave discordant 206Pb/238U ages of 225 Ma and 222 Ma. Discordant ages were interpreted as a minimum age for the pluton (Walker, 1986). The 40Ar/39Ar hornblende cooling ages for intrusive rocks from the Oxbow complex range from 228 to 214 Ma and are interpreted as the time of peak metamorphism (Phelps, 1978; Avé Lallemant et al., 1980, 1985; Balcer, 1980). In this section, we present new U-Pb zircon crystallization ages for one deformed quartz diorite pluton from the Oxbow complex, and a quartz diorite pluton located ∼10 km south of Oxbow, Oregon, along the Brownlee-Oxbow Highway (Fig. 2).

Quartz Diorite (OX08-02)

Ten zircon grains from a deformed quartz diorite (OX08-02) collected in the OXC (Fig. 2) are concordant with 206Pb/238U ages that range from 261.30 ± 0.50 Ma to 258.98 ± 0.17 Ma (Fig. 6A). Six analyses provide a weighted mean crystallization age of 259.14 ± 0.18 Ma (MSWD = 2.8, n = 6; Table 1). Three older analyses are interpreted as either antecrysts from earlier phases of crystallization or as xenocrysts.

Quartz Diorite (OX08-08)

Eight zircon grains from a quartz diorite (OX08-08) located to the south of the OXC (Fig. 2) are concordant with 206Pb/238U ages that range from 258.83 ± 0.25 Ma to 258.62 ± 0.25 Ma (Fig. 6B). All eight crystals yield a weighted mean crystallization age of 258.74 ± 0.6 Ma (MSWD = 0.29, n = 8; Table 1).

Wallowa Arc Terrane: Salmon River Canyon

Walker (1986) provided U-Pb ages for two samples collected in the Salmon River canyon. Two multigrain analyses from a mylonitic tonalite pluton (SC80-1) yielded ages of 259 Ma and 258 Ma. Two multigrain analyses from a second deformed tonalite pluton (LU80-1) gave U-Pb ages of 261 Ma and 259 Ma.

Tonalite (SAL09-01)

Eight zircon grains from a tonalite pluton (SAL09-01) exposed within the Salmon River canyon ∼3 km south of Slate Creek, Idaho, are concordant with 206Pb/238U ages that range from 269.09 ± 0.16 Ma to 268.43 ± 0.21 Ma (Fig. 6C). Seven analyses give a weighted mean crystallization age of 268.57 ± 0.07 Ma (MSWD = 1.5, n = 7; Table 1). One older zircon grain is interpreted as either an antecryst or a xenocryst.

Olds Ferry Terrane

Previous geochronologic investigations in the OF provided one U-Pb crystallization age of 235 Ma for the Brownlee pluton (HU79-1 of Walker, 1986) and five K-Ar ages that range from 200 Ma to 171 Ma (Armstrong and Besancon, 1970; Bruce, 1971; Hendricksen et al., 1972). In this section, we present new U-Pb zircon crystallization ages for nine intrusive units to help constrain the timing and duration of magmatism in the OF.

Brownlee Trondhjemite (HT04-04)

Seven zircon grains from the Brownlee trondhjemite (HT04-04; Fig. 3) yielded concordant analyses with 206Pb/238U dates ranging from 237.85 ± 0.16 Ma to 237.54 ± 0.14 Ma (Fig. 7A). All seven grains provide a weighted mean crystallization age of 237.68 ± 0.07 Ma (MSWD = 1.8, n = 7; Table 1).

Tonalite (HT07-02)

Nine zircon grains from a small tonalitic body (HT07-02) located just southeast of the Brownlee trondhjemite (Fig. 3) yielded concordant analyses with 206Pb/238U dates ranging from 237.72 ± 0.13 Ma to 220.23 ± 0.15 Ma (Fig. 7B). Seven consistent analyses give a weighted mean crystallization age of 237.60 ± 0.05 Ma (MSWD = 1.9, n = 7; Table 1). Two younger zircon ages of 236.28 ± 0.15 Ma and 220.23 ± 0.15 Ma imply variable Pb loss. The statistically identical ages for the Brownlee trondhjemite and this tonalite support the interpretation that they are portions of a single pluton.

Granodiorite (DC08-07)

Eight zircon grains from a biotite granodiorite (DC08-07; Fig. 4) yielded concordant analyses with 206Pb/238U dates ranging from 228.67 ± 0.12 Ma to 224.26 ± 0.13 Ma (Fig. 7C). Three analyses provide a weighted mean crystallization age of 228.61 ± 0.08 Ma (MSWD = 0.81, n = 3; Table 1). Younger zircon ages from 227.97 ± 0.17 Ma to 224.25 ± 0.13 Ma imply variable amounts of Pb loss, possibly related to the intrusion of the nearby Iron Mountain granodiorite (DC07-06; Fig. 4).

Gabbro (RC07-06)

Eight zircon grains from a gabbro (RC07-06) collected in Rush Creek canyon (Fig. 5) yielded concordant analyses with 206Pb/238U dates ranging from 221.29 ± 0.12 Ma to 220.80 ± 0.13 Ma (Fig. 7D). Six analyses provide a weighted mean crystallization age of 220.88 ± 0.05 Ma (MSWD = 1.17, n = 6; Table 1). Two slightly older zircon ages of 221.29 ± 0.12 Ma and 221.13 ± 0.13 Ma are interpreted as antecrysts.

Quartz Diorite (RC07-03)

Nine zircon grains from a quartz diorite (RC07-03) collected in the Rush Creek canyon (Fig. 5) are concordant with 206Pb/238U ages ranging from 221.20 ± 0.13 Ma to 220.22 ± 0.12 Ma (Fig. 7E). Three separate clusters of analyses are recognized, indicating distinct pulses of zircon crystallization. The youngest group of crystals gives a weighted mean age of 220.29 ± 0.06 Ma (MSWD = 1.3, n = 4), constraining the time of final crystallization (Table 1). The two older clusters of analyses are interpreted as antecrysts and represent earlier phases of crystallization at 221.18 ± 0.11 Ma (MSWD = 0.25, n = 2) and 220.78 ± 0.08 Ma (MSWD = 0.001, n = 3). The oldest group of analyses for this sample temporally correlates with the two oldest grains from sample RC07-06, while the younger group of antecrysts from this sample matches the crystallization age of the gabbro (Fig. 7D).

Porphyritic Granodiorite (RC07-04)

Five zircon grains from a porphyritic granodiorite (RC07-04) collected in Rush Creek canyon (Fig. 5) are concordant with 206Pb/238U dates ranging from 219.73 ± 0.23 Ma to 219.03 ± 0.13 Ma (Fig. 7F). Three grains give a weighted mean crystallization age of 219.64 ± 0.09 Ma (MSWD = 0.43, n = 3; Table 1), while two younger analyses may indicate small amounts of Pb loss, or younger crystallization.

Biotite Granite (RC07-05)

Nine zircon grains from a biotite granite (RC07-05) collected in Rush Creek canyon (Fig. 5) are concordant with 206Pb/238U ages ranging from 219.17 ± 0.12 Ma to 216.50 ± 0.12 Ma (Fig. 7G) and do not form any discernible groups that may be used to calculate a mean crystallization age. The youngest single-crystal zircon age of 216.50 ± 0.12 Ma is interpreted as the minimum age for the final solidification of this intrusive unit (Table 1).

Quartz Diorite Boulder (CUD09-01)

Seven concordant zircon grains from a quartz diorite boulder (CUD09-01), collected from the Early Jurassic McCord Butte Conglomerate (basal unit of the Weatherby Formation), located ∼10 mi (16 km) west of Rush Creek along Idaho State Highway 71, provide ages that range from 216.55 ± 0.13 Ma to 216.22 ± 0.12 Ma (Fig. 7H). Six analyses give a weighted mean crystallization age of 216.48 ± 0.06 Ma (MSWD = 0.81, n = 6; Table 1). The youngest analysis may have experienced a small amount of Pb loss and was not included in the weighted mean calculation.

Iron Mountain Granodiorite (DC07-06)

Five zircon grains from a hornblende-biotite granodiorite (DC07-06) collected near Iron Mountain (Fig. 4) are concordant with 206Pb/238U ages ranging from 210.20 ± 0.12 Ma to 209.93 ± 0.12 Ma (Fig. 7I). A weighted mean crystallization age of 210.04 ± 0.12 Ma was calculated for this pluton (MSWD = 2.8, n = 5; Table 1).

Dennett Creek Andesite Porphyry (DC08-01, DC08-14)

Two hypabyssal andesite porphyry intrusions sampled in the Dennett Creek drainage intrude the McCord Butte Conglomerate and Big Hill Shale members of the Lower to Middle Jurassic Weatherby Formation. Igneous titanite crystals were separated from both porphyry intrusions and give middle Eocene 206Pb/204Pb-238U/204Pb isochron ages of 47.78 ± 0.47 Ma (DC08-01) and 46.5 ± 2.6 Ma (DC08-14) (Fig. 8).

TRACE-ELEMENT GEOCHEMISTRY

Geochemical data for intrusive and extrusive rocks (and associated hypabyssal units) from the WA and OF arc terranes have been broadly used to characterize tholeiitic versus calc-alkaline magma series, determine potential source reservoirs, and interpret likely tectonic settings (Vallier, 1995). Kurz et al. (2012) supplemented these data with new major- and trace-element compositions for plutonic rocks of the CCC in the WA terrane. Vallier (1995) presented limited major- and trace-element data for intrusive units of the OXC, providing the only geochemical description of these basement rocks. In this section, we present new trace-element data for plutonic basement rocks of the OF, OXC, and Salmon River corridor (Data Repository materials and Table S3), and we illustrate these data with respect to fields for the Late Permian–Early Triassic and Late Triassic lithologies of the CCC published by Kurz et al. (2012). These data are presented in concert with U-Pb age constraints to document temporal trends in the composition of recorded magmatism. Samples for which there are both geochemical and U-Pb age data are uniquely symbolized to illustrate temporal variations in composition.

Wallowa Arc Terrane

Intrusive rocks of the Oxbow complex show primitive-mantle–normalized trace-element patterns indicative of subduction origins, including negative high field strength element (HFSE) anomalies and accompanying enrichments in large ion lithophile elements (LILEs; Fig. 9A). Chondrite-normalized rare earth element (REE) patterns for the three dated or posited Middle Permian samples (OX08-01, OX08-02, OX08-08) illustrate light REE enrichment, deep HFSE anomalies, and LILE enrichments and are very similar to the patterns for Middle Permian to Early Triassic felsic plutons of the CCC. Other intrusive diabasic samples of the OXC have flatter REE and extended trace-element patterns, more similar to the Late Triassic mafic intrusives of the CCC. Chondrite-normalized La/SmCN ratios range from 0.50 to 2.71, also demonstrating variable light (L) REE enrichment and depletion (Fig. 10A). A Ta versus Yb discriminant diagram for granitic rocks (Pearce et al., 1984) demonstrates that all samples plot within the volcanic arc granite (VAG) field (Fig. 10B). A Th/Yb versus Ta/Yb discriminant diagram for volcanic arc rocks (Pearce, 1982) shows that samples from the Oxbow complex plot within both the island-arc tholeiite (IAT) and calc-alkaline (CA) fields (Fig. 10C). In all of these discriminant diagrams, the trace-element distinctions between Permian–Triassic felsic intrusives and Late Triassic mafic intrusives are paralleled in both the CCC and OXC.

Samples collected along the Salmon River also show negative primitive-mantle–normalized HFSE anomalies and positive spikes in LILEs, signifying an arc environment (Fig. 9B). Chondrite-normalized REE patterns for the Middle Permian tonalite and its interpreted comagmatic quartz diorite show LREE enrichment and extended trace-element patterns reminiscent of the Permian rocks of the CCC and OXC. The intrusive diabase dike in the Salmon River canyon is geochemically more similar to the Late Triassic rocks of the CCC and OXC.

Olds Ferry Arc Terrane

Primitive-mantle–normalized spider diagrams of incompatible trace elements for OF basement rocks show patterns typical of magmas formed in a subduction environment (Figs. 9C and 9D). Positive spikes in normalized LILEs, including Ba, U, Pb, and Sr, are observed in most of the samples, as are negative HFSE anomalies. Chondrite-normalized REE patterns for OF rocks illustrate variable amounts of LREE enrichment and depletion. Some of the older rocks (e.g., Brownlee trondhjemite) show slight convex-up REE patterns, indicating fusion of a previously depleted source. One sample (HT07-04), a gabbroic boulder collected from a conglomerate unit within the lower member of the Huntington Formation, shows a pronounced depleted and convex-up REE pattern, as well as a strong positive Eu anomaly consistent with plagioclase accumulation. Other mainly younger granitoids exhibit relatively flat REE patterns (Rush Creek area), or LREE enrichment (Dennett Creek area). Negative Eu anomalies are also observed for samples that exhibit both LREE enrichment and depletion, and indicate plagioclase fractionation during petrogenesis. Chondrite-normalized La/SmCN ratios for OF basement rocks range from 0.31 to 3.28, further illustrating the variable LREE enrichment and depletion in these rocks (i.e., La/SmN < 1; Fig. 10A). A Ta versus Yb discriminant diagram for granitic rocks (Pearce et al., 1984) shows all samples from the OF fall within the field for volcanic arc granites (VAG; Fig. 10B). A Th/Yb versus Ta/Yb discriminant diagram for volcanic arc rocks (Pearce, 1982) illustrates that samples of plutonic basement for OF plot within both the island-arc tholeiite (IAT) and calc-alkaline (CA) fields (Fig. 10C).

ISOTOPE GEOCHEMISTRY

Until recently, Rb-Sr, Sm-Nd, and common lead (Pb) isotopic data for accreted arc terranes of the Blue Mountains Province were virtually nonexistent, posing a significant information gap and hindrance to the understanding of the tectonic evolution of the region. Schwartz et al. (2010) provided the first published Sr and Nd isotopic data for igneous and sedimentary rocks from the WA and Baker terranes and used them to test hypotheses related to the origin of sedimentary rocks from the WA and the Baker terranes. In this section, we report new Sr, Nd, and Pb isotopic data for intrusive basement rocks from the WA and OF arc terranes (Data Repository materials and Table S4).

Wallowa Arc Terrane

Intrusive rocks of the CCC (i.e., the WA) exhibit initial 87Sr/86Sr ratios that range from 0.7026 to 0.7032, and positive initial εNdi values that range from +6.74 to +9.53 (Figs. 11A–11C). Sr and Nd isotopic compositions do not show any systematic variation with respect to corresponding U-Pb crystallization ages (Kurz et al., 2012). Initial 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb common Pb ratios range from 18.213 to 18.728, 15.507 to 15.567, and 37.626 to 38.059, respectively (Figs. 11D–11F). Measured 206Pb/204Pb ratios do not show any significant variation with regard to known U-Pb crystallization ages. However, 207Pb/204Pb and 208Pb/204Pb ratios show a slight increase with decreasing age.

Igneous rocks of the OXC exhibit initial 87Sr/86Sr ratios that range from 0.7026 to 0.7040, and positive initial εNdi values that range from +6.95 to +8.55 (Figs. 11A–11C). Initial 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb ratios range from 18.470 to 19.324, 15.528 to 15.592, and 37.912 to 38.293, respectively (Figs. 11D–11F). The 206Pb/204Pb ratios for Oxbow samples show increased variation and are more radiogenic compared to 207Pb/204Pb and 208Pb/204Pb ratios, likely the result of either excess radiogenic 206Pb that had not been completely mitigated during consecutive leaching steps, or later feldspar recrystallization and alteration of original Pb isotopic compositions.

Intrusive lithologies collected along the Salmon River, north of Riggins, Idaho, exhibit initial 87Sr/86Sr ratios that range from 0.7029 to 0.7033, and positive initial εNdi values that range from +7.33 to +7.89 (Figs. 11A–11C). Initial 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb ratios range from 19.177 to 19.837, 15.553 to 15.592, and 38.188 to 38.205, respectively (Figs. 11D–11F). Similar to data from the Oxbow complex, 206Pb/204Pb ratios for Salmon River samples show increased variation and are more radiogenic relative to 207Pb/204Pb and 208Pb/204Pb ratios, and this is also attributed to either excess radiogenic 206Pb that had not been completely mitigated during consecutive leaching steps, or later recystallization of feldspar.

Olds Ferry Arc Terrane

Intrusive basement rocks of the OF exhibit initial 87Sr/86Sr ratios that range from 0.7031 to 0.7050, and positive initial εNdi values that range from +2.04 to +7.65 (Figs. 11A–11C). OF samples show a pronounced systematic decrease in their Nd isotopic compositions with respect to known U-Pb crystallization ages (Table S2). Sr isotopic compositions decrease with decreasing age, although this pattern is less pronounced. Initial 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb common Pb ratios range from 18.512 to 18.805, 15.574 to 15.626, and 38.231 to 38.522, respectively (Figs. 11D–11F). All Pb isotopic ratios show a systematic increase relative to decreasing U-Pb crystallization age. The 208Pb/204Pb ratios show the most pronounced increase with progressively younger U-Pb ages. Samples collected from Rush Creek canyon consistently show less-radiogenic Sr and Pb isotopic characteristics, and more-radiogenic Nd initial compositions relative to other analyzed basement rocks of the OF.

DISCUSSION

High-precision U-Pb zircon ages presented in this study help to define the timing and duration of magmatism in the OF and WA–Seven Devils arc terranes. New trace-element and Sr, Nd, and Pb isotopic data further characterize and discriminate pulses of arc magmatism and highlight fundamental geochemical differences. In this section, we begin by describing the magmatic histories of the OF and WA arc terranes by combining new and existing data to build a temporal and geochemical framework for comparison. We then integrate our results into a hypothesized tectonic model to explore the geologic evolution of the Blue Mountains Province.

Magmatism in the Olds Ferry Arc

New and existing U-Pb ages for intrusive lithologies and supracrustal rocks from the Huntington Formation, combined with documented field relationships, distinguish three pulses of magmatism and two unconformities in the OF from the Middle Triassic to the Early Jurassic. Prior to the late Middle Triassic (ca. 237.6 ± 0.1 Ma), the OF produced mafic to felsic tholeiitic magmatism typified by the Brownlee trondhjemite. From 237.6 ± 0.1 Ma to prior to 228.6 ± 0.1 Ma, the OF underwent a period of uplift and erosion with no preserved record of igneous activity. In the early Late Triassic from 228.6 ± 0.1 Ma to 210.0 ± 0.1 Ma, a second cycle of mafic to felsic, tholeiitic to calc-alkaline magmatism became isotopically more enriched with time. After the emplacement of the Iron Mountain granodiorite (ca. 210.0 ± 0.1 Ma) until the deposition of the upper member of the Huntington Formation, the OF records a second episode of erosion and hiatus in igneous activity. The initiation of explosive felsic volcanism associated with the upper Huntington Formation (Brooks, 1979; Vallier, 1995; Tumpane, 2010) represents the third episode of igneous activity in the OF, which began sometime after 210.0 ± 0.1 Ma and lasted until the Early Jurassic (ca. 187.1 ± 0.1 Ma; Tumpane, 2010).

The oldest episode of magmatism in the OF (older than 237.6 Ma) is constrained by the age of the Brownlee trondhjemite and the nature of its contact with the overlying Huntington Formation. Field relationships reported by Tumpane (2010) show that the Brownlee trondhjemite is the depositional basement for both members of the Huntington Formation. The maximum age for this cycle of igneous activity is unknown. Thus far, the Brownlee trondhjemite is the only mappable unit associated with this episode of magmatism; however, additional evidence for older crystalline basement of the OF includes cobble- to boulder-sized plutonic clasts preserved within lower Huntington Formation conglomerates. A gabbro boulder (HT07-04) was collected from one such conglomerate for trace-element and isotopic analyses, and it indicated the occurrence of older mafic intrusive basement related to the OF.

Trace-element data for the Brownlee trondhjemite (HT04-04) and the smaller exposure of tonalite to the south (HT07-02; Fig. 3) illustrate convex-upward chondrite-normalized REE patterns indicating depletion in the LREEs and derivation from a previously depleted source (Figs. 9C and 10). The gabbro clast collected from the lower Huntington Formation is tholeiitic and exhibits overall REE depletion relative to all other samples from the OF and average depleted mid-ocean-ridge basalt (D-MORB) of Salters and Stracke (2004). This gabbro displays a convex-upward chondrite-normalized REE pattern that illustrates extreme depletion of LREEs (i.e., La < chondrite) and a strong positive Eu anomaly, suggesting plagioclase accumulation. Unradiogenic initial 87Sr/86Sri ratios (0.7035–0.7040) and radiogenic initial εNdi values (+6.55–7.15) support the geochemical evidence for derivation from a relatively incompatible trace-element–depleted mantle source (Fig. 11).

Intrusive rocks associated with the second cycle of magmatism are mafic to felsic in composition and are dominantly calc-alkaline. The maximum age for this episode of magmatism (ca. 228.6 ± 0.1 Ma) is constrained by the U-Pb age of a biotite granodiorite (DC08-07) collected east of Iron Mountain (Fig. 4). Ammonite and bivalve fossils from a limestone unit in the lower Huntington Formation, which unconformably overlies the Brownlee trondhjemite and belongs to the second phase of magmatism, are late Carnian to early Norian in age (i.e., ca. 228 Ma, roughly the same age as the granodiorite) and support the placement of this pluton in the second phase of magmatism. Trace-element data for this granodiorite also distinguish it from the Brownlee trondhjemite (Figs. 9C and 10). U-Pb zircon ages for two samples from the lower Huntington Formation (ca. 221.7 ± 0.1 Ma and 220.7 ± 0.2 Ma) are within the temporal range defining this magmatic cycle (Tumpane, 2010). The minimum age for this cycle of magmatism is constrained by the Late Triassic (ca. 210.0 ± 0.1 Ma) Iron Mountain granodiorite (DC07-06), which is overlain by the upper Huntington Formation in the Dennett Creek area (Fig. 4; Payne and Northrup, 2003; Tumpane, 2010). Two undated basaltic dikes (HT07-01 and HT07-03) that crosscut the Brownlee trondhjemite have flat, slightly convex-upward chondrite-normalized REE patterns and are tholeiitic. Field relationships documented by Tumpane (2010) indicate that these dikes feed extrusive units of the lower Huntington Formation. Intrusive rocks collected in Rush Creek canyon are tholeiitic to calc-alkaline and exhibit chondrite-normalized REE patterns that show varying degrees of LREE enrichment (Figs. 9B and 10). New U-Pb ages also show that samples collected from Rush Creek canyon are chronologically related to the second cycle of magmatism (Table 1). The 87Sr/86Sri ratios for intrusive rocks of the second cycle of magmatism range from relatively unradiogenic values (0.7031) to more radiogenic values (0.7045; Table S2). The εNdi values range from relatively radiogenic values (7.65) to more unradiogenic values (2.04). Sr and Nd isotopic data for apatite mineral separates from two lithologies collected from the lower member of the Huntington Formation fall within the range of analyzed intrusive rocks (Fig. 11A; Tumpane, 2010).

The third cycle of magmatism is expressed only as volcanic flows and tuff breccias of the upper member of the Huntington Formation, the maximum age of which, 210.0 ± 0.1 Ma, is only broadly constrained by the age of the Iron Mountain granodiorite (DC07-06). The age of the Iron Mountain granodiorite also provides a maximum age for the regional-scale unconformity that separates the lower and upper members of the Huntington Formation. This third magmatic cycle continued through the Early Jurassic, as constrained by the ca. 187.1 ± 0.1 Ma U-Pb age of a rhyolite tuff (DC07-05) that defines the top of the upper member (Payne and Northrup, 2003; Tumpane, 2010). The upper Huntington Formation is overlain by the Lower to Middle Jurassic Weatherby Formation (Brooks et al., 1976; Brooks, 1979; Imlay, 1986; Dorsey and LaMaskin, 2007), which is defined at its base by a nonmarine conglomerate that grades into a shallow marine limestone (i.e., the Jett Creek member near Huntington, Oregon, and the McCord Butte conglomerate and Dennett Creek limestone in the Dennett Creek area in Idaho). The conglomerate and limestone units are overlain by a thick succession of marine shale and thin-bedded turbidites (Brooks, 1979). U-Pb zircon ages of 180.6 ± 0.2 Ma to 173.9 ± 0.1 Ma for volcanic tuff units collected just above the Dennett Creek limestone show that volcanic activity continued into at least the earliest Middle Jurassic (Tumpane, 2010; Northrup et al., 2011). Sr and Nd isotopic data for apatite mineral separates from five samples collected in the upper member of the Huntington Formation and two from the lower Weatherby are consistent with values measured for plutonic rocks from the second cycle of magmatism (Fig. 11A; Tumpane, 2010).

Magmatism in the Wallowa Arc

The CCC manifests the midcrust of the WA oceanic-island arc and records two temporally and compositionally distinct episodes of magmatic activity (Kurz et al., 2012). From Middle Permian to Early Triassic time (ca. 265–249 Ma), the WA was dominated by intermediate to felsic calc-alkaline magmatism. From the Early to Late Triassic, an evident hiatus in magmatic activity exists in the lithologic record of the arc. In the Late Triassic (229.43 ± 0.08 Ma to 229.13 ± 0.45 Ma), renewed magmatism was dominated by mafic to intermediate tholeiitic compositions. The Middle Permian to Early Triassic and Late Triassic episodes of magmatism in the WA correspond, both compositionally and temporally, with the Hunsaker Creek and Wild Sheep Creek Formations of the Seven Devils Group (Vallier, 1967, 1977), respectively. The Wild Sheep Creek Formation overlies the Hunsaker Creek Formation above a regional unconformity that is reported throughout the WA (Vallier, 1977,1995). This large-scale feature temporally coincides with the Early Triassic to Late Triassic hiatus in intrusive magmatism.

Trace-element data for intrusive rocks from the CCC also clearly distinguish the two cycles of magmatism (Kurz et al., 2012). Chondrite-normalized REE patterns for felsic Middle Permian to Early Triassic units are enriched in LREEs and show variable depletion in the heavy (H) REEs. Late Triassic mafic to intermediate dikes and small plutons from the CCC show flat to convex-upward REE patterns, indicating variable degrees of LREE depletion, and a previously depleted source reservoir. The recorded shift from dominantly felsic calc-alkaline magmatism to later mafic and intermediate tholeiitic igneous activity, after a clear hiatus in volcanism, was interpreted by Kurz et al. (2012) as the result of spreading ridge subduction.

The 87Sr/86Sri ratios for intrusive rocks from both cycles of magmatism cluster between 0.7026 and 0.7032 and are ubiquitously unradiogenic (Fig. 11; Table S4). The εNdi values are consistently radiogenic and range from +6.74 to +9.53. The field for Sr and Nd whole-rock isotopic data for magmatic rocks from the CCC shows a slight amount of overlap with that from the OF, but values are still decisively more depleted compared to data for the OF. Common Pb isotopic data also illustrate a clear distinction between magmatic rocks of the WA and the OF.

Magmatic rocks exposed in the Salmon River canyon show similar lithologic and crosscutting relationships to those in the CCC (Walker, 1986; Vallier, 1995). A metamorphosed tonalite pluton (SAL09-01) dated at 268.57 ± 0.07 Ma confirms previous Middle to Late Permian U-Pb pluton ages from the Salmon River canyon (Walker, 1986). This tonalite is calc-alkaline, shows chondrite-normalized REE patterns enriched in LREEs, and plots within fields for similar-aged tonalitic rocks from the CCC (Figs. 9B and 10). An undated basalt dike (SAL09-02) that crosscuts the tonalite is tholeiitic, exhibits a flat chondrite-normalized REE pattern (Fig. 9B), and consistently plots in or adjacent to fields for Late Triassic intrusive units from the CCC. The 87Sr/86Sri ratios and εNdi values for these samples also plot within the field determined for intrusive rocks from the CCC (Fig. 11; Table S4). Our new U-Pb geochronologic, trace-element, and isotopic data confirm the supposition that rocks exposed in the Salmon River canyon represent an eastern extension of WA basement.

The OXC is lithologically and structurally similar to the CCC, containing numerous, variably deformed felsic dikes and small plutons that are often crosscut by later mafic intrusive units (Schmidt, 1980; Avé Lallemant, 1995; Vallier, 1995). New U-Pb zircon ages for two quartz diorite plutons (OX08-02 and OX08-08) characterize Late Permian (ca. 259.14 ± 0.18 Ma and 258.74 ± 0.06 Ma) felsic magmatism. Several younger crosscutting mafic units were sampled for U-Pb age determination, but they contained no zircon. Previously published 40Ar/39Ar analyses for hornblende separated from these deformed mafic units indicated Late Triassic cooling ages (Phelps, 1978; Avé Lallemant et al., 1980; Balcer, 1980). We note that Middle to Late Triassic hornblende cooling ages for lithologically and structurally similar mafic rocks from the CCC were shown to overlap with active tholeiitic magmatism (Kurz et al., 2012); thus, we tentatively assign a Late Triassic age to these mafic intrusives.

Trace-element data show that the large majority of these mafic rocks are tholeiitic, with the exception of one that lies on the dividing line between the island-arc tholeiite and the calc-alkalic fields (Fig. 10C). Chondrite-normalized patterns for mafic rocks are flat to convex-upward, showing slight depletions in the LREEs, and they plot consistently within the field for Late Triassic rocks from the CCC (i.e., the WA). The two Late Permian plutons and one deformed felsic dike are calc-alkaline and show chondrite-normalized REE patterns that are LREE enriched and plot within the field for felsic rocks from the CCC. Initial 87Sr/86Sri ratios and εNdi values for intrusive rocks from the OXC overlap with values for intrusive rocks from both the WA and OF (Figs. 11A–11C; Table S4); however, initial 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb ratios coincide primarily with common Pb data from the CCC (Figs. 11D–11F). These geochemical characteristics, as well as the two cycles of Late Permian calc-alkaline and Late Triassic tholeiitic magmatism, support correlation of the OXC with the WA arc terrane.

Isotopic Assessment of Potential Magma Sources

In this section, we discuss possible source reservoirs for magmatic rocks that characterize the WA and OF arc terranes and how this assessment may contribute to paleogeographic interpretations. Decades of research on ocean-island basalts (OIB) have delineated geochemical and isotopic characteristics of end-member mantle sources. Several mantle reservoir end members and/or intermediate sources have been defined, including, but not limited to, “depleted MORB mantle” (DMM), “enriched mantle I” (EMI), “enriched mantle II” (EMII), “high-µ mantle” (HIMU, where µ = 238U/204Pb), “prevalent mantle” (PREMA), “primitive mantle” (PRIMA), and the “focus zone” (FOZO) mantle source (Hart, 1984, 1986, 1988; White, 1985; Zindler and Weaver, 1991; Eisele et al., 2002; Stracke et al., 2003, 2005; Salters and Stracke, 2004; Workman et al., 2004; Workman and Hart, 2005). In this discussion, we use the DMM, EMI, EMII, and HIMU mantle components to examine potential reservoirs for magmatic rocks of the WA and OF arcs.

In this section, we provide a brief description of the source reservoirs used in our analysis, and their present-day isotopic characteristics, which are based on compiled end-member OIB isotopic data. DMM is the source for MORBs, which exhibits average Sr, Nd, and Pb isotopic ratios of 87Sr/86Sr = 0.7026, 143Nd/144Nd = 0.51311, 206Pb/204Pb = 18.00, and 208Pb/204Pb = 37.7 (Salters and Stracke, 2004). The EMI source is characterized by low 87Sr/86Sr = 0.7053, low 143Nd/144Nd = 0.51236, low 206Pb/204Pb = 17.65, and higher 208Pb/204Pb = 38.14 (Eisele et al., 2002; Stracke et al., 2003). Geochemical modeling suggests that the EMI component incorporates a mixture of DMM with subducted pelagic sediments and other components such as recycled MORB and gabbro, or primitive mantle (Zindler and Hart, 1986; Eisele et al., 2002; Stracke et al., 2003). The EMII source exhibits high 87Sr/86Sr = 0.7090, low 143Nd/144Nd = 0.5125, higher 206Pb/204Pb = 19.00, and high 208Pb/204Pb = 38.86 (Zindler and Hart, 1986; Workman et al., 2004). Geochemical modeling recreates this enriched mantle end member from a mixture of depleted upper mantle with ancient recycled oceanic crust and continentally derived terrigenous sediments (Zindler and Hart, 1986; Workman et al., 2004). The HIMU source is named for OIB that exhibits the highest Pb isotopic characteristics, requiring high time-integrated 238U/204Pb ratios or “μ” values. This mantle component shows low 87Sr/86Sr = 0.703, high 143Nd/144Nd = 0.5129, high 206Pb/204Pb = 21.00, and high 208Pb/204Pb = 39.75, and it is thought to represent contributions from ancient recycled oceanic crust (Zindler and Hart, 1986; Stracke et al., 2005).

When using OIB-derived end-member mantle components in the context of studying OIB, they may be plotted using their present-day compositions because little time has passed to modify their composition by continued production of respective daughter nuclides. However, when comparing older (i.e., Paleozoic or Mesozoic) magmatic rocks, time-integrated isotopic evolution must be accounted for in order to make accurate comparisons. Thus, in this study, we compiled modeled parent/daughter (P/D) ratios for the DMM, EMI, EMII, and HIMU reservoirs from the literature and used them to evolve the isotopic compositions of the respective reservoirs back to the Middle Triassic (i.e., 235 Ma), so that a valid comparison with intrusive rocks of the WA and OF could be made. Data Repository Table S5 provides present-day Sr, Nd, and Pb isotopic ratios, P/D ratios, and the back-evolved Middle Triassic isotopic compositions for each component (see footnote 1).

Sr, Nd, and Pb isotopic data for magmatic rocks from the OF are dispersed along trajectories that suggest a possible mixing relationship between the DMM and EMII mantle components (Figs. 11A–11F). To explore this observation, binary mixing models between the DMM and EMII reservoirs were constructed using calculated Sr, Nd, and Pb concentrations and isotopic ratios for each end member (Table S5). Nd and Pb concentrations for the EMII component were varied slightly (well within their respective uncertainties) when generating the binary mixing models. A second binary mixing model using parameters for DMM and average North American miogeocline (Ghosh and Lambert, 1989; Schwartz et al., 2010; Table S5) was constructed to assess the possibility of interactions between magmatic rocks of the OF arc terrane and continentally derived material rather than the EMII mantle component. In addition to considering average North American sedimentary material as a binary end member (Ghosh and Lambert, 1989), we also calculated an average composition for Proterozoic crust that underlies much of SW Montana and central Idaho (i.e., the Selway terrane; Mueller et al., 1995; Foster et al., 2006). However, the average Sr and Nd trace-element concentrations and isotopic compositions for Proterozoic crust of the Selway terrane are identical to those determined for average miogeocline, and so the resulting binary model would be identical to that constructed with average miogeocline parameters.

A three-stage evolution model was used to explore Pb isotopic characteristics of magmatic rocks from the WA and OF (Figs. 11D–11F). Terrestrial Pb isotope evolution was calculated to 1.8 Ga using the two-stage model of Stacey and Kramers (1975). For the third stage (1.8 Ga to the present), μ and κ (i.e., 238U/204Pb and 232Th/204Pb, respectively) were varied to model the Pb isotopic composition of possible crustal reservoirs that may have served as sources for magmatic rocks of the WA and OF. A third stage beginning at 1.8 Ga was used in this study to simulate possible crustal reservoirs isolated during Proterozoic time, similar to the known ages of some cratonal blocks of western North America such as the Selway terrane of SW Montana and central Idaho (Foster et al., 2006, and references therein).

In 143Nd/144Nd-87Sr/86Sr space (Fig. 11A), isotopic compositions of magmatic rocks from the WA are described by an ∼2% to ∼6% contribution of the EMII component, while OF isotopic data require ∼8% to ∼25% of the EMII end member. Binary modeling in εNdi-206Pb/204Pb isotopic space (Fig. 11B) illustrates similar mixing relationships for WA samples but calls for higher proportions of the EMII component in OF rocks (e.g., ∼6% to ∼60%). However, 87Sr/86Sr and 206Pb/204Pb mixing relationships are consistent with those modeled in 143Nd/144Nd-87Sr/86Sr space (Fig. 8C). We varied Nd and Pb concentration parameters in our mixing model to assess what values would be required to obtain similar EMII proportions, but we found that this would necessitate variation beyond the given uncertainties for the modeled EMII source from Workman et al. (2004). In 207Pb/204Pb-206Pb/204Pb space, WA data show that <∼6% of the EMII component is required to describe these samples, while OF rocks need a contribution of ∼6% to 10% (Fig. 11D). Our three-stage Pb evolution model also demonstrates that WA Pb may be characterized by a depleted Proterozoic source that evolved with μ values ranging between 5 and 7.25, and κ values that range from 14 to 24 (Figs. 11E and 11F). The Pb compositions of magmatic rocks from the OF may be explained by a more fertile Proterozoic source with μ values ranging from 7.25 to 9.74, and κ values that range between 24 and 36.84.

Sr, Nd, and Pb isotopic data for magmatic rocks of the OF consistently display more enriched compositions compared to intrusive units of the WA. Binary mixtures between DMM and the EMII mantle component provide a dependable explanation for the more isotopically enriched nature of OF magmatism. Recall that the EMII mantle component is inferred to be a mixture of an enriched mantle source and subducted, continentally derived, terrigenous sediment (Zindler and Hart, 1986; Workman et al., 2004). Thus, we interpret the genesis of magmatic rocks of the OF to have involved some degree of interaction with enriched continental material. This interaction may have occurred as either contribution of enriched subducted sediments and subsequent modification of the mantle wedge, and/or through physical intermingling and assimilation of continental crust during magma emplacement. Contrasting with our findings for OF intrusive units, magmatic rocks of the WA are consistently depleted, indicating derivation from a predominantly DMM source within an intra-oceanic setting.

Arc-Arc Collisional Model for the Blue Mountains Province

The tectonic evolution of the Blue Mountains Province has been described within the framework of two differing tectonic models. The first model illustrates the WA and the OF as two temporally distinct phases of volcanic activity within a single composite intra-oceanic arc system (Vallier, 1977, 1995; Brooks and Vallier, 1978; White et al., 1992). The second tectonic model interprets the contemporaneous development of two distinct magmatic arcs, a continental-fringing OF and an intra-oceanic WA, within a doubly convergent Molucca Sea–type arc-arc collision (Dickinson and Thayer, 1978; Dickinson, 1979; Avé Lallemant et al., 1980, 1985; Mortimer, 1986; Follo, 1992; Avé Lallemant, 1995; Dorsey and LaMaskin, 2007; Schwartz et al., 2010). In this section, we integrate the geochemical, isotopic, and geochronologic frameworks for the OF and the WA into a tectonic model (Fig. 12) that follows the arc-arc collision scenario.

Stage One

Stage one illustrates the time interval from the Middle Permian to the Early Triassic (ca. 268–248 Ma). During this stage, the WA existed as an intra-oceanic arc dominated by extensive felsic calc-alkaline volcanism manifested by the Windy Ridge and Hunsaker Creek Formations of the Seven Devils Group and their intrusive equivalents exposed within the CCC, the Salmon River canyon, and the OXC (Fig. 12A; Vallier, 1967, 1977, 1995; Kurz et al., 2012). Sections of the Clover Creek Greenstone of Gilluly (1937) correlate with the Hunsaker Creek Formation (Vallier, 1977). During this time, magmatism in the WA is interpreted to have been produced through NW-directed subduction beneath a southeast-facing arc that was located in the Northern Hemisphere at a paleolatitude of 24° ± 12°N (Harbert et al., 1995; Vallier, 1995). The nature of this earliest cycle of magmatism in the OF is unconstrained in the studied outcrop belts, although one could speculate that an older component of the OF has been buried by subsequent phases of arc construction. This hypothesis requires further investigation.

Stage Two

Stage two includes the time interval after the Early Triassic and prior to the early Late Triassic (ca. 248–229 Ma; Fig. 12B). In this interval, the WA was magmatically inactive. Penetrative deformation recorded in the Permian Hunsaker Creek Formation is also broadly constrained within this period of time (Kurz et al., 2012). Regional uplift and erosion associated with volcanic inactivity and deformation are manifested by a well-developed regional unconformity that truncates the Permian Hunsaker Creek Formation of the Seven Devils Group (Vallier, 1967, 1977). The concurrent hiatus in magmatism, deformation of Permian supracrustal rocks, and regional uplift and exhumation of the WA were interpreted by Kurz et al. (2012) to have resulted from the initial stages of spreading ridge subduction beneath the WA.

Following intrusion of the ca. 238 Ma Brownlee pluton, the OF also underwent a period of volcanic inactivity, as well as extensive regional uplift and erosion, resulting in the development of the unconformity developed on the Brownlee trondhjemite (Tumpane, 2010). The exhumation and erosion of crystalline basement during this interval are also reflected by abundant plutonic clasts in conglomerate units of the lower Huntington Formation. The relatively synchronous exhumation of the WA and OF during this interval of time perhaps indicates that these two arcs were responding to the same tectonic drivers.

Stage Three

Stage three involves the time period from the Late Triassic to the late Early Jurassic (ca. 229–187 Ma; Fig. 12C). In the Late Triassic, magmatism was renewed within the WA, producing voluminous mafic to intermediate magmas manifested by the Wild Sheep Creek and Doyle Creek Formations of the Seven Devils Group (Vallier, 1977, 1995), and their intrusive equivalents exposed in the CCC (Kurz et al., 2012). Portions of the Clover Creek Greenstone of Gilluly (1937) are also Triassic in age and are perhaps correlative with Middle to Late Triassic supracrustal rocks of the Seven Devils Group (Vallier, 1977, 1995). Major- and trace-element data for this younger cycle of magmatism are tholeiitic and contrast with the calc-alkaline Middle Permian to Early Triassic cycle of volcanic activity (Kurz et al., 2012). This distinct change from felsic calc-alkaline magmatism to voluminous mafic and intermediate tholeiitic volcanism occurred after significant deformation, uplift, and erosion of the WA. This combination of phenomena supports the interpretation of Kurz et al. (2012), which calls upon the subduction of a spreading ridge as a common mechanism for these events (see also Cole and Stewart, 2009, and references therein). Cole and Stewart (2009) described magmatism in the forearc and accretionary prism as a hallmark in documented cases of spreading ridge subduction. Although this type of magmatism is not specifically reported by Kurz et al. (2012), numerous early Late and Late Triassic intrusive bodies are reported within portions of the Baker terrane accretionary complex, mostly as fault-bounded tectonic slices (Ferns and Brooks, 1995; Vallier, 1995; Walker, 1995; Schwartz et al., 2010). In fact, Schwartz et al. (2010) reported a fine-grained diorite that intruded metavolcaniclastic rocks of the Elkhorn Ridge Argillite from the Bourne subterrane of the Baker terrane. These plutonic bodies could have been associated with forearc magmatism during spreading ridge subduction, many of which (though not all) were structurally overprinted during the amalgamation and accretion of the Blue Mountains Province. Furthermore, trace-element characteristics of Late Triassic intrusive rocks from the CCC are comparable to the late Paleocene to Eocene Caribou Creek volcanics of southern Alaska (Cole et al., 2006; Cole and Stewart, 2009). The Caribou Creek volcanic field is related to slab window volcanism that resulted from spreading ridge subduction beneath the Talkeetna arc (Cole et al., 2006).

Late Triassic magmatism in the CCC was synchronous with the development of left-lateral, midcrustal, intra-arc mylonitic shear zones produced during sinistral-oblique convergence, and these are constrained by U-Pb crystallization ages from crosscutting Late Triassic intrusive bodies (Kurz et al., 2012). Left-lateral motion related to left-lateral oblique subduction supports the southward transport of the WA relative to the OF arc and North America at this time (Avé Lallemant and Oldow, 1988). Late Triassic 40Ar/39Ar hornblende cooling ages from the CCC coincide with magmatism and deformation and suggest fairly rapid exhumation of the WA (Balcer, 1980; Snee et al., 1995). Plutonic clasts within conglomerate units of the Wild Sheep Creek and Doyle Creek Formations also indicate uplift and erosion of the WA (Vallier, 1977; Follo, 1992).

Late Triassic volcanogenic rocks of the Wild Sheep Creek and Doyle Creek Formations are gradationally and unconformably overlain by the Late Triassic Martin Bridge Limestone (late Carnian to early Norian; Nolf, 1966; Vallier, 1977; Stanley et al., 2009; Riguad et al., 2010). The Martin Bridge Limestone is overlain by siliciclastic marine turbidites, shale, locally abundant carbonate- and chert-clast conglomerate, and sedimentary breccia of the Late Triassic to Early Jurassic Hurwal Formation (Nolf, 1966; Follo, 1992). Middle to Late Jurassic marine rocks of the Coon Hollow Formation overlie the Wild Sheep Creek and Doyle Creek Formations along an angular unconformity that removes the Martin Bridge Limestone and the Hurwal Formation (Morrison, 1963; Vallier, 1977; White and Vallier, 1994). At Pittsburg Landing, Idaho, the lower Red Tuff unit of the Coon Hollow Formation, which was recently dated at 196.82 ± 0.06 Ma, marks the end of Triassic–Early Jurassic magmatism in the WA (LaMaskin et al., 2015).

In the OF arc, this interval of time corresponds to the second cycle of magmatism (ca. 229–210 Ma), defined by the deposition of the lower Huntington Formation over the erosional Brownlee trondhjemite, and the emplacement of the intrusive bodies characterized in this study (Fig. 12C). Trace-element data for this cycle indicate predominant calc-alkaline characteristics, and they contrast with geochemical data from the first cycle (Figs. 6 and 7). Later within this time period, volcanic centers in the OF either shifted or igneous activity ceased altogether, and the area experienced a second episode of exhumation (Tumpane, 2010). This event is illustrated by the truncation of the lower Huntington Formation and the development of a regional unconformity, documented in both the Huntington and Dennett Creek areas, that formed after 210 Ma and before 188 Ma (Tumpane, 2010). The upper Huntington Formation manifests the third cycle of magmatism identified in the OF, overlying the lower Huntington Formation and the Brownlee trondhjemite along an angular unconformity near Huntington, Oregon, and the Iron Mountain granodiorite in the Dennett Creek area (Figs. 3 and 4).

The final arc-arc collision of the OF and WA terranes is not recorded in the plutonic rocks that were the subject of this study. Nonetheless, we view the progression of magmatic, tectonic, and exhumation events outlined here as likely culminating in amalgamation of the OF and WA arcs within the next few tens of millions of years of the Middle Jurassic, similar to models promoted by recent studies (Schwartz et al., 2010; Ware, 2013; LaMaskin et al., 2015).

Terrane Correlations

Links between constituent terranes of the Blue Mountains Province and Paleozoic to Mesozoic lithotectonic elements along the western North American Cordillera have been difficult due to a lack of geochronologic, geochemical, and isotopic constraints. However, the generation of new and recent data offers new opportunities for comparison. In the past, the Wallowa, Baker, and Olds Ferry terranes have been correlated with accreted terranes to the north as a collective tectonostratigraphic assemblage. For example, the Wallowa, Baker, and Olds Ferry comprise a west-to-east triad consisting of magmatic arc–argillite-matrix mélange–magmatic arc, similar to that documented in the Intermontane superterrane of the Canadian Cordillera (e.g., from west to east: the Stikinia arc terrane–Cache Creek argillite-matrix mélange terrane–Quesnellia arc terrane). Lithologic, faunal, structural, and temporal similarities between these northern arc-related terranes and those in the Blue Mountains Province might allow a north-to-south continuation of the group (Mortimer, 1986; Stanley and Senowbari-Daryan, 1986).

The Stikinia and Quesnellia arc terranes have been variously interpreted as a single magmatic belt (Church, 1975; Monger, 1977; Monger and Church, 1977; Dostal et al., 1999, 2009), and as separate volcanic arcs (Mortimer, 1986). These interpretations have led to a range of tectonic models that attempt to explain their present-day positions in relation to one another and relative to the intervening Cache Creek mélange. Among these are (1) the right-lateral offset of a single arc (Wernicke and Klepacki, 1988; Beck, 1991, 1992; Irving et al., 1996), (2) the oroclinal closure of a single arc and consumption of the intervening ocean basin (Nelson and Mihalynuk, 1993; Mihalynuk et al., 1994; Nelson et al., 2006), (3) Middle Jurassic thrusting of the Cache Creek terrane over the arc and subsequent synclinal folding (Samson et al., 1991; Gehrels et al., 1991), (4) the extrusion of high-pressure rocks of the Cache Creek terrane into the central portion of the arc (Dostal et al., 2009), and (5) the collision of two distinct arcs, an outboard Stikinia arc terrane and an inboard Quesnellia arc terrane (Mortimer, 1986).

Sr, Nd, and Pb isotopic data presented in this study support a pericratonic setting for the Olds Ferry arc (Fig. 11; Dickinson, 1979; Miller, 1987; Dorsey and LaMaskin, 2007; Schwartz et al., 2010), similar to interpretations for the Stikinia and Quesnellia arcs (Breitsprecher et al., 2007; Dostal et al., 2009, and references therein). Magmatic rocks from the Olds Ferry terrane are isotopically similar to Late Triassic extrusive and intrusive rocks from both Stikinia and Quesnellia (Ghosh, 1995; Smith et al., 1995; Dostal et al., 1999, 2009). The lower member of the Huntington Formation is lithologically and temporally similar to the Middle to Upper Triassic stratigraphy of Stikinia and Quesnellia, which includes arc-related, mafic to intermediate volcaniclastic units, massive flows, and epiclastic rocks deposited in subaerial and submarine environments (i.e., the Takla, Stuhini, and Nicola Groups; Mortimer, 1987; Monger et al., 1992; Ferri and Melville, 1994; Pantaleyev et al., 1996; Dostal et al., 1999, 2009; MacIntyre et al., 2001; Beatty et al., 2006; MacIntyre, 2006; Breitsprecher et al., 2007). Middle to Late Triassic volcanogenic assemblages from Stikinia, Quesnellia, and the Olds Ferry arc terranes are also bounded by similar-aged unconformities (Dostal et al., 2009). Stratigraphic, lithologic, and temporal similarities between Late Triassic assemblages from these different arc terranes provide reasonable evidence for their association and representation of a single continuous fringing arc system in the Late Triassic and possibly into the Early Jurassic.

Provided the interpretation that Stikinia and Quesnellia represent portions of the same arc, then promoting a north-to-south correlation between the WA–Baker–OF and Stikinia–Cache Creek–Quesnellia groups requires that the WA and the OF are also portions of a single arc. However, U-Pb ages, and geochemical and isotopic data from the WA and OF show that these arcs are fundamentally different. For example, in Late Triassic time, trace-element data illustrate dominant tholeiitic magmatism in the WA contemporaneous with predominantly calc-alkaline magmatism in the OF (Fig. 9). Furthermore, Sr, Nd, and Pb isotopic data clearly distinguish the Middle Permian to Late Triassic intra-oceanic WA arc from the isotopically enriched fringing OF arc (Fig. 11). This geochemical and isotopic distinction appears to preclude any model that employs a single, contemporaneous WA-OF arc system.

Exploring possible scenarios of separate WA and OF arcs leads to previous interpretations of the Wallowa terrane as a fragment of Wrangellia, which is exposed in the Vancouver Islands of British Columbia and in the Wrangell and St. Elias Mountains of southeastern Alaska (Jones et al., 1977; Hillhouse et al., 1982; Wernicke and Klepacki, 1988; Dickinson, 2004). This comparison has been questioned based on differing lithologic and stratigraphic characteristics, and aspects of basalt geochemistry (Mullen and Sarewitz, 1983; Sarewitz, 1983; Scheffler, 1983; Mortimer, 1986; Richards et al., 1991; Greene et al., 2008, 2009). However, previous work involving the geochemistry, geochronology, and paleontology for Middle to Late Triassic rocks of Wrangellia revealed important and useful data that clarify its potential relationship to other accreted terranes (Smith and MacKevett, 1970; Read and Monger, 1976; MacKevett, 1978; Mortensen and Hulbert, 1991; Lassiter, 1995; Muttoni et al., 2004; Israel et al., 2006; Bittenbender et al., 2007; Schmidt and Rogers, 2007; Brack et al., 2008). Kurz et al. (2012) described similar lithostratigraphic characteristics between Wrangellia and WA, such as late Paleozoic arc and marine sequences that were subsequently overlain by voluminous, predominantly mafic volcanic sequences above a regional unconformity. Middle to Late Triassic basalt sequences from both WA (e.g., Wild Sheep Creek Formation) and Wrangellia (e.g., Nikolai and Karmutsen Formations) are also nearly synchronous and were erupted over a relatively short interval of time. Primitive mantle– and chondrite-normalized trace-element patterns of extrusive rocks of WA and Wrangellia are similar and may be linked through common mechanisms of spreading ridge subduction (Kurz et al., 2012) or hotspot volcanism (Richards et al., 1991; Greene et al., 2008, 2009). Following this interpretation, WA represents an outboard intra-oceanic arc, distinct from the OF arc, that began initial interactions possibly as early as the Late Triassic, prior to amalgamation with the OF and eventual accretion to the North American craton by the late Early Cretaceous (Walker, 1986; Vallier, 1995).

Lithospheric Structure

Regional mapping east of the Oregon-Idaho border by Hamilton (1963) differentiated cratonal rocks of North America from what later became known as accreted arc terranes of the Blue Mountains Province, and contrasting isotopic compositions between the craton and arc-related rocks were first characterized by Armstrong et al. (1977) via the initial 87Sr/86Sr analyses of primarily late Mesozoic and younger igneous rocks. Initial 87Sr/86Sr isopleths (Fig. 13) delineate an abrupt transition between isotopically juvenile arc terranes in the west (e.g., 87Sr/86Sri < 0.704) and isotopically enriched cratonal material to the east (e.g., 87Sr/86Sri > 0.706; cf. Kistler and Peterman, 1973). The work of Armstrong et al. (1977) spurred numerous studies that have helped to characterized the lithologic, structural, and geochemical nature, and tectonic significance of this fundamental lithospheric boundary (cf. Fleck and Criss, 1985; Criss and Fleck, 1987; Lund and Snee, 1988; Strayer et al., 1989; Leeman et al., 1992; Manduca et al., 1992, 1993; McClelland et al., 2000; Giorgis et al., 2005, 2008; Gray and Oldow, 2005; King et al., 2007; McClelland and Oldow, 2007; Benford et al., 2010).

Rocks that spatially coincide with this transitional boundary record multiple episodes of deformation; the orientations of major structural features related to these events are also generally concordant with respect to the inferred subvertical geometry of the boundary. The Salmon River suture zone is located in west-central Idaho and consists of a broad thrust belt that records the suturing of arc terranes to North America in the Early Cretaceous (Selverstone et al., 1992; McClelland et al., 2000; Giorgis et al., 2007). The western Idaho shear zone is a relatively narrow mylonite zone that overprints the Salmon River suture zone recording Middle to Late Cretaceous dextral-oblique intra-arc shearing (McClelland et al., 2000; Giorgis et al., 2005, 2007, 2008). The western Idaho shear zone extends for ∼350 km and may represent the northern extension of similar-aged shear zones found within the Sierra Nevada batholith (Busby-Spera and Saleeby, 1990; Benford et al., 2010).

North of Cambridge, Idaho, focused geochemical transects show that the transition from low to high initial 87Sr/86Sr compositions occurs over distances ranging from <1 km to ∼15 km (Fig. 14; Fleck and Criss, 1985; Criss and Fleck, 1987; Manduca et al., 1992; King et al., 2007). However, to the south, isotopic gradients are less steep. An isotopic profile across southeastern Oregon and into the Snake River plain (∼43°N lat.) shows the transition from primitive oceanic initial 87Sr/86Sr values to enriched cratonic compositions occurring over a distance of >150 km, based on the initial Sr isotopic compositions of Neogene basalts and rhyolites (Leeman et al., 1992).

We compiled initial Sr isotopic and spatial reference data for late Paleozoic through Cenozoic igneous rocks from eastern Oregon, western Idaho, northern Nevada, and southeastern Washington from the literature and the North American Volcanic and Intrusive Rock Database. These data were combined with recent Sr, Nd, and Pbc isotopic data for late Paleozoic and early Mesozoic intrusive rocks from the WA and OF of the Blue Mountains Province (Schwartz et al., 2010; Kurz et al., 2012; this study). Point data for ∼1600 samples were mapped using ArcGIS software (ArcMAP) and symbolized based on age and 87Sr/86Sri values, providing a comprehensive isotopic image of the lithospheric mantle (Fig. 13). The 87Sr/86Sri data in 50 km swaths were projected onto three cross-sectional profiles using ArcMAP to illustrate isotopic gradients orthogonally across inferred boundaries that separate the constituent arc terranes of the Blue Mountains and the western Idaho shear zone (Fig. 14). All data references are provided in the Data Repository.

Initial Sr isotopic data along profile A-A′ illustrate an abrupt change between accreted arc lithosphere of the Blue Mountains to the west and cratonic material of North America to the east (Figs. 13 and 14A). Inferred boundaries that separate the WA, Izee, and OF terranes, as well as the position of the western Idaho shear zone, all closely coincide with this abrupt isotopic transition. The 87Sr/86Sri analyses for samples collected to the west of the inferred boundary between the WA and OF arcs are predominantly less than 0.7035. Along profile B-B′, 87Sr/86Sri data located to the west of the inferred boundary for the WA terrane are again predominantly less than 0.7035 (Fig. 14B); however, 87Sr/86Sri compositions located between this boundary and the western Idaho shear zone increase to values up to ∼0.7053, and mainly correspond to intrusive rocks collected from the OF arc. The position of the western Idaho shear zone on transect B-B′ marks a second pronounced increase in 87Sr/86Sri isotopic composition to values greater than 0.706.

Along the western side of the southernmost profile C-C′, initial Sr isotopic compositions are dominantly less than 0.7045, but they show a sharp increase that coincides with the location of the aeroradiometric (K/eTh) boundary identified by Evans et al. (2002) (Figs. 13 and 14C). The northern end of the K/eTh line intersects with the southwesterly projection of the OF arc terrane boundary. Between the K/eTh line and the western Idaho shear zone, 87Sr/86Sri isotopic compositions range from ∼0.7035 to ∼0.7075 (Fig. 14C). East of the western Idaho shear zone, 87Sr/86Sri data show a second prominent increase to values of 0.705 and higher. With the exception of a small number of samples located to the south of the Oregon-Idaho graben, all 87Sr/86Sri isotopic values greater than 0.706 occur on or to the east of the western Idaho shear zone.

The 87Sr/86Sri compositions along these three geochemical profiles (Fig. 14) show sharp west-to-east increases that spatially correspond with the inferred WA-OF arc terrane boundary, the western Idaho shear zone (i.e., the OF arc–continent transition), and prominent geophysical gradients (i.e., the K/eTh line; Evans et al., 2002). We propose that the lithospheric column between the K/eTh domain boundary (Evans et al., 2002) and the western Idaho shear zone (Figs. 13 and 14C) represents the crust and mantle components of the OF terrane. Although Evans et al. (2002) interpreted the K/eTh boundary as the western extent of cratonal North America, a northeast projection of the K/eTh line rather intersects the inferred contact between the Mesozoic Izee basin and the OF terrane, thus constraining the presence of these arc-related rocks to the east of this geophysical boundary. Furthermore, millimeter- to centimeter-sized radiolarian chert and argillite xenoliths occur in Miocene volcanic rocks near Westfall Butte, Oregon (Fig. 13), suggesting the presence of Baker terrane rocks directly beneath the K/eTh line, which also implies that the OF arc lies to the east of the K/eTh line.

The 87Sr/86Sri compositions for Triassic intrusive bodies from the OF arc are transitional between isotopically primitive rocks of the WA arc and evolved material located to the east of the western Idaho shear zone (Figs. 13 and 14B), and they may be used to explain spatially related isotopic variability observed in younger volcanic rocks in the region. Leeman et al. (1992) proposed that transitional isotopic compositions for Neogene basalts and rhyolites from southeastern Oregon, southwestern Idaho, and northern Nevada may be explained via derivation from oceanic and continental reservoirs that were juxtaposed along a west-dipping basal décollement associated with Cretaceous Sevier thrusting, which placed depleted oceanic source rocks above subcontinental lithospheric mantle (Fig. 14D). In this model, juxtaposed oceanic and cratonal mantle would coincide with the zone between the K/eTh line and the western Idaho shear zone. However, based on the 87Sr/86Sri data for Triassic igneous rocks from the OF arc terrane and its possible extension to south, this intermediate isotopic zone has been present since the early Mesozoic. We argue that a lithospheric-scale décollement is not needed to explain isotopic variation in younger volcanic rocks. Rather, the spatially systematic isotopic variation of Cenozoic volcanic rocks may be explained by partial melting across three compositionally distinct lithospheric columns established in the Triassic, prior to Sevier shortening (Fig. 14E). Abrupt increases of 87Sr/86Sri values that correspond to crustal boundaries support a picture of vertical to subvertical boundaries between compositionally distinct lithospheric blocks.

Sr, Nd, and Pbc isotopic data for late Paleozoic and early Mesozoic intrusive rocks from the WA and OF arc terranes also provide insights for the geochemistry of lithospheric components involved with the Cenozoic tectonomagmatic evolution of the region. Of particular interest is the petrogenesis of chemically distinct low-K, high-alumina olivine tholeiites, which are interpreted as being derived from a depleted MORB mantle end member with varying degrees of contamination by small-volume melts of mafic lithosphere (Hart et al., 1984, 1997; Shoemaker and Hart, 2002). Sr and Nd isotopic compositions of high-alumina olivine tholeiites overlap with fields for intrusive rocks from the WA and OF arc terranes (Fig. 15A). In Sr-Pbc space, compositions of high-alumina olivine tholeiites occur almost entirely within the field for OF intrusive rocks (Fig. 15B). Based on these cursory isotopic observations and our proposed lithospheric model (Fig. 14D), we infer that depleted arc mantle associated with the OF arc terrane may have had a strong and direct influence on the petrogenesis of younger igneous rocks in the region.

CONCLUSIONS

New Sr, Nd, and Pb isotopic data for intrusive rocks from the basement intrusive rocks of the Blue Mountains Province help to clarify paleogeographic interpretations for the WA and OF arc terranes. Isotopic data from both cycles of magmatism recorded in the WA show that these intrusive rocks formed in an intra-oceanic island-arc setting and were derived from depleted mantle sources. High-precision U-Pb ages for intrusive and extrusive rocks from the OF distinguish three pulses of igneous activity and two previously undocumented unconformities in the OF from the Middle Triassic to the Early Jurassic. Sr, Nd, Pb isotopic data for intrusive rocks of the OF show that they were generated from a more isotopically enriched source compared to those from the WA, and they establish a clear distinction between the two arcs. This distinction strengthens current paleogeographic interpretations of the OF representing a fringing continental arc, and it provides a basis for comparison and correlation to other Cordilleran arcs as well.

Initial Sr isotopic compositions for late Paleozoic and early Mesozoic intrusive rocks from the WA and OF arc terranes help to develop alternative interpretations of lithospheric architecture in southeastern Oregon and southwestern Idaho. Sharp increases in 87Sr/86Sri compositions coincide with the inferred boundary between the WA and OF arc terranes, and with the position of the western Idaho shear zone. The WA-OF terrane boundary and the increase in 87Sr/86Sri compositions also spatially correspond with the K/eTh line (Evans et al., 2002). A lithospheric-scale basal décollement is not needed to explain the spatial variability in the isotopic compositions of Neogene volcanic rocks in the region (e.g., Leeman et al., 1992). Variation in isotopic compositions for Neogene volcanic rocks in southeastern Oregon and southwestern Idaho may be explained by partial melting across three distinct lithospheric columns. Arc-related lithospheric mantle associated with the OF terrane may have played a significant role in the Cenozoic tectonomagmatic evolution of the region, specifically with regard to the petrogenesis of high-alumina olivine tholeiites.

This study was partially funded by a Geological Society of America Graduate Research Grant, and a Willis and Rose Burnham Graduate Student Research Grant of the Department of Geosciences, Boise State University. Funding for the analytical infrastructure of the Boise State Isotope Geology Laboratory was provided by the National Science Foundation (NSF) Major Research Instrumentation grant EAR-0521221, and NSF EAR Instrumentation and Facilities Program grant EAR-0824974. Detailed comments by Reed S. Lewis and J.K. Mortensen greatly improved aspects of the manuscript.

1GSA Data Repository Item 2016220, Tables S1–S5 (age, trace-element geochemistry, Sr-Nd-Pb isotope geochemistry, isotope mixing model parameters), sample preparation and analytical techniques, and literature data references, is available at www.geosociety.org/pubs/ft2016.htm, or on request from editing@geosociety.org.