Widespread base- and precious-metal anomalies, oxidized sulfide veins, silicified calcareous shales and carbonates, and altered porphyry intrusions occur in the northeastern Sulphur Spring Range, Nevada, 80 km south of important gold deposits in the Carlin trend. The small historic mines and prospects in the area are spatially and perhaps genetically related to a suite of variably altered dikes, small lava flows, silicic domes, and related pyroclastic rocks. New major- and trace-element data and U-Pb zircon ages show that the East Sulphur Spring volcanic suite is Eo-Oligocene in age (36–31 Ma) and ranges in composition from high MgO-basaltic andesite to peraluminous rhyolite. The major- and trace-element compositions of the volcanic rocks are characteristic of continental margin subduction zone magmas and form a high-K, calc-alkaline suite with low Fe/Mg ratios. In addition, the rocks have negative Nb and Ti anomalies and elevated Ba, K, and Pb on normalized trace-element diagrams. Crustal melting is indicated by the eruption of a peraluminous garnet-bearing ignimbrite and as a component in hybridized andesite.

The nature of this suite and its potential for mineralization is elucidated via comparisons to other Eocene age volcanic rocks associated with much larger gold and copper deposits in the Great Basin. The East Sulphur Spring suite is more similar to Eocene igneous rocks found along and near the Carlin trend than it is to those erupted while the Bingham porphyry copper deposit developed 300 km farther to the east. For example, the East Sulphur Spring suite and the Eocene magmatic rocks along the Carlin trend are less alkaline than the Bingham suite and lack its unusual enrichment of Cr, Ni, and Ba in intermediate composition rocks (58–68 wt% silica). Nonetheless, the Bingham and East Sulphur Spring volcanic suites both preserve evidence of mixing that created intermediate compositions. For example, an andesite has obvious mineral disequilibrium with plagioclase, biotite, clinopyroxene, orthopyroxene, olivine, and amphibole coexisting with extensively resorbed megacrysts of quartz, K-feldspar, and garnet—indicative of mixing basaltic andesite or andesite and largely crystallized garnet-bearing rhyolite. On the other hand, we found no evidence for mixing with a mafic alkaline magma like that in the Bingham Canyon magma-ore system.

We conclude that: (1) an unusual tectonic setting prevailed during the Eocene and Oligocene of the western United States that promoted the production of oxidized mafic magma in an arclike setting, but far inland as a result of the rollback of the Farallon slab; (2) the mafic magmas intruded or erupted separately, or mixed with more silicic magma generated by fractional crystallization and assimilation of crustal materials; and (3) these mafic magmas may have delivered significant amounts of sulfur and chalcophile metals to upper crustal magma chambers and eventually to Paleogene ore deposits in the eastern Great Basin.

The Great Basin of the western United States contains a multitude of ore deposits and associated igneous rocks. Studies of the ages and compositions of the volcanic and intrusive rocks have shown that many of the deposits are not only spatially associated with magmatism, but temporally and genetically linked to igneous processes as well. In many cases, magmas and their solidified equivalents were important sources of heat to drive hydrothermal convection, of sulfur used as a complexing agent in the fluids and then deposited in sulfides and sulfates, and of the ore metals themselves. In this paper, we consider this paradigm in light of the relationships between Paleogene magmatism and ore deposition in north-central Nevada.

The Sulphur Spring Range is ∼80 km south of large gold deposits in the Carlin trend (Fig. 1). The Mineral Hill district, on the west side of the range, and the Union district, on the east side, were mined in the late 1800s to early 1900s for gold, silver, and copper (Lincoln, 1923). Mineral Hill is best described as a small polymetallic replacement deposit (19A of Cox and Singer, 1986) and is probably related to the distal effects of middle Cenozoic magmatism according to McKee and Moring (1996). Small base metal–silver replacement bodies, which would now be identified as carbonate replacement deposits (Megaw, 1999), are found in the Union district. Prospects containing Au and As were explored in the 1980s.

The Sulphur Spring Range is underlain by a thick sequence of east-dipping Paleozoic sedimentary rocks. Prior to recent mapping, a 2 km2 area of undifferentiated Tertiary volcanic rock on the east side of the range (Carlisle and Nelson, 1990) contained the only known outcrops of volcanic rock. Our mapping identified numerous small exposures of igneous rocks that either intrude or overlie the Paleozoic deposits.

Some of the igneous rocks are spatially associated with mineralization and exhibit key characteristics of porphyry deposits (cf. Beane and Titley, 1981; Richards, 2003; Seedorf et al., 2005), including substantial amounts of phyllic and argillic hydrothermal alteration, pebble dikes, breccia pipes, and disseminated and vein-related mineralization. Altered mafic and intermediate composition dikes have geochemical anomalies of As, Ba, Bi, Cr, Cu, Mo, Ni, Pb, Sb, Tl, and Zn. We have also identified evidence of magma mixing in the intermediate composition volcanic rocks. This may be an important feature of porphyry copper deposits such as the enormous Eocene Bingham Canyon deposit 300 km to the east (Maughan et al., 2002) and is reexamined here.

In addition, the Sulphur Spring Range has several features in common with Carlin-type gold deposits (Fig. 1), which contain the most productive gold mines in North America (Jensen et al., 1995). The geology and origin of Carlin-type gold deposits are described in detail by Hofstra and Cline (2000) and Cline et al. (2005). Mineralized rocks in the Sulphur Spring Range, like most Carlin-type gold deposits, occur below the Roberts Mountains thrust at intersections of a complex array of structures with permeable and reactive strata, usually Devonian carbonate rocks or calcareous clastic sediments. Small bodies of jasperoid have anomalous concentrations of Au, As, Hg, Sb, and Tl. The deposits are spatially associated with volcanism that is part of a southward sweep of magmatism that passed through this area in the Eocene (Seedorf, 1991; Hofstra et al., 1999; Cline et al., 2005; Ressel and Henry, 2006). However, the exact nature of the relationship between Carlin-type deposits and Eocene magmatism is the subject of debate.

Because of its broad similarities to much larger Eocene porphyry copper and Carlin-type gold deposits, the deposits in the Sulphur Spring Range have recently been the site of grass-roots exploration for base metals and Au. To help assess the potential for and further our understanding of these important ore deposits, we compare the igneous rocks of the Sulphur Spring Range to those from other Eocene magmatic centers related to mineralization in the Great Basin—those near the Carlin trend in Nevada and the Bingham porphyry system in Utah. We conclude that there are several important similarities in these ore-related magma systems, including their ages, tectonic settings, and the potential for mafic magmas to have delivered metals and sulfur to upper crustal hydrothermal systems.

During the late Proterozoic, central Nevada lay on the western edge of the rifted North American continent (Wooden et al., 1998; Grauch et al., 2003). The former continental margin, defined by rifting in the Neoproterozoic and identified by the 87Sr/86Sr = 0.706 line, lies just west of the Carlin and Battle Mountain–Eureka mineralization trends (Fig. 2; Tosdal et al., 2000; Grauch et al. 2003). Neoproterozic and Cambrian siliciclastic sediment were deposited during the rift phase. From the Ordovician through Devonian, thick shelf-type sediments accumulated on the continental margin and silty carbonate rocks accumulated on the slope to the west (Madrid, 1987; Finney et al., 2000; Cook and Corboy, 2004; Cline et al., 2005). In the late Devonian through late Mississippian, the Antler orogeny affected the western margin of the North American plate (Carpenter et al., 1994; Dickinson, 2006). This contractional orogeny produced the Roberts Mountains thrust (Fig. 2) that juxtaposes the Ordovician Vinini Formation and some Mississippian clastic rocks over Devonian carbonate rocks. In the Sulphur Spring Range, these are typically capped by the Devils Gate Limestone or the Telegraph Canyon Formation (Carlisle and Nelson, 1990). The structural contact between carbonate rocks and clastic rocks in the Roberts Mountains allochthon is one of the typical settings for gold mineralization along the Carlin trend, including the Rain mine, the nearest of the Carlin-type mineral deposits (Longo et al., 2002). The Roberts Mountains thrust fault has also been delineated in various locations throughout the Sulphur Spring Range (Carlisle and Nelson, 1990; Johnson and Visconti, 1992). The Sonoma orogeny in the Triassic created the Golconda thrust (Dickinson, 2006), which lies to the west (Fig. 2).

During the Mesozoic, subduction beneath western North America created a protracted series of contractional events (Dickinson, 2006). The Sevier orogeny thickened the crust beneath Nevada and western Utah, and a series of thrust sheets formed to the east (e.g., DeCelles, 2004). In the Great Basin, Jurassic magmatism included metaluminous to peraluminous granitoids and sparse lamprophyre dikes (e.g., Ressel and Henry, 2006; Cline et al., 2005). By the end of the Cretaceous, the Farallon plate was subducting at a very low angle under North America and arc magmatism essentially shut off in Nevada, but small volumes of strongly peraluminous granite were intruded along the Utah-Nevada border (e.g., Kistler et al., 1981). During the early Cenozoic, the Farallon plate apparently detached from the lithosphere in a southward-sweeping fashion allowing hot asthenospheric mantle to flow between the subducting slab and the lithospheric mantle (Best and Christiansen, 1991). This created a continental magmatic arc that extended far inland (Lipman et al., 1972; Severinghaus and Atwater, 1990). Middle Tertiary magmatism related to this event may have resulted from (1) dehydration of the Farallon plate, which induced hydrous melting of the mantle wedge, (2) heating lithospheric mantle by hot asthenospheric mantle or by wedge-derived magma, and (3) decompression melting of hot mantle associated with the pattern of flow in the wedge. These mafic, mantle-derived magmas rose and powered crustal magma systems in which more felsic magmas evolved by fractional crystallization and by partial melting and assimilation of continental crust (Ressel and Henry, 2006).

Cenozoic arc magmatism in the northern Great Basin began in the Eocene, between 40 and 36 Ma and then the front swept farther south (Ressel and Henry, 2006; du Bray, 2007). Magma compositions ranged from basaltic andesite to rhyolite (or their intrusive equivalents), and volcanism was widespread throughout northern Nevada (Brooks et al., 1995; Henry and Boden, 1998; Henry and Faulds, 1999; du Bray, 2007). Apparently several magma systems were episodically active beneath the Carlin trend during the Eocene and Oligocene (Grauch et al., 2003; Ressel and Henry, 2006). Concurrent magmatism occurred at similar latitudes in Utah, including that responsible for the Bingham, Clayton, and Alta stocks (Vogel et al., 2001; Deino and Keith, 1997).

Although still controversial, some geologists have concluded that Eocene stress relaxation and extension accompanied this southward sweep of arc magmatism across western North America (Gans et al., 1989; Feeley and Grunder, 1991; Seedorf, 1991; Hofstra et al., 1999). For example, Henry et al. (2001), Haynes (2003), Satarugsa and Johnson (2000), Cline et al. (2005), and Henrici and Haynes (2006) conclude that the Eocene-age Elko Formation accumulated in an extension-related basin (Fig. 3). The formation consists of a lower conglomerate unit, lacustrine limestone, and organic-rich shale, but also has interlayered volcanic rocks. The extension that created the Northern Nevada rift (Fig. 2) and continued to form the present-day Basin and Range topography probably began in the Miocene (Zoback et al., 1994; Ressel and Henry, 2006). Bimodal basalt-rhyolite volcanism is typical of this time (John, 2001).

Relationship of Paleogene Mineralization to Magmatism

Although the close spatial and temporal relationships of Eocene igneous rocks to Carlin-trend gold mineralization has suggested a probable link (Ressel et al., 2000; Ressel and Henry, 2006), the presence or absence of magmatic fluids in these meteoric water-dominated hydrothermal systems is controversial. In some Carlin-type districts (e.g., Getchell and Gold Bar), Eocene intrusions are absent (Hofstra and Cline, 2000). Cline et al. (2005) conclude that Eocene magmatism, in conjunction with deep crustal melting and metamorphism, and a pulse of extension-released ore fluids were channeled upward along basement-penetrating faults and mixed to varying degree with exchanged meteoric water. In contrast, Emsbo et al. (2006) suggest that the ore fluid was meteoric water that extracted gold and sulfur from sedimentary rocks already enriched in gold by Devonian exhalative processes.

The Bingham porphyry Cu-Mo-Au system has a much clearer connection to igneous rocks and processes with much of the sulfur, metals, and fluids being of magmatic origin. The magmatism related to the deposit is 38–36 Ma (Deino and Keith, 1997) and is comparable in age to the dikes and gold mineralization in the Carlin trend (Ressel et al., 2000; Ressel and Henry, 2006). Moreover, the porphyry system is associated with distal disseminated gold deposits that have much in common with Carlin-type gold deposits (Cunningham et al., 2004). The role of deep, basement-penetrating faults may not be as important as along the Carlin trend, but Bingham is near an old east-trending continental margin to which Proterozoic terranes accreted (Whitmeyer and Karlstrom, 2007).

More than 200 samples of igneous rocks were collected and a geologic map (Fig. 4) was constructed for this study. Hand sample descriptions and locations are presented by Ryskamp (2006). Thin sections cut from 25 samples of the main lithologic units were studied. Garnet and olivine phenocryst compositions were determined with an upgraded Cameca SX50 electron microprobe. A 20 micron beam and 15 kV acceleration voltage were used for analyses, which are reported in Ryskamp (2006).

Major- and trace-element analyses for 87 of the freshest samples were obtained by X-ray fluorescence (XRF) analysis at Brigham Young University using a Siemens SRS-303 spectrometer. Analyses of representative international standards together with our estimates of precision and accuracy are available from the authors. Trace-element analyses (including rare-earth elements) of 50 of these samples were also performed by ALS Chemex using inductively coupled plasma mass spectrometry (ICP-MS). Rocks were dissolved using a four-acid “near-total” digestion method for most trace elements, but were fused with a flux before digestion for rare-earth element (REE) analyses. Zr concentrations reported from the four-acid dissolution technique were much lower than the XRF analyses and are not used here. Other element concentrations (including REE, Nb, and Th) agreed favorably by the various techniques. Representative analyses are presented in 01, and the complete data set is available in Supplemental Table S11.

U-Pb zircon ages were determined by laser ablation ICP-MS (LA-ICP-MS) in the GeoAnalytical Lab at Washington State University. All zircon samples were processed and separated using standard gravimetric and magnetic techniques at Brigham Young University. Zircon grains, both standards and unknowns, were mounted in a 1-inch-diameter epoxy puck that was ground and polished to expose the interiors of the grains. Cathodoluminescence images acquired at the University of Idaho were used as base maps for recording laser spot locations and to reveal growth and compositional zonation, inclusions, and to look for inherited cores. Chang et al. (2006) present a comprehensive overview of the laser ablation techniques, and a brief overview is given below. The analytical results are summarized in 02, and the complete data set is available in Supplemental Table S22.

New U-Pb zircon ages were acquired for six samples from the eastern Sulphur Spring Range. To avoid problems associated with alteration and zircon inheritance from older rock units, we obtained U-Pb ages on zircons using LA-ICP-MS. Zircon ages were determined using a New Wave UP-213 laser ablation system in conjunction with a Thermo Scientific Element2 single collector, double-focusing magnetic sector ICP-MS. Zircons were analyzed using a 30-µm-diameter beam operating at 10 Hz; the ablated material was delivered to the torch by a mixed He and Ar gas. Laser-induced time-dependent fractionation was corrected by normalizing measured ratios in standards and samples to the beginning of the analysis using the intercept method. Static fractionation, including that caused by laser ablation and due to instrumental discrimination, was corrected using external zircon standards. In our case, we used FC1 and Peixe (Paces and Miller, 1993; Dickinson and Gehrels, 2003). Weighted average ages and Tera-Wasserburg concordia were calculated using IsoPlot 3.0 (Ludwig, 2003).

Total uncertainty for each spot analysis of an unknown was combined quadratically with the uncertainty in the measured isotope ratios and the uncertainty in the fractionation factors calculated from the measurement of standards. For individual analyses we estimate that the accuracy and precision are better than 4% at the 2 sigma level, with the largest contribution in uncertainty from the measurement of the standards. Based on a comparison of LA-ICP-MS ages with ages determined by thermal ionization mass spectrometry, Chang et al. (2006) estimated the accuracy of age determinations using this technique to be on the order of 1% or better. However, there are unresolved contributions to uncertainty from the lack of a common Pb correction and from potential matrix effects between standards and unknowns. Consequently, we analyzed the Temora zircon as an independent check on the accuracy and precision. The Temora zircon has been proposed as a zircon standard by Black et al. (2003), who reported a weighted average 206Pb/238U age of 416.8 ± 1.1 Ma based on 21 isotope dilution-thermal ionization mass spectrometric (ID-TIMS) analyses and 416.8 ± 1.8 Ma based on 50 sensitive high-resolution ion microprobe (SHRIMP) analyses. Chang et al. (2006), using the same instrument and analytical conditions as used here, report an age of 416 ± 9 Ma for Temora. During the course of our analyses, 12 LA-ICP-MS analyses on seven grains of Temora were collected in two separate analytical sessions. All analyses, corrected for fractionation and incorporating fractionation factor uncertainty, give a weighted mean 206Pb/238U age of 416.9 ± 5.6 Ma (mean square of weighted deviates [MSWD] = 0.49), which is within error of the ID-TIMS age 02.

Four samples of the principal volcanic units from the Sulphur Spring Range gave middle Tertiary 206Pb/238U ages ranging from 35.9 ± 0.5 to 31.4 ± 0.5 Ma (02; Fig. 5) that correlate with the stratigraphic sequence where it can be interpreted (Fig. 6). Two samples show no evidence of older zircon grains, while the other two have inherited zircons with Precambrian ages. A dacitic sample (04EB86) from the biotite porphyry unit had one zircon that yielded 207Pb/206Pb ages of ca. 1.8 Ga. An andesite lava (04EB123) with considerable evidence for contamination, as described below, had multiple grains with “anomalous” ages. One grain was large enough to yield three 207Pb/206Pb ages of ca. 2.8 Ga; two other zircon grains gave four separate 207Pb/206Pb ages of ca. 1.8 Ga; two other zircon grains gave three separate 207Pb/206Pb ages of ca. 1.4–1.2 Ga; and one grain yielded late Cretaceous 206Pb/238U ages. We interpret the eruption age of the andesite to be best represented by ten 206Pb/238U ages from four separate zircon grains that range from 33.3 to 30.7 Ma. Because the age range is larger than expected for analytical error alone (as indicated by an elevated MSWD value), we used the TuffZirc routine in IsoPlot to estimate an age of 31.4 +1.3/-0.5 Ma. Thus, this lava is most likely Oligocene in age and significantly younger than the other Paleogene volcanic rocks in the volcanic field. For the other samples, our preferred age was the weighted mean of the 206Pb/238U ages 02. A weighted mean age for 16 analyses yielded an age of 35.1 ± 0.5 Ma for one of the plagioclase dacite domes (Fig. 5). A biotite porphyry intrusion and a biotite dacite tuff have indistinguishable ages of 35.9 ± 0.5 and 35.5 ± 0.4 Ma, respectively (Fig. 5).

Zircon separated from a distinctive rhyolitic clast (sample 04JA156) of what we have called the “square quartz porphyry” was analyzed in an attempt to constrain the eruptive age of a tuffaceous interval low in the volcanic section (the Union tuff described below). However, the 206Pb/238U age of the zircons in the clast is 157.4 ± 2.2 Ma or Late Jurassic 02. Jurassic plutonic rocks are exposed farther south in the Sulphur Spring Range and have similar textures to these porphyritic clasts. This sample also contained inherited zircons with Paleozoic and Proterozoic ages.

Zircon from a nonwelded pyroclastic deposit of rhyolitic composition (sample 03JA122) yielded a Miocene age of 14.0 ± 0.3 Ma, the same age as rhyolitic volcanism associated with the Basin and Range province. Although the zircons are compositionally zoned, we found no evidence of age inheritance in the zircons of this sample 02.

Below we use mapping, structural, stratigraphic, and petrological information, together with the new U-Pb ages, to reconstruct the history of magmatism in the northeastern part of the Sulphur Spring Range. We have informally grouped the Paleogene volcanic rocks into the East Sulphur Spring suite to distinguish them from potentially distinctive igneous rocks to the west. Figures 4 and 6 show the geologic and stratigraphic relationships between the units as determined by superposition, cross-cutting relations, and isotopic ages. Figures 7 and 8 show the elemental compositions of the volcanic rocks.


A deformed succession of Paleozoic sedimentary rocks is in high-angle fault contact with a complex suite of east-dipping volcanic units in the studied area (Fig. 4). The Paleozoic strata dip eastward and are cut by thrust faults related to the Roberts Mountains thrust (Fig. 4; Carlisle and Nelson, 1990). The eastward tilt is probably the result of displacement on range-bounding normal faults that developed during the Miocene. Within the map area, the most prominent faults are the north-trending West Graben and East Graben faults, both of which have apparent normal displacements of hundreds of meters (Fig. 4; Carlisle and Nelson, 1990). Post-Oligocene movement on the East Graben Fault cut the Paleozoic thrusts and dropped the Paleogene volcanic section against lower Paleozoic sedimentary rocks. However, kinematic indicators in the fault zones demonstrate that the faults experienced oblique and strike-slip displacement as well as normal displacement. Two other observations suggest that these faults have a protracted history. Facies changes in some of the Devonian carbonate units across the East Graben Fault suggest that it may have controlled sedimentation patterns in the Paleozoic, as described in the northern Carlin trend (Volk et al., 1995; Emsbo et al., 1999). In addition, the East Graben Fault may have localized the emplacement of Eocene dikes and plugs (Fig. 4). An extended history of activity along these faults suggests that they are major structures, possibly of crustal-scale. These faults may have guided the emplacement and eruption of the middle Tertiary volcanic rocks and provided conduits for hydrothermal fluids. A similar interpretation has been made for north- to northwest-trending structures in the southern Carlin trend (Longo et al., 2002).

The northern Sulphur Spring Range lies on a strong gradient in the Bouguer gravity anomaly. The gradient trends northeast and separates a broad gravity low on the south from higher values to the north (Grauch et al., 2003). The gradient is interpreted to separate Proterozoic continental basement on the east from young accreted terranes on the west. An irregular cluster of aeromagnetic highs marks the northern Sulphur Spring Range; the largest is ∼10 km across and is centered in the valley just east of the volcanic outcrops (Grauch et al., 2003). The magnetic highs are probably due to the higher magnetic susceptibility of the volcanic and subjacent intrusive rocks and may reveal the extent of the shallow intrusive system beneath the volcanic field. The size and amplitude of the anomaly are similar to others along the Carlin trend (Ressel and Henry, 2006).


Elko Formation (Te)

We correlate the oldest Paleogene strata in the eastern Sulphur Spring Range with the Elko Formation. These clastic sedimentary rocks were mapped by Carlisle and Nelson (1990) as the Permian Garden Valley Formation, but the presence of Jurassic quartz porphyry cobbles 02 in the Union tuff member shows that this age assignment is not correct. Instead, the overall lithologic character, including the presence of volcanic rocks, suggests correlation with the Elko Formation, a prominent clastic unit of Eocene age in central Nevada. Elsewhere, the Elko Formation is a fining-upward series of fluviolacustrine beds. The base is primarily red-brown pebble to cobble conglomerate and locally arkosic sandstone, but the middle and upper parts are dominated by fine-grained clastic sediments, lacustrine limestone, and oil shale (Henrici and Haynes, 2006). The conglomerate clasts in the lower member are predominantly reworked Paleozoic chert and quartzite, but also include rare (<1%) porphyritic rhyolite (Smith and Ketner, 1978; Henrici and Haynes, 2006). Locally, pyroclastic fallout tuffs and ignimbrites are interlayered with the sedimentary strata and give ages that range from 46 to 38 Ma.

In the eastern Sulphur Spring Range, the beds we correlate with the Elko Formation crop out east of the East Graben Fault (Fig. 4) and consist of red-brown conglomerate and arkose with pebbles and cobbles of quartzite and green chert. Lava flows and tuff are interlayered with the clastic sedimentary rocks, as described below. The combination of nonvolcanic conglomerate, arkose, and tuff is very similar to the lower member of the Elko Formation as described by Ketner and Alpha (1992) and Henrici and Haynes (2006). The overlying Indian Well Formation is dominated by volcaniclastic sedimentary rocks (Ketner and Alpha, 1992).

Flow-banded rhyolite and latite lava flows (Tbr). The basal part of the Elko Formation in the Sulphur Spring Range includes a crystal-poor rhyolite lava flow with distinctive alternating dark-gray and light-gray to pink flow-layers ∼10 cm thick. Although it is only found in one location in the mapped area (Fig. 4), a very similar rhyolite lava flow crops out ∼11 km to the south. The rhyolite has extremely small phenocrysts of quartz, plagioclase, and magnetite in a glassy matrix containing flow-aligned microlites.

A crystal-poor, black, scoriaceous latitic lava flow crops out near the banded rhyolite. The lavas have phenocrysts of pyroxene, sieved plagioclase, quartz with reaction rims, and corroded K-feldspar, indicative of disequilibrium. In some locations, lobes of latite appear to have flowed down paleo-streambeds and banked against the flow-banded rhyolite.

The flow-banded rhyolite flow is older than the nearby plagioclase dacite lava domes; contacts are not exposed, but it crops out on the western edge of the east-dipping volcanic sequence butted against the East Graben Fault. It is correlative with and older than part of the Elko Formation, as there are outcrops of the conglomeratic facies of the Elko Formation stratigraphically above the banded rhyolite (Fig. 4). The latites are somewhat younger than the rhyolite, but are also within the lower part of the Elko Formation in this area.

Union tuff (Tu). Exposures immediately south and southeast of the Union district contain a poorly welded, orange to maroon, dacite ashflow tuff, herein informally named the Union tuff. Plagioclase and quartz are the most prominent phenocrysts in this dacitic tuff. The tuffaceous interval also includes quartzite cobbles and well-rounded pebbles and cobbles of rhyolite with quartz phenocrysts that are probably the same as the porphyritic rhyolite identified by Henrici and Haynes (2006). The porphyritic clasts contain bipyramidal quartz phenocrysts along with plagioclase, potassium feldspar, and sparse biotite. The U-Pb zircon ages of these clasts are Late Jurassic 02. A porphyritic granitic intrusion very similar to the clasts crops out in the southeastern Sulphur Spring Range ∼40 km away. Clast sizes and the apparent source of the clasts are consistent with paleo-current indicators found by Henrici and Haynes (2006), indicating the source area of the sediments was southeast of the Elko Hills.

The Elko Formation is Eocene to Oligocene according to Smith and Ketner (1978) and middle to late Eocene according to Henrici and Haynes (2006). Radiometric ages on tuffs in the unit range from 46.1 to 38.6 Ma (Solomon et al., 1979; Haynes, 2003; Henrici and Haynes, 2006) and suggest that the Union tuff is Eocene in age.

Biotite Porphyry (Tbp) and Biotite Dacite Tuff (Tbd)

A shallow dike or vent-filling intrusion of porphyritic dacite is found near the East Graben Fault (Fig. 4). It contains abundant phenocrysts of coarse biotite, plagioclase, sanidine, quartz, and altered magnetite. The large (∼2 mm) booklets of biotite are particularly distinctive. Some quartz phenocrysts are resorbed. Phyllic alteration has converted most sanidine to sericite. The medium- to fine-grained matrix consists of intergranular quartz, plagioclase, biotite, and small amounts of glass.

A biotite-bearing ash-flow tuff is the most widespread unit in the mapped area. However, it crops out poorly as porphyritic fragments that litter the ground. Roadcuts provide the best exposures, where the unit is seen to be a non-welded ignimbrite with abundant lithic clasts. In most places, it is argillically altered, whitish-yellow, and laced with small oxidized veins containing pyrite. The tuff contains phenocrysts of plagioclase, coarse biotite (∼2 mm across), quartz, Fe-rich garnet, and megacrysts of quartz. Its major-element composition ranges from dacite to rhyolite (Fig. 7).

The biotite porphyry intrusions and the biotite dacite tuff have indistinguishable ages and elemental compositions and similar textures featuring coarse biotite. U-Pb zircon ages of these units are 35.9 ± 0.5 and 35.4 ± 0.4 Ma, respectively (02 and Fig. 5), and their compositions suggest the two units may be comagmatic (Figs. 7, 8, and 9). As such, the biotite dacite tuff may be an explosive product, whereas the biotite porphyry may form a ventfilling dome or intrusion.

The biotite dacite tuff overlies the Union tuff in the Elko Formation (Fig. 6). Thus, the tuff and units overlying it may be correlative with the Indian Well Formation, an Eocene-Oligocene series of ash-flow tuffs interbedded with volcaniclastic sediments mapped in the valley east of the Pinion Range to the north of the Sulphur Spring Range (Smith and Ketner, 1978). The U-Pb ages are similar to the ages of tuffs within the Indian Well Formation.

Dacite Lava Domes (Tpd)

A group of dacite lava domes with distinctive large plagioclase phenocrysts overlies the biotite dacite tuff. These “plagioclase dacite” domes comprise the majority of the knolls in the eastern portion of the volcanic field (Fig. 4). This is the major Tertiary volcanic unit described and mapped by Carlisle and Nelson (1990).

The dacite domes consist of pink or orange phenocryst-rich, flow-foliated or brecciated lava. A vitrophyre is present in some locations along dome margins. Matrix-supported flow breccias developed along some shear planes. Plagioclase is the primary phenocryst with lesser quartz and clinopyroxene, along with sparse amphibole, biotite, and tiny euhedra of magnetite as inclusions and in the groundmass. Chlorite, clay minerals, and iron-stains are characteristic of altered rocks.

The U-Pb zircon age of 35.1 ± 0.5 Ma obtained from one dome (Fig. 5) is consistent with field relations showing that the dacite lavas erupted onto or through older biotite dacite tuff.

Basaltic Andesite Dikes (Tba)

The most mafic of the Paleogene igneous rocks found in the East Sulphur Spring suite are dikes that generally trend north-northeast (Fig. 4). The dikes cut the biotite dacite tuff (Tbd) as well as the dacite domes (Tpd). Most of the dikes are dense, black porphyritic basaltic andesite containing phenocrysts of clinopyroxene, orthopyroxene, olivine (Fo84), and sparse plagioclase. The matrix is composed of plagioclase, pyroxene, Fe-Ti oxides, and minor glass. A distinct trachytic flow foliation is defined by elongate plagioclase microphenocrysts. Dikes with shoshonite and biotite-bearing latite compositions are rare (Fig. 7). Whole-rock compositions range from basaltic andesite with MgO as high as 13 wt% to as low as 3 wt% in the latite; Cr concentrations follow suit ranging from nearly 500 to 15 ppm 01.

Andesite Dikes and Lava Flows (Ta)

The youngest volcanic unit in the East Sulphur Spring suite is a series of crystal-rich andesite dikes and near-vent flows with unique textural aspects. The exposed dikes trend NNE, like most other dikes in the northern Sulphur Spring Range. Rounded, sieve-textured plagioclase is the dominant phenocryst. Other prominent phenocrysts include biotite, clinopyroxene, orthopyroxene, olivine (Fo78), and magnetite. These mafic phenocrysts coexist with megacrysts of sanidine and unstrained quartz as much as 3.5 cm long. Smaller quartz phenocrysts are extensively resorbed, and many of these crystals also have reaction rims of clinopyroxene (Fig. 10). The andesite also contains grains of resorbed and oxidized garnet. The garnet is Mn-rich compared to that found in the biotite dacite (Fig. 11). Dike and flow margins have perlitic glass in their matrices.

We interpret the mineral assemblage (e.g., forsteritic olivine coexisting with quartz), the resorption of the felsic phases, and the reaction rims to be the result of mixing silicic magma (containing quartz and sanidine megacrysts) and mafic magma (containing olivine and pyroxene). Clinopyroxene halos around resorbed quartz are also diagnostic of magma mixing (Fig. 10; Coombs and Gardner, 2004).

We interpret the U-Pb zircon data to show that the andesite has an Oligocene eruptive age of ca. 32 Ma, although the error and the MSWD are rather large 02. The 2 sigma uncertainty does not overlap with the ages of the other igneous rocks in the East Sulphur Spring suite (Fig. 12). Stratigraphic relationships also suggest that this is the youngest exposed unit; it appears to overlie the dacite domes (Tpd) because it outcrops farthest to the east in this east-dipping sequence of strata (Fig. 4). The large spread in U-Pb ages of zircon xenocrysts may show that some of the crystal cargo was derived from already solid rock by assimilation of the Paleoproterozoic (2.5 and 1.7 Ga) basement 02.

Mafic Dikes (Tmd)

In addition to the dikes in the East Sulphur Spring suite, a series of weakly to extremely altered dikes crops out west and northwest of the mapped area (Fig. 1). The overwhelming majority of the dikes are basaltic andesite, but sho-shonite and latite have also been found (Fig. 7). The dikes occur largely in two east-northeast- trending swarms that transect the northern Sulphur Spring Range where they cut pre-Cenozoic rocks. Most of the dikes are altered to argillic or phyllic mineral assemblages and crop out as light-orange to tan fragments. Several propyliti-cally altered dikes preserve phenocrysts of olivine and clinopyroxene.

The ages of these dikes are uncertain because they do not cut any of the Paleogene units. However, the elemental compositions and mineralogy of the freshest rocks show that these dikes are not like the Jurassic lamprophyres in the region. The basaltic andesites from the Sulphur Spring Range have lower alkalis, P, Rb, Sr, and Ba; their light REE (LREE) patterns are also different than those of the Jurassic lamprophyres (Fig. 9). The basaltic andesites also lack mica and amphibole phenocrysts, but they have plagioclase phenocrysts. Moreover, most of the Jurassic lamprophyres along the Carlin trend strike NW, not NE to ENE like most of the Paleogene dikes. Even the latite samples are unlike the Carlin lamprophyres; they have significantly lower MgO, CaO, Cr, Ni, Sr, La, and Ce and higher SiO2, Al2O=, K2O, Rb, Nb, and Ba than the Jurassic lamprophyres described by Ressel and Henry (2006). These differences suggest that the latites are more fractionated rocks and that they have different parentage than the Jurassic lamprophyres. The Sulphur Spring dikes are also unlike Neogene basalt flows that crop out to the south—the dikes are magnesian and Nb concentrations than and have lower TiO2 the young basalts 01. These mafic dikes are similar to the Paleogene basaltic andesite dikes on the east side of the range in their orientations, major- and trace-element compositions, mineral assemblages, and alteration styles. Consequently, we presume that they are Paleogene in age.

The altered mafic dikes have anomalous concentrations of As, Ba, Cr, Cu, Ni, Pb, Sb, and Zn. These dikes are believed to lie within the hydrothermally altered outer carapace of the buried Eocene intrusive system. The high concentrations of Cr are especially noteworthy. Cr concentrations range from 60 to 630 ppm with an average of 355 ppm. Since Cr is relatively immobile during hydrothermal alteration, these high concentrations show that even the highly altered dikes were mafic.

Mineralization: Gossan Veins and Jasperoid

A distinctive set of veins is found within and surrounding the core of the northern Sulphur Spring Range and includes those in the Union Pass and Mineral Hill districts (Fig. 1). Their compositions show that large quantities of S, Au, Ag, As, Cu, Mo, Pb, Sb, and Zn were deposited by hydrothermal fluids. Most veins are oxidized, sulfide-rich quartz veins or gossans. Veins range in width from 1 to 100 cm, are often banded, and have been brecciated and then healed by younger deposits of silica; some veins contain unoxidized fragments of sulfide minerals, including pyrite or galena. These are highly sulfidic polymetallic veins (now completely oxidized in most cases), and might be equivalent to “D” veins in the “alpha” system originally established for the El Salvador deposit by Gustafson and Hunt (1975). The lack of sericite-rich margins is due to their formation in carbonate rocks.

The dominant trend of the gossan veins is northeast, which is the same as the dated Paleogene, and presumed Paleogene, mafic dikes described above. Their relationship to the Paleogene magmatism is not certain because they are not directly adjacent to any dikes, but because the gossan veins are found in the same part of the range as the dikes and share a common orientation, we speculate that they are distal to but contemporaneous with the major intrusive cluster.

The east side of the Sulphur Spring Range also has a number of small bodies of jasperoid, particularly along the contact between carbonate and clastic rocks with silty carbonates and calcareous shales. These silicified bodies have anomalously high concentrations of Au, As, Hg, Sb, and Tl and are very similar to those found in Carlin-type systems (e.g., Wilson et al., 1994). The Paleozoic host rocks are complexly faulted, and the mineralization is spatially associated with the Eocene igneous rocks. These structural, geochemical, and stratigraphic characteristics are similar to those of Carlin-type gold deposits elsewhere in Nevada. (Indeed, Carlin-style alteration hosted in calcareous shales and carbonates has been intercepted in a drill hole just north of the mapped area, but the results of these investigations are still proprietary.)

During its ∼4 million year lifetime, the Eocene-Oligocene magma system beneath the northern Sulphur Spring Range erupted basaltic andesite to rhyolite as lava flows, domes, and small pyroclastic deposits, apparently fed by a series of dikes. Dacite is the dominant erupted volume. Much of the volcanic section is interlayered with contemporary sedimentary deposits. Modal and whole-rock chemical compositions of the igneous rocks provide some general insights into the nature and evolution of the magma system.

Subduction Zone Origin

The Eocene-Oligocene volcanic rocks of the Sulphur Spring Range have much in common with other subduction-related continental margin suites (Fig. 7). They are largely calc-alkalic using the classification of Frost et al. (2001). Only some of the basaltic andesite dikes and the biotite dacite tuff are calcic. The suite is overwhelmingly magnesian (using the dividing line of Miyashiro [1974] and the terminology of Frost et al. [2001]); this lack of Fe-enrichment is characteristic of crystallization at relatively high f O2.

Likewise, even the most primitive basaltic andesite dikes have relatively low TiO2, less than ∼1.4 wt%, similar to volcanic rocks found in arcs. Most of the volcanic units form a high-K series, but the hybridized andesite and some of the intermediate composition dikes range widely from medium K2O to shoshonitic (Fig. 7). REE patterns of the rocks in the East Sulphur Spring suite are similar to those of igneous rocks from continental margin subduction zones (Ewart, 1979) with relatively steep slopes and small negative Eu anomalies (Fig. 9). The latites have the highest REE concentrations, and the most silicic rhyolite has lower LREE concentrations and a larger Eu anomaly. On the tectonic discrimination diagrams of Pearce et al. (1984), which are based on Rb, Nb, and Y abundances, the rhyolite lavas and more voluminous dacite lavas and tuffs are similar to subduction-related volcanic arc granites (Fig. 13). Based on Zr-Ti-Ce-P systematics, mafic rocks in the Sulphur Spring suite are also of a continental arc-type (Fig. 13). Finally, like volcanic rocks erupted in other subduction settings, all of the rocks have negative Nb and Ti anomalies and positive anomalies for Pb on primitive-mantle normalized diagrams (Fig. 8). We conclude that the East Sulphur Spring suite is related to subduction and that the parental magmas were hydrous and oxidized. In addition, arc magmas like these are typically rich in S and chalcophile metals (e.g., Richards, 2003).

Crustal Melting and Felsic Peraluminous Magmas

An important characteristic of the felsic members of the East Sulphur Spring suite is the inclusion of peraluminous silicic rocks. Two of the volcanic units contain Fe-rich magmatic garnet—a mineralogical indicator of the excess of Al over Ca + Na + K. The andesite lavas (Ta) have rimmed crystals of Mn-rich garnet, and the biotite dacite tuff (Tbd) has sparse Mn-poor grains of euhedral garnet. Most of the individual samples of the tuff are peraluminous with the alumina saturation index ranging from 0.95 to over 1.1. Although peraluminous silicic rocks in the Great Basin are typically Cretaceous in age (Miller and Bradfish, 1980; Lee and Christiansen, 1983), Eocene granites (ca. 39–34 Ma) in the nearby Ruby Mountains are strongly per-aluminous and locally contain garnet (Kistler et al., 1981; Barnes et al., 2001), as do Eocene rhyolites at Mount Hope (Westra and Riedell, 1995) to the south (Fig. 1). Strongly peraluminous silicic magmas like these are most likely generated as a result of partial melting of pelitic or semipelitic metasedimentary rocks in thickened continental crust. They show that many of the Paleogene magma systems of north-central Nevada include large proportions of crustal magma. This implies that sufficient mantle-derived magma was inserted into the crust to induce melting by breakdown of hydrous minerals. Some of these crustal melts were probably assimilated in the more mafic magmas as well.

Fractional Crystallization

Fractional crystallization appears to have been an important differentiation process for both felsic and mafic magmas in the East Sulphur Spring suite. Even though the alkalis have been perturbed by slight alteration, it is apparent that the biotite dacite tuff (unit Tbd) ranges from dacite to rhyolite in composition (Fig. 7). Its chemical variation is consistent with fractional crystallization of the observed phases—feldspars and mafic silicates and oxides. As SiO2 increases, Al2O3, CaO, TiO2, Fe2O3, and MgO all decline in concentration. Increases in the incompatible elements (Rb, Nb, Pb, and Th) suggest that 25% to 35% fractional crystallization could have created these changes. REE concentrations change little or decrease, perhaps as a result of the removal of garnet.

The intermediate composition dikes (Tba) range in composition from olivine-bearing basaltic andesite (with as much as 13% MgO, 175 ppm Ni, and 500 ppm Cr) to latite (with 2% MgO, 20 ppm Ni, and 15 ppm Cr). These steep declines in compatible element concentrations (Fig. 14) suggest fractional crystallization of mafic mineral phases that have high partition coefficients for these elements. Other compatible elements, such as Ti, Fe, Mg, Ca, Sc, and V, increases. On decrease in concentration as SiO2 the other hand, incompatible elements like P, Zr, LREE, Pb, Rb, and Ba increase from the basaltic andesites to the latites. The curved trends and sharp decreases of compatible element concentrations are not typical of magma mixing, which produces linear trends on two-element variation diagrams. Consequently, we conclude that the basaltic andesite, shoshonite, and latite dikes are probably related to one another by fractional crystallization of pyroxene, olivine, plagioclase, and oxides. The extent of fractionation from parental basaltic andesite to derivative latite was ∼50% based on the enrichments of incompatible elements (Nb, Zr, Pb, Th, and Rb).

Magma Mixing

Even though fractional crystallization seems to have been the dominant process leading to latite, a few of the basaltic andesite dikes have trace-element compositions that are consistent with mixing of mafic and silicic magmas. Anomalously high concentrations of Cr and Ni are found in at least two dikes (Fig. 14). However, this remains a tentative conclusion because it is difficult to distinguish fractionation from mixing using other elements because compositional trends produced by both processes are linear and overlapping. In fact, Ba variations do not clearly show the effects of mixing basaltic andesite and rhyolite, which would pull hybridized magma off the fractionation trend to lower concentrations of Ba (Fig. 14).

A less ambiguous example of magma mixing is found in the andesite flows and dikes (Ta). They form a tight compositional array on most variation diagrams. Their high K2O contents render them the most consistently shoshonitic unit in the area (Fig. 7). On the other hand, the andesites have lower concentrations of Al2O3, Zr, and Ba (Figs. 14 and 15). Their anomalous positions and linear trends on these variation diagrams suggest that they formed by mixing of mafic and silicic magmas. The disequilibrium mineral assemblage, which includes forsteritic olivine and quartz, is compelling evidence for magma mixing. This is substantiated by reaction rims, and extensive resorption of phases that may have come from the silicic magma—quartz, sanidine, and Mn-rich garnet. The silicic end member must have been a high-silica rhyolite to explain the presence of quartz, sanidine, and Mn-rich garnet. (Garnet in the older dacite tuff is not as rich in Mn [Fig. 11].) In addition, mixing of mafic magma with a highly evolved rhyolite could explain the low Zr, Ba, and Sr in the andesite compared to other intermediate composition rocks from the volcanic field (Fig. 15). Although peraluminous, garnet-bearing silicic magmas are not common in the Great Basin, it is clear that such magmas were generated during the Eocene as noted above. Identifying the mafic end member is more problematic, but mineral assemblages, olivine compositions, and trends on most silica variation diagrams (Ti, Al, Fe, Mg, Ca, Na, and K) suggest that it was a member of the basaltic andesite to latite suite described above. Elemental trends require that the mafic end member was moder- (>0.5%) and low ately evolved, with high P2O5 Cr (<50 ppm) and Ni (<25 ppm). Based on these end members, major-element mass balance calculations show the proportion of rhyolite in the hybridized andesite ranged from ∼15% to 35% by weight.

The potential for mineralization, styles of alteration, ages, and overall compositional characteristics of the Paleogene volcanic rocks in the Sulphur Spring Range invite comparison with other volcanic suites associated with mineralization in the Great Basin. Below, we compare the East Sulphur Spring suite with Eo-Oligocene volcanic rocks at the Bingham Canyon Cu-Mo-Au porphyry deposit farther east in Utah and then to magmatic rocks associated with gold deposits in the Carlin trend of northern Nevada.

Comparison with Bingham Canyon Volcanic Rocks

The igneous rocks, structures, and alteration styles of the Sulphur Spring Range are somewhat similar to those of porphyry copper systems, such as that at the giant Bingham Canyon deposit in the eastern Great Basin (Fig. 3). Dike swarms, vent-facies volcanic rocks, pebble dikes, local bleaching and marbleization of the carbonate rocks, and oxidized sulfidic veins (gossan veins) straddle and surround an aeromagnetic anomaly. Collectively these features cover an area of at least 5 km × 7 km, comparable in size to that associated with the Bingham deposit, and suggest that the northern Sulphur Spring Range overlies a shallow intrusive center. Similar tectonic regimes and structural histories also shaped the features of the hydrothermal and magmatic systems. Both are near paleocontinental margins. The Sulphur Spring Range (and the Carlin trend) are near an accretionary boundary on the western edge of the Proterozoic basement in central Nevada (e.g., Emsbo et al., 2006) marked by the 87Sr/86Sr = 0.706 line and the Paleozoicage Roberts Mountains and Golconda thrusts (Fig. 2). Bingham lies near a hypothetical suture between Archean and Proterozoic terranes (e.g., Whitmeyer and Karlstrom, 2007), and Mesozoic thrust faults related to the Sevier Orogeny are cut by the Bingham intrusions. Thus, deep, crust-penetrating faults may have formed anciently in both areas. On the other hand, igneous rocks in the Sulphur Spring Range are ca. 2–3 Ma younger than (Fig. 12) those associated with the Bingham porphyry copper deposit.

Both volcanic suites have compositions that are generally consistent with a subduction zone origin—low Fe/Mg ratios, high oxygen fugacities, high K2O, and similar “spiky” trace-element patterns, for example. Apparently, both regions were affected by the same Paleogene detachment or roll back of the Farallon plate, which contributed to continental arc magmatism over a wide area of western North America. These magmas appear to have played a central role in the mineralization in both areas. Fractionation from basaltic andesite to shoshonite and latite also occurred in both suites, although latite is rare and found only in a small lava flow and a few dikes in the East Sulphur Spring suite (Fig. 16). A similar fractionation trend is also important in the Eocene Absaroka volcanic field of Wyoming and Montana (Bray, 1999; Feeley and Cosca, 2003) and appears to be a common process in these continental interior magma systems.

In spite of these similarities, the major- and trace-element compositions of the East Sulphur Spring suite define arrays that are distinct from those of the Bingham volcanic suite (Figs. 14 and 16). The Bingham suite has high total alkali contents and includes melanephelinite, minette, shoshonite, latite, trachyte, and rhyolite (Maughan et al., 2002), whereas the Paleogene East Sulphur Spring suite is dominated by basaltic andesite, andesite, dacite, and rhyolite; latite is rare as noted above (Fig. 16). Silica-under-saturated magmas are unknown in the East Sulphur Spring suite, but melanephelinite intruded as dikes and stocks and erupted as lava flows at Bingham and appears to have been an important source of S, Cu, and Au in the deposits.

Characteristically high Cr and Ni contents of the intermediate composition (58%–65% SiO2) volcanic rocks from Bingham (Maughan et al., 2002) are not found in the unaltered rocks of the East Sulphur Spring suite (Fig. 14). The volcanic suite at Bingham has elevated concentrations of Cr and Ni across the full range of silica content. Only the least silicic rocks from the Sulphur Spring Range contain high Cr and Ni contents and their concentrations decrease to very low values (>50 ppm) in the more evolved andesites and dacites. For example, dacites from the Bingham have as much as volcanic center with 65% SiO2 200 ppm Cr and 75 ppm Ni. Dacites from the Sulphur Spring Range (and from the Carlin trend) have less than 40 ppm Cr and 20 ppm Ni. The low Cr and Ni concentrations are typical of many continental calc-alkaline suites; the high Cr and Ni in the Bingham volcanic suite is a fairly unique feature. Another distinguishing characteristic of the Bingham magmas is the very high Ba concentration (Fig. 14). In the East Sulphur Spring suite, only the fractionated latite dikes have Ba contents (>2500 ppm) approaching, but not equaling, the maximum Ba contents of the volcanic rocks at Bingham. Moreover, the Ba content of the Bingham suite is high across the silica range, but is the highest in the mafic and intermediate composition rocks (<65 wt% SiO2). In the East Sulphur Spring suite, Ba increases as SiO2 increases, whereas in the Bingham suite Ba generally decreases as SiO2 increases.

Both the Bingham and the East Sulphur Spring suites preserve petrographic and chemical evidence of mafic magma having mixed with silicic magma to create intermediate compositions. At Bingham, the high concentrations of Cr, Ni, and Ba in the intermediate composition volcanic and intrusive rocks have been traced to mixing with a mafic alkaline magma—silica-undersaturated melanephelinite (or minette) with high concentrations of Ni, Cr, Ba, Cu, Au, and S. Altered olivine and pyroxene rimmed by amphibole in intermediate composition rocks at Bingham also demonstrate mixing of mafic alkaline magma with more silicic magma. Other evidence includes dacite clasts in block and ash flows that contain cuspate mafic clots, large resorbed potassium feldspar phenocrysts in silicic rocks with elevated Cr and Ni concentrations, and adjacent, same-aged minette and quartz latite dikes (Pulsipher, 2000; Maughan et al., 2002). The Sulphur Spring unit that most vividly expresses magma mixing is the andesite (unit Ta), with its disequilibrium mineral assemblage and anomalous elemental composition. More subtle chemical evidence of mixing is found in a few dikes in the basaltic andesite suite (Tba) as well.

In spite of the evidence for magma mixing, in the East Sulphur Spring suite, the mafic component was not silica-undersaturated, nor was it especially rich in Ni, Cr, or Ba compared to the melanephelinite and minette at Bingham. Instead, the “mafic” component must have been a moderately evolved (>55% SiO2) basaltic andesite or shoshonite, similar to those found in a few dikes. As a consequence, the igneous rocks of the East Sulphur Spring Range have lower concentrations of Ni, Cr, Cu, and probably S and Au (which follow Cu concentrations) than the intermediate composition stocks and lavas associated with the Bingham intrusion.

Comparison with Igneous Rocks along the Carlin Trend

The setting of the Sulphur Spring Range is similar to that of nearby, intensely mineralized localities to the north along the Carlin trend. The Roberts Mountains thrust is an important element of mineralization along the Carlin trend; it is also exposed in the northern Sulphur Spring Range where it juxtaposes Ordovician strata on Devonian carbonate rocks and is in turn cut by high-angle, N-S–, NW-SSE–, NE-, and E-W–trending normal faults (Carlisle and Nelson, 1990). These structural and stratigraphic elements appear to be conducive to genesis of Carlin-type gold deposits (Hofstra and Cline, 2000; Grauch et al., 2003; Cline et al., 2005; Ressel and Henry, 2006). In addition, we have noted the presence of jasperoid and altered rocks with anomalous concentrations of As, Hg, Sb, and Tl, which are also features of mineralization along the Carlin trend.

Ages, as revealed by new U-Pb data and stratigraphic correlations, show that the magmatic history of the northern Sulphur Spring Range is also similar to that along the main Carlin trend as outlined by Ressel and Henry (2006). They identified Jurassic plutonic rocks (158 Ma), Cretaceous granite (112 Ma), abundant Eocene dikes, lavas, tuffs, and stocks (40–36 Ma) thought to be directly related to Au mineralization, and Miocene rhyolite (15 Ma). Inherited and primary zircons from the igneous rocks in the Sulphur Spring Range record each of these episodes. Clasts in the Elko Formation are derived from a Late Jurassic granitic rock; inherited zircon in the andesite lava has a Cretaceous age; most of the volcanic rocks and dikes are Paleogene in age but slightly younger (36–32 Ma) than those to the north along the Carlin trend (40–36 Ma; Fig. 12); and finally, a rhyolite pyroclastic deposit has a Miocene age.

Igneous rocks in the Sulphur Spring Range are compositionally and genetically similar to those associated with the Carlin trend, which we take to include the Emigrant Pass and Tuscarora volcanic fields (Henry et al., 1999; Ressel and Henry, 2006). For example, both suites consist of subalkaline basaltic andesite to rhyolite, and include only minor latite and trachyte (Fig. 16). Like those from the Sulphur Spring area, the Carlin-trend rocks are dominantly calc-alkalic, magnesian, and high-K. REE patterns of the Sulphur Spring suite are also similar to those of igneous rocks along the Carlin trend (Ressel and Henry, 2006), which also have relatively steep slopes and small negative Eu anomalies (Fig. 9). In other words, both regions erupted magmas with subduction zone characteristics—low Fe/Mg ratios, oxidized, hydrous, and high large-ion lithophile to high-field-strength element (LIL/HFSE) ratios—produced during Paleogene rollback of the Farallon slab. Reduced, ilmenite-dominated silicic magmas are probably rare in the region, but the strongly peraluminous, muscovite- and garnet-bearing Harrison Pass intrusion in the Ruby Range is one example (Barnes et al., 2001).

The igneous rocks of the Carlin trend also have concentrations of key trace elements—Cr, Ni, and Ba—much more similar to those of the Sulphur Spring suite than to those at Bingham (Fig. 14; Ressel and Henry, 2006; C.D. Henry, 2005, written commun.). Low concentrations of Cr and Ni in the intermediate magmas and the association with peraluminous magmas imply that fractional crystallization and assimilation of pelitic crustal materials were the predominant magmatic processes in both areas (Fig. 14). Magma mixing played a lesser role and did not involve mafic alkaline magma as an end member. Rather, it involved mixing of basaltic andesite or andesite with rhyolite. Mixing of compositionally similar magmas along the fractionation trends is not ruled out and is in fact quite likely.

The only differences between the East Sulphur Spring volcanic suite and the magmatic rocks along the Carlin trend that we have identified are that the volcanism in the Sulphur Spring Range is a few million years younger (Fig. 12). Some of the intermediate magmas in the East Sulphur Spring suite have higher K2O contents and are more shoshonitic, if the handful of potassic samples from Sulphur Spring is compared with the small set of “representative” analyses of Carlin-trend igneous rocks published by Ressel and Henry (2006). However, a larger compilation of igneous rock compositions from north-central Nevada (C.D. Henry, 2005, written commun.) has a few Eocene latites, but no shoshonites. There are also a few potassic Eocene plutonic rocks from northeastern Nevada in du Bray's (2007) database (eight of 643 samples after removing a few obviously altered rocks) in the 50% to 65% range, and three of those are lamprophyre dikes from the Fish Canyon Range whose age is not clearly established as Eocene. Thus, it appears that potassic igneous rocks of Eocene age are sparsely found across northern Nevada.

Given the similarities in structure, stratigraphy, alteration styles, geochemical anomalies, magmatic history, and petrology of the Eo-Oligocene igneous rocks, it is reasonable to infer that economic Carlin-type mineralization might exist in the Sulphur Spring Range.

Potential Importance of Mafic Magma and Magma Mixing for Mineralization

Intermediate and felsic rocks, such as those that typically host porphyry Cu-Au systems, have low concentrations of sulfur and of chalcophile ore metals compared to more mafic magmas. Sulfur concentrations are limited by the solubility of sulfur in felsic magmas, which decrease as silica concentrations increase causing sulfide and sulfate minerals to precipitate. A sulfur-saturated andesite with 60% SiO2 may have less than 200 ppm S, whereas mantle-derived basalt may have over 1000 ppm S (Liu et al., 2007). Gold and copper are also enriched in mafic magmas compared to silicic magmas; copper concentrations in alkaline mafic magmas can be as much as 120 ppm and in rhyolite concentrations are only a few ppm (e.g., Maughan et al., 2002). In fresh igneous rocks of the East Sulphur Spring suite, Cu concentrations are as much as 50 ppm in the mafic dikes and basaltic andesite dikes, but typically less than 20 ppm in the andesites and more silicic rocks. Gold concentrations are likely to be on the order of 1.7–0.5 ppb based on a copper:gold ratio of ∼30,000 (Rudnick and Gao, 2004). Copper and gold behave as compatible elements—in differentiated magma systems their concentrations usually covary with strongly compatible elements like Ni and Cr. Immiscible sulfide melts and minerals and magnetite have high partition coefficients for chalcophile elements like gold and copper and, once evolving magmas become sulfide-saturated, the residual melt becomes strongly depleted in these elements (Jugo et al., 1999: Simon et al., 2003). Thus, the potential for the development of a porphyry Cu-Au system hosted by felsic intrusions may be strongly linked to the extent of mixing with mafic magma to yield higher than “normal” concentrations of ore metals and sulfur. On the other hand, disseminated Carlin-type gold deposits are typically hosted by sedimentary rocks and not by intermediate composition igneous stocks. In most deposits, the S, O, H, and C isotopic compositions are closely linked to sediments and meteoric fluids (e.g., Cline et al., 2005). Therefore, if elements from mafic magmas are important to their generation, mixing may not be as critical as the mere presence of mafic magma that in the process of crystallizing and “degassing” can give off significant sulfur and gold to upper crustal hydrothermal systems.

Waite et al. (1997), Hattori and Keith (2001), and Maughan et al. (2002) concluded that porphyry and Carlin-like mineralization in the Bingham district could not have formed without involvement of mafic alkaline magmas rich in S, Cu, and Au. Intermediate “calc-alkaline” magmas at Bingham were simply too poor in these elements for any reasonable volume of magma to have served as a source of metal in the deposits. Although only small amounts of mafic rock are known near Bingham, evidence of magma mixing involving alkaline olivine-bearing magmas is widespread. Apparently during the Eocene, mafic alkaline magmas intercepted and mixed with evolved calc-alkaline magmas in shallow subvolcanic settings. This mafic alkaline magma was richer in compatible ore elements (e.g., Cu and Au) and was probably the main source of the ore metals and sulfur.

In light of these conclusions, the obvious question is: Did mafic magmas play such a role in the Sulphur Spring Range or for that matter along the Carlin trend?

While there is no direct evidence for the involvement of mafic magma in mineralization along the Carlin trend (e.g., Cline et al., 2005; Emsbo et al., 2006), the presence of olivine-bearing basaltic andesites along the Carlin trend (Ressel and Henry, 2006) and in the Sulphur Spring Range shows that mafic magma was present. MgO contents as high as 13% have been found in dikes from the Sulphur Spring Range. The mafic component of the andesite of the Sulphur Spring Range is also clear evidence that mafic magmas were involved in this magma system. These mafic magmas form coherent compositional trends relating them to the rest of the volcanic rocks, demonstrating that mafic mantle-derived magmas were at the “roots” of these Eocene magma systems, as in all subduction related magmatic arcs. Finally, the gold deposits and the igneous rocks have similar ages.

Other geologists emphasize the role of Paleogene magmas as simple heat sources to drive fluid flow (e.g., Tosdal, 1998). However, the close spatial association of the basaltic andesite dikes in the Sulphur Spring Range with mineralized veins suggests an even closer genetic link between the magmas and ore deposits. These dikes are not volumetrically significant at the current level of exposure, but they may have dominated at deeper levels of the Eo-Oligocene magma system. Many of the dikes and lava flows are Cu-, Cr-, Ni-, and MgO-rich, indicating little fractionation occurred since they left their mantle sources (01; Fig. 14). Consequently, if they are like other oxidized mafic magmas, they would have been enriched in Au and S compared to more silicic magmas. “Degassing” of mafic magma in dikes and volcanic conduits or as underplated and mixed magma may have contributed large quantities of sulfur, fluids, and metals to the ore-forming systems, in addition to any heat they would have released to drive convecting fluids. This concept has been proposed for multiple magma-ore systems, including porphyry Cu deposits at Santa Rita in New Mexico (Audetat and Pettke, 2006), Questa Mo deposit in New Mexico, Nukay Au-Cu deposit in Mexico, Las Bambas porphyry Cu deposit in Peru (Jones, 2002), the Farallón Negro Cu-Au deposit in Argentina (Halter et al., 2005), as well as the Bingham porphyry copper system emphasized here. The presence of mafic volcanic and/or dike rocks (Cr- and MgO-rich dikes) in the Sulphur Spring Range and along the Carlin trend during the Paleogene is permissive evidence for the operation of this process here as well. These mafic rocks, with their higher concentrations of compatible elements (including chalcophile elements), could have been an additional source of ore metals and fluids in the deposits. At most Carlin-type deposits, the isotopic heritage of the magmatic components may have been overwhelmed by later meteoric water interaction (Cline et al., 2005). In any case, such mafic magmas would be much better sources of Au and S than the andesites and dacites that dominate the eruptive record. Thus, from the evidence found in the northern Sulphur Spring Range and that previously published for the Carlin trend, we concur with the hypothesis that Paleogene magmas may have served as important sources of S and Au (Cline et al., 2005; Ressel and Henry, 2006) in addition to the isotopically identified crustal sources of these elements (e.g., Arehart et al., 1993; Hofstra and Cline, 2000; Emsbo et al., 2006).

A suite of Eo-Oligocene lava flows, domes, and pyroclastic rocks is interlayered with clastic sediments and cut by dikes in the northern Sulphur Spring Range of central Nevada. These rocks form a dominantly high-K, calc-alkalic suite, with low Fe/Mg ratios similar to those found in subduction settings worldwide. The volcanic and subvolcanic rocks range from olivine-bearing, high-MgO basaltic andesite to garnet-bearing dacite, and high-silica rhyolite. The intermediate to silicic rocks have spiky trace-element patterns with Nb-Ti depletions and enrichment of Pb, also similar to those formed at convergent margins. The oldest Paleogene volcanic rocks are rhyolite and latite lava flows interlayered with fluvial conglomerates and dacitic tuff that probably correlate with the Eocene Elko Formation. This interpretation is based on lithologic similarities and on new U-Pb zircon ages of overlying units, which show the basal unit is probably middle Eocene in age. A porphyritic dacite intrusion (35.9 ± 0.5 Ma) and probably cogenetic rhyolite to dacite tuff (35.5 ± 0.4 Ma) overlie this succession. The tuff is per-aluminous and has sparse phenocrysts of garnet. A series of dacite lava domes (35.1 ± 0.5 Ma) caps the Eocene sequence. The youngest volcanic unit is an extensively hybridized Oligocene andesite (31.4 ± 0.5 Ma). It has a disequilibrium phenocryst assemblage of plagioclase, biotite, clinopyroxene, orthopyroxene, amphibole, and olivine, along with resorbed megacrysts of quartz, potassium feldspar, and garnet. Linear trends on variation diagrams are consistent with mixing between intermediate composition magma and peraluminous rhyolitic magma to form the andesite. The dated units probably correlate in time with the Indian Well Formation of Eo-Oligocene age that is mapped in adjacent areas. Dikes (dominantly of basaltic andesite) cut this unit and trend NNE. We correlate these mafic dikes, which cut Paleogene volcanic and sedimentary rocks, with dikes of similar orientation and composition that cut only Paleozoic rocks in the central part of the range. These mafic dikes are also dominated by basaltic andesite but include shoshonite and latite as well. The dikes are parallel to a distinctive set of oxidized quartz-sulfide veins in the central part of the range. The veins have anomalous concentrations of Au, Ag, As, Cu, Mo, Pb, Sb, and Zn and surround two historically mined polymetallic vein deposits. Jasperoid bodies on the margins of the range also have Carlin-like trace-element signatures.

These characteristics suggest the magma system was rooted in a subduction zone that formed as the Farallon plate steepened during the Paleogene (Fig. 17). As hot asthenosphere came in contact with the slab, dehydration of the plate produced an oxidized aqueous fluid enriched in S, Cu, Au, Pb, and other soluble elements. The fluid lowered the melting temperature of the overlying mantle wedge, generating distinctly arclike magmas as a result of hydrous partial melting. This hot, hydrous, oxidized magma rose buoyantly into the crust, stagnated because of density differences, and promoted partial melting of the continental crust to form peraluminous silicic magma. These disparate mantle- and crust-derived magmas mixed, rose into shallow chambers, differentiated by fractional crystallization, and eventually erupted or filled fractures to form dikes. As a consequence of crustal trapping, only a small fraction of the mafic magma was able to erupt. Structural boundaries (including Proterozoic-age basement-penetrating faults, Paleozoic and Mesozoic thrust faults, and magma-generated fractures) guided magma emplacement routes, controlled levels of stagnation, promoted hydrothermal fluid flow, and placed reactive wall rocks in the flow paths. “Degassing” of mafic magma may have contributed sulfur and chalcophile metals to the mineralized veins and small ore deposits.

The Eo-Oligocene igneous rocks of the East Sulphur Spring suite are compositionally akin to Eocene igneous rocks associated with large gold deposits in the Carlin trend. When other similarities in structure, stratigraphy, and alteration are considered, we conclude that the Sulphur Spring Range is prospective for Carlin-type gold deposits. Johnston and Ressel (2004) suggested that porphyry Cu-Au deposits, Carlin-type deposits, and distal disseminated deposits are all part of a continuum with differences depending mostly on spatial relations to the magmatic hydrothermal system. We agree and suggest that, like Sulphur Spring and Bingham, Carlin-related Eocene magmatic systems include relatively mafic magmas that are vital to generating the large quantities of gold in the ore deposits. We readily acknowledge that the mere presence of mafic magmatic rocks does not demonstrate conclusively that they contributed to the mineralization, but combined with the close spatial and temporal association of mafic lava flows, mafic dikes, and mineralized veins in the Sulphur Spring Range (and in the broader Carlin trend), we suggest that this is a viable hypothesis that merits further investigation.

If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00113.S1 (Table S1) or the full-text article on www.gsajournals.org to view Supplemental Table S1.

If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00113.S2 (Table S2) or the full-text article on www.gsajournals.org to view Supplemental Table S2.

*Corresponding author

We are grateful for the assistance of Greg Melton in the field and Michael Dorais in the electron micro-probe laboratory. The comments of W. Bagby and the assistance of C.D. Henry are also appreciated. The thorough reviews of Edward du Bray and John Dilles and the editorial assistance of Albert Hofstra were vital. Their careful examination of our manuscript focused our ideas on the essentials and tremendously improved the presentation. The research was supported by funds from Brigham Young University and Golden Gryphon Explorations.