A reassessment of Mojavia and a new Cheyenne Belt alignment in the eastern Great Basin

A number of studies have proposed the location and character of isotopic, and by inference, crustal provinces in the western U.S. This paper focuses on one of these provinces and its boundaries, a major terrane, “Mojavia” (e.g., Bennett and DePaolo, 1987) or the “Mojave province” (e.g., Wooden et al., 1988) that has been suggested to underlie much of the eastern and southern Great Basin (Fig. 1).

Mojavia, the criteria upon which it may (or may not) be discriminated, and the manner in which it persists in the literature comprise the subject of this study. Our recent work, combined with a survey and reinterpretation of existing data, question the basis upon which Mojavia has been defined and whether it exists at all outside of the Mojave Desert region. Of particular interest is the extension of a relatively small region of “outcrop” Mojavia in the Mojave Desert (e.g., Wooden et al., 1988) to “covered” Mojavia in the larger Great Basin.

This is an important topic for three reasons. First, although North America is arguably the best-studied continent on Earth in terms of crustal structure, and detailed studies of its assembly have been published (e.g., Whitmeyer and Karlstrom, 2007), this study shows that considerable uncertainty remains regarding some of its most fundamental crustal elements. Second, this study shows that such terranes and their boundaries can become entrenched in the literature once proposed. The problem is not with the proposal of such boundaries, but the failure to fully reevaluate them as new information becomes available. Finally, some quantitative approaches are presented that may aid in the recognition of crustal terranes and associated boundaries, especially where basement exposures are limited or absent as is the case in the eastern Great Basin.

Bennett and DePaolo (1987) defined Mojavia on the basis of Nd isotope systematics in crystalline rocks, whereas Wooden et al. (1988) defined it on the basis of common Pb isotopes. It is variably conceived as comprising 2.0–2.3 Ga basement rock (Wooden and Miller, 1990), or ∼1.7 Ga basement that includes components of isotopically evolved and older material (Bennett and DePaolo, 1987; Wooden et al., 1988) that is lacking in Yavapai crust.

Wooden and Miller (1990), for example, described the presence of 2.0–2.3 Ga euhedral zircons in a silicic metamorphic rock interpreted as metarhyolite. The implications are clear: if the zircon ages represent the time of emplacement of a volcanic rock, then this body must have been erupted onto basement of equal or greater antiquity.

Van Schmus et al. (1993) provided a broader view of Mojavia, which is considered to have a “different character” from the Yavapai province, or that they are “distinct isotopic provinces” involved in a single orogenic event at ∼1.7 Ga. While acknowledging the presence of 2.0–2.3 Ga zircons, they further note that Pb and Nd isotope data do not require the presence of an “Archean component.” In essence, there are older zircons in basement rocks of the Mojave Desert accompanied by older Nd model ages and high ²⁰⁷Pb/²⁰⁶Pb ratios indicative of the involvement of material >1.7 Ga.

Whitmeyer and Karlstrom (2007) provided a complicated view of Mojavia. In their figures, Mojavia is limited to southeastern California and southern Nevada and is mapped as Archean. Those areas of the eastern Great Basin attributed to Mojavia by Karlstrom et al. (2002) and Bennett and DePaolo (1987) (Fig. 1) are mapped as 2.0–1.8 Ga juvenile arcs and 1.9–1.8 Ga reworked Archean crust. In their text they describe Mojavia as ∼1.7 Ga gneisses containing isotopic evidence for older material and offer a number of scenarios for how it may have been incorporated into the basement.

For the purposes of this study, we retain the broad view of Mojavia expressed by Van Schmus et al. (1993) for our evaluation of the broader province in the eastern Great Basin. This is true whether the boundary considered is that of Bennett and DePaolo (1987) or Karlstrom et al. (2002). This is necessary because exposures of basement in the eastern Great Basin are extremely limited (Fig. 1).

Yavapai crust, by definition, should lack >2.0 Ga material within its juvenile basement (Van Schmus et al., 1993; Karlstrom and Humphreys, 1998). Otherwise, there may be little basis for differentiating it from Mojavia. However, older material was shed from the Archean Wyoming craton and perhaps other unknown sources onto Paleoproterozoic terranes that were metamorphosed as they docked to North America at ∼1.7 Ga. Thus, an older isotopic signature easily could be imparted to the isotope systematics of Phanerozoic igneous rocks that are potential probes in defining crustal boundaries. In this case, assimilation of ∼1.7 Ga juvenile crust and Archean clastic materials may yield mean apparent model ages of 2.0–2.3 Ga. In such instances there is no reason to classify the basement as belonging to Mojavia.

In summary, in order for crust in the eastern Great Basin to be excluded from the Yavapai province, geochronological analysis of juvenile basement should contain direct evidence of 2.0–2.3 Ga or older components. Because Mojavia was originally defined on the basis of 2.0–2.3 Ga model ages, its basement should contain a large fraction of material of this age or older. From this perspective, and given sufficient data, inherited zircons >2.0 Ga should be a diagnostic feature of Mojavia, whereas Yavapai crust should lack them.

Where basement rocks cannot be directly observed and their ages determined, the Sr, Nd, and/or Pb isotope systematics of igneous rocks emplaced through the crust should strongly infer interaction with crust of an average age of 2.0–2.3 Ga. If contamination by supracrustal rocks or metamorphosed accretionary wedges is at least as likely an explanation of the attendant observations as the involvement of juvenile basement, then classification of the crust as belonging to the Yavapai province is a simpler explanation.

The western boundary of Mojavia is commonly defined as the initial ⁸⁷Sr/⁸⁶Sr_i 0.706 line (Kistler and Peterman, 1973, 1978), which is thought to represent the western limit of Precambrian basement in the western U.S. However, the exact location of the western boundary of Mojavia is not particularly important to this study because we confine our analysis to its eastern and northern boundaries in the eastern Great Basin.

The eastern and northern boundaries are difficult to locate precisely due to sparse outcrop of basement (Fig. 1). In those few locations that occur, basement beneath thick supracrustal cover has been tectonically exhumed via incorporation into the upper plate of Sevier thrust faults or in the footwall of large normal faults, or both. Such exposures typically are small and rare.

The Santaquin and Farmington Canyon complexes are allochthonous, having been translated tens of kilometers in the upper plate of east-verging Sevier thrust sheets (e.g., Constenius et al., 2004; Horton et al., 2004; Yonkee and Weil, 2010) and subsequently relaxed and segmented by east-west orogenic collapse (Constenius, 1996) and Basin and Range extension. However, these translations are parallel to the crustal boundaries we examine in this paper; thus past motions of these bodies have relatively little effect on our interpretations.

Because basement outcrop in the eastern Great Basin is sparse, the historical placement of crustal boundaries has always been inherently speculative. The northern boundary of Mojavia is frequently referred to as the Cheyenne Belt, separating the Archean Wyoming province to the north from Paleoproterozoic basement to the south (Karlstrom and Houston, 1984). It could be argued that the term Cheyenne Belt should strictly apply to exposures of the suture in southern Wyoming, although we retain the use of this term in the Great Basin as a matter of convenience.

The easternmost Cheyenne Belt alignment is well established (Fig. 1) in exposures in the Sierra Madre and Medicine Bow Mountains of southeastern Wyoming (e.g., Karlstrom and Houston, 1984). Westward, this boundary has traditionally been placed through the Owiyukuts complex–Red Creek Quartzite (ORCQ) of northeastern Utah (Sears et al., 1982). The Cheyenne Belt is usually drawn along or near the axis of the Uinta Mountains, presumably based on an inference that the structural grain of this range reflects a deep-seated crustal boundary. Farthest west, Wright and Snoke (1993) placed the Cheyenne Belt in the Ruby Range of Nevada, south of a reported exposure of Archean basement (Angels Lake orthogneiss; Lush et al., 1988). The Ruby Range is an allochthonous mass with eastward vergence (Hudec, 1992), so these exposures are not in place.

The southern boundary separates Mojavia from the Yavapai province. The existence of a boundary between Mojavia and the Yavapai province is well established in the Mojave Desert region (Wooden and Miller, 1990; Dueben dorfer et al., 2001, 2006), although its exact location and orientation has evolved over time (e.g., Bryant et al., 2001). However, this region is far from the eastern Great Basin of western Utah and eastern Nevada. As shown in Figure 1, there are contrasting placements for this boundary. Karlstrom et al. (2002) greatly expand the area in the eastern Great Basin and Colorado Plateau underlain by Mojavia compared to the placement of Bennett and DePaolo (1987). This expansion may be justified as the Elves Chasm and Tuna Creek plutons of the Grand Canyon contain older zircons and exhibit Pb isotope characteristics similar to Mojavia (Hawkins et al., 1996).

In critically reevaluating the existence of a greater Mojavia in the eastern Great Basin, this study aims to satisfy several goals. First, we pre sent new geochronological results in parallel with a compilation of existing ages of basement rocks within the eastern Great Basin and surrounding regions, attempting to constrain the extent of Mojavia directly. Second, we have compiled and presented new Sr, Nd, and Pb isotope data that bear on Mojavia and its borders. A subset of these data has been subjected to statistical scrutiny to evaluate the location of historically proposed boundaries. We also consider whether it is more likely that a Mojavian signature is a characteristic of eastern Great Basin juvenile basement or has been inherited from supracrustal sequences. Finally, we propose a refined Cheyenne Belt alignment between Paleoproterozoic and Archean crust in the northern Great Basin.

New U/Pb ages for key basement exposures have been obtained for this study and are augmented with existing data compiled from the literature. Zircons were separated from their matrices by standard heavy-mineral separation procedures. Laser-ablation inductively coupled plasma–mass spectrometry U/Pb analytical methods were conducted at Washington State University, chiefly on the cores of zircon crystals following the methods of Chang et al. (2006) for three samples from the Beaver Dam Mountains and one sample from the Owiyukuts complex. U/Pb data for zircons from a biotite orthogneiss of the Farmington Canyon complex at the south end of Antelope Island was obtained by sensitive high-resolution ion microprobe–reverse geometry (SHRIMP-RG) at Australian National University, Canberra, Australia, following procedures described in Williams (1998) and references therein. Data were evaluated with the assistance of ISOPLOT (Ludwig, 1999).

We also report the results of two new ⁴⁰Ar/³⁹Ar ages on hornblende in amphibolite. Hornblende separates were step heated in a Mo-resistance furnace and gases analyzed with a Mass Analyzer Products 215-50 mass spectrometer at New Mexico Institute of Mining and Technology. Irradiation was monitored with Fish Canyon sanidine with an assigned age of 28.02 Ma (Renne et al., 1998).

New major element and trace element data were collected for samples from the Farmington Canyon complex from the footwall of the northern Wasatch Range and Antelope Island. Samples include both metaigneous rocks and pelitic gneisses from Antelope Island, south Ogden, and Farmington Canyon. Major, trace, and rare-earth element (REE) analyses were performed at XRAL (now SGS Minerals Services), Don Mills, Ontario, Canada. Major elements and Rb, Sr, Zr, Y, Nb, and Ba were analyzed by X-ray fluorescence (XRF). Rare-earth elements were determined by inductively coupled plasma–mass spectrometry (ICP-MS). Detection limits for major elements are 2 ppm for Rb, Sr, Zr, Y, Nb, and Ba. Detection limits for REE analyzed by ICP-MS are 0.1 ppm except for Eu, Ho, and Lu, which are 0.05 ppm.

New Nd isotopic data are reported for the samples from Farmington Canyon as described above. Approximately 80–100 mg of bulk rock sample was dissolved in HF-HNO₃. After conversion to chlorides, one-third of the sample was spiked with ¹⁴⁹Sm and ¹⁴⁶Nd. Rare-earth elements were isolated by conventional cation-exchange procedures followed by di-ethyl-hexyl orthophosphoric acid separation. All isotopic measurements were made on a VG Sector multi collector mass spectrometer at the University of Wyoming. An average ¹⁴³Nd/¹⁴⁴Nd ratio of 0.511846 ± 11 (2σ) was measured for the LaJolla Nd standard. Uncertainties in Nd isotopic ratio measurements are ±0.00001 (2σ). Blanks are <50 pg for Sm and Nd, and no blank correction was made. Uncertainties in Nd and Sm concentrations are ±2% of the measured value, and uncertainties on initial ε_Nd = ±0.5 epsilon units.

For consistency, Nd isotope, depleted-mantle model ages (T_DMs) for new and compiled data have been calculated or recalculated for a depleted-mantle reservoir (¹⁴⁷Sm/¹⁴⁴Nd ratio of 0.222, ¹⁴³Nd/¹⁴⁴Nd 0.513114; Michard et al., 1985). Secondary Pb isochron ages have been calculated by simple linear regression and are reported only in cases where correlation is good R² ≥0.88. New data reported in this study can be obtained in the Supplemental Data File¹.

Much of this study relies on the analysis of a compilation of published Sr and Pb isotope data for silicic igneous rocks of Phanerozoic age. Differences in the distributions of iso topic values from various proposed provinces were evaluated by two statistical methods. The underlying purpose of these tests was to estimate the probability that data representing two candidate provinces, when compared, may come from the same crustal age province, where a P-value of 1 indicates that they may. First, we used a student's T-test (two tail, unequal variance) in Microsoft Excel. However, the T-test assumes that the data are normally distributed, and since this cannot always be assumed, the Mann-Whitney U-test, was executed on line at http://elegans.swmed.edu/∼leon/stats/utest.cgi. Unlike the T-test, the U-test makes no assumption regarding the underlying statistical distribution of data. It is useful to treat the data in this fashion as Sr and Pb isotope ratios are dependent variables that are controlled by a host of independent factors such as crustal age and heterogeneity, degree of crustal contamination of magmas, parent/daughter ratios, etc.

Evaluation of the spatial patterns for all data, regardless of source, was facilitated by the use of ArcGIS. Shape files for the geology of western states were obtained from Ludington et al. (2005) and Stoeser et al. (2005).

As mentioned in the Introduction, in addition to new data, this study relies on a large compilation of published information. In the following presentation of these data, new and compiled information is described together. The reader should note, however, that we have taken care to clearly indicate data that are new.

Figure 1 shows the location of ages compiled for the Grand Canyon, Beaver Dam Mountains, Mineral Mountains, Santaquin complex, Owiyukuts complex–Red Creek Quartzite, Farmington Canyon complex, and Raft River–Grouse Creek–Albion Ranges, as well as crustal xenoliths from eastern Snake River Plain basalts and the Leucite Hills. The distribution of age data compiled for these locations is shown on Figure 2 along with results for new samples discussed below and illustrated in Figures 3–7.

Basement rocks in the inner gorge of the Grand Canyon generally range in age from 1680 to 1750 Ma, with two important exceptions. The ∼1450 Ma Quartermaster Pluton is clearly related to an episode of Mesoproterozoic regional magmatism and plutonism that affected much of the western U.S. (e.g., Karlstrom and Humphreys, 1998; Ferguson et al., 2004; Van Schmus et al., 1993), whereas the 1840 Ma Elves Chasm Pluton has been interpreted to be “basement” to other rocks of the rest of the inner gorge (Karlstrom et al., 2001).

The previously undated Beaver Dam Mountains include a relatively large exposure of basement rocks in extreme southwestern Utah (Fig. 1). New sample WMP-11306-1 is an orthogneiss with a weighted mean ²⁰⁷Pb/²⁰⁶Pb age of 1733.3 ± 4.8 (2σ standard error) Ma after filtering the data to only include samples with <10% discordance (68 of 102 analyses, Fig. 3). Overall, two of 68 total individual analyses have ²⁰⁷Pb/²⁰⁶Pb ages of 1943 and 1902 Ma, respectively. The rest are <1800 Ma.

New sample CC-11306-2 is a garnet para gneiss with a mean ²⁰⁷Pb/²⁰⁶Pb age of 1744 ± 3 (2σ standard error) Ma after application of the discordance filter (60 of 71 analyses, Fig. 4). Of 60 filtered analyses, two have ²⁰⁷Pb/²⁰⁶Pb ages >1800 Ma, and one is between 1880 and 1900 Ma.

New sample WMP-11306-3 is a leucogranite that exhibits little to no foliation. In general, the zircons in this rock are highly discordant (Fig. 5). Only 15 of 130 analyses are <10% discordant. However, these 15 grains produce a mean age of 1720 ± 3 (2σ standard error) Ma, which agrees reasonably well with the upper intercept age of 1709 ± 12 (2σ standard error) Ma shown in Figure 5. Assuming that the elimination of discordant analyses is an appropriate filter, peak metamorphism appears to have predated ∼1710 Ma in this area.

A few very small basement exposures are found along the central portion of the western flank of the Mineral Mountains (Fig. 1). Aleinikoff et al. (1986) reported a range of ²⁰⁷Pb/²⁰⁶Pb ages from 1708 to 1723 Ma for one paragneiss and one orthogneiss. Each rock was analyzed for two size fractions of zircon, and combined the four analyses yielded an age of 1716 ± 31 Ma on a conventional concordia plot. Overall, four Rb/Sr analyses are consistent with a model isochron of ∼1750 Ma. Thus, although the data are sparse, basement rocks in the Mineral Mountains region also appear to have formed at approximately the same time as the Grand Canyon and Beaver Dam Mountains crystalline rocks (Fig. 2).

A small exposure of basement rock in the footwall of the Wasatch fault, or Santaquin complex (Fig. 1), was described by Nelson et al. (2002), who reported an imprecise mean microprobe age on monazite of 1704 ± 90 (1σ) Ma for garnet gneiss. A weakly to nonfoliated mafic syenite (postdating peak metamorphism) yielded highly correlated, but strongly discordant zircons with an upper intercept of 1673 ± 23 (2σ) Ma. ⁴⁰Ar/³⁹Ar ages of hornblende from amphibolite range from 1623 to 1657 Ma (Nelson et al., 2002; Kurt Constenius, 2005, personal commun.), although recalculation of these ages using a new decay constant (Min et al., 2000) and updated ages for the flux monitors (MMhb-1 = 523.1 Ma; Fish Canyon sanidine = 28.201 Ma) produces a range from 1645 to 1679 Ma (Nelson et al., 2009). The upper end of this range of ages is concordant with U/Pb results in syenite.

The Farmington Canyon complex comprises orthogneiss, migmatitic paragneiss, quartzite, amphibolite, pegmatite, and schists with arkosic, graywacke, silty, or tuffaceous protoliths (Bryant, 1988). Barnett et al. (1993) reported six ²⁰⁷Pb/²⁰⁶Pb ages for monazite that range from 1.64 to 1.71 Ga. Monazite U/Pb ages avoid the pitfalls of inheritance during initial episodes of prograde metamorphism as this phase is at least as retentive as zircon during subsequent reheating and recrystallization from sedimentary phosphate (Nelson et al., 2002, and primary references therein). Although monazite may record multiple episodes of metamorphism, the oldest age domains should record the time of first metamorphism. Nelson et al. (2002) report a mean monazite age of 1705 ± 90 (1σ) Ma for 34 analyses of seven grains from paragneiss. Although this age is imprecise, no single analysis yielded an apparent age in excess of 1900 Ma.

Earlier, Hedge et al. (1983) reported an Rb/Sr isochron and U/Pb age of ∼1800 Ma for a quartz monzonite gneiss. In that study, 18 discordant zircon separates yielded ²⁰⁷Pb/²⁰⁶Pb ages from 1.70 to 2.30 Ga. Stroud et al. (2007) reported in an abstract that an orthogneiss body from the Farmington Canyon complex contained 2.44 Ga and 1.70 Ga zircons, interpreted to represent crystallization and metamorphic ages, respectively. Alternatively, it is possible that these dates may reflect 1.70 Ga anatexis and 2.44 Ga inheritance.

New data for banded gneiss from Antelope Island outcrops are reported in Figure 6. Antelope Island exposes a variety of metamorphic rocks, including orthogneiss, which was separated from the remainder of the Farmington Canyon complex by extension along the Wasatch fault (Yonkee et al., 2000). This biotite orthogneiss was collected at the south end of Antelope Island. According to the classification scheme of Frost et al. (2001) and Frost and Frost (2008), the orthogneiss is calcic, ferroan, and marginally peraluminous. It is siliceous (SiO₂ = 72 wt%) and its REE pattern is gull-wing shaped, with marginally enriched, light rare-earth elements (LREEs), moderate negative Eu anomaly, and flat, heavy rare-earth elements (HREEs). Zr is higher than typical granitoids (600 ppm as opposed to 100–200 ppm). Zircons from this rock yield ²⁰⁷Pb/²⁰⁶Pb ages that range from 1615 to 2137 Ma for all grains (n = 21). Eleven grains passing the concordance filter range from 1632 to 1762 Ma (Fig. 6) and have a weighted mean age of 1691 ± 26 (2σ). Two grains, with ²⁰⁷Pb/²⁰⁶Pb ages of 1960 and 2173 Ma, respectively, show evidence of inheritance of older material, but despite the spread in ages, these data clearly indicate that this unit of the Farmington Canyon complex is Paleoproterozoic because the youngest grains in the orthogneiss represent a maximum age. Taken together, these data for the Farmington Canyon complex suggest that it is Paleoproterozoic in age, not Archean as was suggested earlier (e.g., Hedge et al., 1983; Bryant, 1988).

The Owiyukuts complex has been considered Archean based on an unpublished 2.70 Ga Rb/Sr age and lithologic correlation to Cheyenne Belt exposures in southern Wyoming, whereas the Red Creek Quartzite was considered Protero zoic (Sears et al., 1982). Hansen (1961), however, mapped the Owiyukuts complex as structurally overlying the Red Creek Quartzite. So far, our fieldwork has not allowed us to determine relative ages of these two bodies.

New ages suggest that both are Paleo protero zoic, however. Two ⁴⁰Ar/³⁹Ar ages on hornblende in tholeiitic orthoamphibolite from the Red Creek Quartzite are ∼1650 Ma (Fig. 7). These samples are from the easternmost mapped block of the Red Creek Quartzite and occur as abundant tabular bodies from a few centimeters to several meters in width. In turn, these are intercalated with pure quartzite as amphibolites represent tholeiitic dikes and sills injected into pure quartz sand.

The ⁴⁰Ar/³⁹Ar ages probably approximate the timing of peak metamorphism. Electron microprobe traverses show that hornblende is not zoned, and Red Creek Quartzite garnet lacks significant retrograde zoning, especially for Fe and Mg, indicating that postpeak-metamorphic cooling was probably rapid.

U/Pb ages for zircon from an adjacent exposure of the Owiyukuts complex vary widely (Fig. 7). Although Sears et al. (1982) referred to this body as granite gneiss, it is clearly a quartz and potassium feldspar paragneiss with up to 90 wt% SiO₂. Rounded detrital zircons range from 1439 to 2748 Ma, with a mean value of 2156 Ma. If individual analyses are filtered to exclude data with >10% discordance, two <1600 Ma ages are eliminated. Filtered ages range from ∼1740 to 2750 Ma. These data constrain the deposition of this part of the Owiyukuts complex to postdate the age of the youngest detrital zircon, that is, ∼1740 Ma.

The Albion–Raft River–Grouse Creek Ranges, by contrast, are considered to contain Archean material (Premo et al., 2008). Egger et al. (2003) reported a 2.62 Ga upper intercept on a standard concordia plot for zircons from a mafic orthogneiss from this area, reinforcing the interpretation of Rb/Sr data by Armstrong and Hills (1967) and Compton et al. (1977) that late Archean rocks are present.

In the Snake River Plain, Wolf et al. (2005) reported Archean U/Pb ages for zircons in crustal xenoliths (2.5 to >3.5 Ga) entrained in basalt, indicating that Archean crust underlies parts of southern Idaho (Fig. 1). The Leucite Hills of south-central Wyoming also contain ∼2.6 Ga zircons (Farmer et al., 2005). The presence of Archean rocks in both of these localities provides important regional constraints on the extent of the Archean Wyoming province and where its southern boundary, the Cheyenne Belt, may be located.

We note the use of secondary Pb isochrons as tools to identify the age of basement rock with a caveat. Because of a paucity of data, the following discussion includes feldspar, ore lead, and in a few cases, whole-rock data to maximize available information. We recognize that these materials sample crustal lead, but perhaps over different spatial scales, so these results should be interpreted with this in mind.

Ore lead from the Milford region (Fig. 8A) yields a secondary Pb isochron of 1730 Ma, which is concordant with U/Pb ages in basement rock from the nearby Mineral Mountains as well as the Yavapai province as a whole. As for Mojavia, whole-rock Pb isotopes from the Santaquin complex yield an unusually young age of 1370 Ma. Ore lead from the nearby Tintic mining district yields an age of 2070 Ma. Ore leads produce a 2040 Ma age for Gold Hill and 2340 Ma for the Park City–Alta area. However, Gold Hill feldspar lead produces much younger ages (1020 and 1220 Ma), as do Bingham ore leads (1640 Ma) (Fig. 8A). The reader should note that Bingham ore leads are very close to the 2340 Ma Park City–Alta ore leads, especially after Basin and Range extension is removed. The same relationship holds for the Santaquin complex and Tintic ages.

Neodymium T_DMs for the Yavapai province (Mineral Mountains only; Fig. 8B) range from 1000 to 1440 Ma. Within Mojavia of the Great Basin, Nd T_DMs of the Santaquin complex range from 1250 to 1810 Ma, and 1280 to 2180 for the Gold Hill area. Most of the remainder of the samples are near the historical alignment of the Cheyenne Belt, or boundary between Mojavia and the Wyoming province. Data for the intrusions underlying the Bingham, Alta, and Park City areas range from 1420 to 1670 Ma. Variability in the Ruby Mountains area is much larger, ranging from 840 to 3200 Ma. Wright and Snoke (1993) call on a northern Archean and southern Proterozoic terrane beneath the Ruby Mountains–East Humboldt range along with mantle sources to explain this range in model ages.

North of the traditional Cheyenne Belt in the Farmington Canyon complex, Hedge et al. (1983) reported three Archean Nd T_DMs. In this study, we report ten new Nd model ages, nine of which range from 2.48 to 2.71 Ga, and include a variety of lithologies including metasediment, metaigneous, and igneous rocks. One sample, a medium-grained granite from Antelope Island, has an unusual Nd T_DM of 1.09 Ga, and an unusually high initial ¹⁴³Nd/¹⁴⁴Nd of 0.511155 (at 1680 Ma).

The Neoproterozoic Uinta Mountain Group and Big Cottonwood Formation exhibit model ages in excess of 2.00 Ga. However, these are unmetamorphosed to weakly metamorphosed supracrustal rocks with isotope systematics that reflect sedimentary provenance rather than the age of underlying basement. These samples are discussed here because they contribute to the interpretation of other samples, which may be affected by contamination by supracrustal sequences, or highly metamorphosed equivalents.

Wooden et al. (1999), augmented by data from Armstrong and Hills (1967) and Compton et al. (1977), provide a large set of Sr and common Pb isotope data that permits the testing of previously published alignments of the northern and southern boundaries of Mojavia (Fig. 1). These data are from silicic Phanerozoic plutons and volcanic rocks that have been emplaced through the crustal column. We have selected a subset of their data (shown in Fig. 1), bracketing the Utah-Nevada border by approximately ±2° longitude and excluding other data. Use of this spatial filter limits the data set to lie on or west of the Basin and Range–Colorado Plateau boundary and east of the 0.706 line to minimize inferences being drawn based upon the influence of other known or suspected crustal features. The remaining data were submitted to statistical scrutiny.

We conducted a separate statistical test case using the common lead data of Wooden and DeWitt (1991) and Wooden and Miller (1990) that span the Yavapai-Mojave boundary in western Arizona. That such a boundary exists in this location is well established but is not evaluated in this study. Rather, we use it as a test case for the validity of the statistical approach. For ²⁰⁶Pb/²⁰⁴Pb and ²⁰⁷Pb/²⁰⁴Pb ratios, the boundary is distinguished at a ≥98% probability for both T- and U-tests. ²⁰⁸Pb/²⁰⁴Pb ratios appear to be less discriminatory and are therefore not considered further.

The proposed boundary of Bennett and DePaolo (1987) between the Yavapai province and Mojavia is east-west and nearly perpendicular to the Utah-Nevada border (Fig. 1). Thus, examination of latitudinal trends in the isotopic composition of igneous rocks may support the existence of the boundary, as each block should be composed of basement with a contrasting average age. Crustal contamination of igneous rocks emplaced through these blocks may be expected to produce, on average, more radiogenic Sr and Pb isotopic compositions within Mojavia.

The larger Mojave and Yavapai provinces seem to be well discriminated (>99% confidence interval; P <0.01) on the basis of Pb isotopes (Table 1). Mean lead isotope ratios are more radiogenic for the Mojave block relative to the Yavapai province, as expected (Table 1).

With Sr, it is also reasonable to assume that the Mojave province will be, on average, radiogenic. This appears to be the case. In fact, ⁸⁷Sr/⁸⁶Sr ratios are recognized at a 90% confidence interval for the Yavapai-Mojave boundary for the U-test and >99% confidence for the T-test (Table 1).

Results for the northern boundary of Mojavia only discriminate ²⁰⁶Pb/²⁰⁴Pb ratios for the T-test at a >95% confidence level. Average ²⁰⁶Pb/²⁰⁴Pb ratios are less radiogenic north of the Cheyenne Belt alignment, counter to the expectation that Archean crust should be more radiogenic, unless U loss occurred during metamorphism or U was partitioned during partial melting into younger magmas. ²⁰⁷Pb/²⁰⁴Pb and ⁸⁷Sr/⁸⁶Sr_i ratios are effectively identical (Table 1). A simple post hoc power analysis (see the Supplemental Text File²) clearly shows that given the mean and range of isotope ratios from both candidate provinces, the existing data set is grossly insufficient for discriminatory purposes. In fact, it may not be possible to gather sufficient data to employ any meaningful statistical test.

The three categories of data discussed above, crystallization ages of basement rocks, Pb and Nd model ages, and spatial trends of Pb and Sr isotopes, are evaluated in terms of which category of data is best suited to discriminate between candidate basement provinces. As discussed below, we ultimately conclude that it is at least as likely that the Yavapai province underlies the entire eastern Great Basin, and that no discrete Mojavia exists here. In addition, we conclude that the traditional Cheyenne Belt alignment is in error and should be located farther north. However, for the sake of discussion of the evidence for these conclusions, we have retained traditional terminologies for the regions illustrated in Figure 1.

There is very little or no direct geochronological evidence for Mojave-age material (>2.0 Ga) in the western Grand Canyon, the Beaver Dam Mountains, or the Mineral Mountains. This is a vital observation because these localities are within Mojavia as drawn by Karlstrom et al. (2002). In fact, the oldest known body from these sequences is the 1.84 Ga Elves Chasm gneiss in the Grand Canyon (Hawkins et al., 1996).

The Farmington Canyon complex contains evidence for zircon inheritance of Mojavian age, both in previously published and new SHRIMP data presented here. New SHRIMP data include concordant zircons that are as young as 1632 Ma in high-grade banded ortho gneiss from Antelope Island, a body which experienced metamorphic pressures of up to 4 Kbar and temperatures as high as 785 °C (Barnett et al., 1993). The best evidence available suggests that the Farmington Canyon complex, as a coherent metamorphic body, is Paleoproterozoic rather than Archean.

The 1.84 Ga Elves Chasm pluton, interpreted to be “basement” to the metasedimentary rocks of the rest of the inner gorge of the Grand Canyon (Karlstrom et al., 2001), may place an upper limit to the age of crust in this area. Thus, Yavapai basement may be at least 100 Ma older than its overlying supracrustal sequences. Based on present knowledge, 1.84 Ga appears to be an upper bound to the age of Yavapai and Mojavian basement outside the Mojave Desert.

The only evidence for inheritance of materials >1.84 Ma for localities south of the traditional Cheyenne Belt alignment is found in the new ages for the Beaver Dam Mountains. Because these ages were determined by laser-ablation methods on several tens to >100 individual analyses per sample, they afford the opportunity to examine for inheritance of older grains. No grains have apparent ages old enough to qualify as true “Mojavian” (>2.0 Ma), although there is weak evidence for inheritance of >1.84 Ga material as three individual analyses range between ∼1890 and 1950 Ma.

Although fairly sparse within the Yavapai province, the available data are consistent with the presumed age of basement in the case of the single ²⁰⁷Pb/²⁰⁶Pb secondary isochron for Miners ville leads, and with a mixture of mantle- and crustal-derived Nd for the nearby Mineral Mountains (Fig. 8).

For Mojavia, ²⁰⁷Pb/²⁰⁶Pb and Nd T_DM model ages are varied (Fig. 8). Some young Nd T_DMs (<1.6 Ga) may result from combined mantle and crustal inputs, such as the younger samples from Gold Hill, the Mineral Mountains, etc. Bingham, by contrast, exhibits Yavapai ages for both Pb and Nd systems and lies essentially along the conventional Cheyenne Belt alignment. Samples of apparent Mojavian antiquity (>2000 Ma) are found, but they may simply result from partial inheritance of material shed from the Wyoming province, rather than representing the true crustal residence time of juvenile basement. Many of the Nd T_DMs of apparent Mojavian age, in fact, are clustered near the traditional alignment (Fig. 8B), suggesting that these model ages may reflect an older (on average) provenance, because these localities are relatively close to an Archean source from the Wyoming province, even if the boundary is not well located.

Inheritance is well exemplified by the Archean Nd T_DMs of the Farmington Canyon complex, which lies north of the traditional Cheyenne Belt alignment. Coupled with the strong evidence for ∼1.7 rather than >2.5 Ga metamorphism, the Archean model ages are best explained through incorporation of sedimentary input from the Wyoming province into Paleoproterozoic gneisses. The weakly metamorphosed and supracrustal Neoproterozoic Big Cottonwood Formation and Uinta Mountain Group, both of which have T_DMs in excess of 2.0 Ga, also suggest that considerable sediment was shed from the Wyoming craton onto Mojavia from Paleoproterozoic through Neoproterozoic time.

In summary, Pb and Nd isotope model ages are highly varied, and given the possible isotopic end members of mantle as well as both basement and supracrustal Paleoproterozoic and Archean sources, it is difficult to recognize an area of true Mojavian basement affinity in the eastern Great Basin.

All but one statistical test suggest, at >99% confidence, that the Yavapai province and Mojavia are isotopically distinct. The remaining test discriminates the two provinces at a 90% confidence interval (Table 1). We conclude that whatever the origin, there is a fundamental difference between the distribution of Pb and Sr isotopes in the Phanerozoic igneous rocks of these two candidate provinces.

The two crustal blocks should show a marked change in the distribution of isotope values across their common boundary. Figure 9 illustrates what may be a major change as the crustal boundary is crossed. For example, the lower end of the distribution of ⁸⁷Sr/⁸⁶Sr_i values is the same (∼0.705) in both provinces, but the upper range is much higher in Mojavia. As the traditional alignment of the Cheyenne Belt is reached, highly radiogenic values of nearly 0.740 are achieved.

Remembering that the Sr isotope signature is recorded largely in Mesozoic and Tertiary igneous rocks, two end-member conditions could be responsible for the trend observed in Figure 9, including: (1) an increased average age and heterogeneity in the Rb/Sr ratio in true Mojave basement, or (2) a tendency toward greater northward involvement of Archean sediment shed from the Wyoming craton and emplaced on juvenile Paleoproterozoic basement to be sampled at a later time by Phanerozoic igneous rocks.

If the first case applies, then the statistical tests strongly support the existence of a distinct change in the character of the basement between the Yavapai province and Mojavia. If the second case applies, then the increased heterogeneity indicates that no fundamental crustal boundary need be present at this location. Concealed Mojavia in the eastern Great Basin may merely be juvenile Yavapai basement upon which a wedge of sediment, largely derived from the Wyoming Craton, has been deposited and metamorphosed, including accretionary mélanges such as the Farmington Canyon complex (Shervais, 2006). Crustal contamination during the ascent of these magmas could produce a highly heterogeneous character by variably sampling juvenile Yavapai lower crust and overlying material of Archean heritage (Fig. 9), which may include both high-grade Paleoproterozoic metamorphic sequences and overlying Neoproterozoic sediments.

There is some precedent for our second and favored interpretation. Jones et al. (2009) detected sparse 2.1–2.8 Ga detrital zircons in Paleoproterozoic quartzites and metaconglomerates in southern Colorado and northern New Mexico, even though these localities are much farther from a known sediment source of Archean age such as the Wyoming province. It seems reasonable that an accretionary mélange such as the Farmington Canyon complex that formed adjacent to an Archean cratonic block would be strongly influenced by material of Archean heritage.

The Albion–Raft River–Grouse Creek Ranges preserve what are perhaps the most complete sections of Proterozoic rocks, where <4 km and generally <1 km of supracrustal sediment are exposed overlying Archean rocks (Hintze and Kowallis, 2009) north of the traditional Cheyenne Belt. Southward, sequences of supracrustal Neoproterozoic rocks thicken considerably, although only minimum thicknesses can be deduced because the base of supracrustal rocks is not exposed. In the Deep Creek Range on the Utah-Nevada border nearly 6 km are exposed, whereas up to nearly 10 km are found in the west Tintic Mountains of Utah.

In the eastern Uinta Mountains on the order of 10 km of Proterozoic supracrustal rocks are exposed in the Uinta Mountain Group and the Red Creek Quartzite (Hintze and Kowallis, 2009). The thickness of the Owiyukuts complex and other concealed supracrustal lithologies is also unknown, but could be considerable. The thickness of the Farmington Canyon complex protolith sequence cannot be known, but given the size of the body and the variations in metamorphic grade (Bryant, 1988), this ∼1.7 Ga accretionary mélange (Shervais, 2006) lying above slightly older juvenile basement may have been very thick.

It is tempting to consider the Owiyukuts complex, Red Creek Quartzite, and Farmington Canyon complex as “basement” given that at least portions of them were metamorphosed at amphibolite facies or higher. However, correlative metamorphic rocks very likely comprise the upper part of the crystalline sequence in the northeastern Great Basin south of our proposed Cheyenne Belt (Fig. 1) and represent a body of radiogenic material of largely Archean parentage (Fig. 9) that also may be overlain by Neoproterozoic sediments with significant Archean input.

We favor the contamination of igneous rocks by sequences with a large component of Archean heritage overlying juvenile Yavapai arc basement (Fig. 9). The docking of the Yavapai to the Wyoming province would have resulted in the emplacement of Archean sediment onto Paleoproterozoic terranes to the south. In fact, the Nd isotopic character of the Farmington Canyon complex and the broad population of zircons in the Owiyukuts complex are direct evidence that this is the case. As mentioned, Nd T_DMs of the Big Cottonwood Formation and Uinta Mountain Group and zircon age distributions of these same units indicate continued input from Archean and Paleoproterozoic sources (Dehler et al., 2010), as well as the reworking of uplifted and eroded Farmington-equivalent materials.

Although the use of statistical methods provides a powerful objective tool for testing differences in the isotopic character of candidate crustal provinces, all they can really establish are differences or similarities between candidate data sets. Ascribing meaning to the differences or similarities remains an inductive task based on other factors.

The new ages of the Owiyukuts complex and Red Creek Quartzite invalidates the suggestion that Archean crust is present at this location. Zircon analysis suggests that the metamorphism of the Owiyukuts complex is at least as young as ∼1740 Ma.

⁴⁰Ar/³⁹Ar ages of the Red Creek Quartzite suggest that the age of metamorphism of this body is probably ∼1650 Ma. Such ages reflect cooling through ∼500 °C and therefore cannot preclude an episode of protracted cooling since Archean time. Quantitative reconnaissance estimates of the pressures and temperatures of metamorphism of three amphibolites from the type locality at Red Creek range from 630 to 660 °C and 5.5 to 7.4 Kbar (Zenk and Schulz, 2004). Such estimates place these rocks at mid-crustal levels, yet they lack zoning in hornblende and garnet, as described above. It seems unlikely to us that temperatures would cool 100–150 °C over nearly 1 Ga of Earth history without producing strong retrograde zoning. In summary, there is no solid evidence for Archean rocks at this locality, and the Cheyenne Belt may lie in a concealed position somewhere between this locality and the Leucite Hills (Farmer et al., 2005).

Well to the east of this area near outcrop of the Cheyenne Belt in southern Wyoming the CD-ROM experiment (Karlstrom et al., 2005) shows evidence of a high-velocity slab remnant, or north-dipping subduction scar. Just to the east of the Red Creek area, however, the Deep Probe experiment lacks a prominent subduction scar. Rather, low-velocity mantle underlies the crust immediately north of the Cheyenne Belt (Karlstrom et al., 2005). Thus, there is little convincing evidence from geophysical data for the location of the Cheyenne Belt in this area.

A comparison of our new results and compiled data for the Farmington Canyon complex reveals two important features. First, Nd T_DMs are nearly all Archean, suggesting extensive involvement of ancient material in the formation of this body. Second, there is no direct evidence for metamorphism of this body prior to 1.8 Ga. The best interpretation for these observations is that the Farmington Canyon complex is located near the Wyoming Craton yet south of the Cheyenne Belt. Thus, it represents the Paleoproterozoic metamorphism of a sedimentary sequence of chiefly Archean provenance. The boundary is probably buried north of the Farmington Canyon complex and south of Archean xenolith localities of the Snake River Plain (Wolf et al., 2005).

With respect to geophysical constraints, Figure 10 illustrates the Snake River Plain tomographic section with both the traditional and our revised location of the Cheyenne Belt. Karlstrom et al. (2005) interpret a north-dipping, high-velocity zone as a subduction scar, similar to the CD-ROM section. We believe this interpretation is consistent with our revised placement of the Cheyenne Belt, because the crustal suture should be placed at the leading (northward) rather than the trailing (southward) edge of the scar. We favor this interpretation, but note it is also possible to draw a southward dipping subduction scar that would return the Cheyenne Belt to its original alignment. In fact, a third north-dipping scar can be drawn to the southeast of both of these features. All three could represent subduction scars because subduction polarity changed, and new arcs accreted to North America. On the other hand, apparent subduction scars could be artifacts of other sources of mantle heterogeneity beneath North America. In summary, although the interpretation of Karlstrom et al. (2005) supports our model, tomographic data may not have unique interpretations or sufficient resolution to unambiguously identify the location of crustal sutures in the eastern Great Basin.

The region of the Utah-Nevada border is devoid of basement exposures, except at its north end in the Albion–Raft River–Grouse Creek area, where true Archean crust appears to be present (Premo et al., 2008; Egger et al., 2003; Armstrong and Hills, 1967; Compton et al., 1977). Assuming this body is connected at depth to the remainder of the Wyoming province (see Foster et al., 2006, for an alternate view), the Cheyenne Belt must lie to the south.

As discussed above, the Cheyenne Belt, at least in its historical position (Fig. 1), is poorly discriminated by Pb and Sr data. In fact, mean isotopic values across the boundary are sufficiently similar and the standard deviations about these means are sufficiently large that it may not be tractable to gather sufficient data to rigorously test this boundary (Table 1; Fig. 10). In summary, the only constraint that can be placed on the Cheyenne Belt in this region is that it probably lies southward of the Grouse Creek Range.

Premo et al. (2008) cast doubt on the existence of Archean basement in northern Nevada. Lush et al. (1988) proposed the existence of Archean crust in the Angel Lake area of the East Humboldt Range of northern Nevada (Fig. 1). Recent in situ analysis of zircon grains, however, suggests inheritance of Proterozoic to Archean cores during a Cretaceous magmatic event and that Archean crust is absent (Premo et al., 2008). It appears that the Cheyenne Belt must lie to the north of the Angel Lake area as well.

Although prior studies suggested that Archean rocks were represented in the Owiyukuts complex, Farmington Canyon complex, and East Humboldt Range, no rocks of this age are clearly present. This infers the Cheyenne Belt lies to the north of all of them. By contrast, Archean rocks in the Grouse Creek Range combined with Archean xenoliths in Snake River Plain basalts and ultrapotassic rocks of the Leucite Hills require that the Cheyenne Belt lie to the south.

The best evidence is that the Cheyenne Belt must lie within a broad region north of its historical alignment (Fig. 1). Foster et al. (2006) suggested that the Archean rocks of the Albion–Raft River–Grouse Creek block might not be attached at depth to the Wyoming province. Speculatively, the Albion–Raft River–Grouse Creek area could be a small block of Archean rocks embedded in the Paleoproterozoic mass that accreted to the Wyoming province. Thus, the Cheyenne Belt could conceivably lie farther north than our proposed alignment (Fig. 1). Although much uncertainty remains, existing evidence casts considerable doubt on the historical placement of the Cheyenne Belt.

This study shows that there is no direct geochronological evidence for material of 2.0–2.3 Ga in basement rocks exposed within greater Mojavia of the eastern Great Basin. Lead and Nd model ages, by contrast, range from Archean to Neoproterozoic, even over small spatial scales. Archean model ages, where they occur, tend to be located in the northern portion of greater Mojavia. This observation suggests that sediments derived from the Wyoming province and deposited onto Paleoproterozic basement terranes south of the Cheyenne Belt may influence isotope systematics, imparting an apparent Mojavian affinity. In summary, there is no strong evidence that Mojavian basement exists outside of the Mojave Desert region.

An evaluation of an apparent boundary between the Yavapai province and Mojavia was conducted based largely on common Pb and Sr isotopes in Phanerozoic silicic plutons and volcanic rocks. Statistically, there is a distinct difference in these parameters north and south of the boundary as first drawn by Bennett and DePaolo (1987). However, the statistics only demonstrate that the data are different, which is a necessary condition for recognizing a boundary, but insufficient by itself. Close examination of the data similarly suggests that supracrustal material shed from the Wyoming province is likely responsible for the difference.

A review of new and existing isotopic ages indicates the Cheyenne Belt, representing the northern boundary of Paleoproterozoic basement in the eastern Great Basin, is commonly located too far south. Rocks of the Owiyukuts complex and Red Creek Quartzite have a Paleoproterozoic age of metamorphism, as does the Farmington Canyon complex, although both show inheritance of older zircons. Ortho gneisses in the East Humboldt Range of northern Nevada have recently been shown to be Cretaceous in age (Premo et al., 2008). All of these rock bodies were previously assigned to the Archean. Thus, the Cheyenne Belt must lie in a poorly defined location north of these localities and south of the Grouse Creek Range, Snake River Plain, and Leucite Hills. Common Pb and Sr isotope data similarly fail to identify the boundary in the position of its historically drawn alignment.

Our analysis was based in part upon statistical tests of terrane boundaries (Table 1; Supplemental Text File [see ^{footnote 2}]). Where mean differences in isotopic parameters are large and variances about those means are small, such as between Yavapai and Mojavia data sets, boundary hypotheses can be tested with surprisingly small data sets. However, where the opposite conditions exist, as between Mojavia and Wyoming provinces, the analysis can become intractable due to large data requirements.

The Utah Geological Survey provided funding for U/Pb ages of the Beaver Dam Mountains. Lehi Hintze and Janice Hayden are also thanked for field assistance there. Our appreciation is also extended to Dennis Eggett of the Brigham Young University Statistical Consulting Office. C. Frost thanks R. Frost, K. Chamberlain, and A. Yonkee for field assistance, and K. Chamberlain is thanked for overseeing mineral separation. Previous versions of this text have benefited from comments by David Foster, John Shervais, Ernest Duebendorfer, Adolph Yonkee, Marion Bickford, Jamey Jones, and two anonymous reviewers.