Abstract

Trace-element analyses of zircons from ash-flow tuffs and granitic rocks in the east-central Sierran Nevada, combined with U-Pb zircon ages and trace-element abundances in whole rocks, allow us to interpret a shared magmatic heritage for widespread eruption of ash-flow tuffs and assembly of a typical Sierran granodioritic to granitic intrusive suite. The tuffs have 5%–15% phenocrysts of quartz, feldspar, biotite, and amphibole, silica contents ranging from 70% to 74%, and have variable alkali abundances but are consistently light rare-earth enriched with weak negative Eu anomalies, suggesting chemical classification as hydrothermally altered, arc-type, low-silica rhyolites. The tuffs yielded U-Pb zircon crystallization ages of 232, 224, and 219 Ma, indicating that explosive volcanism began before and continued throughout emplacement of granodioritic to granitic plutons in the underlying 226–218 Ma Scheelite Intrusive Suite, one of the oldest and largest intrusive suites in the Sierra Nevada batholith. Zircon crystals from ash-flow tuffs and contemporaneous felsic granodiorite and granite show distinct chemical zoning and extensive overlap in trace-element abundances, indicating crystallization of zircons from very similar melt compositions despite dissimilar late cooling histories. These data provide evidence for a unified volcano-plutonic model for Sierran arc magmatism that describes a long-lived, multi-stage magma system composed of an incrementally emplaced, granodioritic to granitic intrusive suite in the upper crust, overlain by widespread deposits of ignimbrite eruptions generated by thermal rejuvenation of felsic crystal mushes.

INTRODUCTION

Silicic metavolcanic rocks surrounded by granitic rocks in the western U.S. Cordillera provide a record of Mesozoic volcanism and construction of an intermediate composition batholith in a regional convergent margin magmatic arc setting (Kistler and Swanson, 1981; Saleeby et al., 1990; Bateman, 1992; Fiske and Tobisch, 1994; Schermer and Busby, 1994; Schweickert and Lahren, 1999; Tobisch et al., 2000; Fohey-Breting et al., 2010; Fig. 1). The temporal and compositional linkages between volcanism and plutonism are discontinuously exposed to view in the exhumed California Cordillera, and describing them is fundamental to understanding both arc magmatic processes and the potential coupling between the styles of magmatism and the geodynamics of the evolving plate margin.

Precise geochronology and geochemistry of silicic ash-flow tuffs and adjacent intrusive suites can shed light on the question of whether large magma chambers existed at all in the Mesozoic Cordilleran arc, or existed only during transient states of high heat flux through arc upper crust. Coleman et al. (2004) suggested the Late Cretaceous Tuolumne Intrusive Suite in the Sierra Nevada was assembled by relatively small increments of magma over several million years. This model does not require the existence of a large volume of eruptible magma in the crust during assembly of that intrusive suite, consistent with its lack of identified cogenetic extrusive rocks. However, other parts of the California Cordillera suggest closer links between plutonic and volcanic processes. Saleeby et al. (1990) and Bateman (1992) identified other large intrusive suites in the Sierra Nevada for which broadly contemporaneous extrusive rocks are likely preserved, including the mid-Cretaceous Buena Vista, Middle Jurassic Palisades Crest, and Triassic Scheelite Intrusive Suites. Similarly, links between cogenetic Middle Jurassic intrusive suites and large-volume ignimbrites have been established in the Mojave Desert to the south (Schermer et al., 2002; Fohey-Breting et al., 2010). Comparing the age and petrogenesis of extrusive rocks to individual intrusive increments and broader, average pluton compositions sheds light on larger-scale magmatic processes and the assembly of these intrusive suites.

The timing of pluton assembly and emplacement of silicic ash-flow tuffs can also yield insights into the structural evolution of the Cordilleran arc. Thick ash-flow tuff sequences suggest the former existence of calderas (Busby-Spera, 1983, 1988; Saleeby et al., 1990; Fiske and Tobisch, 1994; Schermer and Busby, 1994; Schweickert and Lahren, 1999). However, most thick volcanic sequences in Sierra Nevada pendants dip steeply, and it has been suggested that such volcanic sequences were preserved by a combination of shortening in a compressional upper-crustal strain field and downward flow of upper-crustal framework rocks associated with pluton emplacement (Schweickert and Lahren, 1987; Saleeby et al., 1990; Tobisch et al., 2000; Dunne and Walker, 2004). If so, rather than directly linking plutonic to volcanic processes, the preserved volcanic rock record may then be biased toward magmatic processes during plutonic lulls or volcanic episodes immediately prior to assembly of large upper-crustal plutons.

The relationships of silicic ash-flow tuffs to contemporaneous intrusive suites are therefore fundamental to understanding both the genesis and evolution of arc magmas and the structural evolution of arc crust. In this report, we present whole-rock and zircon geochemical evidence linking ignimbrite eruptions and the assembly of one of the oldest and most areally extensive intrusive suites in the Sierra Nevada magmatic arc. Here, we follow previous workers who have suggested that these intrusive and extrusive rocks provide an example of broadly synchronous plutonism and volcanism in the California Cordillera (Schweickert and Lahren, 1987, 1999; Bateman, 1992; Wonderly et al., 2009). The ash-flow tuffs are weakly to moderately deformed, hydrothermally altered, and metamorphosed, but we propose that a combination of U-Pb zircon ages, immobile element abundances in whole rocks, and trace-element abundances in zircons allows us to interpret a shared magmatic heritage between individual tuffs and assembly of components of an underlying Mesozoic granodioritic to granitic intrusive suite.

GEOCHRONOLOGIC SETTING OF THE CORDILLERAN ARC IN CALIFORNIA

In landmark summaries of early geologic and geochronologic work in the central Sierra Nevada batholith and its pendants and framework rock, Kistler (1978; cf. Schweickert, 1978) and Bateman (1992) concluded that Mesozoic volcanism alternated in time with plutonism, and suggested that periods of enhanced volcanic activity marked distinct tectonic episodes of weak convergence or intra-arc extension. However, support for this hypothesis, of necessity, relied on a variety of geochronometers of dubious reliability, particularly when these geochronometers were applied to metamorphosed and hydrothermally altered volcanic rocks or to magmatic rocks formed in the early and medial evolutionary stages of such a long-lived arc. More recent U-Pb zircon geochronologic studies in both the Sierra Nevada and adjacent Mojave Desert (Fig. 1) suggest a much closer correspondence in age between silicic volcanism and intermediate to felsic intrusive suites in the Mesozoic arc. Great caution is called for when interpreting these data in terms of relative magmatic volumes, because: (1) geochronologic data are generally not collected in a spatially systematic manner; (2) these data represent incomplete records of a largely plutonic arc that is obliquely exhumed, with associated volcanic and shallow crustal plutonic rocks having been preferentially stripped by erosion; and (3) the data set for volcanic rocks is still relatively small. However, at the scale of the entire arc in California, it is apparent that silicic metavolcanic rocks with Late Triassic, Middle to Late Jurassic, and mid-Cretaceous U-Pb zircon crystallization ages are commonly preserved, and their ages are broadly similar to the zircon ages of the most areally extensive (and probably the most voluminous) plutons in the batholith. These silicic metavolcanic rocks then may be the erupted equivalents of arc batholiths, and their precise age and origin in relation to intrusive rocks make them important indicators of melt compositions (e.g., Glazner et al., 2008; Bachmann and Bergantz, 2008b) and important markers of strain fields during deposition and later regional tilting and shortening (e.g., Busby-Spera, 1988; Tobisch et al., 2000; Dunne and Walker, 2004).

EARLY CORDILLERAN ARC PLUTONIC AND VOLCANIC ROCKS

In east-central California, Cordilleran arc initiation and early arc evolution is recorded in both plutonic and volcanic rocks. The Late Triassic, granodioritic to granitic Scheelite Intrusive Suite is exposed in the eastern range front of the central Sierra Nevada and in adjacent ranges surrounding northern Owens Valley. Along the western margin of the Scheelite Intrusive Suite, the Saddlebag Lake and northern Ritter Range pendants preserve early Mesozoic volcanic rocks, including three widespread ash-flow tuff units of Late Triassic age.

The volcanic section overlying Paleozoic to Lower Triassic metasedimentary rocks in the Saddlebag Lake pendant belongs to the Koip sequence (Fig. 2), as originally defined in the Mono Craters quadrangle by Kistler (1966), and further mapped and described by Keith and Seitz (1981), Kistler and Swanson (1981), and Bateman et al. (1983). We studied these rocks principally in the central and southern part of the pendant between Virginia Lakes and Tioga Pass, where they have been mapped and described in detail by Schweickert and Lahren (1999, 2006). Mesozoic metasedimentary and metavolcanic rocks of the Koip sequence have greenschist-facies mineral assemblages and typically have a moderate to strong composite planar fabric (bedding and foliation) that strikes northwest and dips steeply (Brook, 1977). In metamorphosed ash-flow tuffs, this fabric is best defined by flattening of pumice fiamme and rarely by enhanced compositional layering surrounding angular lithic clasts and subhedral phenocrysts. In more deformed sections, both lithic clasts and phenocrysts are moderately to strongly flattened parallel to the pumice fabric. In weakly deformed sections, lithic clasts and phenocrysts of feldspar, quartz, and rarely mafic minerals are well preserved, and these localities provide the basis for petrographic description of the tuffs (we omit the prefix “meta” in the discussion that follows, in the interest of brevity).

The three Koip sequence tuffs analyzed in this study are petrographically similar. At the base of the sequence, tuff of Black Mountain is a light gray, relatively pumice-poor ash-flow tuff with 5%–15% phenocrysts of biotite (and possibly amphibole), feldspar, and quartz. Biotite phenocrysts are least common, subhedral to euhedral, and largely recrystallized. Feldspar phenocrysts are euhedral to subhedral up to 1 mm in long dimension, with alkali feldspar > plagioclase. Quartz phenocrysts are usually the most abundant and are euhedral to subhedral up to 3.5 mm across, locally broken and weakly embayed. Overlying tuff of Black Mountain and intervening conglomerate, tuff of Saddlebag Lake is a very light tan to light-gray ash-flow tuff with dark-gray flattened pumice fiamme up to 15 cm long, and 5%–10% phenocrysts of a mafic mineral (possibly both amphibole and biotite), feldspar, and quartz. The mafic phenocrysts are entirely recrystallized, but appear to have been mostly subhedral to euhedral amphibole up to ∼1 mm across. Feldspar phenocrysts are broken, anhedral crystals up to 3 mm in long dimension, and are typically partially to completely replaced by epidote + calcite. Quartz phenocrysts up to 4 mm across are rarely euhedral, but more commonly are rounded, broken, and locally deeply embayed. At or near the top of the sequence, tuff of Greenstone Lake is a tan to medium-gray and usually relatively pumice-poor ash-flow tuff with dark-gray flattened pumice fiamme up to 5 cm long, and ∼10% phenocrysts of a recrystallized mafic mineral, feldspar, and quartz. Phenocrysts of both plagioclase and alkali feldspar are anhedral, broken crystals up to 2 mm in long dimension. The most abundant phenocrysts are quartz crystals up to 3 mm across that are rarely euhedral, but most commonly rounded, broken, and locally deeply embayed.

Despite the preservation of clastic and less commonly euhedral textures, electron microscope images indicate that fracturing and fluid-assisted alteration of quartz phenocrysts accompanied alteration of feldspar phenocrysts and the tuff matrix (Fig. 3). Quartz phenocryst in silicic tuffs typically exhibit oscillatory zoning in cathodoluminescence (CL) images, e.g., in crystals and fragments from rhyolitic ignimbrites of the Bishop Tuff (Peppard et al., 2001; Wark et al., 2007) and Taupo Volcanic Zone (Shane et al., 2008). In marked contrast, cathodoluminescence images of quartz phenocrysts and crystal fragments from all three ash-flow tuffs in the Saddlebag Lake pendant reveal patchy and diffuse, irregular zonation, commonly crosscut by networks of dark veins. Such quartz textures are similar to quartz formed by fracture-controlled dissolution and reprecipitation in pre-main stage veins at Butte, Montana (e.g., Rusk and Reed, 2002), suggesting fracture-controlled fluid flow dissolved and reprecipitated quartz within these relic phenocrysts. Only rare faint traces of oscillatory igneous zoning are preserved, and these show relatively CL-bright rims, as observed in middle and late stage flow units of the Bishop Tuff (Wark et al., 2007).

The three Koip sequence ash-flow tuffs appear to record large-volume (tens of km3) and explosive ignimbrite eruptions (Schweickert and Lahren, 1999). Based on mapping by Keith and Seitz (1981) and Schweickert and Lahren (2006) and our field observations, tuff of Saddlebag Lake is ∼100–150 m thick over much of its strike length north of Tioga Pass. Based on the cross sections of Schweickert and Lahren (2006), restored outcrop area suggests a minimum preserved dense rock equivalent volume on the order of ∼30 km3 for this ash-flow tuff. The tuffs of Black Mountain and Greenstone Lake are typically thinner and not as continuous in map view, and we estimate each of these two tuffs to have had smaller preserved volumes, on the order of ∼10 km3.

To the east and structurally beneath the Koip sequence are plutonic rocks of identical age (Fig. 4). Bateman (1992) grouped most granitic rocks intruding the eastern side of the Saddlebag Lake pendant as the Scheelite Intrusive Suite, the oldest and one of the largest intrusive suites in the central Sierra Nevada batholith. As recently described and slightly revised by Barth et al. (2011), the Scheelite Intrusive Suite comprises the Wheeler Crest Granodiorite, including the Mount Olsen pluton (226 Ma) and granodiorites of the Benton Range (225 Ma) and Wheeler Crest (218 Ma), the granite of Lee Vining Canyon (218–220 Ma), and the Pine Creek Granite (>215 Ma). Comparison of zircon U-Pb data indicates that assembly of this intrusive suite was very closely linked in time to eruption of ash-flow tuffs now preserved in the adjacent Saddlebag Lake pendant. The three tuffs we studied yielded U-Pb zircon crystallization ages of 232, 224, and 219 Ma. The 232 Ma tuff of Black Mountain is older than any plutonic sample, although intriguingly two antecrystic zircons with ages of ca. 232 Ma were identified in the 225 Ma granodiorite of the Benton Range. The 224 Ma tuff of Saddlebag Lake is identical in age to the older granodiorite and granite in the northern part of the Scheelite Intrusive Suite (granodiorite of the Benton Range and Mount Olsen pluton). The 219 Ma tuff of Greenstone Lake is identical in age to the younger granite in the same area (granite of Lee Vining Canyon). Furthermore, both the tuff of Greenstone Lake and the granite of Lee Vining Canyon contain an older population of antecrystic zircons that yield ages of ca. 225 Ma, identical in age to the tuff of Saddlebag Lake and the older granodiorite and granite. Thus, the zircon U-Pb data suggest a very close correspondence in age between extrusive and intrusive rocks of the Late Triassic magmatic suite, including both magmatic ages and the ages of individual antecrystic zircons recycled into younger magma batches. We explore these correspondences in more detail below using whole-rock compositions and the trace-element compositions of zircons.

ANALYTICAL METHODS

Major elements and Rb, Sr, and Zr abundances of tuff whole rocks were measured by X-ray fluorescence (XRF), and additional trace-element analyses by inductively coupled plasma–mass spectrometer (ICP-MS) at Michigan State University. In-run precision was monitored using JB-1a basalt and JA3 andesite rock standards, and is better than 1% for most major elements and trace elements analyzed by XRF that are present in abundances >100 ppm, and is mostly 2%–5% for trace elements analyzed by ICP-MS. Ignition loss was measured after heating at 1000 °C on powders dried at 105 °C. H2O and CO2 were measured in ash-flow tuffs using a combustion analyzer. Whole-rock major and trace-element analyses were reported by Barth et al. (2011), and all analyses are available for download at www.navdat.org.

Zircons were separated from whole rocks using standard crushing and density separation techniques, and quartz was handpicked from crushed whole rocks. Zircon and quartz were mounted in epoxy and imaged with a cathodoluminescence detector on a scanning electron microscope. These images were used as a guide for selection of analysis points. Trace-element concentrations in zircons (Supplemental Table 11) were measured using the U.S. Geological Survey sensitive high-resolution ion microprobe–reverse geometry (SHRIMP-RG) ion microprobe at Stanford University, during the same magnet cycles as used for U, Th, and Pb isotopic measurements. Zircons were ablated using a ∼30-micron diameter, 5–6 nA O2 primary beam, and count rates for Hf, Y, and rare-earth elements were measured during the beginning of each mass cycle. Count rates for these elements and Th and U were normalized to 90Zr2O, and concentrations were standardized against Madagascar Green (MAD) zircon (Mazdab and Wooden, 2006; Barth and Wooden, 2010). Th and U concentrations in Supplemental Table 1 (see footnote 1) are not identical to those calculated for geochronologic analyses, because these trace-element analyses utilize a larger set of standards for comparison and were not reduced with isotopic data.

RESULTS

Whole-Rock Geochemistry of AshFlow Tuffs

Classification of the ash-flow tuffs is based on a combination of phenocryst abundances, silica content, and major-element abundances. Silica contents range from 70% to 74% (Fig. 5), and quartz phenocrysts are abundant, suggesting chemical classification of the tuffs as Group 5, arc-type, low-silica rhyolites (Hildreth, 1981). Silica contents are similar to the Group III ignimbrites of Bachmann and Bergantz (2008a), also suggesting an affinity to arc-type, low-silica rhyolites, but Na contents are distinctly lower, and Ca ranges to higher values than observed in modern arc-type, Group III ignimbrites such as Holocene tuffs at Crater Lake in the Cascade Range and the Taupo caldera in New Zealand. Sodium abundances decrease with increasing SiO2, and lower Na is accompanied by lower Ba and Sr but little change in K and Rb. The net result of these changes is highly variable whole-rock Rb/Sr, which ranges from 0.3 to 2.4. Because whole-rock Rb/Sr isochrons are in agreement with U-Pb zircon ages of individual tuffs (Kistler and Swanson, 1981; Barth et al., 2011), it is likely that the volcanic section was hydrothermally altered during or soon after deposition, by sodium depletion locally accompanied by variable enrichment in carbonate and silica. The hypothesis of hydrothermal alteration is supported by petrographic evidence for fracturing and recrystallization of quartz and replacement of feldspar phenocrysts by epidote + calcite.

Relatively fluid-immobile, high field strength element abundances support classification of the ash-flow tuffs as arc-type, low-silica rhyolites (Fig. 6). Silica contents are distinctly lower than medium- to high-silica rhyolites such as Bishop Tuff (Hildreth and Wilson, 2007), yet much higher than phenocryst-rich dacites such as Fish Canyon Tuff (Bachmann et al., 2002). Silica contents and titanium and aluminum oxide abundances indicate the three ash-flow tuffs have very similar overall ranges in bulk composition, and a close compositional affinity to relatively felsic granodiorites and mafic granites of the Scheelite Intrusive Suite. All three tuffs have ranges of rare-earth element abundances that are similar (Fig. 7), being characterized by smooth overall trends on chondrite-normalized plots, light rare-earth enrichment, and weak negative Eu anomalies. These trends also support correlation of the tuffs to modern low-silica, arc-type rhyolites rather than high-silica rhyolites such as the nearby Bishop Tuff. Titanium, zirconium, and rare-earth element abundances further illustrate that the ash-flow tuffs have whole-rock compositions that are indistinguishable, despite their distinct crystallization ages, and provide support for the similarity in bulk compositions of ash-flow tuffs to felsic granodiorites of the Scheelite Intrusive Suite (Fig. 8).

Zircon Trace-Element Geochemistry

Zircon crystals from granitic rocks and ash-flow tuffs show distinct geochemical zoning in cathodoluminescence images (Fig. 9). Cathodoluminescence zoning ranges from larger (≥25 μm across) homogeneous core regions to thin (≤10 μm) oscillatory zones, and similar zoning patterns are seen in zircons from both granitic rocks and tuffs. Many grains in each sample show a characteristic internal stratigraphy characterized by an interior with euhedral, oscillatory zoning that is variably truncated along a dissolution surface often decorated by an irregular, bright luminescent zone. This truncation surface is overgrown by an oscillatory-zoned, euhedral rim. This crystal zoning pattern is evident when examining the geochemistry of the zircons by ion microprobe. We analyzed the larger core regions and groups of oscillatory bands using a ∼30- micron diameter ion beam, and the minor- and trace-element abundances in these spots are shown in Figures 10 and 11.

Previous studies have shown that minor- and trace-element abundances in zircons from granitic rocks and tuffs analyzed at ∼30-micron scale show regular variations with the overall pattern of crystal growth and with falling temperature (e.g., Claiborne et al., 2006, 2010; Grimes et al., 2009; Fohey-Breting et al., 2010; Barth and Wooden, 2010). In general, crystal growth with falling temperature is accompanied by an increase in Hf solid solution. This increasing Hf solution is correlated to other trace-element variations, especially a decrease in Th/U, relative enrichment in heavy rare-earth elements, and constant or increasing depth of the Eu anomaly. These coupled solid-solution and trace-element variations are useful to characterize patterns of melt compositional evolution, and when combined with U-Pb isotopic data can be used to recover petrogenetic information independent of hydrothermal and/or metamorphic alteration of the phenocrysts and matrix of the host igneous rock.

Hafnium concentrations in zircons from granodiorites and granites range from 7000 to 13,000 ppm (Fig. 10). As was observed in the earlier studies cited above, relatively Hf-enriched zircons have lower Th/U and are enriched in heavy rare-earth elements (higher Yb/Gd). The overall compositional ranges and trends are similar in all samples, yet regional differences are apparent. Samples from the granodiorite of Wheeler Crest and the Pine Creek Granite, both from the southern part of the Scheelite Intrusive Suite, have Hf-richer zircons, and early crystallized, low Hf zircons have somewhat higher Th/U. Accounting for these regional differences, zircons from granodiorite and granite in the northern part of the Scheelite Intrusive Suite show largely overlapping trace-element concentrations. Zircons in the relatively mafic granodiorite sample do not show consistently less evolved trace-element concentrations when compared to coeval northern granites, as might be expected if these rocks were related by closed-system fractionation (Barth and Wooden, 2010). These observations indicate that there were not large differences in interstitial melt compositions over the later zircon crystallization interval in the granodiorite and granites.

Comparison of the trace-element concentrations of zircons from granitic rocks and zircons from contemporaneous tuffs yields insights into similarities and differences in melt evolution between intrusive and extrusive components of the magma system (Fig. 11). Hafnium concentrations in zircons from the tuffs of Saddlebag Lake and Greenstone Lake are very similar, ranging from 8000 to over 10,000 ppm, and increasing Hf is correlated with lower Th/U and higher Yb/Gd. The trace-element concentrations of the zircons from the tuffs show complete overlap with zircons from intrusive rocks of the northern Scheelite Intrusive Suite, the granodiorite of the Benton Range, Mount Olsen pluton, and the granite of Lee Vining Canyon. These geochemical similarities support petrogenetic linkage of tuffs with intrusive rocks, as described above based on whole-rock compositions and zircon U-Pb systematics. However, in detail, average tuff whole-rock compositions are more fractionated than average granodiorite, and more similar in composition to relatively felsic granodiorite and mafic granite. Zircon compositions reinforce this conclusion; Eu anomalies in zircons from the tuffs are deeper than those that are typically observed in zircons from the granodiorites, and overlap anomalies observed in late crystallized, high Hf zircons from the granites (Fig. 12). These observations indicate that zircons in both the tuffs and granites crystallized from compositionally similar late interstitial melts.

DISCUSSION

New petrologic data for granitic rocks and ash-flow tuffs provide firm evidence for the presence of a long-lived, multi-stage, upper-crustal magma system. The magma system was composed of an incrementally emplaced, granodioritic to granitic intrusive suite in the upper crust, overlain by remnants of its metasedimentary framework rocks and by widespread deposits of ignimbrite eruptions. The intrusive and extrusive rocks that formed in this magma system are compositionally and texturally similar to those produced in later episodes of Sierran arc magmatism. Later rotation of the volcanic section to steep dips, along with at least some of the underlying granitic rocks, has exposed this magma system in a near-vertical section. This superposition of high degrees of tectonic rotation on a long-lived intermediate to felsic magma system makes this tilted section through the upper-crustal portions of a Triassic magma system particularly useful for establishing a unified volcano-plutonic model for Sierran arc magmatism.

Timing and Setting of Magmatism

The early Mesozoic volcanic rock sequence in the Saddlebag Lake pendant is comparable in age and structure to sequences in a series of pendants along and east of the Sierra Nevada range crest, including the Ritter Range, Mount Morrison, and Oak Creek pendants. All of these pendants appear to be remnants of a broad, west-facing, steeply dipping volcanic carapace of the east Sierran early Mesozoic magmatic arc (Longiaru, 1987; Saleeby et al., 1990; Schweickert and Lahren, 1993; Tobisch et al., 2000). Although the precise timing and mechanics of rotation of the volcanic sequences is still uncertain, the steeply dipping sections provide a temporally extended record of volcanism during arc construction over a long strike length of the central east Sierran magmatic arc.

In the Saddlebag Lake–Tioga Pass region, Late Triassic plutons of the Scheelite Intrusive Suite intrude a Paleozoic shelf to basinal sequence of metasedimentary rocks that is unconformably overlain by the steeply dipping early Mesozoic volcanic sequence (Kistler, 1966; Brook, 1977; Brook et al., 1974; Bateman et al., 1983; Schweickert and Lahren, 2006; Fig. 2). Regional down-to-the-west tilting of ∼80° in this region has therefore exposed a near-vertical crustal section though the volcanic and at least some part of the upper-crustal plutonic components of the magmatic system. We have shown that the basal, Late Triassic part of the volcanic sequence contains widespread silicic ash-flow tuff units that are compositionally equivalent to silicic granitic rocks of the underlying batholith, and most of the Triassic sequence was deposited during batholith construction. The volcanic sequence and underlying batholithic rocks of the steeply tilted, upper-crustal section preserved here can then be projected into a cross section of the extrusive and intrusive components of the magma system (Fig. 13). This cross section of the magma system is schematic, due to lack of continuous exposure and uncertainty as to the magnitude of younger east-west extension in the region, which has extended the middle part of the projected section.

The 232 Ma tuff of Black Mountain is older than any dated plutonic rocks, a relationship similar to that observed by Fohey-Breting et al. (2010) in Jurassic intrusive and volcanic rocks farther south in the arc. We should be cautious in interpreting these results; they may be coincidental, and older plutonic rocks are yet to be identified in both regions. Indeed, it is also possible to envision missing volcanic rocks lost to erosion in the relatively limited exposure in pendants, so there could be a bias against preservation of yet older volcanic rocks. Nevertheless, there is some theoretical evidence that might bear on the existing observation, wherein models of the evolving rheology of initially cool wallrocks in the runup to an intrusive pulse might favor eruption over buildup of plutons (Jellinek and DePaolo, 2003; Annen, 2009). In these models, early and relatively smaller volume volcanic eruptions would precede later and relatively larger magma bodies, larger eruptions, and associated plutons. In a sense these may not be mutually exclusive relationships, because plutons associated with early magmatism would be of small volume and easy to miss in the field, especially with only partial exposure of the underlying intrusive parts of a long-lived magma system.

Generation of Intermediate to Felsic Magmas

The ultimate origin of intermediate to felsic magmas that intruded the upper crust of the Sierran arc remains uncertain, and a complete review is beyond the scope of this paper. The 8–12 m.y. life cycle of this Late Triassic magma system, similar to what has been observed in nearby and better known Cretaceous Sierran intrusive suites (Kistler et al., 1986; Coleman et al., 2004; Gray et al., 2008), favors a model with periodic transfer of magma batches formed by mixing of mantle-derived magmas with lower crustal partial melts (Barth et al., 2011). However, developing a more specific model for the generation of parental magmas delivered to the upper-crustal system is hampered by several factors. First, a lack of clarity exists as to the composition and isotopic character of the mantle-derived mafic primary magmas; several lines of evidence suggest an important role for relatively radiogenic lithospheric mantle. These include both direct sampling of mantle xenoliths with enriched isotopic compositions in younger volcanic rocks in this region (Beard and Glazner, 1995), and indirect analysis of compositional trends in younger Sierran arc intrusive suites (Coleman and Glazner, 1997). Second, the isotopic character of the crustal framework is not well established. In particular, isotopic studies suggest that the average crustal component may have been similar to, yet locally either more or less radiogenic than the mantle-derived mafic parental magmas (Kistler et al., 1986; Coleman et al., 1992; Coleman and Glazner, 1997). Neodymium and strontium isotopic data for the early Sierran arc suite of Late Triassic age in the east-central Sierra implicate partial melts of relatively unradiogenic crustal source rocks compared to any likely parental mafic magmatic end member (Barth et al., 2011).

Upper-Crustal Silicic Magmatism

The whole-rock and zircon geochemical data presented here comprise three complementary data sets relevant to constructing an integrated volcano-plutonic model for silicic magmatism in the upper crust:

(1) Ages of individual zircons indicate a long-lived (∼8–12 m.y.) magma system that was periodically active, with active phases characterized by both intrusion and periodic eruption of ignimbrites, and with recycling of older system components (including zircons) into younger intrusive magma batches and into rhyolitic eruptive rocks.

(2) Bulk-rock compositions of rhyolitic ash-flow tuffs diverge from intrusive rock compositions for low-charge, fluid-mobile elements, but high field strength element abundances show broad regions of overlap between silicic eruptives and felsic granodiorite to mafic granite components of the intrusive suite, suggesting comparable degrees of upper-crustal fractionation of plutonic rocks and the magma batches that supplied ignimbrite eruptions.

(3) Trace-element compositions of zircons from intrusive rocks show broad overlap with zircons from volcanic rocks, also suggesting comparable degrees of fractionation, an observation seemingly at odds with the clear petrographic distinction between phenocryst-poor ash-flow tuffs and intrusive rocks.

The tilted nature of the crustal section, together with the whole-rock and zircon compositional data, allows us to sketch out a model for the volcano-plutonic magma system. We envision an upper-crustal magma system consisting of volumetrically dominant intrusions of granodiorite and granite, which constitute magma batches generated in and transported from a lower crustal source region. The scale and pattern of the magma batches is derived from the map pattern in the tilted section, but the size range and number of batches is not well constrained by the degree of exposure and sampling scale of our regional geochronologic data set; individual batches could be significantly smaller. The map pattern indicates that granitic plutons are concentrated in the uppermost part of the system, implying a crude compositional stratification to the intrusive suite, but no obvious spatial pattern to the distribution of ages of magma batches has been recognized.

Contemporaneous with the emplacement of silicic magma batches, mafic satellite plutons were emplaced in the uppermost part of the crust. We infer that these plutons represent small magma chambers beneath mafic eruptive centers, because mafic flows and breccias are widely exposed within the Koip sequence (Bateman et al., 1983; Schweickert and Lahren, 2006). The compositions and petrogenesis of these satellite plutons are not well known, and warrant further study to better constrain the mafic inputs into the system.

All three data sets presented here suggest that the Saddlebag Lake and Greenstone Lake silicic ash-flow tuffs are compositionally and temporally the erupted equivalents of intermediate magma batches that formed parts of the underlying intrusive suite. Correlation of whole-rock and zircon trace-element compositions suggest compositionally similar magma batches, whether forming either granodiorites or tuffs, had similar zircon crystallization histories at the same times of heightened magmatic activity in the system. The observation of similar zircon compositions implies broadly similar melt compositions and melt evolution, yet silicate phenocryst abundances suggest very different late thermal histories for the tuffs compared to the granitic rocks. This apparent contradiction between thermal histories recorded by zircon and rock-forming silicate phases may be explained in two ways: (1) zircons crystallized early, at comparatively high temperatures, before the majority of the rock-forming silicates, or (2) magmas erupted to form the tuffs typically were generated by thermal and/or compositional rejuvenation of cool and largely crystallized, near-solidus mushes or plutonic rocks. Early, comparatively high-temperature crystallization of zircon seems unlikely. These felsic and zirconium-poor rock compositions have low estimated zircon saturation temperatures (<800 °C), which is a common characteristic of arc plutons in this region. Regionally, magmatic zircons from compositionally similar arc plutonic rocks also predominantly yield low calculated crystallization temperatures from Ti thermometry (Anderson et al., 2008; Miller et al., 2008; Barth et al., 2010), consistent with growth of zircon at or below nominal zircon saturation. Crystallization of zircon at these relatively low temperatures is also in accord with the propensity of these plutons to yield high zircon age dispersion among magmatic grains and to preserve antecrysts and premagmatic zircons (Miller et al., 2007; Anderson et al., 2008; Barth and Wooden, 2010; Barth et al., 2010). These observations are inconsistent with early high-temperature growth of zircon, but are consistent with comparatively low-temperature growth in cool, hydrous magmas that were never heated above zircon saturation (∼800 °C) for very long periods of time.

We suggest, therefore, that zircon age and compositional data are compatible with a model for generation of ash-flow tuffs by thermal and/or compositional rejuvenation of minimally molten crystal mushes, by the introduction of heat and/or volatiles from crystallizing mafic intrusions. These mush rejuvenation episodes either did not require temperatures high enough to melt zircon, or at least did not require such high temperatures for a long enough period of time to dissolve previously crystallized zircons (e.g., Miller et al., 2007). We model this process in Figure 13 by injection of cool, hydrous, and buoyant mafic magma into a silicic crystal mush prior to eruption. Although we have not yet documented such mafic intrusions in this system, our model is based on younger synmagmatic mafic intrusions described in the eastern Sierra Nevada (Moore, 1963; Wiebe et al., 2002; Frost and Mattinson, 1993; Coleman et al., 1995). These intrusions are characterized by the presence of dioritic intrusive rocks derived from crystallization of cool and hydrous mafic magma that intruded upper-crustal granitic magma chambers (Sisson et al., 1996). These volatile-rich mafic magmas formed sheeted intrusions in the partially crystallized granodioritic to granitic magma batches. Final crystallization of the mafic rocks would evolve abundant volatiles, and the heat and evolved volatiles passing into stored felsic crystal mush would provide the thermal and compositional trigger to generate the low crystal fraction necessary to drive explosive eruption of phenocryst-poor ash-flow tuff.

CONCLUSIONS

Petrogenetic connections between intrusive and extrusive rocks in the east-central Sierra Nevada define an upper-crustal magma system—a calc-alkalic arc intrusive suite and ignimbrites with overlapping U-Pb zircon ages and bulk-rock compositions. Ranges of measured ages of single zircon crystals indicate an 8–12 m.y. life span for this system, and trace-element concentrations indicate similar crystallization histories for zircons in granitic rocks and in ash-flow tuffs. These observations suggest thermal and compositional cycling commonly led to recycling and rejuvenation of intrusive rocks, and occasionally led to eruptions of widespread ignimbrites.

Modern volcanic systems exhibit geochronologic characteristics analogous to those described herein, where some or all zircon ages are older than eruption age estimated by other means, including Geysers (Schmitt et al., 2003), Coso (Miller and Wooden, 2004), southwest Nevada (Bindeman et al., 2006), and modern convergent margin arc settings such as the Taupo Volcanic Zone (Charlier et al., 2005) and Aucanquilcha (Walker et al., 2010). Recycling of older, partially to completely solidified parts of these magma systems is inferred from the persistence of these age differences, and from the magnitude of the age differences in comparison to estimated solidification rates. At Taupo and Aucanquilcha, zircon compositions also suggest crystallization to near-solidus temperatures, and therefore that rejuvenation of crystal mushes is an important process in modern arc volcanic systems. This of course also indicates that the growth of zircons found in these Sierran ash-flow tuffs may actually have predated eruption and deposition by some 104 to 105 years (e.g., Simon et al., 2008), but this difference is indiscernible given the uncertainty of our calculated U-Pb ages. We therefore conclude that the volcanic record of the Koip sequence and the plutonic record of the Scheelite Intrusive Suite are in concert, indicating a long-lived and episodically active felsic magma system in the Cordilleran arc that behaved in manners similar to many modern felsic volcanic systems.

Funding for this study was provided by the National Science Foundation through grants EAR-0711115 and EAR-0711119 to Barth, EAR-0711541 to Riggs, and EAR-0711107 to Walker, and by the Blaustein Fund at Stanford University. We thank the University of California White Mountain Research Station for accommodations, John Barth, Mat Beshears, and Ali Szymanski for assistance in the field, Brad Ito for expert maintenance of the ion microprobe, and Rita Economos, Nicole Fohey-Breting, and Sarah Needy for helpful conversations. APB thanks the faculty and students at Stanford University for their generous hospitality and in particular Elizabeth Miller, Gail Mahood, Keith Putirka, and Richard Schweickert for several enlightening discussions. Thoughtful comments by journal reviewers Jason Saleeby and Benjamin Andrews led to improvements in the final manuscript.

1Supplemental Table 1. PDF file of zircon geochemistry for igneous rocks from the east-central Sierra Nevada. If you are viewing the PDF of this paper or reading it offline, please visit http://dx.doi.org/10.1130/GES00737.S1 or the full-text article on www.gsapubs.org to view Supplemental Table 1.