Abstract

We conducted a (U-Th)/He zircon thermochronology study of the southern part of the Idaho batholith (central Idaho, USA) to constrain cooling through ∼200 °C and exhumation of the batholith. Samples were collected adjacent to the Idaho-Oregon (IDOR) seismic transect and at localities where U-Pb zircon, geochemical, and fabric analyses were conducted. The rocks affected by the western Idaho shear zone and associated border zone suite of the batholith cooled through the closure temperature for He in zircon prior to ca. 60 Ma, before or during emplacement of the voluminous Atlanta lobe. In contrast, the Atlanta lobe (Atlanta peraluminous suite) records a relatively constant cooling rate, in which the (U-Th)/He zircon ages are systematically ∼30 m.y. younger than the U-Pb zircon ages. We interpret this data to reflect post-magmatic isobaric cooling with little or no unroofing. The only deviation from a smooth regional cooling pattern occurs near Sawtooth Valley, where samples from the Sawtooth Range on the west side of the valley show distinctly younger ages than those from the White Cloud Peaks to the east. We interpret this difference to reflect recent cooling and exhumation associated with extensional deformation. The regionally consistent pattern of cooling and hence exhumation indicates that the current exposure level of the Idaho batholith was <5 km deep (assuming a geothermal gradient of 40 °C/km) at 50 Ma during the initiation of Challis magmatism. Our data are consistent with the existence of a crustal plateau during formation of the Atlanta lobe of the Idaho batholith.

INTRODUCTION

A diagnostic attribute of the North American Cordillera is the presence of large igneous batholiths that intruded the continental margin from ca. 125 Ma to 50 Ma (e.g., Ducea, 2001; DeCelles et al., 2009; Gehrels et al., 2009; Miller et al., 2009; Gaschnig et al., 2010; Paterson and Ducea, 2015; Premo et al., 2014) (Fig. 1). These batholiths—Coast Mountains, Idaho, Sierra Nevada, and Peninsular Range—are typically thought to have resulted from eastward-dipping subduction from the Pacific basin under the western margin of North America. The timing and geochemistry of these igneous rocks provide a framework for the growth of North America and the influence of pre-existing structures on subsequent tectonism. Further, constraining the cooling and exhumation of batholiths provides a record of orogenic crust stabilization through time (e.g., DeCelles et al., 2009).

The age and petrogenesis of the Idaho batholith (central Idaho, USA) has recently been constrained through detailed U-Pb geochronology (Gaschnig et al., 2010) and isotope geochemistry (Gaschnig et al., 2011). Gaschnig et al. (2010) subdivided the batholith into five main pulses of magmatism: the border zone suite, the early metaluminous suite, the Atlanta peraluminous suite, the late metaluminous suite, and the Bitterroot peraluminous suite (Fig. 2). Geochemical analyses of the Atlanta peraluminous suite, which forms the bulk of the geographically defined Atlanta lobe of the Idaho batholith, indicate that it is largely a result of crustal melting, unlike any other batholith of the North American Cordillera.

There are additional features that distinguish the Idaho batholith from the other Cordilleran batholiths. First, the Idaho batholith is notable due to its relative structural and bulk geochemical homogeneity (e.g., Gaschnig et al., 2011; Byerly et al., 2016). Second, the Idaho batholith is the only large batholith of the North American Cordillera that intrudes entirely into North American crust, and it is located farther east than other subduction-related batholiths in the North American Cordillera (e.g., Armstrong et al., 1977; Burchfiel et al. 1992). Third, the batholith is dominated by two-mica–bearing granites; its sources are predominantly recycled Precambrian crust (Gaschnig et al., 2011), and its zircons show pervasive xenocrystic components (e.g., Chase et al., 1978; Bickford et al., 1981; Toth and Stacey, 1992; Foster and Fanning, 1997; Gaschnig et al., 2010, 2013). Fourth, Giorgis et al. (2005) hypothesized that a ∼80–100-km-wide, subduction-related magmatic arc was active from ca. 125 to 91 Ma and was located entirely west of the current Idaho batholith; this arc was shortened to a ∼8-km-wide zone by subsequent tectonism associated with the western Idaho shear zone (see also Gaschnig et al., 2017). The other coastal batholiths lack this type of major shear zone with intense shortening on their margins, suggesting that Cretaceous igneous rocks in Idaho might also record fundamentally different intrusive and cooling paths, reflecting a different tectonic environment relative to other coastal batholiths.

In this contribution, we provide 46 new (U-Th)/He zircon data that help constrain the post-intrusion cooling and exhumation of the Idaho batholith. These (U-Th)/He zircon data were co-located with geochronology and geochemistry sites of Gaschnig et al. (2010, 2011) and Braudy et al. (2016) and the fabric studies of Byerly et al. (2016). This coordination was made possible through the Earthscope IDOR project, which—in addition to active-source (Davenport et al., 2017) and passive-source (Stanciu et al., 2016) seismic surveys—funded integrated structural, geochronological, and geochemical data collection and interpretation. Together, these data allow us to provide temperature-time (T-t) paths documenting the evolving thermal structure of the batholith and its margins. These data indicate distinct patterns of cooling across the Idaho batholith, with systematic differences in cooling within the Atlanta lobe, along the western Idaho shear zone, and in the Sawtooth Valley region. Within the Atlanta peraluminous suite, which constitutes the majority of the Atlanta lobe, the (U-Th)/He zircon ages are ∼30 m.y. younger than the U-Pb zircon ages. The presence of a relationship between age and elevation in the Atlanta lobe indicates that the (U-Th)/He zircon cooling ages reflect post-magmatic isobaric cooling, with little or no tectonic unroofing. The western Idaho shear zone exhibits systematically older (U-Th)/He zircon ages than the adjacent Atlanta lobe, with a west-to-east younging trend. Finally, on the eastern edge of the batholith, there is a distinct step in regional cooling patterns in the vicinity of the Sawtooth Valley (Fig. 1), which we interpret as recent extensional deformation resulting in the exhumation of the Sawtooth Range (Fig. 1). Overall, the pattern of cooling is consistent with formation of a Cretaceous–Paleogene crustal plateau in central Idaho.

PREVIOUS WORK ON THE COOLING AND EXHUMATION OF THE IDAHO BATHOLITH

Previous studies on the cooling and exhumation of the Idaho batholith document variable cooling as a function of position within the batholith. Earlier studies (Criss et al., 1984; Sweetkind and Blackwell, 1989) show older ages obtained along the margins, while younger ages were observed within the central part of the batholith. Criss et al. (1982) obtained K-Ar cooling ages as old as 95 Ma along the margins. In the central portions of the batholith, cooling ages are significantly younger, ca. 35 Ma. Sweetkind and Blackwell (1989) observed a similar pattern in apatite and zircon fission-track ages. They reported older apatite fission-track ages (>45 Ma) along the margins of the batholith and younger ages at lower elevations and in close proximity to Challis suite exposures. The Challis magmatic event succeeded Idaho batholith emplacement and continued for ∼10 m.y. resulting in elevated thermal gradients during the Eocene. Based on their results, Sweetkind and Blackwell (1989) concluded that the batholith cooled below ∼110 °C by ca. 50 Ma and has undergone ∼3 km of erosion since 10 Ma. Sweetkind and Blackwell (1989) inferred that hydrothermal activity affected the apatite fission-track ages, particularly in the deeply incised valleys, resulting in young ages.

Other studies focusing on the cooling of the Idaho batholith have considered the evolution or exhumation of the deep to mid-crust along low-angle detachment faults, forming metamorphic core complexes (e.g., Silverberg, 1990; Foster et al., 2001; Foster and Raza, 2002; Vogl et al., 2012). Foster et al. (2001) documented magmatism, extensional deformation, and exhumation of the Bitterroot metamorphic core complex occurring in regions of thickened crust in the northeastern section of the Idaho batholith, within the Bitterroot lobe. Another metamorphic core complex in the vicinity of the Idaho batholith is the Pioneer metamorphic core complex, located east of the Idaho batholith (Fig. 1; O’Neill and Pavlis, 1988; Silverberg, 1990; Vogl et al., 2012). Vogl et al. (2012) have constrained the timing of the syn-extensional Pioneer intrusive suite to 50–48 Ma, similar to that of the early phases of Challis magmatism (Gaschnig et al., 2010).

More recent studies have focused on the cooling and exhumation of the border zone suite adjacent to the western Idaho shear zone (Fig. 1; Giorgis et al., 2008; Braudy et al., 2016; Gaschnig et al., 2017; Montz and Kruckenberg, 2017). U-Pb monazite ages of ca. 90 Ma were obtained within a migmatite domain immediately east of the western Idaho shear zone (Montz and Kruckenberg, 2017). Within the western Idaho shear zone, Giorgis et al. (2008) documented that cooling below the 40Ar/39Ar biotite closure temperature (∼350 °C) and the apatite fission-track annealing temperature (<120 °C) occurred between 85–70 Ma and ca. 40 Ma, respectively. The 40Ar/39Ar ages from within the western Idaho shear zone are generally consistent with regional data (e.g., Snee et al., 1995) that indicate increasingly younger ages from west to east within the western Idaho shear zone (Fig. 1).

We present new (U-Th)/He zircon data across the western Idaho shear zone and Idaho batholith to constrain the cooling and exhumation of the batholith relative to the IDOR seismic transect. The T-t history—from crystallization to <200 °C—has not been considered in the context of new divisions of the Idaho batholith based on U-Pb geochronology and geochemistry (Gaschnig et al., 2010, 2011) and new seismic data (Davenport et al., 2017) acquired through the EarthScope IDOR project. Combining thermochronometric data with new seismic models (Davenport et al., 2017) and structural and geochronologic data (Braudy et al., 2016; Byerly et al., 2016; Gaschnig et al., 2017) gives a broader understanding of the crustal growth and stabilization of the northern North American Cordillera.

ZIRCON (U-Th)/He DATING METHODS

(U-Th)/He zircon thermochronometry is based on the formation and retention of He in zircon and provides the time at which rocks cool below ∼130–250 °C (e.g., Reiners, 2005). Closure temperature variability is a function of grain size and cooling rate, with larger grain sizes and faster cooling rates corresponding to higher closure temperatures. Recent investigations on the (U-Th)/He system in zircon have revealed the relationship between radiation damage, ages, and cooling rates. Accumulated radiation damage in zircon can result in either inhibited or enhanced diffusion (e.g., Nasdala et al., 2004; Reiners, 2005; Guenthner et al., 2013; Ketcham et al., 2013). Diffusion is inhibited when the alpha damage acts as a trap for He, preventing the atom from readily diffusing through the lattice. The net result is a higher closure temperature and an older apparent cooling age. Conversely, as the amount of accumulated radiation damage increases, the alpha tracks form a connected network thereby providing fast diffusion pathways for He. This results in a lower closure temperature and a younger apparent cooling age. Modeling of these different scenarios has allowed for more detailed interpretation of (U-Th)/He zircon ages. The closure temperatures predicted by the age-eU (effective Uranium) model of Guenthner et al. (2013) are 140–220 °C. At higher alpha doses the closure temperature decreases, resulting in a younger apparent cooling age.

Samples used in this study were obtained by Gaschnig et al. (2010, 2013) and Braudy et al. (2016); supplemental samples were collected at strategic locations relative to the EarthScope IDOR active-source seismic survey (Fig. 2; Davenport et al., 2017). Supplemental samples were processed at the University of Minnesota using standard mineral separation techniques—milling, water, heavy liquid, and magnetic separation. Zircon grains were analyzed and selected for clarity, morphology, and size at University of Arizona (USA) Radiogenic Helium Dating Laboratory using a Leica MZ16 stereo-zoom microscope. Three grains per sample were imaged and measured for alpha-ejection corrections (Farley et al., 1996; Farley, 2002) and packed in Nb tubes. Grains were lased using a CO2 laser for 12 min for the first extract, followed by 15 min re-extractions until the He extracted was <1% of the total He. Gas concentrations were measured using a quadropole mass spectrometer; U, Th, and Sm were analyzed on a high-resolution Element 2 inductively coupled plasma-mass spectrometer (ICP-MS) using methods outlined in Reiners and Nicolescu (2006). Ages are reported in Table 1 and shown in Figures 2 and 3.

COOLING OF THE SOUTHERN IDAHO BATHOLITH: RESULTS

The 46 new (U-Th)/He zircon data presented here illustrate the variability of cooling and hence exhumation recorded by the Idaho batholith as exposed in the vicinity of the IDOR seismic surveys. All ages are reported as the mean and standard deviation of three single-grain ages, unless otherwise noted. We report the ages and standard deviations as strict averages to the nearest ±0.1 m.y., although we acknowledge that the true analytical precision is likely ±1–2 m.y. Ages range from Late Cretaceous to Early Miocene (Table 1; Figs. 2, 3). Projecting all ages onto a latitudinal transect reveals a pattern in the cooling data, with the western and eastern margins of the batholith cooling before the interior of the Atlanta lobe (Fig. 3). The data also show variability in the vicinity of the Sawtooth Valley.

Western Idaho Shear Zone

Samples within the western Idaho shear zone near McCall and Riggins, Idaho, range in age from ca. 76 to 63 Ma. The variability of ages is a function of position relative to the boundary with the Atlanta lobe to the east; ages decrease with increasing proximity to the Atlanta lobe (Fig. 3). Near McCall, ages range from 76.2 ± 3.1 (sample 10RMG009) to 63.2 ± 2.5 Ma (07RMG47). Along strike northward within the western Idaho shear zone near Lava Buttes (Fig. 2), ages are slightly younger ranging from 69.2 ± 2.4 (13AF03) to 63.1 ± 2.7 Ma (13AF04). These results document cooling below ∼200 °C by ca. 63 Ma (Fig. 3) and agree with previously published 40Ar/39Ar ages (Giorgis et al., 2008). Samples collected in accreted terrane rocks exposed near Riggins yield slightly older ages of 73.1 ± 14.6 (13AF06) and 71.1 ± 12.9 Ma (13AF10). However, both ages are consistent with other cooling ages within the western Idaho shear zone. The large standard deviations on these ages reflect the spread in the single-grain ages for these two samples.

The single-grain age-eU patterns for samples from within the western Idaho shear zone show no variation in age as a function of eU concentration (Fig. 4). Single-grain ages range between 60 and 85 Ma with a range of eU concentration from 100 to 1200 ppm (Figs. 4B, 4C). This trend is consistent with rapid cooling through the He retention temperature in zircon (e.g., Guenthner et al., 2013).

Atlanta Lobe

Immediately east of the western Idaho shear zone (east of longitude 116°W), rocks cooled below ∼200 °C ∼7–30 m.y. after those rocks exposed within the western Idaho shear zone. Adjacent to the western Idaho shear zone, the Atlanta lobe yields (U-Th)/He zircon ages of 57.2 ± 1.7 (sample 98IB53), 49.4 ± 3.4 (10RMG44), and 36.8 ± 6.5 Ma (07RMG46). Further east into the Atlanta lobe, ages vary and are as young as 20.2 ± 1.4 Ma (13AF14) in the Sawtooth Range adjacent to Sawtooth Valley. The overall age pattern shows no distinct variation with elevation, but there is a variability with longitude, with younger ages in the center of the transect and older ages toward the margins (Fig. 3). This pattern is also evident in the U-Pb zircon age distribution. The average (U-Th)/He age for the Atlanta lobe is 43 Ma.

The eU concentrations for samples from the Atlanta lobe range from ∼100 to 2400 ppm, yet there is no correlation between eU concentrations and ages obtained (Fig. 4). If eU concentration affected the (U-Th)/He age, younger ages would be expected for samples with higher concentrations. This pattern is not observed. Rather, some samples with higher eU concentrations yield older ages (Fig. 4). Further, we interpret these samples as recording slow cooling in contrast to the rapid cooling observed in the western Idaho shear zone. A rapid cooling event would result in similar ages regardless of eU concentrations.

Sawtooth Valley

Samples collected in the Sawtooth Range and White Cloud Peaks west and east of the Sawtooth Valley, respectively, yield ages ranging from 20.2 ± 1.4 to 68.5 ± 2.1 Ma (Kahn, 2014). Younger ages were obtained from higher elevations in the Sawtooth Range, whereas older ages were obtained from high elevations in the White Cloud Peaks. Samples from this region yield ages with the highest variability.

Ages obtained for the Sawtooth batholith, an Eocene Challis intrusion that cross-cuts the Atlanta lobe exposed in the Sawtooth Range, are 20.2 ± 1.4 (sample 13AF14) and 29.3 ± 7.5 Ma (13AF16). The early metaluminous suite in the Sawtooth Range yields a slightly older age of 33.3 ± 3.1 Ma (13AF13), whereas the same phase of the batholith exposed at high elevations in the White Cloud Peaks yields ages of 68.5 ± 2.1 (13AF23) and 64.2 ± 4.2 Ma (13AF22). Lower-elevation samples include a Challis volcaniclastic rock dated at 44.5 ± 3.7 Ma (13AF21) and a granite related to the Atlanta lobe dated at 48.8 ± 5.7 Ma (Z14BT14).

Certain samples from the Sawtooth Valley have very high eU concentrations, exceeding 6000 ppm (Fig. 4C). With such high concentrations, the expectation is a lower closure temperature and a younger age. However, the data reveal a trend of slight increase in age with increasing eU concentrations (Fig. 4C). Evaluating the single-grain ages for sample 13AF11, the youngest age is obtained from the grain with the lowest eU concentration. Therefore, the young ages obtained are not likely a result of radiation damage, but do record recent cooling below ∼200 °C.

Lemhi Range

The oldest age obtained along the IDOR transect is located beyond the eastern edge of the batholith within the Basin and Range region. A Proterozoic quartzite from the easternmost margin in the Lemhi Range, near Pahsimeroi Valley, yields an age of 87.7 ± 5.4 Ma (Z14BT11). This age indicates that the Proterozoic rocks in the eastern Lemhi range cooled during the Sevier orogeny and were not thermally perturbed by subsequent Atlanta lobe or Challis magmatism.

MODELING OF THERMOCHRONOLOGY DATA

We used HeFTy inverse modeling software (Ketcham, 2005) to determine the best-fit T-t paths for the (U-Th)/He ages obtained. We concentrated on specific locations along boundaries between different structural/magmatic domains or areas that exhibited changes in cooling patterns. The HeFTy program tests T-t paths based on the input of independent constraints. Here we used geochronologic (Giorgis et al., 2008; Gaschnig et al., 2010; Braudy et al., 2016), thermochronologic (Giorgis et al., 2008; Braudy et al., 2016), and petrologic (Dutrow et al., 2014) data as model constraints. Specifically, U-Pb zircon ages and 40Ar/39Ar ages provide first-order age constraints for the HeFTy modeling. For the U-Pb zircon ages, we used a range of ages based on all results from that particular intrusive suite (e.g., border zone suite), as given by Gaschnig et al. (2010). For the 40Ar/39Ar ages, we use a range of ages based on the closure temperatures for both hornblende and biotite from all nearby samples of the relevant intrusive suite (Giorgis et al., 2008; Braudy et al., 2016). We applied the Guenthner et al. (2013) He diffusion model, which accounts for eU concentrations, for each sample modeled. Each model result is based on 10,000 paths. The acceptable and good-fit paths calculated along with the mean and best-fit T-t paths are shown in Figures 5 and 6. Results clearly demonstrate the different thermal histories recorded by different domains within the Idaho batholith.

Western Idaho Shear Zone

We modeled T-t paths for two samples from within the western Idaho shear zone, samples 10RMG011 (Braudy et al., 2016; Byerly et al., 2016) and 13AF03 (this study). The protolith for 10RMG011 is from the suture zone suite (Braudy et al., 2016). The sample is modeled using two T-t constraints (Fig. 5A). The high-temperature constraint of 700–750 °C at a time of 84–93 Ma is based on a 93 Ma U-Pb zircon age from Braudy et al. (2016). A second constraint of 300–400 °C at 72–84 Ma is applied based on an 40Ar/39Ar age on the same sample (Byerly et al., 2016). The resulting model predicts two stages of cooling. The first stage involves moderate to rapid cooling event, from T > 700 °C to T < 150 °C in ∼20 m.y., corresponding to a cooling rate of ∼30 °C/m.y. The second stage is a relatively slow cooling event, from T ∼150 °C to near-surface temperatures, at a rate <<10 °C/m.y. (Fig. 5A).

Sample 13AF03 is located near Lava Buttes within the western Idaho shear zone. The easternmost part of the Hazard Creek intrusive suite is affected by the western Idaho shear zone and is presumed to have intruded through accreted terrane material (Manduca et al., 1993; Giorgis et al., 2008). This sample is modeled with constraints from Giorgis et al. (2008), as follows: T = 700–750 °C at 90–110 Ma (U-Pb age), T = 300–500 °C at 70–80 Ma (40Ar/39Ar hornblende and biotite ages), and T = 50–120 °C at 40–50 Ma (apatite fission-track ages) (Fig. 5B). The resulting paths are similar to those calculated for 10RMG011 in that two stages of cooling are consistent with the observed (U-Th)/He zircon ages (Fig. 5B). Differences are apparent in the higher-temperature part of the cooling path because the Hazard Creek suite here has an intrusive age ∼15 m.y. older than the border zone suite to the south. The two samples, however, cool below 300 °C at approximately the same time (ca. 80 Ma) and therefore share a common T-t history from moderate to near-surface temperatures (Figs. 5A, 5B). Regardless of location along strike, the western Idaho shear zone rocks record two stages of cooling and were below 150 °C by 70 Ma.

Atlanta Lobe

Byerly et al. (2016) reported two 40Ar/39Ar biotite ages for samples within the Atlanta lobe. Combining these data with previously published U-Pb zircon ages (Gaschnig et al., 2010), we modeled the cooling paths for the Atlanta lobe. The Cape Horn summit sample (07RMG56) constraints are T = 700–750 °C between 70 and 80 Ma and T = 300–400 °C at 40–50 Ma, based on a U-Pb zircon age of 71.9 ± 2.7 Ma (Gaschnig et al., 2010) and a 40Ar/39Ar biotite age of 46.9 ± 0.2 Ma (Byerly et al., 2016), respectively. We modeled this sample using three different scenarios. In the first case, the model determines best-fit T-t paths based on simple cooling from crystallization temperatures through the Ar closure temperature in biotite (Fig. 5C). Because of the sample’s close proximity to Challis-age rocks, we test two further scenarios in which heating from Challis magmatism is introduced (Figs. 5D, 5E). Each model predicts acceptable T-t paths given the (U-Th)/He single-grain ages. Regardless of constraints used, this sample cooled below 200 °C by 40 Ma (Figs. 5D, 5E).

The second sample from the Atlanta lobe with an age of 39.7 ± 5.3 Ma (10RL896) is located near the Boise Basin dike swarm at the southeast corner of Deadwood (Fig. 2). The T-t constraints are T = 700–750 °C between 70 and 80 Ma and T = 300–400 °C at 65–75 Ma, based on a U-Pb zircon age of 71.9 ± 2.7 Ma (Gaschnig et al., 2010) and a 40Ar/39Ar biotite age of 69.0 ± 0.2 Ma, respectively (Byerly et al., 2016). The best-fit T-t paths represent rapid cooling through the Ar closure temperature for biotite, followed by slow cooling from ca. 65 Ma to present, passing through the He closure temperature in zircon at ca. 35 Ma (Fig. 5F).

Sawtooth Valley

Samples from the Sawtooth Valley area yield the largest differences in ages obtained in the study, varying from ca. 20 to 68 Ma. We model samples from the early metaluminous suite as exposed in the Sawtooth Range (sample 13AF13) and White Cloud Peaks (13AF23), as well as the Eocene Sawtooth batholith (13AF16). Sample 13AF13 is from a high elevation in the Sawtooth Range near the contact of the Idaho batholith with the Sawtooth batholith. This sample contains hornblende and is therefore likely part of the early metaluminous suite. We present three models for this sample, each with different constraints to assess which thermal history best explains the observed (U-Th)/He ages (Figs. 5G–5I). In all three models, the high-temperature constraint is the same: T = 700–750 °C at 80–100 Ma, based on the U-Pb age for the early metaluminous suite (Gaschnig et al., 2011). In the first scenario, a moderate- to low-temperature constraint similar to the thermal conditions of the suture zone suite within the western Idaho shear zone is added (Fig. 5G). In the second model, we assume the sample remained at elevated temperatures prior to the intrusion of the Sawtooth batholith. The T-t conditions for the Sawtooth intrusive event are T = 600–700 °C at 40–50 Ma (Fig. 5H; Dutrow et al., 2014). In the third scenario, we assume the sample cooled to <200 °C and was subsequently reheated by the Sawtooth batholith intrusion (Fig. 5I). Each scenario can explain the observed (U-Th)/He single-grain ages for sample 13AF13. Regardless of model conditions, the early metaluminous suite as exposed in the Sawtooth Range did not cool below 200 °C before ca. 30 Ma.

The White Cloud Peaks sample 13AF23 yields a (U-Th)/He age of 68.5 ± 2.1 Ma (Figs. 2, 3), which is the oldest age obtained east of the western Idaho shear zone. Using the range of ages for the early metaluminous suite given by Gaschnig et al. (2010), we constrain the high-temperature range of the T-t model at 80–100 Ma. We further constrain the model at T = 300–400 °C at 75–85 Ma based on a published 40Ar/39Ar age from a quartz monzonite exposed in the White Cloud Peaks (Taylor et al., 2007). Results indicate the sample likely cooled from T > 700 °C to T < 150 °C along a monotonic cooling path defined by a cooling rate of ∼25 °C/m.y., similar to the border zone suite within the western Idaho shear zone (Fig. 5K). These rocks were likely at or near the Earth’s surface prior to Challis magmatism.

DISCUSSION

Exhumation Patterns in the Western Idaho Shear Zone and Atlanta Lobe

Cooling in the Western Idaho Shear Zone

Within the western Idaho shear zone, (U-Th)/He zircon ages clearly decrease from west to east with no correlation to elevation (Figs. 6, 7, 8): samples near the western margin of the western Idaho shear zone yield (U-Th)/He zircon ages of ca. 76 Ma, whereas sample ages in the east are ca. 64 Ma. We note that the spatial pattern observed in the new data mimics that of both the U-Pb zircon ages (120–90 Ma; Manduca et al., 1993; Giorgis et al., 2008) and Ar-Ar hornblende and biotite ages (110–70 Ma; Snee et al., 1995; Giorgis et al., 2008) from the McCall region. It does not, however, reflect the apatite fission-track ages within the western Idaho shear zone, which show little spatial variation in cooling through ∼120 °C (Giorgis et al., 2008).

We interpret this record to result from a combination of west-to-east migration of magmatism and/or deformation within the western Idaho shear zone. The U-Pb zircon data indicate that there is sequential west-to-east emplacement of igneous bodies, with the eastern edge of the western Idaho shear zone intruded by the syn-tectonic Payette River tonalite (ca. 91 Ma; Giorgis et al., 2008). Braudy et al. (2016) documented the existence of syn-tectonic intrusions in West Mountain, south of McCall, ID (Fig. 1), indicating a relation between deformation and magmatism. Giorgis et al. (2008) suggested that deformation in the western Idaho shear zone ceased at ca. 90 Ma, constrained by the presence of an undeformed, cross-cutting dike in the Little Goose Creek complex. Because of the complex interplay between magmatism and deformation in the western Idaho shear zone, we cannot uniquely attribute our cooling ages to either effect.

Cooling in the Atlanta Lobe of the Idaho Batholith

Within the Atlanta lobe, samples show a diffuse pattern of increasing age with elevation, consistent with cooling and erosion (Fig. 8). It should be noted that all Atlanta lobe samples from the study area are projected onto a single transect line. It is therefore likely that other trends are present within this population, but—given the projection—all trends are not visible. Nonetheless, the presence of two age populations is observed: (1) samples from elevations between 1900 and 1000 m; and (2) samples from elevations >1900 m.

The first population of ages consists of the majority of samples from within the Atlanta lobe, located at elevations between 1900 and 1000 m. Within this population, samples from higher elevations cooled through the He retention temperature in zircon at ca. 60 Ma. In contrast, samples at lower elevations cooled by 40 Ma, ∼20 m.y. after samples at higher elevations. The second population of ages, from elevations >1900 m, range in age from 60 to 40 Ma and come exclusively from north of Yellow Pine, Idaho. These samples have young ages relative to the overall age-elevation trend for the majority of Atlanta lobe samples. There are two possible reasons that these ages might be systematically younger in age. First, the samples are adjacent to the Johnson Creek–Profile Gap shear zone (Lund, 2004). Second, there is a relatively large set of Challis-aged intrusions near the sample localities. We hypothesize that the discrepancy results from the presence of the nearby Challis intrusions, which likely caused thermal resetting of the samples starting at ca. 50 Ma.

The age-elevation correlation (Fig. 8) is based on ages from exposures in the interior of the Atlanta lobe except: (1) samples 98IB53 and 07RMG46; and (2) samples within the Sawtooth Range area. The latter samples are excluded from discussion of the cooling history from the main Atlanta lobe because they are affected by exhumation of the Sawtooth Range, as explained below. The first two samples (98IB53 and 07RMG46) are located in the west-southwest part of the Atlanta Lobe, adjacent to the border zone suites of Gaschnig et al. (2010). These samples occur at lower elevations and reveal older ages than the Atlanta lobe, but are similar in cooling age to the adjacent border zone suites. We interpret these samples to have a similar cooling pattern to the border zone suites because they lie on the “edge” of a crustal plateau, as discussed in the final section of the discussion.

Cooling in the Western Idaho Shear Zone versus Atlanta Lobe

A first-order observation of our study is that there is a major difference between cooling in the western Idaho shear zone relative to the Atlanta lobe. The western Idaho shear zone appears to have cooled through 200 °C during emplacement of the voluminous Atlanta lobe to the east. This lack of thermal resetting is also consistent with the absence of Atlanta lobe intrusions anywhere along the extent of the western Idaho shear zone. As such, the western Idaho shear zone seems to have acted both as a major thermal and structural barrier to subsequent magmatism.

Cooling in the Sawtooth Valley Area: A Sawtooth Metamorphic Core Complex?

Samples from the Sawtooth Valley area of the Idaho batholith record a unique history of exhumation within the largely homogeneously cooled Atlanta lobe of the Idaho batholith. The north-northwest–oriented Sawtooth Valley separates the Sawtooth Range on its west side from the White Cloud Peaks on its east side (Fig. 9). Given the consistent emplacement ages of Atlanta lobe rocks in both mountain ranges, the variation in cooling history requires interpretation of structural features across the valley. The present geometry of Sawtooth Valley is a graben defined by range-bounding faults. The Sawtooth Range is bounded on the east side by the active, east-dipping Sawtooth fault (Thackray et al., 2013). This fault is presently exposed on the west side of Sawtooth Valley. In the southeast corner of Sawtooth Valley, previous mapping indicates the presence of the west-dipping, normal Obsidian fault (Witkind, 1975; Fisher et al., 1992). We suggest that the Obsidian fault runs along the entire east side of Sawtooth Valley, although it is likely inactive and its location is cryptic on the northeast side of the valley.

(U-Th)/He zircon ages from the early metaluminous suite in the White Cloud Peaks indicate cooling below 200 °C significantly earlier than the same suite exposed in the Sawtooth Range. The oldest ages in the Sawtooth Range is 40.5 ± 2.0 Ma, while the youngest is 20.2 ± 1.4 Ma at low elevations in the Sawtooth Range (Table 1). In contrast, (U-Th)/He zircon ages from Atlanta lobe in the White Cloud Peaks indicate ages of ca. 50 Ma. In particular, the (U-Th)/He zircon ages from the base of the White Cloud Peaks (in Sawtooth Valley) are ∼10 m.y. older than the sample from the top of the Sawtooth Range.

There is also a different relation to Challis-aged magmatism between the two ranges. The Sawtooth Range contains the intrusive Sawtooth batholith, which is Challis aged (ca. 50–45 Ma) and intruded at a depth of ∼5 km (Gaschnig et al., 2010; Dutrow et al., 2014). In contrast, the White Cloud Peaks contain dikes and volcaniclastic rocks of Challis age. Thus, portions of the White Cloud Peaks were at or near the surface at ca. 50 Ma, while the Sawtooth Range was at least 5 km deep.

Two tectonic models can explain the (U-Th)/He zircon data from the Sawtooth Valley region: (1) an asymmetric graben (Fig. 9A); and (2) a low-angle normal fault followed by an asymmetric graben (Fig. 9B). In the first tectonic model, Sawtooth Valley lies in an asymmetric graben with both the Sawtooth Range and White Cloud Peaks exhumed along oppositely dipping, high-angle normal faults. Within the Sawtooth Range, younger (U-Th)/He zircon ages indicate cooling and exhumation along a normal fault active in the Neogene. Assuming a steep geothermal gradient of 50 °C/km, the depth to the 200 °C isotherm would have been at 4 km prior to exhumation. A simple geometric analysis shows that exhumation along a high-angle normal fault would require ∼1.6 km of throw to explain the cooling age variability. This magnitude of slip has occurred on other normal faults adjacent to the Idaho batholith (Long Valley fault; e.g., Giorgis et al., 2006). A more typical geothermal gradient of 25 °C/km would require ∼3.2 km of fault throw to explain the cooling age variability. The current relief between our sample in the Sawtooth Range and the valley floor is 600 m, which is a minimum slip amount. Recent scarps indicate active fault activity along the Sawtooth fault, suggesting that normal faulting is a viable exhumation mechanism (e.g., Thackray et al., 2013).

Another possible model is exhumation of the Sawtooth Range along a low-angle normal fault (Fig. 9B). In this model, the White Cloud Peaks restore to a position structurally overlying the Sawtooth Range. This model satisfies the (U-Th)/He zircon ages, as the White Cloud Peaks cooled first because they were higher in the crustal column. In addition, the petrological record suggest that in the Eocene the Sawtooth Range was at depth while the White Clouds Peaks were at or near the surface. The 48 Ma Sawtooth batholith in the Sawtooth Range intruded at ∼5 km depth (Gaschnig et al., 2010; Dutrow et al., 2014), while Challis-aged volcaniclastic rocks were deposited at the Earth’s surface in the White Cloud Peaks. In this scenario, the Sawtooth Range cools only after low-angle normal faulting removes the overlying White Cloud Peaks (Fig. 9A). Then, at a later time, graben formation occurs with high-angle normal faults on both sides of the Sawtooth Valley (e.g., Thackray et al., 2013). The low-angle fault is removed by erosion from on top of the Sawtooth Range and buried by a high-angle normal fault below Sawtooth Valley and the White Cloud Peaks. This scenario requires significantly less throw on the high-angle faults in Sawtooth Valley. Movement on the low-angle fault must have occurred during or slightly after intrusion of the ca. 48 Ma Sawtooth batholith, consistent with studies of the nearby Pioneer core complex (Vogl et al., 2012).

If this second model is correct, then the Sawtooth Range represents an unrecognized metamorphic core complex in the U.S. Cordillera. The Bitteroot metamorphic core complex occurs further north (Foster et al., 2001), on the eastern edge of the Bitteroot lobe, in a structurally equivalent location. The uplift of deep rocks after Challis magmatism (e.g., Dutrow et al., 2014) and the presence of Eocene magmatism at the locus of young uplift ages are both consistent with a core-complex model.

Given our data and the lack of corroborating structures, we cannot resolve between these two models. However, any possible model must address the observed cooling age pattern that indicates that the Atlanta lobe rocks currently exposed in the Sawtooth Range were at elevated temperatures after the White Cloud Peaks were exhumed to shallow crustal levels. As such, we favor the low-angle fault model.

The Existence of a Plateau: The Atlanta Lobe as the Northern Extension of the Nevadaplano

The cooling history presented here is consistent with the formation of a Late Cretaceous–Paleogene plateau in Idaho. The cooling data alone indicate that the region was thermally stable for ∼40 m.y. The self-similar pattern of U-Pb and (U-Th)/He zircon ages from the Atlanta lobe suggests that the region was at a thermal steady state prior to initiation of the Challis magmatic event in early Eocene time and that cooling occurred isobarically across the entire crustal region (Figs. 7, 8). The general correlation of cooling age and elevation supports this contention. Further, it appears that this post-magmatic isobaric cooling occurred with little or no unroofing, except in the Sawtooth Valley region. The only other variations in cooling path that are younger than Early Eocene occur as a result of Challis magmatism.

We hypothesize that the Atlanta lobe was emplaced into a crustal plateau based on the cooling history described above and: (1) geochemistry of the Atlanta lobe; (2) the lack of strong fabrics within the Atlanta peraluminous suite; (3) crustal thickness determined by seismic data; and (4) the presence of detrital zircons derived from the Idaho batholith in basins throughout the North American Cordillera. The Atlanta lobe is an extensive belt of two-mica granites, which geochemical analysis indicates formed exclusively as a result of crustal melting (Gaschnig et al., 2011). Thus, the most likely timing for the formation of the crustal plateau is during the intrusion of the Atlanta peraluminous suite. The lack of any consistently oriented tectonic fabrics within the Atlanta lobe intrusions, which were emplaced during regional contraction in the Sevier orogenic belt, led Byerly et al. (2016) to conclude that intrusion must have occurred in a “neutral” to extensional tectonic setting. They concluded that a crustal plateau was the most likely tectonic environment for Atlanta lobe emplacement. Davenport et al. (2017) indicated that the base of the crust below the Idaho batholith is located at about ∼43 km below current exposure levels. Accounting for ∼12 km of erosion (e.g., Jordan, 1994), this estimate suggests a crustal thickness of ∼55 km in central Idaho during Atlanta lobe emplacement. This estimate of crustal thickening is consistent with the palinspastic reconstruction of the Sevier orogeny hinterland (Coney and Harms, 1984).

Basin evolution, however, is also critical for interpreting the Idaho batholith as a crustal plateau. The Idaho batholith is the source of major detrital material throughout the North American Cordillera, including the Franciscan Basin of California (e.g., Dumitru et al., 2013), the Green River Basin of Wyoming (e.g., Chetel et al., 2011), and the Tyee forearc in coastal Oregon (e.g., Heller et al., 1985). Dumitru et al. (2016) also noted that Proterozoic detritus linked to unique sources in Idaho was found in these basins as well as other basins in Washington (Swakane) and Alaska (Yakutat). These data, taken together, suggest that the early metaluminous suite and Atlanta lobe were at the Earth’s surface at least 20 m.y. before the current exposures in the Idaho batholith cooled below ∼200 °C. This pattern is consistent with continuous, slow cooling accompanied by slow exhumation of the highland. For example, in their study of detrital zircons from the Franciscan Basin, Dumitru et al. (2013, 2015) identified a clear Idaho batholith signal with zircon ages ranging from 85 to 60 Ma. The ages of these zircons uniquely tie them to the peraluminous Atlanta lobe, although deposition occurred while the current exposures of the Atlanta lobe were at T > 200 °C. If we are correct about the geothermal gradients (a maximum of 50 °C/km), the rocks currently exposed at the surface were at depths >4 km.

If Dumitru et al. (2013, 2015) have correctly identified the source of the sediment in a variety of Paleogene basins as the Idaho batholith, it suggests that the Idaho batholith was topographically high and provided abundant source material for basin deposition. As such, a minimum of 4 km of erosion of the top surface of the batholith must have occurred, and likely significantly more. Given the lack of focused tectonic unroofing during the Paleogene, it suggests that the entire region must have been a high crustal plateau. Finally, in the last 40 m.y., the Atlanta lobe was exhumed an additional 4 km, indicating an average exhumation rate of 1 mm/yr or less. In aggregate, all of the above data support the existence of a crustal plateau that ultimately resulted in the formation of the Atlanta peraluminous suite.

CONCLUSIONS

In general, cooling ages are older along the margins and younger in the interior of the Idaho batholith, consistent with results from previous thermochronologic studies (Criss et al., 1984; Sweetkind and Blackwell, 1989). (U-Th)/He zircon ages and HeFTy modeling of the data, however, illustrate the variability of cooling of the Idaho batholith as a function of position. First, exhumation in the western Idaho shear zone differs significantly from that in the Atlanta lobe of the Idaho batholith. The western Idaho shear zone exhibits no correlation between (U-Th)/He zircon age and present elevation. Rather, there is a west-to-east younging observed in the data, reflecting spatial migration in magmatism. Second, samples from within the Atlanta lobe of the Idaho batholith exhibit a clear pattern of cooling age as a function of elevation. This pattern applies to the majority of the batholith and suggests isobaric cooling in the absence of tectonic activity. We suggest that the distribution of cooling ages is consistent with a crustal plateau that corresponds to the extent of the Atlanta lobe. Third, a discrete break in cooling histories occurs across the Sawtooth Valley, with 30–20 Ma (U-Th)/He zircon ages recorded in the Sawtooth Range and 50–40 Ma (U-Th)/He zircon ages in the White Cloud Peaks. We attribute this difference to be the result of focused tectonic extension and speculate the presence of a core complex and associated low-angle detachment below this section of the Idaho batholith.

The (U-Th)/He zircon thermochronometry was conducted in the University of Arizona Radiogenic Helium Dating Laboratory, led by Dr. P. Reiners. We gratefully acknowledge Erin Abel and Uttam Chowdhury for help in the laboratory. Reviews by J. Vogl, an anonymous reviewer, and the editor were extremely helpful in improving the presentation of the data and interpretations. Funding for the (U-Th)/He zircon thermochronometry was made available through GeoEarthScope. Rock collection, sample preparation, and manuscript preparation was supported by National Science Foundation grants EAR-0844260 and EAR-1251877 to Tikoff.