Abstract

Processes linked to shallow subduction, slab rollback, and extension are recorded in the whole-rock major-, trace-element, and Sr, Nd, and Pb isotopic compositions of mafic magmatic rocks in both time and space over southwestern United States. Eocene to Mio-Pliocene volcanic rocks were sampled along a transect across the west-central Great Basin (GB) in Nevada to the Ancestral Cascade Arc (ACA) in the northern Sierra Nevada, California (∼39°–40° latitude), which are interpreted to represent a critical segment of a magmatic sweep that occurred as a result of subduction from east-northeast convergence between the Farallon and North American plates and extension related to the change from a convergent to a transform margin along the western edge of North America.

Mafic volcanic rocks from the study area can be spatially divided into three broad regions: GB (5–35 Ma), eastern ACA, and western ACA (2.5–16 Ma). The volcanic products are dominantly calc-alkalic but transition to alkalic toward the east. Great Basin lavas erupted far inland from the continental margin and have higher K, P, Ti, and La/Sm as well as lower (Sr/P)pmn, Th/Rb, and Ba/Nb compared to ACA lavas. Higher Pb isotopic values, combined with lower Ce/Ce* and high Th/Nb ratios in some ACA lavas, are interpreted to come from slab sediment. Mafic lavas from the GB and ACA have overlapping 87Sr/86Sr and 143Nd/144Nd values that are consistent with mantle wedge melts mixing with a subduction-modified lithospheric mantle source. Eastern and western ACA lavas largely overlap in age and elemental and isotopic composition, with the exception of a small subset of lavas from the westernmost ACA region; these lavas show lower 87Sr/86Sr at a given 143Nd/144Nd. Results show that although extension contributes to melting in some regions (e.g., selected lavas in the GB and Pyramid Lake), chemical signatures for most mafic melts are dominated by subduction-related mantle wedge and a lithospheric mantle component.

INTRODUCTION

The contrast between magmas derived at destructive plate margins and those at within-plate settings has long been explored (e.g., Hofmann et al., 1984; Ellam and Hawkesworth, 1988; Ormerod et al., 1991; Pearce and Peate, 1995; Pearce and Stern, 2006). The extensive tectonomagmatic geologic record of the North American Pacific margin includes orogenic events, passive and Andean-style convergent margins, hotspot magmatism, and extension (Dickinson, 2006, and references therein). Between ca. 45 Ma and ca. 3 Ma, a complex tectonic transition from subduction- to extension-related magmatism occurred in southwestern United States (e.g., Dickinson, 2006). Regionally, igneous activity in both the Great Basin (GB) and Ancestral Cascades Arc (ACA) consists of extensive magmatic events that migrated southwestward from western Utah to eastern California, commonly referred to as a “magmatic sweep” (Coney, 1978; Christiansen et al., 1992; Humphreys, 1995; Dickinson, 2006). The sweep of Cenozoic magmatism across the southwestern United States is thought to be due to the “rollback” of the shallow-dipping Farallon slap (i.e., sinking of the negatively buoyant slab into the asthenosphere and migration of the hinge line). Slab rollback caused the continental arc to retreat oceanward, inducing the southwestward magmatic sweep as the slab angle continued to steepen and volcanism migrated across the GB and transitioned into the ACA (Cousens et al., 2008; John et al., 2012). Volcanic eruptions began far from the continental margin (as far east as the San Juan volcanic field in Colorado, USA), include both mafic to intermediate lava flow complexes and rhyolitic caldera centers, and to date the mafic-intermediate lava complexes have received little attention in studies relevant to the origin and evolution of this migrating sweep of volcanism compared to caldera studies (e.g., Best et al., 2013; Henry and John, 2013). A segment spanning central western Nevada and eastern California was selected as the study area in order to examine the geochemical nature of mafic lavas extruded along the path of this magmatic sweep to address three issues: (1) origin of magmatism in the GB compared to the ACA, (2) changes in geochemistry over space and time, and (3) subduction- versus extension-related magmatism.

Geochemical studies of neighboring volcanic regions outside the study area are well represented in the literature, including the modern Cascade arc (e.g., Baker et al., 1994; Bacon et al., 1997; Green and Harry, 1999; Harry and Green, 1999; Borg et al., 2002; Strong and Wolff, 2003; Leeman et al., 2004; Green and Sinha, 2005), Yellowstone and the Snake River Plain (e.g., Camp, 1995; Christiansen et al., 2002; Jordan et al., 2004), the western Great Basin (WGB; Ormerod, 1988), the Mojave Desert and Basin and Range (e.g., Glazner et al., 1991; Perry et al., 1993; Farmer et al., 1995; Yogodzinski et al., 1996), and the San Andreas fault and offshore California (e.g., Cole and Basu, 1995; Davis et al., 1995; Dickinson, 1997). The ACA, located to the south of the modern Cascades, is also known as the Miocene–Pliocene Cascade Arc (Priest, 1990; Christiansen et al., 1992; Lipman, 1992; Dickinson, 1997, 2002, 2004; John 2001; Cousens et al., 2008; du Bray et al., 2009), with particular emphasis on the ACA within the Sierra Nevada (e.g., Putirka et al., 2012, and references therein). Late Cenozoic uplift of the southern Sierra Nevada is linked to ongoing delamination of the underlying lithospheric mantle based on geophysical (Wernicke et al., 1996; Wernicke and Snow, 1998; Boyd et al., 2004; Zandt et al., 2004), geological, and petrological work (Ducea and Saleeby, 1996, 1998; Manley et al., 2000; Ducea, 2001; Lee et al., 2001; Farmer et al., 2002; Lee, 2005; Putirka and Busby, 2007, 2011; Putirka et al., 2012). It should be noted that “lithosphere” is defined by McKenzie (1989) as “the mechanical boundary layer beneath the continental crust which remains physically isolated for geologically long periods from the convecting mantle beneath because it is relatively cold (so that heat passes through it entirely by conduction) and buoyant (due to chemical differentiation by melt extraction).” Thus the “asthenosphere” is the hot convecting upper mantle and is for the most part chemically separate from the lithosphere.

A lingering issue that plagues this region is whether or not the Eocene to late Pliocene “sweep” of magmatism across the GB was solely the result of subduction and rollback of the Farallon plate underneath the North American plate, or does extensional tectonism within the Basin and Range province after ca. 17 Ma also contribute to melt generation (e.g., Fitton et al., 1988; Ormerod et al., 1991; Asmerom et al., 1994; Yogodzinski et al., 1996)? Previous research from tectonomagmatic provinces within southwestern United States has determined specific geochemical signatures that represent subduction- versus extension-related processes. For example, lavas from the modern southern Cascades arc are calc-alkaline, dominantly intermediate in composition with enrichment in large-ion lithophile elements (LILEs) and light rare-earth elements (LREEs) and depletion in high field strength elements (HFSEs) (e.g., Borg et al., 1997). In contrast, eruptions due to extension in eastern California are commonly bimodal (basalt and rhyolite), high-K (typically) alkaline lavas, with higher 87Sr/86Sr and lower 143Nd/144Nd due to melting of the modified lithospheric mantle (e.g., Big Pine volcanic field, Ormerod et al., 1991), though proposed extension-related intermediate, trachyandesitic eruptions from volcanic centers such as the ca. 6 Ma Ebbetts Pass region (e.g., Hagan et al., 2009; Busby and Putirka, 2009) and Lake Mead (e.g., Weber and Smith, 1987) have been reported. In some cases, extension has thinned the lithosphere sufficiently that asthenosphere-derived magmatism allows alkali basalts with within-plate trace-element signatures to be emplaced (Reno-Fallon region, Buffalo Valley volcanic field, Lunar Craters volcanic field, and the Mojave Desert; Farmer et al., 1995; Cousens et al., 2012; Cousens et al., 2013; Rasoazanamparany et al., 2015).

Since primitive lavas (Mg# > 0.7, where Mg# = Mg/(Mg + Fe2+)) are rare in the GB-ACA region, the data have been filtered to include lavas with SiO2 < 57 wt% and Mg# > 0.55, which limits the eastern reach of the east-west transect to the Stillwater Mountain Range within the central GB. This study offers a comprehensive geochemical investigation of the most “primitive” lavas erupted in an east-west transect across the GB to help understand the forces behind magma generation and tectonic evolution during 35 Ma to 3 Ma magmatic sweep (Fig. 1). A total of 98 basalts, basaltic andesites, and basaltic trachyandesites from the GB and the ACA were analyzed for petrography, detailed major- and trace-element contents as well as Sr, Nd, and Pb isotopic ratios to evaluate (1) mantle sources, (2) potential lithospheric and/or continental crustal contamination, and (3) variation in magmatic processes during the transition between subduction- and extension-related tectonic settings.

GEOLOGICAL SETTING

Regional Geological Setting

The tectonic history of southwestern North America has been described by several authors, including Lipman (1992), Atwater and Stock (1998), Sonder and Jones (1999), Henry and Ressel (2000), Dickinson (2002, 2006), and DeCelles (2004). Pre-Cenozoic rocks range in age from Precambrian to Mesozoic and almost all are meta-volcanic and meta-sedimentary (Bonham and Papke, 1969); all were subjected to the Antler (Late Devonian into the Mississippian), Sonoma (during the Permian/Triassic transition), Sevier (Jurassic to Eocene), and Laramide (Late Cretaceous to Eocene) orogenies (Dickinson, 2006; Mann, 2007). Subduction of the Farallon plate under the North American plate began during the Jurassic (Dickinson, 2006; Mann, 2007), and the North American coast underwent Andean-style arc volcanism centered in the Sierra Nevada arc until ca. 74 Ma. Subsequently, the arc front migrated several hundred kilometers inboard of the trench (Lipman, 1992), which is commonly referred to as the Laramide migration (ca. 70 to ca. 40 Ma) (e.g., Dickinson, 2006) because it accompanied the Laramide orogeny. The Laramide orogeny (ca. 80–35 Ma) is characterized by uplift of Precambrian crystalline basement due to the rapid, northeast-directed convergence between the Farallon and North American plates, whereas the Late Cretaceous Sevier orogeny is primarily characterized by thrusting and folding. During Laramide time, the relatively high convergence rate meant that the North American plate overrode the buoyant Farallon plate faster than the Farallon plate was able to sink. As a result, the Farallon plate was subducted beneath the North American plate at a very shallow angle, increasing the distance between the trench and the volcanic front (Lipman, 1992). In the literature, it is also proposed that uplift during the Laramide orogeny is the result of the shallow slab segment carrying a fragment of an aseismic ridge that was the counterpart of a large igneous province of the northwest Pacific basin named the Hess-Shatsky conjugate (Saleeby, 2003; Liu et al., 2010). Its thickened mafic crustal section relative to abyssal lithosphere is presumed to have rendered a greater buoyancy leading to slab segmentation into the respective shallow domain (Saleeby, 2003).

By the late Cretaceous, the thickened and uplifted interior of western North America was tilted toward what is now referred to as the Sierra Nevada (Mitrovica et al., 1989; Humphreys et al., 2003), while the slab of younger and more buoyant lithosphere of the Farallon plate continued to subduct (Humphreys et al., 2003). This descending “flat slab” would have had a subduction angle of ∼5° (Dumitru, 1991), creating traction on the base of the lithosphere, pulling it down, which caused the opening of the Cretaceous Seaway (Mitrovica et al., 1989), also called the Western Interior Seaway or North American Inland Sea. By mid-Eocene, the Farallon slab detached (Schellart et al., 2010) from the overriding lithosphere and “rolled back” (Best and Christiansen, 1991; Henry and Ressel, 2000; Dickinson, 2004, 2006; Best et al., 2016) along an east-northeast–trending axis, resulting in a southwestward magmatic “sweep” through Nevada before finally steepening to a “normal” (∼30°) subduction angle (Coney and Reynolds, 1977; Dickinson and Snyder, 1978; Best and Christiansen, 1991; Dumitru, 1991; Seedorff, 1991; Hofstra et al., 1999; Henry and Ressel, 2000; Cline et al., 2005; Ressel and Henry, 2006). The magmatic sweep involved the emplacement of andesitic and dacite to rhyolitic ash flows and dome complexes (John, 2001), as well as small volumes of basaltic, andesitic, and dacitic lava as far east as the San Juan volcanic field in southwestern Colorado (Lipman et al., 1978). A northwest-trending belt of ash-flow calderas that developed through central Nevada is referred to as the “ignimbrite flare-up” (Best and Christiansen, 1991; John et al., 2008; Best et al., 2009; Best et al., 2016). Magmatic activity reached westernmost Nevada and eastern California between 20 and 17 Ma, forming major volcanic centers (e.g., Busby et al., 2008; Cousens et al., 2008; John et al., 2012). Between 17 and 14 Ma, subduction-related igneous activity in the Sierra Nevada and western Nevada was overlapped by rift or hotspot volcanism of the Northern Nevada Rift, Oregon High Plains, and Columbia River Flood Basalt provinces (Dickinson, 2006).

The latest geological chapter of the GB and ACA is occasionally referred to as an extensional orogeny (e.g., Elston, 1984). At ca. 28 Ma, the East Pacific Rise impacted the North American plate, bringing the Pacific plate into contact with the North American plate. A plate margin triple junction marked the initial point of contact of the Pacific, North American, and Farallon plates. The Farallon plate later split into smaller plates, which include the Juan de Fuca to the north (Mendocino Triple Junction [MTJ]) and Cocos plate to the south (Rivera Triple Junction [RTJ]). The MTJ has migrated northward such that the southern end of the volcanic (Cascade) arc associated with Cenozoic subduction beneath the North American plate has migrated northward since the Miocene (Atwater and Stock, 1998). Behind the arc, lithospheric extension began at ca. 17 Ma (Colgan et al., 2006a, 2006b) to form the GB, and migration of the GB into the Sierra Nevada mountains is an ongoing process (e.g., Manley et al., 2000; Kent et al., 2005).

Local Geological Setting

Samples of mafic rocks were collected over several field excursions from 1997 to 2011 between latitudes 39° and 40°. The area can be divided longitudinally into three broad regions (Fig. 1): (1) the GB at approximate longitudes 117.00° to 119.52° (includes Stillwater Mountain Range, Truckee Range, Hot Springs Mountain, and Churchill Group), (2) the eastern ACA at approximate longitudes 119.52° to 120.00° (includes Sparks, Silver City, Pyramid, I-80 Suite, and Lousetown), and (3) the western ACA at approximate longitudes 120.00° to 121.00° (includes Stampede Reservoir, Ladybug, Portola, Mount Lincoln, Dog Valley, Stanford/Twin Peaks, Pond Terrace, Andesite Ridge, Devil’s Peak, Boreal Ridge, Needle Peak, Squaw, Lowell Ridge, Sawtooth, Henness Pass, and Susanville) (Timmermans, 2015). Also included are geochemical and isotopic data for all three regions from Henry and Sloan (2003), Sloan et al. (2003), and du Bray et al. (2009). Petrographic descriptions for samples with the identifiers 04-LT and 11-CN are summarized in Table 1.

The Great Basin (GB)

The Stillwater Range is a ridge ∼110 km long that trends north-northeast along the east side of Carson Sink, northeast of Fallon, Nevada. The southern Stillwater Range is composed of Tertiary igneous and non-marine sedimentary rocks (Fig. 2), that unconformably overlie, are faulted against, or intrude into faulted pre-Tertiary metasedimentary and metavolcanic basement rocks (Page, 1965; John, 1995). The region was later subjected to Miocene structural tilting and extensional faulting (Page, 1965; John, 1995). Miocene basalt and basaltic andesite occur at various sites including the volcanic fields near Mountain Well Road and La Plata Canyon. Jurassic lavas from the Stillwater Range show higher levels of hydrothermal alteration causing extreme saussuritization of plagioclase to primarily sericite, calcite, and chlorite compared to Tertiary lavas. Seriate texture is present in sample 04-LT-41, whereas other lava flows show at least two stages of feldspar growth. Basalt from La Plata Canyon is aphyric and plagioclase rich. Bell Canyon is situated southeast of the Stillwater Range and encompasses most of a range that includes Fairview Peak and the northern part of Slate Mountain (Henry, 1996). Rock units include Jurassic to Triassic metamorphic rocks intruded by Cretaceous granodiorite, a wide range of Oligocene to Miocene volcanic and volcaniclastic rocks (most of which are related to a ca. 19 Ma caldera), and upper Neogene to Quaternary sedimentary deposits. Vitrophyric basaltic andesitic lava flows and dikes have 25% phenocrysts of plagioclase and olivine in a hyalocrystalline groundmass (Fig. 3A). The basaltic unit (04-LT-51) has basal scoria breccia that develops into a massive but highly vesicular lava with 20% phenocrysts of plagioclase, clinopyroxene, and iddingsitized olivine (Henry, 1996). A sample from Bell Canyon (04-LT-48) has a holohyaline groundmass.

The Hot Springs Mountain and the Truckee Ranges lie in west-central GB abutting against the Walker Lane (Bonham and Papke, 1969). They comprise a series of hills that form part of the northwestern boundary of the large Carson Sink depression. Exposed rocks in the region range from Triassic to Holocene (Willden and Speed, 1974). Cenozoic rocks are primarily volcanic and consist of alternating sequences of flows, tuffs, and shallow intrusive rocks. Lavas are basaltic trachyandesites that commonly contain abundant plagioclase and minor olivine and/or augite phenocrysts (Fig. 3C). Olivine is generally euhedral and mildly to completely altered to iddingsite. Subhedral augite commonly shows subophitic texture with plagioclase, and most include plagioclase phenocrysts with sieve texture and oscillatory or normal zonation (Fig. 3B).

Eastern and Western Ancestral Cascades Arc (Eastern and Western ACA)

The pre-Cenozoic rocks are mainly Mesozoic granitoids with minor Paleozoic metasedimentary and metavolcanic rocks (Ashley et al., 1979). A series of intermediate magmatic events that occurred between 17 and 12 Ma resulted in several formations including Kate Peak, Alta, and Truckee formations (Ashley et al., 1979). Magmatism that began again at ca. 10 Ma (Cousens et al., 2008) consisted of more mafic compositions ranging from pyroxene andesite to olivine-pyroxene basalt. Basalts from the eastern and western ACA are primarily porphyritic with olivine and/or augite phenocrysts. Basaltic andesites are mineralogically similar to the basalts, although plagioclase dominates over olivine and sparse microphenocrysts of orthopyroxene and/or clinopyroxene (Cousens et al., 2008). Plagioclase phenocrysts (as much as 15% by volume) typically occur as glomerocrysts that are either zoned with resorbed interiors or, more commonly, are zoned and subhedral. Smaller feldspar grains are skeletal to euhedral and generally lack zoning.

Sparks is located to the north of the town of Sparks, just east of Reno. Mafic lavas from Sparks are basaltic andesites that contain phenocrysts of clinopyroxene, plagioclase, and iddingsitized olivine. The extensive Lousetown Formation, southwest of Reno, covers ∼40 km2 in the Virginia City quadrangle, and extends into the Mount Rose, Wadsworth, and Churchill Butte quadrangles (Ashley et al., 1979). The I-80 Suite refers to lavas with similar ages (<10 Ma) and chemical and physical characteristics that are located along the Interstate-80 corridor between north Lake Tahoe and Reno (Henry and Perkins, 2001; Cousens et al., 2008). Ladybug Peak, at the north end of Verdi Range, represents an explosive and effusive episode of mafic volcanism (Cousens et al., 2008) that may be related to the region’s 12–10 Ma extensional event (Henry and Perkins, 2001; Henry et al., 2011). These basalts and basaltic andesites contain as much as 15% olivine grains up to 4 mm in size, 10% augite up to 2 mm in size, and minor plagioclase set in a fine-grained matrix of feldspar and pyroxene (Cousens et al., 2008).

The Pyramid Sequence (ca. 13 Ma), located to the southeast of Pyramid Lake, consists primarily of interbedded basalt to basaltic andesite, sparse andesitic lavas and rare rhyolitic lava, coarse- to fine-grained clastic rocks, and a dacitic ash-flow tuff (Henry et al., 2004a). The mafic lavas considered range from aphyric to coarsely porphyritic. The coarsely porphyritic rocks are characterized by prominent, tabular plagioclase phenocrysts up to 20 mm long and 2–4 mm wide (Henry et al., 2004a).

Younger (ca. 3 Ma) and mildly phyric, olivine-clinopyroxene basaltic andesites, and basaltic trachyandesites were extruded atop the northern Carson Range, just south of Reno. The northern end of the Carson Range is composed primarily of tilted, porphyritic lavas capped by gently dipping to flat-lying, poorly phyric lava flows (Thompson and White, 1964; Latham, 1985). Samples 01-LT-10, -11 (basaltic andesites), and -12 (basaltic trachyandesite) are from a set of flows near Fuller Lake. Samples 01-LT-10 and -11 contain altered olivine and rare plagioclase phenocrysts, whereas sample 12 is much fresher with minor olivine and pyroxene phenocrysts. Locations for samples 01-LT-46–53 are included in Cousens et al. (2011). These samples commonly include minor olivine phenocrysts, with or without minor plagioclase and pyroxene phenocrysts, in a trachytic matrix. The basaltic and basaltic andesitic lava flows of Dog Valley (4–3 Ma) are located just west of the Carson Range.

Samples from Pond Terrace, Andesite Peak, Devil’s Peak, Mount Lincoln, Twin Peaks, Stanford Peak, Squaw Peak, Sawtooth, and Henness Pass are from highly eroded volcanic complexes emplaced 8–6 Ma and 5–3 Ma that cap high Sierra Nevada granitoid intrusions from the west side of Lake Tahoe north to Donner Pass (Harwood, 1981; Saucedo and Wagner, 1992; Henry et al., 2004b; Cousens et al., 2008). The volcanic rocks of this area include lava flows, agglomerates, tuffs, and plug intrusions of lava and breccia (Hudson, 1951), as well as volumetrically minor amounts of olivine-clinopyroxene basalt and basaltic andesite lava flows (Henry et al., 2004b; Cousens et al., 2008). At Squaw Peak and Mount Lincoln, debris lava flows and dome-collapse deposits are composed of fragments of plagioclase-phyric volcanic rock set in a mineralogically similar volcanic matrix. Subordinate lava flows are commonly >10 m thick, massive, and highly plagioclase porphyritic. Some basaltic andesites, particularly those in the Squaw–Twin Peaks area, include abundant anhedral sieve-textured plagioclase grains that may be xenocrystic. Trachytic texture in the groundmass is less common in the basaltic andesites compared to the basalts. Farther north into the Sierra Nevada along Highway 70, basalts and basaltic andesites from Portola contain phenocrysts of embayed olivine and rare augite in a hypocrystalline matrix of plagioclase and olivine.

In general, the mafic lavas with Mg# > 0.55 closely follow the idealized olivine-clinopyroxene fractionation trend (Harker, 1909). Many of the lavas contain phenocrystic olivine and clinopyroxene along with plagioclase, implying that the initial magma evolution was at relatively high pressure under hydrous and/or non-hydrous conditions (>8 kbar, Green and Hibberson, 1970; Green et al., 2010).

The western ACA is located almost entirely in the unextended Sierra Nevada, whereas the eastern ACA includes volcanic rocks that were emplaced in terranes adjacent to the Sierra Nevada that either were subsequently faulted and rotated or were emplaced into terrane that was undergoing extensional faulting and deformation. Did extension in the eastern ACA allow for different mantle sources to be tapped than did the unextended Sierra Nevada? Or do the similar basement rocks in both western and eastern ACA control mantle sources and petrogenetic evolution in lavas emplaced in either part of the arc?

METHODOLOGY

Field Work and Data Sources

Locations of mafic volcanic rocks between 40 and 3 Ma in age were determined using existing geologic maps: Page (1965), Stewart and Carlson (1976), Bell and Bonham (1987), Henry (1996), Jones (1997), Stewart (1999), Henry et al. (2004c), Saucedo et al. (2005), Castor et al. (2006), Hudson et al. (2009), and Bell et al. (2010). Ice sheets from Pleistocene glacial events reached well into the southwestern United States, leaving only remnants of lava flows, debris flows, and volcanic necks. Bedrock is for the most part well exposed, though outcrops within the GB suffer from higher degrees of weathering compared to the ACA (Fig. 2). Figure 1 shows the sample locations; a complete list of sample names and locations (note: samples were located using a handheld global positioning system [GPS] and GPS coordinates) is presented in Timmermans (2015) and Supplemental Table S11.

Samples with identifiers 04-LT and 11-CN were collected by the authors during two field seasons in 2004 and 2011, respectively, for the purpose of filling data gaps along the east-west transect. Sampling was occasionally obstructed by middle to late Pleistocene lacustrine and interlacustrine deposits—the result of several glacial episodes that also eroded the Sierra Nevada Mountain landscape (Morrison, 1964). Outcrops surrounding the Carson Sink (e.g., Lahontan Mountains, Stillwater Mountain Range) are commonly shoreline deposits encrusted by tufa (carbonate deposits formed by precipitation from saline lake water due to significant departure from ambient temperature), deposited along the margin of a large endorheic lake named Lake Lahontan (Morrison, 1964). Samples from these regions were collected from outcrops above the ancient shorelines.

Existing published data from the region were included in the database for this study. Samples with identifiers LT and Towle were collected between 1997 and 2008 by Brian Cousens, Julie Prytulak, and Russell Towle. Hand samples, thin sections, and geochemistry data from these specimens were all processed and analyzed at Carleton University, Ottawa, Ontario. Sample data with all other identifiers were obtained from Henry and Sloan (2003), Sloan et al. (2003), du Bray et al. (2009), and geochemical databases, for which only the geochemical data were adopted.

Geochronology

Ages for the mafic samples used are based on radiometric ages determined for samples or sample localities. Radiometric ages were primarily derived from 40Ar/39Ar geochronology from the Nevada Bureau of Mines and Geology (NBMG) using methodology as described in detail by Henry et al. (1996) and Henry et al. (1997). Ages are rounded to the nearest million years (du Bray et al., 2009).

Geochemistry

Geochemical data are listed in Supplemental Table S1 (footnote 1). Collected samples from the GB and ACA were prepared at Carleton University, Ottawa. Rock samples were slabbed, crushed in a Bico chipmunk jaw crusher, and ground to a fine powder in either an agate or a chromium-steel ring mill. Between samples, the jaw crusher and the ring mill were cleaned thoroughly using water, compressed air, and ethyl alcohol. Replicate analyses for 11-CN-1 and 11-CN-11 through a chrome-steel and an agate ring mill have been compared, indicating only minor Cr contamination for splits processed through the chrome-steel head. Whole-rock major- and trace-element contents were determined by fused-disc X-ray fluorescence spectrometry (XRF; University of Ottawa) and solution mode inductively coupled plasma mass spectrometry (ICP-MS; Ontario Geological Survey, Sudbury). The precisions of the data, based on replicate analyses of samples and blind standards, are listed with the data presented in Supplemental Table S1 (footnote 1). Detection limits indicate the minimum concentration that can be distinguished from a blank specimen; values were calculated by personnel at the Ontario Geological Survey Geoscience Laboratories.

All Pb, Sr, and Nd isotopic ratio analyses and sample preparations, including weighing, acid washing, dissolution, column chromatography, filament loading, and mass spectrometry, were performed at the Isotope Geochemistry and Geochronology Research Centre (IGGRC) at Carleton University, Ottawa, Canada, using techniques presented in Cousens (1996). For samples processed before 2002, samples were run on a Finnigan-MAT 261 thermal ionization mass spectrometer run in static mode. Strontium was loaded onto a single Ta filament with H3PO4. Isotope ratios were normalized to 86Sr/88Sr = 0.11940. Two standards were run: National Institute of Standards and Technology (NIST) standard reference material (SRM) 987 (87Sr/86Sr = 0.710257 ± 0.000022) and the Eimer and Amend (EandA) SrCO3 (87Sr86Sr = 0.708035 ± 0.000010). Neodymium was loaded with H3PO4 on one side of a Re double-filament assembly. Isotope ratios are normalized to 146Nd/144Nd = 0.72190. Analyses of the La Jolla standard average 143Nd/144Nd = 0.511874 ± 0.000019. Lead was loaded onto single Re filaments with H3PO4 and silica gel. All mass spectrometer runs were corrected for fractionation by monitoring NIST SRM 981 (Todt et al., 1996). The average measured ratios for SRM 981 are 206Pb/204Pb = 16.890 ± 0.012, 207Pb/204Pb = 15.429 ± 0.014, and 208Pb/204Pb = 36.502 ± 0.048. Lead concentrations are precise to ±0.5%. Analysis of U.S. Geological Survey (USGS) standard BCR-1 yields Pb = 13.56 ppm, 206Pb/204Pb = 18.818, 207Pb/204Pb = 15.633, and 208Pb/204Pb = 38.633.

For samples processed after 2002, samples were run on the Thermo Finnigan Triton TI thermal ionization mass spectrometer. Strontium isotope ratios were normalized to 87Sr/88Sr = 0.11940, and the lab average for NBS 987 was 87Sr/86Sr = 0.710248 ± 0.000026. Neodymium isotope ratios are normalized to 146Nd/144Nd = 0.72190, and runs of the La Jolla standard averaged 143Nd/144Nd = 0.511877 ± 0.000018. All Pb ratio mass spectrometer runs are corrected for fractionation relative to the following accepted values for NIST NBS 981: 206Pb/204Pb = 16.937 ± 0.009, 207Pb/204Pb = 15.492 ± 0.008, and 208Pb/204Pb = 36.722 ± 0.019 (Todt et al., 1996). All quoted uncertainties are 2 standard deviations of the mean.

RESULTS

Low loss-on-ignition (LOI) values (<3 wt%) and petrological observations indicate that secondary processes have not significantly affected the collected samples. Secondary minerals such as sericite and chlorite are rarely seen during petrological examination. LOI data are also good proxies to discriminate between fresh and weathered samples when compared to La/Sm values, for which correlation was observed for the ACA and the GB lavas (not shown) (Labanieh et al., 2012). This lack of correlation implies that weathering did not have a significant effect on the REEs in the studied samples (Yusoff et al., 2013).

Mafic lavas that erupted within the east-west transect exhibit a wide range of compositions and significant variations among and within the eruptive centers. Subalkaline lavas are found throughout the study area (Fig. 4; Le Bas et al., 1986). The Mg/(Mg + Fe2+) ratios (where Fe2+/Fetot = 0.9), represented as the Mg#, range from 0.75 to 0.55, indicating that although some of the mostly subalkaline mafic lavas are near-primary melts (that is, if mantle olivine is Fo86–90, primary magmas must be Mg# = 0.68–0.75) (e.g., mafic samples from the western ACA, Mg# > 0.65; Le Bas et al., 1986), most lavas have undergone some postgeneration modification. Figures 5A–5F and 6 show variation diagrams, where major and trace elements are plotted against Mg#. In general, basalts, basaltic andesites, and basaltic trachyandesites from the GB have an overall lower Mg# (0.55–0.65) than those from the eastern and western ACA (up to 0.75 in the case of the western ACA). Basalts with Mg# > 0.7 also have Cr > 600 ppm, Ni > 180 ppm, and Sc > 32 ppm. The AFM (alkalis-FeO-MgO) diagram (not shown, Irvine and Baragar, 1971) shows that most mafic lavas plot primarily in the mafic field along a calc-alkaline trend. The GB lavas have higher Fe2O3t/MgO values compared to the ACA lavas. The concentrations of Ni, Cr, and Sc all decrease with decreasing Mg#, along with CaO/Al2O3, indicating that olivine, Cr-spinel, and clinopyroxene in the ACA lavas were dominant crystallizing mineral phases. Al2O3 contents range from 13 wt% to as high as 20 wt%, with the eastern ACA containing the highest Al2O3 and the least scatter. K2O values range from 0.5 wt% to as high as 2.5 wt% (sample 01-LT-12 from the Carson Range trends to shoshonitic affinities). The TiO2, P2O5, K2O, REEs, Zr, and Y contents all increase with decreasing Mg# and are notably higher in lavas from the GB. The decrease in transition metal (V, Ni, Co, and Cu) abundances during fractional crystallization is principally a function of the precipitation of olivine, Fe-Ti oxides (crystallization of Fe-Ti oxides is largely governed by the melt redox state), and immiscible sulfides. Vanadium generally decreases only slightly with decreasing Mg#, confirming together with increasing TiO2 that for mafic melts, Fe-Ti oxides are not a dominant fractionating phase, which agrees with experimental studies for primitive lavas at oxygen fugacities typical for arcs (e.g., Lee, 2005). Large-ion lithophile element abundances for different localities generally do not co-vary with Mg#. Isolated flows of basalt and basaltic trachyandesite from the GB (in the Bell Mountains and the Stillwater Range) are characterized by similar SiO2 and K2O but lower Mg#, higher P2O5, TiO2, and Zr relative to the more abundant subalkaline basalts. These slightly alkaline lavas generally have higher REE abundances compared to the subalkaline basalts. The basaltic trachyandesites are very uniform in composition and are generally enriched in TiO2, K2O, P2O5, Rb, Ni, Zr, Nb, Pb, Zr/Y, and the REEs compared to the subalkaline basalt and basaltic andesite groups.

Figure 7 compares (A) Zr/Yb and (B) Th/Yb with Nb/Yb. Mafic melts generally possess only a small range of Yb contents, whereas the Nb contents can reflect variations from the source (Pearce, 2008). All samples plot above the mid-ocean ridge basalt–ocean island basalt (MORB–OIB) array, indicating that they derived from mantle contaminated by either a slab fluid or continental crust (higher Th/Yb) or a continental lithospheric mantle (CLM) component (higher Zr/Yb). The ACA lavas trend parallel to the MORB-OIB array in Figure 7B, indicating the importance of melt extraction during subduction in an arc system (Pearce and Peate, 1995). The GB lavas, however, do not parallel the array, which may indicate that the higher Th/Yb values (Fig. 7B) are derived from another source. Comparing Zr/Yb with Nb/Yb (Fig. 7A), most lavas from the eastern ACA and the western ACA have Zr/Yb values that also lie within and slightly above the MORB array, whereas the GB lavas trend toward the E-MORB or OIB mantle array (Pearce and Peate, 1995; Pearce, 2008). Younger (ca. 3 Ma) basaltic trachyandesites from the Carson Range lie outside the eastern ACA and the western ACA fields (basalts and basaltic andesites) with comparatively higher La/Yb, Th/Yb, Zr/Yb, and Nb/Yb values than other lavas.

Rare-earth and trace-element concentrations (Figs. 8 and 9) have been normalized to chondritic and primitive mantle abundances of Sun and McDonough (1989), respectively. The three regions are represented separately by symbols representing composition. All lavas have primitive mantle normalized La/Sm between 2 and 3.7; however, the GB lavas have a more restricted range (2.4–3.6) than the ACA lavas. All GB lavas show higher abundance in all REEs with no fractionation of the light rare earth elements (LREE) and/or heavy rare earth elements (HREE). The GB lavas also have negative Eu-anomalies that increase with increasing REE abundance to a maximum of Eu/Eu* = 0.77. Eu anomalies are denoted by Eu/Eu* and calculated as (Eu/Eu*)pmn = Eupmn/((Smpmn + Gdpmn)/2). The REE patterns in Figure 8A show that all GB lavas share parallel REE patterns but vary in overall abundance. For the eastern ACA lavas (Fig. 8B), normalized values are variable from La to Sm, but then the REE patterns converge toward the HREEs, forming a fanning set of patterns. The western ACA lavas (Fig. 8C) show a lower range of La to Sm ratios compared to the eastern ACA lavas; however, both groups of ACA lavas show variations in light REE abundance but have relatively flat middle to heavy REE patterns (mean values for GB, eastern ACA, and western ACA shown in Fig. 8D). Lavas with higher silica and alkali contents generally have higher REE abundances and higher La/Sm ratios compared to lavas with lower silica and alkali contents. All samples show a slight negative anomaly (Eu/Eu* = 0.997–0.768, average 0.911) with lavas from the GB containing the largest Eu anomalies (averaging 0.834). Basaltic trachyandesites from Hot Springs Mountain (GB) have the overall largest Eu anomaly of 0.768. A basalt from Portola and a basaltic andesite from Stanford/Twin Peaks, both from the western ACA, have an atypical negative Ce-anomaly (Ce/Ce* = 0.849 and 0.713, respectively), where Ce/Ce* anomalies are calculated Ce/Ce* = Cepmn/((Lapmn + Prpmn)/2). All other samples have Ce-anomalies that hover just under 1 (average 0.954).

Most trace-element patterns show large enrichments in LILEs (e.g., Ba, Rb, K, and Pb) and lesser enrichments in LREEs (e.g., La, Ce, Nd, and Sm) relative to HFSEs (e.g., Nb, Ti, and Ta) and heavy rare-earth (HRE, e.g., Tb, Tm, and Yb) elements, characteristic of arc magmas (Figs. 9A–9C) (Gill, 1981; Pearce, 1983; Rollinson, 2014; Briqueu et al., 1984; Weaver et al., 1987). The subduction signature seen in Figure 9 (i.e., depletion in Nb and Ta and enrichment in LILEs when compared to the primitive mantle) is strongest in the western and selected eastern ACA and weakest in the GB (Fig. 9D). The eastern and western ACA mafic rocks show a larger Sr enrichment relative to Nd compared to the GB.

Isotopic compositions of Sr, Nd, and Pb have been measured for select samples from the GB and the eastern and western ACA, the results of which are shown in Figure 10 and listed in Supplemental Table S1 (footnote 1). Isotopically, the groups are highly heterogeneous with the western ACA containing the widest range: 87Sr/86Sr = 0.70381–0.70612 and 143Nd/144Nd = 0.51242–0.51283. The eastern ACA and the GB have more restricted ranges: 87Sr/86Sr = 0.70411–0.70552 and 143Nd/144Nd = 0.51261–0.51272, 87Sr/86Sr = 0.70453–0.70531 and 143Nd/144Nd = 0.51262–0.51283, respectively (Fig. 10A). The isotopic ratios overlap for the GB, part of I-80 Suite, Sparks, Carson Range, Lousetown, Henness Pass, and Susanville (eastern and western ACA). Note that six samples from the westernmost part of the ACA, along the crest of the Sierra Nevada, plot to the left of the main array of mafic lavas: i.e., at low 87Sr/86Sr given their 143Nd/144Nd. These lavas include Mount Lincoln basalt, Squaw Peak, Pond Terrace, Devil’s Peak, and Lowell Ridge; subsequently, these flows will be referred to as the westernmost part of the ACA (circled in Fig. 10A). However, other lavas from the Sierra Nevada crest plot within the main data array.

Lead-isotope compositions for mafic lavas are also heterogeneous, shown in Figures 10B and 10C relative to the Northern Hemisphere Reference Line (NHRL) for oceanic basalts from the Northern Hemisphere (Hart, 1984). The GB and the ACA lavas range in 206Pb/204Pb between 18.85 and 19.15, 207Pb/204Pb between 15.52 and 15.69, and 208Pb/204Pb between 38.33 and 38.97. For 207Pb/204Pb versus 206Pb/204Pb (Fig. 10B), all lavas plot above the NHRL toward an enriched end member, whereas for 208Pb/204Pb versus 206Pb/204Pb, the values parallel the NHRL trending toward a mantle component with high μ (238U/204Pb) and high Th/Pb (Fig. 10C). The GB mafic lavas commonly have lower 207Pb/204Pb and 208Pb/204Pb values than the ACA mafic lavas.

Figures 10D and 10E show that 206Pb/204Pb correlates with Sr and Nd isotope ratios. Comparing 206Pb/204Pb versus 87Sr/86Sr (Fig. 10D), the westernmost part of the ACA and GB lavas plot in separate clusters, and all lavas trend toward an enriched component in terms of higher Rb/Sr and μ (238U/204Pb) compared to undifferentiated material. Lavas in the westernmost part of the ACA have lower 87Sr/86Sr given their 143Nd/144Nd and 206Pb/204Pb values, as well as lower 206Pb/204Pb values for their given 207Pb/204Pb (not shown) values compared to the rest of the ACA lavas. Sample Towle #1 (Sawtooth) has the highest 206Pb/204Pb and 207Pb/204Pb values of 19.12 and 15.69, respectively. Looking to Figure 10E, GB lavas plot in a distinct cluster of 208Pb/204Pb versus 143Nd/144Nd values compared to ACA lavas.

DISCUSSION

The purpose of this study is to address three issues: (1) origin of magmatism in the GB compared to the ACA, (2) changes in geochemistry over space and time, and (3) subduction- versus extension-related magmatism. A geochemical model for Cenozoic mafic magmatism in southwestern North America should also explain: (1) the chemical variation as lavas decrease in age from east to west, (2) the sparse mafic magmatism within the GB and comparatively higher volumes of mafic magmatism within the western ACA, and (3) the broad variation in trace elements and isotopic ratios over space and time.

The trace-element and isotopic magmatic signatures within mafic lavas indicate variation of source and/or influences on the source melts. The source components in primary continental arc lavas may include: (1) depleted upper-mantle peridotite, (2) enriched mantle similar to the source of ocean-island basalt, (3) the sub-continental lithospheric mantle, (4) melt from the continental crust, (5) hydrous fluids derived from the subducting plate, (6) fluids derived from the subducting sediment, (7) partial melts of the subducting plate, and (8) partial melts of subducting sediment. Before examining source components, possible crustal contamination of the lavas after they left the source region must be addressed.

Evidence for Post-Magma Generation Crustal Contamination

Within the confines of the study area, various authors (e.g., Farmer and DePaolo, 1983, 1984) have recognized a range of different Mesozoic and Cenozoic granitoids and eugeocline sedimentary rocks, many of which are the direct or indirect product of subduction processes. Addition of these Mesozoic crustal components to mantle-derived magmas would increase the SiO2 content and 87Sr/86Sr while lowering the Mg#, Nb/La, and 143Nd/144Nd ratios of those magmas (Brown et al., 2014). Most Sierra Nevada granitoids have high 87Sr/86Sr, low 143Nd/144Nd, and high 207Pb/204Pb (e.g., Barbarin et al., 1989; Cousens et al., 2008, 2009). The Sierran granitoids also have lower Ba/La and (Sr/P)pmn values compared to the mafic lavas. The volcanic rocks from the GB have a lower range in Mg# and higher range in SiO2 values compared to the ACA lavas, possibly due to thicker crust during the Oligocene, when what is now the Great Basin was a high plateau resulting from Mesozoic compressive forces and crustal thickening (Henry, 2009). At the time of their eruptions, GB mafic lavas may have traveled through as much as 100 km of compositionally heterogeneous continental crust compared to lesser thickened crust within the younger, western ACA lavas (Henry, 2009).

Figure 11 tests the possibility that crustal contamination of the mafic melts is a major petrological process by investigating correlations of Nb/La (Fig. 11A) and 143Nd/144Nd (Fig. 11B) with SiO2 values. Nb/La values are low (<1) for all three regions, and both Figures 11A and 11B demonstrate no covariation with increasing SiO2 values. The lack of correlation of Sr (Fig. 11 insert) and Nd isotopic composition with SiO2, combined with the absence of inherited crustal quartz or zircons into the melts (determined petrographically) suggests that the isotopic and trace-element compositions of the ACA and GB mafic volcanic rocks are not strongly affected by upper-crustal contamination.

High-K Magmatism and Degree of Partial Melting

Moving from east to west across the study area, mafic melts generally decrease in their K2O, P2O5, TiO2 contents and Nb/Yb (examples shown in Figs. 5A, 5D, and 12A). Mafic magmas with K2O contents between 1%–3% are most common in the central GB. Mafic lavas from Pyramid Lake and basaltic trachyandesites from the Carson Range (eastern ACA) also show higher ranges in K2O, P2O5, and TiO2, similar to the GB mafic lavas. Ranges in Na2O contents do not change much for GB and eastern ACA mafic lavas (2.6–4.5 wt%), slightly lower in western ACA mafic lavas (2.5–3.8 wt%), and lowest for the westernmost ACA lavas (2.1–2.6 wt%).

Major- and trace-element chemistry shown in Figures 5 (K2O, TiO2, P2O5 versus Mg#; K2O versus SiO2) and 12 (TiO2 versus SiO2, La/Sm versus La) confirms that the proportion of high-K (and shoshonitic) rocks decreases from east to west. Low degrees of partial melting of the mantle can produce melts with enriched K, Ti, and P, as well as LREE and/or HREEs. Plots involving ratios of highly incompatible elements (e.g., La and Ce) to a less incompatible element (e.g., Sm and Yb) versus the concentration of the highly incompatible element can reveal degrees of partial melting in comparison to fractional crystallization. In Figure 12B, crystal fractionation will not have a great effect on the La/Sm ratio; however, low partial melting would result in greater fractionation of La from Sm with greater La abundance, thereby providing a method for distinguishing between partial melting and crystal fractionation.

Fractional crystallization can be partially responsible for the slightly higher range in SiO2 and lower range in Mg# for the GB mafic lavas compared to the ACA lavas; however, it cannot be the only mechanism responsible for the results seen in Figure 12. Another component is required to explain these geochemical variances for mafic lavas (i.e., with minimal range in Mg#). The chondrite-normalized REE patterns for the GB mafic lavas (Fig. 8) exhibit parallel patterns of increasing abundance with slight enrichment in LREEs relative to HREEs. The parallel REE patterns of the GB rocks imply that the mafic melts were derived from a similar mantle source region and underwent similar degrees of partial melting. The parallel REE patterns and lower Mg# values also confirm that the GB mafic melts underwent some degree of fractional crystallization before erupting at the surface. Additionally, the high LREE and incompatible trace-element contents (Fig. 9) suggest that the mantle source must have been previously metasomatically enriched (Cousens et al., 2008). Figure 10E shows that GB lavas plot in a distinct cluster of 208Pb/204Pb versus 143Nd/144Nd values compared to ACA lavas, which also may indicate a basement (lithospheric) signature. Trace-element data indicate that melting occurred at a depth within the spinel peridotite field ((Dy/Yb)pmn <1.6), between 70 and 35 km (Rollinson, 2014), which dismisses the possibility of partial melts generated from garnet peridotites and/or garnet-bearing (eclogitic) subducted slab or garnet-bearing lower crust.

We can consider low degrees of partial melting (low F, where F is melt fraction) due to (1) low non-synchronous tensile stresses or faulting (Putirka and Busby, 2007; Busby and Putirka, 2009; Busby, 2013) and (2) a function of temperature and water (volatile) content released into the mantle source by the subducted slab. High K2O volcanism is represented within the older lavas from the GB (age >14 Ma) and lavas from some regions within the eastern ACA (age <14 Ma) (Pyramid Lake and Carson Range). These high-K2O lavas are also enriched in TiO2, Na2O, and P2O5 at Mg# between 0.45 and 0.65 for GB mafic lavas and 0.58–0.72 for Pyramid mafic lavas (Figs. 5 and 12). This observation may signify differences in the petrogenesis of lavas from these two regions. The high-K mafic lavas from the GB erupted during the initial stages of slab rollback after >30 m.y. of flat-slab subduction. The low angle between the slab and overlying CLM would result in low temperatures and therefore lower volatile release from the slab compared to the “normal” arc system (e.g., Pearce and Parkinson, 1993). In summary, the high-K2O, TiO2, Na2O, and P2O5 lavas are due to either source-region enrichments or to low degrees of partial melting.

The <14 Ma mafic lavas from Pyramid Lake erupted near the north-south–trending Pyramid Lake fault (PLF), which formed during development of the Walker Lane (see figure 3B from Faulds et al., 2005). Although the exact age of the PLF is debated, it has been determined that the PLF has accommodated ∼10 km of dextral offset during the past 9 m.y. (Faulds et al., 2005) indicating a region under considerable tensile stress since before 9 Ma. The generation and composition of the Pyramid mafic lavas are consistent with the idea that tensile stresses in the Walker Lane favored eruption of low melt fraction magmas (Takada, 1994; Putirka and Busby, 2007). The 25–15 Ma GB lavas, however, were erupted in regions that were not heavily faulted at the time of emplacement. Great Basin mafic lavas have lower Ba/Nb and Pb isotopic ratios compared to Pyramid Lake mafic lavas, and so it is suggested that the GB mafic lavas were generated under less hydrous, lower P-T conditions during the early phases of the slab rollback.

Basaltic trachyandesites from the Stillwater Mountains do not show the same partial melting trend as the other central GB mafic lavas. These GB basaltic trachyandesites show a lower-F melting percentage (Fig. 12B), as well as similar chemical behaviors as post-ACA lavas (Ormerod, 1988; Cousens et al., 2011) that are proposed to be generated from the lithospheric mantle: higher 87Sr/86Sr (>0.705) and K2O + Na2O (>5 wt%). Ormerod (1988) and Cousens et al. (2011) suggested that basalts with these chemical signatures imply melting (possibly exclusively) within the CLM. It is therefore suggested that basaltic trachyandesites from the central GB formed with a major melt component from the CLM.

Chemical Contributions from Source Components

Variations over Space and Time

It is proposed that flat-slab subduction of the Farallon plate hydrated the North American CLM, perhaps as far east as the Rocky Mountains (Smith et al., 1999; Humphreys et al., 2003; Smith et al., 2004; Lee, 2005). Unlike chemical contributions from a dehydrating slab (primarily LILEs such as Rb, Ba, and Sr), the CLM adds a wider range of its trace-element content to the melt (Ce, Sm, P, Ta, Nb, Zr, Hf, Ti, Y, and Yb). Lee et al. (2001) and Kempton et al. (1991) suggested that the CLM is enriched in Mg and Ni compared to the continental crust and is also enriched in LREEs compared to the primitive mantle. Also, melts from the CLM have low 143Nd/144Nd and variable 87Sr/86Sr compositions compared to the asthenosphere due to metasomatism and the age of the lithospheric mantle (e.g., Omerod, 1988; Omerod et al., 1988; Cousens et al., 2008). Figure 13 presents major, trace, and isotopic data trends along the east-west transect (left column) and as a function of age (right column). Note that the data symbols in the left column represent composition, whereas symbols in the right column represent location. Figure 13 explores assimilation within the deeper crust (Fig. 13A, K2O) and fractional crystallization (Fig. 13B, CaO/Al2O3), both of which have higher values in the basaltic trachyandesites of the GB and some eastern ACA lavas (Carson Range) compared to other lavas from the area. Mafic lavas show a slight increase in Mg# (Fig. 13C) from east to west, and all three rock types overlap. (La/Sm)pmn values vary between 2.0 and 3.5 with distance and time, showing abrupt peaks at longitude 120° and at ages 12 Ma and 3 Ma (Carson Range) (Fig. 13D). This pattern is mirrored by La/Yb and Cs/Rb (not shown), with the highest values in scoria for Ladybug Peak and Carson Range basaltic trachyandesites ((La/Sm)pmn > 4 for samples 01-LT-46–48). The sum of REEs (not shown) and the depth of negative Eu/Eu* anomalies steadily decrease from east to west, whereas (Sr/Nd)pmn and (Sr/P)pmn values increase from east to west as well as from oldest to youngest (Figs. 13E–13G). LILE and/or HFSE values also increase to the west and with younger age. Figures 13H–13J show proxies for sediment melts, represented by Th/Nb, Pb/Ce, and Ce/Ce* ratios. The presence of sediments can affect these ratios; for example, Th/Nb and Pb/Ce are considerably higher in sediments (e.g., Pb/Ce > 0.3) compared to more uniform values for a MORB-like mantle (e.g., Pb/Ce < 0.04, White, 2013). In a similar manner, sediments may inherit seawater’s negative Ce/Ce* anomaly (Class and le Roex, 2008). Under low P-T conditions, neither Th nor Nb are particularly soluble in slab aqueous fluids; however, under higher P-T conditions, subducted sediment will melt, releasing Th into the mantle wedge, while Nb remains within the sediment residue (Pearce and Peate, 1995). Therefore, fractionation between Th and Nb results from the assimilation of sediment melts in an arc system (Johnson and Plank, 1999; White, 2013). The Th/Nb and Ce/Ce* data show that the sediment signature in the mafic lavas increases from east to west. An example of radiogenic isotopic variation is shown in Figures 13K and 13L: initial Sr and Nd (not shown) isotopic ratios have a more restricted range in values toward the east end of the study area compared to the larger variation in values to the west. Lead isotopic ratios show the highest variation in central GB mafic lavas as well as toward the western ACA lavas, compared to very little variation in westernmost GB lavas.

Depth of Melting

Since heavy REEs and Y are compatible in garnet, partial melts of a garnet-bearing source (i.e., garnet peridotite or eclogite) should have high middle and/or heavy REE ratios relative to melts of spinel peridotite. With the exception of one sample from Carson Range ((Dy/Yb)pmn = 1.6), the (Dy/Yb)pmn ratios for the mafic lavas from all regions are <1.5, supporting a garnet-free source (George et al., 2003). Also considering geochemical evidence of a plagioclase-free source (absence of strong Eu anomaly and high Al2O3 and Sr values), it can be concluded that the mafic magmas originated at a depth within the P-T conditions for spinel-peridotite (35 and 70 km, Kinzler, 1997).

Contributions from the Slab

Aqueous fluid-mobile elements, such as Cs, Sr, Rb, Ba, Pb, and U, would be released from the dehydrating slab (basalt and sediments) into the overlying mantle, increasing their abundance in the mantle compared to the HFSE and the REE. Therefore, aqueous fluid transport should result in higher ratios of Ba to Th, Pb to Ce, and Sr to Nd.

In contrast, enrichments in “fluid-immobile” Th relative to a less mobile element (HFSEs, e.g., Th/Nb) can be attributed to a partial melt of subducted sediment (e.g., Plank and Langmuir, 1988, 1993, 1998; Elliott et al., 1997; Hawkesworth et al., 1997; Johnson and Plank, 1999; Elliott, 2003) along with negative Ce/Ce* anomalies and isotopic characteristics of 87Sr/86Sr (up to 0.7060) and 143Nd/144Nd (as low as 0.51270). Figures 14A and 14B (as well as Figs. 13H and 13J) show less influence from slab fluids and the sediment component to be absent in the east and in older rocks (GB mafic lavas), whereas higher influence from slab fluid and a minor sediment signature is present for the ACA lavas. The sediment signature occurs at 120° longitude at ca. 10–12 Ma, which may indicate either P-T conditions were more favorable for sediment melting (>800 °C at 3 GPa, Nichols et al., 1994; Johnson and Plank, 1999) or that more sediment was present on the slab surface.

Involvement of sediment in subduction-related melts may also have had an influence on the 207Pb/204Pb and 208Pb/204Pb ratios. Components from subducted oceanic sediments can emerge in arc volcanic rocks in the form of chemical tracers (e.g., Johnson and Plank, 1999; Kessel et al., 2005). Ocean sediment is chemically variable from a composition that includes volcaniclastic deposits, chert, silt, biogenic deposits, clay, continental discharge, and interstitial seawater (Church, 1976; Lackschewitz et al., 2000; Kessel et al., 2005; Prytulak et al., 2006). In continental subduction zones, seafloor sediments are either scraped off the subducting plate at the ocean trench and accreted along the continental plate margin or subducted along with the oceanic plate; as a result, they contribute to arc magmatism and/or are recycled in the convecting mantle (Plank and Langmuir, 1993). Melts from subducted sediments should leave a chemical signature in the erupted arc lavas; however, the bulk chemical composition of the original sediments needs to be determined. The terrigenous sediments that have been scraped off along the Coast Range in California and accreted as metasediments are >100 m.y. old (Wakabayashi, 1992) and therefore are too old to be associated with the Cenozoic magmatism. We therefore compare our geochemical data to the bulk compositions of sediments from modern subduction environments, namely drill core acquired from the Gorda plate, which is located off the coast of northern California. Sediments recovered from the Escanaba Trough, the southernmost point of the Gorda plate, are olive-green to gray hemipelagic silty clay. Bulk chemical analyses, including trace elements (Th, Ce, and Nb, Lackschewitz et al., 2000) as well as radiogenic isotopic compositions (Prytulak et al., 2006) are plotted in the graphs presented in Figure 15.

Trace-element and isotopic tracers for sediment melts in arc volcanic rocks include Th and Pb (Plank and Langmuir, 1992, 1993; Hawkesworth et al., 1997), which can be presented as high Th/Nb, 207Pb/204Pb, and 208Pb/204Pb ratios. Lead and Th can once again become mobilized within sediment melts (Plank and Langmuir, 1992; Hawkesworth et al., 1997).

Figures 10B and 10C show that all samples have elevated Pb isotopic ratios compared to MORB and plot above (Fig. 10B) or along (Fig. 10C) the NHRL (Hart, 1984). The GB mafic lavas have slightly lower 207Pb/204Pb and 208Pb/204Pb and plot closer to the NHRL than the western ACA lavas, perhaps due to lower Pb input from the slab and less recycled crustal Pb in the mantle farther away from the plate margin. Figures 10D and 10E compare 206Pb/204Pb to 87Sr/86Sr and 143Nd/144Nd, respectively. Combined with the Ce/Ce* data, the ACA mafic lavas that trend to higher 206Pb/204Pb at equal 143Nd/144Nd values are consistent with more sediment in the ACA mantle source.

All mafic lavas have variable Sr and Nd isotope compositions (Fig. 10). The degree of variation increases from east to west, with the greatest variation occurring in mafic lavas from the eastern and western ACA. Figure 15 compares the Sr, Nd, and Pb isotopic data from this study with data from several neighboring physiographic regions summarized in Cousens et al. (2008). All lavas shift away from the MORB field and trend toward an isotopically enriched source. In Figure 15A, most mafic lavas from all locations plot along a mixing line between the modern south Cascade arc mantle wedge and the enriched CLM that is a typical source for alkali basalts in the western Great Basin (Menzies et al., 1983; Ormerod et al., 1991). Mafic lavas from the westernmost part of the ACA, however, plot to the left of the array, with lower Sr isotopic ratios compared to other lavas with a given 143Nd/144Nd. Sources of higher 87Sr/86Sr include seawater, subducted sediments, upper continental crust, and metasomatized CLM. If higher 87Sr/86Sr within the other ACA and GB lavas is derived from mixing with the metasomatized CLM (Fig. 15A), the lower 87Sr/86Sr values in westernmost ACA lavas could be attributed to (1) the absence of or presence of a chemically different metasomatized CLM, (2) melt generation deeper in the mantle wedge, or (3) mixing with crustal terrane distinct from the CLM. The westernmost ACA mafic lavas also show lower La/Sm and Pb/Ce values. The north-south–trending CLM under the westernmost ACA is possibly older and less enriched in LILE and LREE elements, including Sr.

Earlier, it was noted that high abundances of LILE and LREE and low abundances of HFSE in lavas erupted in an arc system can be attributed to involvement of fluids transported into the mantle wedge from the subducted slab (Hughes, 1990; Leeman et al., 1990; Baker et al., 1994; Borg et al., 1997). Similarly, by confirming correlation between ((Sr/P)pmn and (Sr/LREE)pmn, previous workers have used higher (Sr/P)pmn values as a proxy for slab-derived components in the mantle wedge (Borg et al., 1997). We can compare the results from the GB and ACA lavas to neighboring provinces presented in Borg et al. (1997) and Cousens et al. (2008). Figure 15C shows the variation in Sr relative to P (used to reflect variation in Sr-rich slab-derived fluids added to the mantle wedge) at a given 87Sr/86Sr for mafic rocks compared to other regions in the western United States. Increases in (Sr/P)pmn for mafic lavas from the western ACA, from Lousetown and sample 01-LT-53 from the Carson Range, are consistent with higher slab fluid content of mantle sources overlying the subducted slab. The (Sr/P)pmn values from the eastern ACA samples show a negative correlation with 87Sr/86Sr. Mafic lavas from the GB have very low (Sr/P)pmn values compared to other regions but have 87Sr/86Sr that are elevated compared to all other nearby magmatic suites with the exception of the western Great Basin. To the west, (Sr/P)pmn values increase indicating an increasing contribution from slab fluids to the mantle source, which is consistent with trace-element patterns shown in Figures 9 and 13. For example, the trend in Ba/Nb (not shown) suggests that, although all lavas show influence from the subducted Farallon slab, the subduction signature increases from east (older) to west (younger). The similarities in phenocryst compositions, REE, and isotopic signatures amongst the GB mafic compositions further support the derivation of their primary magmas from similar source regions, with variations in chemistry due to a range of contributions from the CLM and variable degrees of partial melting (F). We can conclude that the mafic lavas that erupted within the GB were generated within drier, low P-T conditions compared to the mafic ACA lavas. The western ACA mafic lavas were generated within highest P-T conditions with the highest slab fluid content compared to eastern ACA and GB lavas. The P-T conditions and volatile content relate to the degree of slab dip; that is, shallow dip would result in lower P-T conditions and less dehydration of the slab. A larger subduction angle would result in higher P-T conditions and greater dehydration, releasing more fluids (Ba, Rb, Sr, and La). Then, as the slab continues to descend, sediment melts and contributes a component to the mantle wedge. Pressure-temperatures conditions were not high enough to melt the basaltic slab itself, therefore retaining the HSFE (Nb and Ta).

The 0.706 Line

The 87Sr/86Sr = 0.706 isopleth (where 87Sr/86Sr is the initial Sr-isotope ratio of plutonic rocks) defined by Kistler and Peterman (1973, 1978), commonly known as the 0.706 line, has been interpreted to mark the transition from Precambrian continentally derived rocks on the east to Phanerozoic volcanic arc–derived rocks on the west (Kistler and Peterman, 1973, 1978; Farmer and DePaolo, 1983; Elison et al., 1990). Kistler and Peterman (1973) interpreted the 0.706 line as the western edge of the Precambrian crystalline basement (Fig. 1). The 0.706 line is recognized in Mesozoic and younger igneous granitoid rocks from western North America (Kistler, 1990) and Cenozoic volcanic rocks, as noted by Kistler and Peterman (1973).

The correlation of chemical characteristics in Cenozoic basalts from this study with their proximity to the 0.706 line (Reid and Ramos, 1996) implies that they are at least partially derived from the CLM (Kempton et al., 1991; Reid and Ramos, 1996). Figure 1 includes the 0.706 line (heavy dashed line) taken from Kistler (1990) and confirms that the sampled mafic lavas mostly lie to the west of the 0.706 line with one exception of a 3.82 Ma basalt from Sawtooth Ridge (sample Towle #1, SiO2 = 48.93 wt%, Mg# = 0.74), which has an 87Sr/86Sr value >0.706, and 143Nd/144Nd value <0.5125.

Subduction- versus Extension-Related Magmatism

Between 35 and 3 Ma, arc volcanism migrated southwestward across what is now the Great Basin. At ca. 17 Ma, Great Basin extension began (Colgan et al., 2006a, 2006b); thus, any volcanic rock in the region younger than 17 Ma in age may have been produced by arc processes, extension-related partial melting of the mantle, or a combination of both (Dickinson, 2006; Cousens et al., 2008; Busby and Putirka, 2009; Busby, 2012). Slab rollback completed its migration across the Great Basin by ca. 16 Ma, when the arc front reached the eastern Sierra Nevada (Busby et al., 2008; Busby and Putirka, 2009; Busby, 2013). Since the timing of both arc and extensional processes overlap, it is therefore difficult to establish by age alone which process generated ACA lavas. We have established that all mafic lavas have strong trace-element subduction signatures, which suggest a role for a subduction-modified mantle source.

Caldera-forming magmatism, known as the “ignimbrite flare-up,” dominated the GB during the magmatic sweep east of the Stillwater Range (Dickinson, 2002, 2006, 2013; Henry, 2008; Best et al., 2009), which was emplaced through >100 km thickened crust referred to as the “Nevadaplano” (DeCelles, 2004). It is possible that east of the Stillwater Range, mafic magmas would pond at depth, partially melt and assimilate lower and upper crust, and thus form a caldera field. To the west of the Stillwater Range, no calderas are present, suggesting that the lithosphere was thin enough that mafic magmas could reach the surface. Assuming that the calderas of the ignimbrite flare-up indicate the position of the continental arc over time, then the only clear expressions of younger magmatism (postarc) are the basalt, basaltic andesite to basaltic trachyandesites of the Stillwater Range (La Plata Canyon, southwestern range margin). Calderas of the southern Stillwater Range are ca. 29–25 Ma in age (Hudson et al., 2000; Henry and John, 2013), but the Stillwater basaltic rocks are 15–14 Ma in age, ∼10 m.y. younger than the caldera events. We infer that extension reached the Stillwater Range at ca. 15 Ma, when normal faulting may have allowed mafic magmas to reach the surface. A backarc origin, however, cannot be discounted. We can compare the ca. 15 Ma mafic rocks from Stillwater Range to other known postarc extension-related mafic lavas from small volcanic fields in the Reno-Fallon area (Cousens et al., 2012) and from the Owens Valley (e.g., Big Pine volcanic field, Fig. 16). Stillwater mafic rocks have primitive mantle-normalized incompatible element patterns very similar to the average for other GB mafic lavas, all having a subduction signature. The Stillwater lavas are also similar to extension-related subalkaline basalts from the Big Pine volcanic field in the Owens Valley, where Great Basin extension is proposed to melt metasomatized lithospheric mantle (Ormerod et al., 1991). The average pattern for lava blocks from Upsal Hogback volcano near Fallon, an example of postarc volcanic rocks in western Nevada, has lower Ba/La, La/Nb, and a more within-plate trace-element composition compared to GB volcanic rocks. We conclude that if the Stillwater mafic lavas are extension related, then they are partial melts of the same mantle source that produced the arc-related GB mafic lavas.

Model for GB-ACA Mafic Magmatism

At 40 Ma, a high rate of convergence of the Farallon plate (10–15 cm/yr, English and Johnston, 2004) resulted in flat-slab subduction (English et al., 2003; DeCelles, 2004; Liu et al., 2008) underneath the North American plate; this subduction thickened the lithosphere and amplified the relief in the Sierra Nevada and in what is now the Great Basin (Fig. 17A) (DeCelles, 2004; Henry, 2009). The East Pacific Rise was ∼200 km from the edge of the North American plate, and the thermal structure of the Farallon plate was influenced by its young age, shear heating from the flat-slab subduction (Peacock et al., 1994; Peacock, 1996), and fast subduction rate, allowing incompatible elements to be released from the slab (i.e., dehydration) and chemically enrich the overlying CLM.

Flat-slab subduction, which currently occurs in 10% of modern convergent margins (Gutscher et al., 2000b), can be caused by (1) a high rate of convergence of the overlying plate and/or (2) the buoyancy of thickened oceanic crust of moderate to young age (Gutscher et al., 2000b). In the flat-slab scenario for the Farallon plate, it has been argued that there is no asthenospheric mantle wedge between the North American lithospheric mantle and the Farallon plate (DeCelles, 2004; Liu et al., 2008). This can cause a delay in the basalt to eclogite transition by up to 8 to10 m.y. (Gutscher et al., 2000b) due to the cool thermal structure of two overlapping plates (Gutscher et al., 2000a; Arcay et al., 2007). Dehydration of metabasaltic oceanic crust directly enriches the CLM; however, no melting took place because of lack of heat. High intraplate coupling may have scraped off much of the sediments at the wedge, possibly explaining the absence of a sediment-sourced chemical signature in the GB lavas. As subduction progressed, the oceanic crust and lithospheric mantle underwent metamorphic transformation far into the North American plate, which would have increased the density of the subducted slab. As the hydrous minerals continued to break down, they would have released substantial amounts of H2O and other fluid-mobile elements into the CLM, thereby altering the rheological properties of the boundary between the slab and CLM, which would have reduced the intraplate coupling above a flat-slab segment (Gutscher et al., 2000a). The density of the slab probably increased by up to 10% once the basalt was transformed to eclogite, causing the lithospheric plate to sink at a steeper angle, at which point slab detachment from the base of the North American plate was possible (Spencer et al., 1995).

As the convergence rate between the Farallon and North American plates decreased, the density of the metamorphosed (eclogitic) slab increased, and the slab hinge point started to roll back (Fig. 17B). As the slab hinge point rolled back, hot asthenosphere likely moved in between the North American and subducted Farallon plates, initiating dehydration of the slab and triggering partial melting (F) of the hydrated CLM and the mantle wedge at the CLM boundary. Low-F mafic melts of the GB were generated within the spinel field of the mantle, enriched in incompatible elements including the REEs, K, P, Ti, and HFSEs. Magmatism migrated southwestward as the Farallon slab hinge point continued to roll back (Fig. 17C), and high-K lavas of Bell Canyon, Stillwater, Truckee Range, and Hot Springs Mountain erupted in the GB at ca. 25–15 Ma. The similarities in phenocryst compositions, REE, and isotopic signatures amongst GB mafic compositions further support the derivation of their primary magmas from similar source regions, with variations in chemistry due to a wide range of contributions from the CLM and mantle wedge and variable degrees of partial melting (F). Alkaline melts (basaltic trachyandesites) within the GB represents melting primarily in the CLM.

To the southwest, the Pacific plate margin collided with the North American plate at ∼34° latitude at ca. 25 Ma, initiating a right-lateral strike-slip fault near the present site of Los Angeles (Schellart et al., 2010), creating the MTJ that began migrating northward relative to stable North America at ∼33 mm/yr (Atwater and Stock, 1998). Intermediate volcanism migrated southwestward across Nevada as the Farallon plate hinge point rolled back and dipped more steeply into the mantle, increasing P, T, and resultant dehydration (aqueous fluid) from the slab. Extensional and transtensional stresses likely aided the upward movement of mafic magmas starting at ca. 15 Ma within what is now identified as the Walker Lane. By 12–10 Ma, the rate of subduction slowed as well as the rate of east-west migration, and higher volumes of mafic magmatism erupted in the eastern ACA (Fig. 17D). As the angle of subduction increased, heating of the slab resulted in larger contributions from the slab-derived fluids into the widening mantle wedge. Concurrently, the slab window was expanding to the south of the study area (latitude ∼31°) as the MTJ continued to migrate northward (Atwater and Stock, 1998; Schellart et al., 2010) at a rate of ∼52 mm/yr (Atwater and Stock, 1998). Extension continued to fracture and thin the lithosphere in the northern and southern Basin and Range (Dickinson, 2002; Henry, 2009; Schellart et al., 2010) as well as to increase the lithospheric transtensional stress between eastern ACA and western GB (from Sparks, around Pyramid Lake, Lousetown, Coal Valley, and Carson Range). The angle of subduction increased west of longitude 120°, and P-T conditions at the slab-mantle wedge interface were high enough to allow sediment melt components from the Farallon slab to infiltrate the mantle wedge. By 5 Ma, the rate of subduction slowed to 8 cm/yr (English and Johnston, 2004), and the slab dip angle equilibrated in the mantle (Fig. 17E). Slab components continued to infiltrate the mantle wedge ((Sr/P)pmn = 2–5), and melting resulted in mafic extrusives in the western ACA (Stampede Reservoir, Ladybug region, Portola, and Mount Lincoln). Up to this point, arc magmatism in the entire ACA shared the same two mantle sources (lithospheric mantle and mantle wedge), thus emplacement in extended versus unextended terrane had no influence on geochemistry. Farther to the west, however, a distinct, less enriched lithospheric mantle component may have contributed to magmas in the westernmost part of the ACA (Sierra Nevada, including lavas of Squaw Peak, Pond Terrace, Devil’s Peak, and Lowell Ridge), resulting in overall lower 87Sr/86Sr isotopic ratios. The mafic melt generated at Sawtooth (3.82 Ma) may have been derived from or mixed with an older, more enriched mantle source mixed with a component possessing an exceptionally high 87Sr/86Sr ratio (0.7061), which suggests that the vent may be over the Precambrian CLM, represented by the 0.706 line (Kistler, 1990; Beard and Glazner, 1995). Between 17 and 2 Ma, the combination of arc and extensional processes resulted in magma production from similar sources, all with a subduction trace-element signature. Only extension-related lavas younger than 2 Ma in age from western Nevada include a component of magma derived from non-subduction modified asthenosphere (Cousens et al., 2012).

SUMMARY

The mafic volcanic rocks from this study were geographically divided into three broad regions: central GB (age range 35–5 Ma) and eastern and western ACA (age range 16–2.5 Ma). Aside from a small subset of western ACA lavas, the eastern and western ACA lavas largely overlap in age and elemental and isotopic composition. Basalt and basaltic andesite lavas follow olivine-clinopyroxene + plagioclase fractionating phases, with clinopyroxene noted as the dominant phase. The effects of plagioclase fractionation are minimal for all mafic lavas, though greater in the GB mafic lavas compared to ACA lavas.

Most mafic melts are dominated by a lithospheric mantle component that is revealed in normalized incompatible element patterns and in Sr and Nd isotopic compositions. Lavas from a subsection of western ACA lavas, referred to as the westernmost ACA, are recognized by their lower 87Sr/86Sr values compared to lavas with equal 144Nd/143Nd values, possibly because the lavas are partial melts of less enriched lithospheric mantle compared to the CLM to the east. Trace-element data show that compositions of all mafic melts possess a subduction zone signature. This subduction signature, which includes negative Ta, Nb, and Ti anomalies and LREE + LILE enrichment, is weakest in the GB (older) mafic lavas where the dip angle and dehydration from the slab was the lowest. The subduction signature is strongest in the western ACA (younger) mafic lavas, when the dip angle and slab dehydration into the mantle wedge was comparatively greater. The GB mafic lavas have likely obtained a subduction signature via melting of the metasomatized lithospheric mantle (i.e., the CLM was modified from exchange of incompatible elements from the slab during flat-slab subduction). While the Farallon slab rolled back across the GB, the shallow dip did not allow for large degrees of fluid release. For the ACA lavas, the slab dip was sufficient for fluid release into the mantle wedge, and thus, melting occurred in both the mantle wedge and the CLM. Increased fluids into the mantle wedge also produced higher volumes of melt, which explains the larger volumes of melt in the ACA compared to the GB.

Geochemical indicators that signify contributions from sediment melts (e.g., higher Th/Nb, Pb/Ce, and lower Ce/Ce*) are present for magmas west of longitude 120° (western and westernmost ACA lavas). The 207Pb/204Pb and 208Pb/204Pb values for the western ACA lavas are higher compared to the GB lavas at a given 143Nd/144Nd value but show no evidence of crustal contamination. It is therefore likely that the more radiogenic Pb was sourced from melted sediment rather than the crust.

Mafic lavas from the GB and selected eastern ACA lavas have higher K2O, TiO2, P2O5, HREE, and Nb contents for equal silica and lower Mg# values compared to the rest of the ACA lavas. These results are consistent with low degrees of partial melting due to (1) lower temperatures and water (volatile) abundances generated from the shallow slab for the GB mafic lavas and (2) faulting due to extension. Basaltic trachyandesites are more alkaline with higher 87Sr/86Sr, suggesting these melts resulted from lower degrees of partially melted CLM. Great Basin extension at ca. 17–14 Ma in the Pyramid Lake region and Stillwater Range triggered lower-F melting that resulted in overall enrichment in incompatible elements but no change in incompatible element patterns. In this region, extension primarily causes melting of the CLM until ca. 2 Ma, when melting of the asthenosphere is first recorded in alkaline lavas from the Reno-Fallon corridor.

ACKNOWLEDGMENTS

This project was financially supported by Natural Sciences and Engineering Research Council (NSERC) grants to Dr. Brian Cousens with additional funding from the Mineralogical Association of Canada. Thank you especially to Dr. Christopher Henry and colleagues from the Nevada Bureau of Mines and Geology. Dr. Henry’s expertise in the region and enthusiasm in the field have been extremely valuable and vital to the success of this project. I have benefited from the discussions and encouragement from the Cousens Community, including Dr. Richard Ernst, Christopher Rogers, Susan Kingdon, Alana MacKinder, Cole Kingsbury, and Seamus Magnus as a wonderfully encouraging group for scientific discussion and learning. Thanks also to Keith Bell, Lizzy Ann Spencer, Mike Jackson, Beth Halfkenny, John Blenkinsop, Claire Samson, and Sharon Carr for training, encouragement, and overwhelming support. Alan Donaldson, Ron Pickerill, Cliff Shaw, and David Lentz provided valuable reviews, support, and counsel. Finally, I am immensely honored and grateful for the time and valuable counsel from Geosphere editors, associate editors, and reviewers: Dr. Shanaka de Silva, Dr. Eric Christiansen, Dr. Keith Putirka, and Dr. Richard Bradshaw.

1Supplemental Tables. Sample locations, GPS coordinates, and geochemical data. Please visit https://doi.org/10.1130/GEOS.S.12233756 to access the supplemental material, and contact editing@geosociety.org with any questions.
Science Editor: Shanaka de Silva
Associate Editor: Eric H. Christiansen
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.