Abstract

Detailed geologic mapping on the ∼1-km-tall, vertical southeast face of El Capitan was completed to determine the chronology and geometry of emplacement. Field relations reveal a complex intrusive history at the boundary between two intrusive suites involving interaction between several granitic units. No resolvable faulting or other postemplacement deformation was observed. New U-Pb zircon geochronologic data (laser ablation and isotope dilution) demonstrate assembly of the El Capitan Granite and diorites of the Rockslides and North America between ca. 106 and 103 Ma. New ages for the Taft (106.6 ± 0.7 Ma), Leaning Tower (104.1 ± 0.10 Ma), and Bridalveil (103.4 ± 0.4 Ma) plutons reveal that they intruded over the same interval as the other plutonic rocks exposed on the face of El Capitan, although field relations and geochronology suggest a distinct order of emplacement.

Two sets of aplite dikes are also exposed. Their whole-rock compositions suggest segregation at depths of 5–6 km and derivation from the intrusive suites of Yosemite Valley or Buena Vista Crest. Chemical analysis of samples collected along ∼1-km-tall vertical transects through the El Capitan and Taft Granites reveals no systematic variations in major or trace elements. Analysis of 78 photographs within the El Capitan Granite also shows no systematic variations in texture or mineralogy with elevation. Lack of resolvable vertical variations in both field and petrologic observations is consistent with incremental assembly, and is hard to reconcile with models that envision magma chambers as large fractionating bodies.

INTRODUCTION

Plutonic rocks are fundamental building blocks of Earth’s crust, yet much about processes associated with their emplacement and evolution remains poorly understood. Plutons were long viewed as the remains of massive, mobile, largely liquid magma bodies that rose into the crust (Daly, 1933; Buddington, 1959; Bateman, 1992) and were emplaced over time scales on the order of 105 years (e.g., Harrison and Clarke, 1979; Spera, 1980). Recent geochronologic studies, however, show that many plutons were likely assembled over time scales from 105 to 106 Ma, in batches that are significantly smaller than the full pluton volume (Coleman et al., 2004; Glazner et al., 2004; Matzel et al., 2006; Bolhar et al., 2010; Davis et al., 2012; Leuthold et al., 2012; Schoene et al., 2012; Frazer et al., 2014). These studies are difficult to reconcile with the view that plutons are frozen remnants of large (∼103 km3), long-lived (∼106–107 m.y.) melt-interconnected reservoirs (e.g., Paterson and Vernon, 1995; Hildreth and Wilson, 2004; Miura and Wada, 2007; Walker et al., 2007; Memeti et al., 2010). Detailed mapping, coupled with geochronologic and geochemical studies of well-exposed plutonic systems, can help to evaluate these models of magma emplacement.

In situ crystal-liquid fractionation of a single intrusion is commonly invoked to explain formation of both lateral and vertical zoning in plutons (e.g., Bateman and Chappell, 1979; Atherton, 1981; Tindle and Pearce, 1981; Sawka et al., 1990; Verplanck et al., 1999; Economos, et al., 2009), and vertical profiles of ignimbrites frequently display compositional patterns interpreted to reflect vertical zonation in magma chambers (Bacon and Druitt, 1988; Hildreth and Wilson, 2007). If a pluton crystallized from a largely liquid body at some point, such fractionation should produce both vertical and horizontal zonation in mineral abundances and whole-rock geochemistry, as early crystallizing minerals form and residual melt segregates from them (e.g., Ragland and Butler, 1972; Baker and McBirney, 1985). Several studies (e.g., Sawka et al., 1990; Bachl et al., 2001; Hirt, 2007) have also reported vertical variations in crystal size and modal abundance that were inferred to result from vertical segregation of crystals and liquid within large magma bodies. Vertical variations in incompatible elements produced by this process could also redistribute heat-producing elements (Sawka and Chappell, 1988).

By contrast, in an incrementally emplaced pluton, compositional variability cannot have been produced wholly by crystal-liquid separation at the level of observation (Glazner et al., 2004). The overall compositional zoning of many intrusive bodies is thus inferred to result from other processes, such as changes in the composition of the magma supplied from a deeper source (Clemens et al., 2010; Tappa et al., 2011; Coleman et al., 2012), superimposed with local variability produced by crystal-liquid separation.

Yosemite Valley, California (Fig. 1), provides opportunities to generate true vertical mapping and geochronologic and geochemical transects through largely undeformed plutons, and thus, to test these contrasting views of pluton genesis. Studies of spatial zoning in granitic pluton geochemistry and mineral composition are generally based on observation of outcrop surfaces that have little relief in the vertical dimension compared to the horizontal (e.g., Tindle and Pearce, 1981), although relief of over 2 km can be produced by deep river incision (Cornejo and Mahood, 1997). Undeformed plutons with vertical cliff exposure comparable to or greater than the face of El Capitan are rare and remote (Michel et al., 2008; Searle and Parrish, 2010). Direct observation of significant vertical pluton extents is thus generally limited to plutons exhumed by tilting (e.g., Flood and Shaw, 1979; John, 1988; Seager and McCurry, 1988; Rosenberg et al., 1995; Bachl et al., 2001). Although these examples may offer much deeper exposure than those found in single cliffs, such studies are subject to uncertainties in structural reconstruction and interpretation, and the deformation that produced the exposure commonly obscures original features in the tilted rocks.

Sawka et al. (1990) examined vertical variability in relatively undeformed Jurassic plutons exhumed by normal faulting along the eastern margin of the Sierra Nevada in California. They interpreted variations in major- and trace-element data with elevation as being consistent with those predicted by crystal-liquid fractionation models. Although elevation in this area reaches 4205 m with 900 m of topographic relief, the horizontal distance over which this elevation is gained is ∼1400 m, resulting in an average slope of 32°. Thus, this ambitious transect is still a relatively low-angle oblique cut through the plutonic system, where differences in elevation are not truly vertical differences in the original system.

In Yosemite Valley, Quaternary glaciation, river erosion, and ongoing rockfall have produced a steep-sided valley with over 1 km of local relief (Fig. 1). The vertical northwest cliff face of Half Dome is 680 m tall, and continuous exposure from the top of Half Dome to Tenaya Creek covers 1340 vertical meters at an average angle of 51°. This entire section is sculpted in one pluton, the Half Dome Granodiorite (Calkins, 1985). El Capitan (Figs. 1 and 2) exposes a 1-km vertical section of plutonic rocks (Calkins, 1985; Peck, 2002) in a massive cliff that is locally overhanging but is typically steeper than 75°.

The southeast face of El Capitan has continuous and steep exposure, fully visible from the valley floor, which is crossed by over 70 named climbing routes and is thus more accessible for observation, mapping, and sampling than other km-tall cross sections of granitic systems around the world. Its south-facing aspect provides rock that is largely unobscured by water staining and lichen growth that covers parts of nearby north-facing cliffs such as Half Dome and Glacier Point. Prior work (Calkins, 1930; Reid et al., 1983; Bateman, 1992; Ratajeski et al., 2001) revealed that the southeast face of El Capitan exposes the intersection of two major intrusive suites and two mafic dike swarms. However, understanding contact relationships among these units and their emplacement chronology was hampered by the extreme vertical nature of the study area. In intervening years, new remote sensing tools were devised, the precision of U-Pb geochronology was significantly improved, and rock climbing techniques were refined, making El Capitan a much more accessible exposure on which to study interactions between intrusive units and to test hypotheses regarding the mechanisms and time scales of pluton emplacement in a near-vertical transect.

In this article, we present the field relations revealed by decimeter-scale mapping of the southeast face of El Capitan (Putnam et al., 2014), supplemented by new geochemical analyses and U-Pb dating of zircon. These data reveal 3 Ma assembly of the plutonic system and show no evidence for gravity-driven separation of crystals and liquid over the 1 km vertical extent of the cliff. In this contribution, we use climbing route designations (italicized) as landmarks in describing the geology, along with both official and unofficial (e.g., North America; The Alcove) local place names (Fig. 3; Putnam and Sloan, 2014).

GEOLOGIC BACKGROUND

El Capitan (Fig. 1) is in the west-central portion of the Sierra Nevada batholith, near the west end of Yosemite Valley (Fig. 2). It was carved in rocks of the intrusive suite of Yosemite Valley (ISYV; Reid et al., 1983; Bateman, 1992; Ratajeski et al., 2001), which is intruded on the east by the massive Tuolumne Intrusive Suite (TIS; Bateman, 1992). Most mapping and sampling of El Capitan has focused on outcrops along the base and on the summit dome and on optical reconnaissance (Calkins, 1930; Bateman 1992; Peck, 2002). Mafic dike swarms cropping out on the face were used to study magma mixing processes by Reid et al. (1983), Ratajeski et al. (2001), and Nelson et al. (2013). A detailed map of the summit was made by Ratajeski et al. (2001; Fig. 4), and a portion of the southeast face was mapped in order to study the genesis of a prehistoric rock avalanche (Stock and Uhrhammer, 2010).

Geomorphologic and geochronologic data indicate that the Sierra Nevada batholith has been tilted westward only a few degrees since the Late Cretaceous (Huber, 1981; House et al., 2001). There are no data suggesting significant postintrusive deformation of the region; so we assume that the near-vertical face of El Capitan represents a vertical transect through the magmatic system.

The Intrusive Suite of Yosemite Valley

Numerous units of the ISYV are exposed on the summit and face of El Capitan. This suite is interpreted to have been emplaced in a continental-arc environment from 105 to 95 Ma (Stern et al., 1981; Bateman, 1992). To the west, the ISYV intrudes Paleozoic metasedimentary rocks and granodioritic units of the 124–105 Ma Fine Gold Intrusive Suite (Lackey et al., 2012). To the east, the ISYV is intruded by the Late Cretaceous TIS (Bateman, 1992), which was emplaced from 95 to 85 Ma (Fig. 2; Coleman et al., 2004; Memeti et al., 2010).

Petrographic characteristics of the ISYV were described by Calkins (1930), Pabst (1938), Smith (1967), Reid et al. (1983), Bateman et al. (1984), and Ratajeski et al. (2001), and chemical analyses of various units were presented by Reid et al. (1983), Bateman et al. (1984), Ratajeski et al. (2001), and Nelson et al. (2013).

Granites

The El Capitan Granite is the dominant unit exposed on the southeast face of El Capitan and was the first unit to be emplaced (Ratajeski et al., 2001). The main body of the pluton is 30 km long and 5 km wide and is a principal member of the ISYV (Huber et al., 1989; Bateman, 1992). It is slightly porphyritic with 1–2 cm K-feldspar phenocrysts and biotite as the principal mafic mineral. Existing U-Pb zircon isotope dilution–thermal ionization mass spectrometry (ID-TIMS) geochronologic data for the El Capitan Granite in the type locality are discordant but suggest an age of 105–102 Ma (Ratajeski et al., 2001).

The Taft Granite, a medium-grained, equigranular biotite granite, intrudes and is more leucocratic than the El Capitan Granite. The El Capitan and Taft Granites overlap on all major- and trace-element trends, although the Taft is generally more felsic (Bateman, 1992). This granite has proven difficult to date: Stern et al. (1981) obtained a discordant U-Pb zircon ID-TIMS age of 95 Ma, but Ratajeski et al. (2001) found that Taft zircons, like those of the El Capitan, plot near concordia at 105–102 Ma.

Mafic Units

The El Capitan and Taft Granites are both cut by a series of compositionally diverse, biotite-rich, hornblende-poor, moderately dipping dikes and pods of dioritic to granodioritic rock previously referred to as pre–North America dikes (Ratajeski et al., 2001) and herein called dikes of the Oceans after their abundance in the outcrops east and west of the North America feature (Fig. 3). These dikes contain partially reacted xenoliths of other plutonic rocks up to 15 m across (Reid et al., 1983; Ratajeski et al., 2001).

A steeply south-dipping tonalitic unit crops out in the center of the Nose. This unit was mapped as diorite separating the El Capitan and Taft Granites by Calkins (1985) and Peck (2002), and it was mapped by Ratajeski et al. (2001) as a marginal facies of the Taft Granite (Fig. 4). Herein, it is referred to as tonalite of the Gray Bands.

The diorite of North America comprises a series of steeply dipping mafic dikes that derives its name from the resemblance of its exposure to the continent of North America (Fig. 3). It displays high compositional and textural variability but is largely Al-rich hornblende gabbro and diorite (Ratajeski et al., 2001). Precise dating of the diorite of North America has proven difficult because of Pb-loss and inherited zircons, but it is spatially associated with the Taft Granite, and they are interpreted to be coeval because the contacts of the diorite with the Taft Granite are diffuse and often grade into schlieren (Ratajeski et al., 2001).

Intrusive Suite of Buena Vista Crest

The intrusive suite of Buena Vista Crest mostly crops out south of the ISYV and, in some areas, intrudes it (Fig. 2). Bateman (1992) recognized this suite as normally zoned, comprising several successively younger inward map units. One unit, the Granodiorite of Ostrander Lake, was discordantly dated at ca. 112 and 107 Ma (Stern et al., 1981), and the granodiorite of Illilouette Creek was dated at 99 ± 1 Ma (Tobisch et al., 1995).

Reconnaissance mapping (B. Law, 2010, personal commun.) suggested that another unit of the intrusive suite of Buena Vista Crest, the Leaning Tower Granite, crops out on El Capitan. It is a medium-gray, medium-grained rock with distinctive biotite and hornblende clusters ∼10 mm in diameter. Bateman (1992) mapped it as a marginal facies of the intrusive suite of Buena Vista Crest and inferred that it was coeval with the granodiorite of Illilouette Creek. The type locality for this unit is located on the south side of Yosemite Valley in the Cathedral Rocks area (Fig. 2), and Calkins (1985) did not map it in the vicinity of El Capitan.

METHODS

Mapping

Mapping was conducted using several data sets and techniques. We employed remote sensing using light detection and ranging (LiDAR) point-cloud and return strength data, high-resolution gigapixel photographs of the southeast face taken by xRez Studios (www.xRez.com) and, as part of this investigation, photographs of rock texture taken by rock climbers as they ascended the cliff, and high-resolution photographs taken of the cliff face taken from El Capitan Bridge by climbing photographer Tom Evans. We made direct examination of the face by rappelling the Prophet and the Nose and climbing the Muir Wall, New Dawn, Zodiac, Freeblast, East Buttress, and lower half of the North America Wall (Fig. 3).

We constructed gigapixel photographs in May 2012 using a Gigapan robotic camera mount with a Nikon D5000 SLR and a 300 mm lens. These photographs were taken from the East Buttress of Middle Cathedral Rock, 1200–1500 m from the mapped face, at look-up angles of –6° to 28°, typically in flat light. Theoretical resolution with this setup is 2.2–2.7 cm/pixel and is consistent with resolution estimated from the photographs (e.g., 10 mm climbing rope is easily visible, and a climber’s helmet is typically 6–7 pixels wide). For photographs by Tom Evans, the theoretical resolution is ∼1 cm/pixel, but these images are significantly more oblique (typical look-up angles of 12°–39°). Mapped contacts are accurate to within 10 cm within the frame of reference of individual photographs, but all photographs are subject to camera distortion. Gradational contacts on El Capitan are on decimeter scale and are visible with this camera technology.

The map was constructed using geographic information system (GIS) software. Contact lines of different units were manually digitized over a gigapixel image of the southeast face. These polygons were assigned rock types by evaluating the rock texture visible in close-up photographs of the rock taken by climbers (Fig. 5). Where contacts were obscured by shadows, lichen, surface encrustation, or photo-stitching errors, Tom Evans’ photographs were examined because they were taken under a variety of lighting conditions, from different vantage points with a different perspective, and often show contacts better than gigapixel imagery.

Mineral Abundance and Grain Size

Photographs taken by climbers on the southeast face presented the means to study vertical changes in rock texture. Using the Exelis ENVI image processing package, mineral types were classified using simple quantitative thresholds in pixel value for 78 photographs taken over much of the extent of the El Capitan Granite (Supplemental Figure1). Images were classified in accordance with the distribution of biotite and hornblende (black-brown), feldspars (white-pink), and quartz (gray). Only photographs of fresh faces with no shadows were selected for this analysis. For each mineral type, contiguous regions representing individual crystals, defined using a 4-connect neighborhood, were automatically identified. One-pixel groups were ignored. From these classifications, the relative mineral group abundances and mean crystal sizes were calculated. To verify the accuracy of this method, a point count was conducted on six of the analyzed photographs. The manually counted relative mineral abundances were within ±5.1% of the automatically calculated values.

Sample Collection

Samples of all units were collected from the base, summit, and face of El Capitan and from the base of Lower Cathedral Rock on the south side of Yosemite Valley (Supplemental Figure [see footnote 1], Supplemental Tables 12 and 23). Samples (typically ∼500 g) from the face were primarily collected from belay stances on climbing routes at sites that had no impact on the climbing route. Specific climbing routes were selected for sampling to test for vertical variation because they passed through large extents of single units. Samples were collected from areas interpreted to be representative of the unit, and areas of significant evidence for mingling and mixing (Reid et al., 1983; Ratajeski et al., 2001) were avoided. The elevation of each sample from the southeast face of El Capitan was determined using a georeferenced terrestrial LiDAR point cloud and are accurate to within a meter. Additional aplite samples were collected along the walls of the valley from El Capitan east to North Dome (Fig. 2; Supplemental Table 2 [see footnote 3]).

Whole-Rock and Trace-Element Geochemistry

Whole-rock and trace-element analyses were performed using wavelength-dispersive X-ray fluorescence (XRF) at the University of North Carolina at Chapel Hill on a Rigaku Supermini XRF spectrometer following powdering in a steel jaw crusher and ceramic shatter box. Loss on ignition was determined by heating ∼2 g of rock powder to 950 °C for 1.5 h. Ignited sample (0.9000 ± 0.005 g) and 64.7% Li2B4O7, 35.3% LiBO2, 0.5% LiBr flux (8.1000 ± 0.005 g) were melted in a Pt crucible and fused into a glass bead. Trace-element analyses were performed with XRF using pressed-powder disks. Unignited samples (6.000 g) were combined with paraffin powder (0.600 g), mixed in a ball mill, and hydraulically pressed into an Al mold. Calibrations were run against standards AGV-1, G-2, QLO-1, RGM-1, GSP-2, NIST 278, MAG-1, BHVO-1, and DNC-1; accuracy and reproducibility are reported in Supplemental Tables 1 and 2 (see footnotes 2 and 3). Extended trace-element analyses were performed by Actlabs (Ontario, Canada), where samples were dissolved by Li2B4O7/LiBO2 fusion and analyzed by ICPMS (Supplemental Table 34).

Tests for significance of correlation of chemical values with elevation were performed using a 2-tailed t-test a significance level p = 0.05.

Geochronology

Eight samples were selected for zircon U-Pb geochronology by both laser ablation–inductively coupled plasma mass spectrometry (LA-ICPMS) and single-grain isotope dilution-thermal ionization mass spectrometry (TIMS). Samples were selected for geochronology from units on and around El Capitan in order to evaluate the timing of the mapped intrusive events. Of particular interest was the Leaning Tower Granite as there was no published date for the unit and knowing its age would help resolve the timing of the intrusions exposed on the southeast face of El Capitan.

All samples were broken down using a jaw crusher and a disc mill. Zircons were separated using standard density (water table and heavy liquids) and magnetic techniques. Samples selected for LA-ICPMS dating include: El Capitan Granite from the Cookie Slide near the west entrance to Yosemite Park (CS720; 63 zircons) and the summit of El Capitan (ECS726; 28 zircons); Taft Granite (ECS-01; 14 zircons); Leaning Tower Granite from the base of Lower Cathedral Rock, the type locality of the unit (LC-01; 27 zircons); Bridalveil Granodiorite (LC-02; 39 zircons); granite from near El Portal (presently mapped as El Capitan Granite [Huber et al., 1989] but texturally distinct at the sample location; RF-01; 26 zircons); diorite of the Rockslides (YOS-23C; 38 zircons); and diorite of North America (YOS-104; 36 zircons). Grains representative of the population’s size and morphology were separated from each sample, mounted in epoxy, polished using standard polishing techniques, and mapped by backscattered electron and monochromatic cathodoluminescence imaging on a scanning electron microscope. Operating procedures for LA-ICPMS follow those outlined in Kylander-Clark et al. (2013). Spot analyses were run at 4 Hz for 15–20 seconds; spot sizes ranged from 15 to 24 µm in diameter and 6–8 µm deep. Isotopic concentrations were measured on the Nu Plasma, with 238U and 232Th on Faraday detectors, and 208Pb, 207Pb, 206Pb, and 204Pb concentrations were measured on ion counters. Isotopic ratios and standard correction were performed using Iolite (Paton et al., 2010), using the 91500 zircon (1065 Ma; Wiedenbeck et al., 1995) as the primary reference material; GJ1 (602 Ma; D. Condon, 2010, personal commun.), Plešovice (337 Ma; Sláma et al., 2008), SL1 (564 Ma; Gehrels, 2000), and Temora2 (417 Ma; Black et al., 2004) were also measured for quality control. We consistently obtained weighted mean 206Pb/238U ages within 1% of the reference value for each of the secondary reference materials (605.5 ± 1.8 Ma (n = 52), 560.8 ± 5.0 Ma (n = 6), 340.2 ± 1.8 Ma (n = 20), 418.9 ± 5.4 Ma (n = 3) for GJ1, SL1, Plešovice, and Temora2, respectively, and thus conservatively add 1% in quadrature to the final age of each sample. The analytical uncertainty is expressed first, followed by the propagated uncertainty in brackets for each sample. All errors are 2σ unless expressed otherwise.

Three samples dated by LA-ICPMS were also dated by ID-TIMS: the El Capitan Granite at the Cookie Slide and El Capitan’s summit (CS720 and ECS726, respectively) and the Leaning Tower Granite (LC-01). Zircon grains were thermally annealed at 900 °C for 48 h and were chemically abraded in 29M HF for 16 h at 220 °C to remove mineral inclusions and zones affected by radiation damage that are subject to Pb-loss (Mattinson, 2005). Fractions were spiked with a 205Pb-233U-236U tracer (Parrish and Krogh, 1987) and dissolved in 29M HF following a procedure modified after Krogh (1973) and Parrish (1987). Anion exchange (HCl) column chromatography was used to isolate U and Pb from the dissolved solution. Analyses of U and Pb were completed using a VG Sector 54 thermal ionization mass spectrometer at the University of North Carolina at Chapel Hill. Uranium was run on single Re filaments either as a metal, after loading in graphite and H3PO4, or as an oxide, after loading in silica gel. Lead was loaded in silica gel on single, zone-refined Re filaments. Both U and Pb were analyzed in single-collector peak-switching mode using a Daly ion-counting system. Data processing and age calculations were completed using the applications Tripoli and U-Pb_Redux developed as part of the EARTHTIME initiative (Bowring et al., 2011; McLean et al., 2011). Decay constants used were 238U = 1.55125 × 10–10 a–1 and 235U = 9.8485 × 10–10 a–1 (Steiger and Jäger, 1977).

Corrections for initial Th/U disequilibrium (Mattinson, 1973; Schmitz and Bowring, 2001) for the ID-TIMS samples were made using U-Pb_Redux (corrections are much smaller than the uncertainties in LA-ICPMS ages). An assumed magmatic Th/U ratio of 4 was used based on an average Th/U ratio of granodiorites in the Sierra Nevada batholith drawn from the NAVDAT database (www.navdat.org). The difference between an uncorrected 206Pb/238U weighted mean age and an age that has been corrected for a magmatic Th/U ratio of 4 is ∼95 ka in these samples.

RESULTS

Field Relations

The new mapping (Fig. 6) reveals field relations that help to resolve an intrusive history previously obscured by problematic geochronology and physical inaccessibility. The oldest and most extensive intrusion, the El Capitan Granite, is homogenous in outcrop, with little obvious modal layering. On the southeast face, magmatic fabric is rare and, where present, poorly developed. Mafic enclaves are uncommon (occupying less than 0.01 area%) and typically small (<10 cm diameter); enclaves increase in abundance dramatically to the west of El Capitan (Coleman et al., 2005).

The Taft Granite primarily crops out in a ∼250-m-tall wedge centered on the upper third of the Nose as mapped by Calkins (1985) and in a large block centered on the East Buttress that extends west along the base to El Capitan Tree (Fig. 3). From El Capitan Tree, the granite intermittently crops out along the base west to The Alcove in dikes up to 2 m thick that cut the El Capitan Granite. These dikes have sharp to gradational margins and tend to be more leucocratic than typical Taft (Fig. 7D). Where the Taft is in contact with the El Capitan Granite on the face (such as 40 m up Zodiac), the contact is gradational with little diking. Much of the unit is modally layered and sometimes contains slightly more abundant mafic bands up to 10 m thick. There is no apparent pattern to the orientation or width of these features.

The next several intrusive episodes following intrusion of the Taft Granite were interpreted in earlier work to be roughly contemporaneous (Reid et al., 1983; Ratajeski et al., 2001). However, crosscutting relations ∼600 m up the Nose (Fig. 3) reveal that the first of these was the dikes of the Oceans. Dikes of the Oceans strike east and dip ∼40° south on the upper part of the Nose, and dip much more steeply on other parts of the cliff. They commonly display dextral sense-of-separation indicators, especially on the western part of the face (Fig. 7B). Contacts of the dikes of the Oceans with the Taft and El Capitan Granites tend to be sharp to gradational over ∼0.25 m. Contacts of the dikes of the Oceans with all other units are sharp, and the dikes commonly contain abundant enclaves of the diorite of North America in mafic pods and swarms (Fig. 8), indicating that the two intrusive episodes overlapped.

The tonalite of the Gray Bands crops out below the Taft Granite between the Salathé Wall and Mescalito (Figs. 3 and 6) and dips steeply to the southwest. It is in gradational contact with the Taft Granite and sharp contact with El Capitan Granite, containing angular xenoliths of El Capitan Granite up to ∼20 m across (Fig. 7A). Locally, the intrusion appears to be a composite of 0.5- to 1-m-thick subparallel dikes.

The Leaning Tower Granite principally crops out at the bottom of the cliff below El Capitan Tree and runs in a large, diagonal, en echelon band up to the east and past the Gray Circle (Figs. 3 and 6). This band is ∼30 m wide at the base of the cliff and gradually tapers out to the east. One smaller band crops out ∼75 m east of Zodiac and, similar to the large band, pinches out up and to the east. The contacts of this unit are typically sharp.

The diorite of North America intruded in three ∼50-m-thick composite dikes that strike southeast and dip nearly vertically (Figs. 2, 4, and 6). These dikes crop out on the southeast face in North America, around The Gray Circle, and just east of East Buttress (Figs. 3 and 4). Because these dikes dip so steeply, their intersection with the subvertical face gives them irregular outlines. Isolated sections of host rock up to ∼20 m across within the dikes superficially appear to be detached blocks but likely were attached to the surrounding host rock (Fig. 9B). The diorite of North America’s contacts with all other units are sharp or gradational over <<1 m (Fig. 7C).

Two series of aplite dikes cut the southeast face. First-series dikes are 0.5–3 m thick and dip steeply to the northeast. They are locally composite and commonly contain 2- to 10-cm-thick pegmatite bands. The second series dikes are 10–20 cm thick and are subhorizontal (Fig. 9A). The horizontal series crops out along the northern walls of the valley and cuts every unit older than the TIS. In a few locations, the second series cuts the first, suggesting separate emplacement events (Fig. 9A). First-series dikes are locally separated across second-series dikes, displaying inconsistent sense-of-separation indicators, whereas the second series is largely continuous (Fig. 9B).

Our mapping revealed no brittle faults of resolvable offset (i.e., >5–10 cm) cutting the southeast face of El Capitan. The only offset markers of any kind are the separated first-series aplite dikes, but these separations are not accomplished by planar faults and may simply represent out-of-plane effects. Given the abundance of sharp planar contacts in the face, this absence of faulting is real.

Xenoliths of host metasedimentary or metavolcanic rock are conspicuously absent from the southeast face of El Capitan; none were found in this study, nor by any of the climbers engaged in the study, either on the face or in talus. The nearest mapped xenoliths of wall rocks to El Capitan are located near Sentinel Dome, ∼5 km to the southeast (Calkins, 1985).

Petrography and Geochemistry

We analyzed 70 samples for major elements by XRF (Supplemental Table 1 [see footnote 2]), 27 aplites for selected trace elements by XRF (Supplemental Table 2 [see footnote 2]), and 17 samples for extended trace elements by ICPMS (Supplemental Table 3 [see footnote 4]). Our sampling and analysis focused on the felsic units to evaluate evidence for vertical variations in composition, but data from more mafic units (Ratajeski et al., 2001; Nelson et al., 2013) are also discussed below.

Samples analyzed for this study range from 61 to 78 wt% SiO2; when combined with data from the mafic units, they form a bimodal pattern, with only one sample plotting between 58 and 66 wt% SiO2 (Fig. 10). El Capitan and Taft Granites overlap in composition, although the Taft is generally higher in SiO2. On most plots, the tonalite of the Gray Bands continues the linear trends defined by the Taft and El Capitan Granites to lower SiO2 concentrations.

Textural and mineralogical data from the Leaning Tower Granite cropping out on El Capitan confirm its assignment to this unit. The Leaning Tower Granite has more hornblende, biotite, and plagioclase than the principal silicic units on El Capitan (Bateman, 1992). Myrmekite occurs along plagioclase–K-feldspar boundaries, and biotite is moderately chloritized (as in most granitoids in the Yosemite Valley area). Accessory minerals are dominantly magnetite, zircon, and titanite, with rare apatite, ilmenite, pyrite, and secondary epidote. Magnetite, zircon, titanite, and apatite tend to be found near or within biotite. Samples of the Leaning Tower Granite from the newly recognized outcrops on El Capitan chemically cluster with samples taken from the type locality on the south side of Yosemite Valley. Leaning Tower Granite samples range from 68.3 to 70.1 wt% SiO2 and have lower Al2O3 and Na2O at comparable SiO2 than rocks of the ISYV (Fig. 10). The samples taken from El Capitan are slightly lower in SiO2 than those from the type locality but cluster with them on major-element plots and possess the same distinctive texture and mineralogy.

Both the dipping and horizontal aplite dike series locally contain trace amounts of garnet and muscovite. Garnet occurs locally in dense aggregations of mm-sized crystals, often in association with graphic texture within the dike as a whole. Both series contain ∼76 wt% SiO2 compared with 75 and 74 wt% SiO2 for the Taft and El Capitan Granites, respectively, with which they overlap on many major-element trends (Fig. 10). Despite the presence of garnet and muscovite, they are barely metaluminous; molar Al2O3/(CaO+K2O+Na2O) averages 1.03.

Vertical Patterns

There are few apparent trends in major-element composition with increasing elevation within individual units (Fig. 11), in contrast to resolvable variations with elevation in major and trace elements noted by Sawka et al. (1990) in the Tinemaha suite. For the El Capitan Granite, none of the major-element oxides yield statistically significant correlation coefficients. For the Taft Granite, two outliers at high elevation cause correlation coefficients to pass the p = 0.05 significance test for Na2O, MgO, K2O, and Fe2O3t. Taken at face value, these correlations indicate that Fe2O3t and MgO increase and Na2O and K2O decrease with elevation, opposite the predictions of crystal-liquid separation.

Even when limited to the area of the cliff with the lowest concentration of mafic dikes (samples from the Muir Wall, the Nose, and New Dawn; Fig. 3), there are no apparent correlations with elevation over the 900 m vertical extent (Fig. 12). The Taft Granite generally crops out at higher elevations than El Capitan Granite and is slightly more silicic, but Taft exposed low on the cliff is compositionally the same as that exposed high on the Nose (Fig. 12), and Taft Granite with the lowest SiO2 is exposed at the highest elevation, high on New Dawn. Rare-earth–element patterns (Fig. 13) and trace-element compositions (Fig. 14) of samples taken along two vertical transects within single rock types also lack a correlation with elevation.

Rare-earth elements (REEs) show weak correlations with elevation (Fig. 13). For the NED series of samples, the three heaviest REEs pass the significance test, whereas the remainder do not; these elements increase in concentration with increasing elevation.

Texture analysis reveals no statistically significant correlations in El Capitan Granite between elevation and crystal size, color index (vol% ferromagnesian minerals), or feldspar abundance (Fig. 15). Our data closely reproduce Bateman’s (1992) color index data, but feldspar abundance data seem somewhat low compared to his.

Geochronology

Geochronologic data (Supplemental Tables 45 and 56; Fig. 16) yield single distributions about the mean. Nearly all of the laser ablation ages are concordant within uncertainty; for those that are discordant, we consider the 207Pb-corrected 206Pb/238U age to be the best estimate, assuming the discordance is caused primarily by minor contamination of common Pb. We accepted all LA-ICPMS results except a few (<10% of the analyses, indicated on Fig. 16) that fall outside what otherwise appear to be normally distributed populations. For the remaining data, mean square weighted deviations (MSWDs) are near unity for six of the eight samples—two samples yield an MSWD between 2 and 3. Samples for which we obtained both LA-ICPMS and ID-TIMS dates yield results that overlap within uncertainty, a testament to the accuracy of the LA-ICPMS data; ID-TIMS results, however, are more precise (Fig. 16).

The zircon grains analyzed by LA-ICPMS for each sample appear to be derived from single populations. The granite from near El Portal (RF-01), the El Capitan Granite at the Cookie Slide (CS720) and the summit of El Capitan (ECS726), and the Leaning Tower Granite (LC-01) include grains that are significantly older than the main population (Fig. 16). The samples of El Capitan at Cookie Slide (CS720), the Bridalveil Granodiorite (LC-02), and the diorite of the Rockslides (YOS-104) each include one zircon that is distinctly younger than the main population. One ID-TIMS grain from the Leaning Tower Granite is distinctly older than the other four analyses and was not included in age calculations.

The granite from near El Portal (RF-01) yields a weighted mean 206Pb/238U age of 114.9 ± 0.8 [1.3] Ma (27 of 29; MSWD = 2.3; LA-ICPMS). The Taft Granite (ECS-01) taken from the base of El Capitan yields an age of 106.6 ± 0.7 [1.3] Ma (ID-TIMS). Two samples from the El Capitan Granite from the summit of El Capitan (ECS726) and Cookie Slide (CS720) yield indistinguishable ages of 106.1 ± 0.5 [1.2] Ma (29 of 32; MSWD = 1.2; LA-ICPMS) and 105.5 ± 0.3 [1.1] Ma (57 of 62; MSWD = 0.7; LA-ICPMS), respectively. Both these samples were also dated by ID-TIMS and yield somewhat more precise age estimates of 105.58 ± 0.58 Ma and 105.45 ± 0.25 Ma. Sample LC-01 of the Leaning Tower Granite yields an LA-ICPMS age of 103.5 ± 0.6 [1.2] Ma (24 of 26; MSWD = 1.2) and an ID-TIMS age of 104.10 ± 0.10 Ma. The Bridalveil Granodiorite (LC-02) yields an age of 103.5 ± 0.4 [1.1] Ma (37 of 37; MSWD = 0.6; LA-ICPMS). Two samples of the diorites associated with the El Capitan Granite yield indistinguishable ages of 104.0 ± 0.4 [1.1] Ma (diorite of North America, taken from the base of El Capitan: YOS 23C; 36 of 36; MSWD = 0.7; LA-ICPMS) and 103.5 ± 0.6 [1.2] Ma (diorite of the Rockslides: YOS-104; 36 of 37; MSWD = 2.1; LA-ICPMS).

DISCUSSION

Timing

Most of the units that crop out on El Capitan have historically been difficult to date by U-Pb TIMS due to inherited grains and Pb-loss. Previous attempts to date the units were also hampered by higher Pb blanks and the necessity of dating large, multiple-grain fractions (Stern et al., 1981; Ratajeski et al., 2001). This problematic geochronology is partially resolved by our new laser ablation and U-Pb TIMS data (Figs. 16 and 17).

The sample collected from near El Portal, the westernmost sample collected, is distinctly older than samples collected around Yosemite Valley at ca. 115 Ma. This age is intermediate between the oldest locally dated portions of the Bass Lake Tonalite ages to the west (∼120–117 Ma, Bateman 1992; Lackey et al., 2012) and the ISYV and likely belongs to a unit other than the El Capitan Granite.

The El Capitan and Taft Granites are somewhat older than the other intrusive rocks in the Yosemite Valley area and were emplaced between 107 and 105 Ma. Ratajeski et al. (2001) found that Taft and El Capitan Granite zircons are 105–102 Ma. This age range is significantly reduced by our TIMS date on the Leaning Tower Granite. Because the Leaning Tower Granite clearly cuts the Taft Granite, it must be younger, which limits the age of the Taft to be older than the Leaning Tower Granite (106.6 Ma versus 104.1 Ma, respectively, in this area; Fig. 16). Alternatively, the sample of Taft Granite in this study could be a separate unit, a possibility supported by the abundance of monazite in our heavy mineral separate; monazite was not reported in other studies (Calkins, 1930; Pabst, 1938; Smith, 1967; Bateman et al., 1984; Ratajeski et al., 2001).

The diorites of North America (104.0 ± 0.4 Ma) and the Rockslides (103.5 ± 0.6 Ma), the Leaning Tower Granite (104.10 ± 0.10 Ma), and Bridalveil Granodiorite (103.5 ± 0.4 Ma) are indistinguishable in age but are consistently younger than the El Capitan (106.1 ± 0.5 Ma) and Taft (106.6 ± 0.7 Ma) Granites. These ages agree well with mapped field relations.

Intrusive History

Contact Relationships

The map reveals a complex intrusive history involving eight distinct and overlapping intrusive episodes. Contacts between the El Capitan and Taft Granites are typically sharp to gradational over ∼0.25 m, whereas contacts between these granites and all other units are typically sharp (Fig. 7D). If gradational contacts are an indication of intrusion into a partially molten system, then our mapping and geochemistry support Ratajeski et al.’s (2001) interpretation that the Taft Granite was generated from a partial melt of the same source as the El Capitan Granite and emplaced at the same approximate time. Geochronology supports the temporal connection, but uncertainty in the age of the Taft Granite (±1.3 Ma) is too large to determine whether it pre- or postdates the El Capitan Granite. However, because dikes of Taft Granite cut the El Capitan Granite (e.g., between The Alcove and The Footstool and at Taft Point on the south rim of the valley; Figs. 2 and 3), it must be at least locally younger.

Contacts between the dikes of the Oceans, the Leaning Tower Granite, and the diorite of North America, with all the units they intrude (including each other), are sharp to gradational over ∼0.25 m (Figs. 7A–7C). This suggests that the system into which they intruded was at low enough temperatures to prevent significant mixing. Ratajeski et al. (2001) noted that contacts between the diorite of North America and the Taft Granite are diffuse and grade into schlieren on the summit, and they interpreted these relations to indicate cogenesis. However, our mapping shows the contact is typically quite sharp, which is consistent with geochronology indicating that the Taft Granite is locally 1–2 Ma older than the diorite of North America. The diffuse contacts observed by Ratajeski et al. (2001) could be the result of localized partial melting rather than cogenesis. Perhaps the voluminous diorite of North America generated high local temperatures that induced local partial melting in the Taft Granite, but not in the other, less silicic units into which it intruded. Furthermore, the diorite of North America is locally found within the dikes of the Oceans and the Leaning Tower Granite as pillows and enclave swarms. It is possible that the diorite intruded along the rheological weaknesses provided by these younger, partially molten, perhaps higher melt-fraction younger intrusions. Such field relations are observed in other silicic plutons (Wiebe and Collins, 1998; Leuthold et at., 2012).

Tonalite of the Gray Bands

Because it is compositionally similar to the biotite-rich, hornblende-poor dikes of the Oceans, the tonalite of the Gray Bands has been interpreted as coeval mafic material that was mixed with the Taft Granite during emplacement (Ratajeski et al., 2001). However, in a few locations, such as at the lowest point of the bands, this unit cuts the dikes of the Oceans, thus suggesting they are separate intrusions and that the tonalite of the Gray Bands is younger. Furthermore, because the tonalite of the Gray Bands is only found in this outlying area of the Taft Granite and nowhere else, we suggest that it is a separate intermediate intrusion rather than a ubiquitous marginal facies of the Taft Granite.

The tonalite of the Gray Bands has a gradational contact with the Taft Granite above and a sharp contact with the El Capitan Granite below. Within the tonalite of the Gray Bands, there are large (up to ∼20 m) xenoliths of the El Capitan Granite but none of the Taft (Figs. 6 and 7A). Taken together, these observations suggest that the tonalite of the Gray Bands intruded along the contact of a partially molten Taft Granite and a cooler, more brittle El Capitan Granite. Because the tonalite of the Gray Bands is unlike any other member of the ISYV, and intruded after all other units in the suite, it may be more closely tied to the magmatic system responsible for construction of the intrusive suite of Buena Vista Crest. The intrusive suite of Buena Vista Crest has several mafic marginal members (Bateman 1992), and the relationships between and possible cogenesis of these units are poorly understood. Given that the Leaning Tower Granite crops out in the same area of the valley, it is not unreasonable to propose that the tonalite of the Gray Bands is better considered as an early-stage magma of the intrusive suite of Buena Vista Crest.

Aplite Dikes

The youngest intrusive rocks exposed on El Capitan are a suite of thin subhorizontal aplites that cut all the units on El Capitan and become less obvious to the east, toward the TIS. These field relations suggest that the aplites radiated out of the TIS and into the older ISYV as observed elsewhere (e.g., Bateman, 1965; Searle et al., 1993). However, the trace-element compositions of aplite dikes in the ISYV indicate that they are locally sourced rather than injected from the adjacent suite (Fig. 18). Glazner et al. (2008) showed that aplites in the TIS are characterized by extremely low middle REE and Y (<5 ppm) concentrations, consistent with separation from a crystalline residue rich in titanite. In contrast, the aplites that cut the ISYV have high Y contents (typically >10 ppm) and lack the scoop-shaped REE patterns that characterize aplites separated from titanite-bearing melts (Fig. 18). This suggests that aplites cropping out on El Capitan did not radiate outward from the TIS. However, the dikes cut all units on the face including units of the intrusive suite of Buena Vista Crest, raising the question of which intrusive episode provided the late-stage liquids that intruded as aplites.

Depth of Emplacement

The depth of emplacement of Yosemite granitoids is poorly known. Ague and Brimhall’s (1988) regional study of emplacement pressures using Al-in-hornblende barometry indicated regional emplacement pressures of 300–400 MPa for western Yosemite Valley, although there is a strong gradient to significantly lower pressures to the east. Estimates based on contact metamorphic assemblages (e.g., Anderson et al., 2007) are sparse owing to a lack of useful (e.g., Al-rich) wall-rock assemblages. Rose (1957) found andalusite in wall rocks of the TIS at May Lake, which indicates pressures <∼400 MPa.

If aplite dikes in the ISYV were segregated from host granites during the late stages of solidification, their compositions can be used to estimate the temperature (T) and pressure (P) at the time of separation, assuming that aplite whole-rock compositions are equivalent to liquids separated from a quartz-, feldspar-, and zircon-bearing crystalline mush, and that the magma was water-saturated with aH2O = 1. We use the method of Blundy and Cashman (2001) to reproject Ca-bearing aplite compositions into the albite-orthoclase-quartz-H2O system to estimate P and the Boehnke et al. (2013) zircon saturation thermometer to estimate T. Excluding one significant outlier, results for eight samples average 230 ± 47 MPa and 564 ± 63 °C (Fig. 19) and are scattered near and below the water-saturated solidus for granitic compositions. Taken at face value, the average indicates a depth of around 5–7 km for typical crustal densities, somewhat shallower than previous estimates.

Zircon saturation temperatures are well below the expected solidus temperatures of water-rich granitic compositions (Whitney, 1988) but well above the solidus temperatures for many pegmatites, in which B, F, and other fluxing components are abundant (Sirbescu and Nabelek, 2003; London et al., 2012). If such components were abundant in El Capitan aplites, then aH2O was less than unity by an unknown amount; how this would affect pressure estimates is as yet unknown. Temperature estimates from the older Watson and Harrison (1983) calibration are systematically 50 °C higher than the newer calibration and are closer to the solidus.

Lack of Faulting

Although faults can be difficult to trace into plutons because of the general lack of markers, the abundant sharp contacts of diverse orientation on the southeast face of El Capitan means that the absence of brittle faulting across the face is real. In many areas, fault systems that are well developed in sedimentary or metamorphic rocks outside a pluton are deflected around the pluton, presumably owing to its strength (e.g., Parry and Bruhn, 1986). Rock fracture by faulting is important in erosion, and Yosemite-like features such as domes and tall cliffs require particularly flaw-free rocks (Molnar et al., 2007). Brittle fractures are abundant east of Yosemite, approaching the eastern frontal fault of the Sierra Nevada, but they are rare in Yosemite Valley. This difference in fracture density between bedrock under the east-draining creeks that flow down the eastern slope and the west-draining rivers that flow through the rest of the park probably contributes to the large size discrepancy between their respective moraines (Glazner and Stock, 2010, p. 264).

Vertical Variations

Many hypotheses for the generation of silicic melts are based on gravitational separation of melt from crystals and predict upward accumulation of liquids rich in silica and incompatible elements (Sawka et al., 1990; Bachl et al., 2001; Hirt, 2007). The compositional time sequence displayed by many large pyroclastic eruptions is interpreted as a record of top-down tapping of a zoned magma body (Ritchey, 1980; Sparks et al., 1984). The vertical scale over which significant zonation develops is estimated at several km (e.g., Bailey et al., 1976; Ritchey, 1980; Fridrich and Mahood, 1987); if the El Capitan or Taft Granites formed in a large magma body of this sort, vertical variations in composition should be observable within the 1 km face. Many plutonic series are interpreted as cumulate piles of crystals and associated liquid (Miller and Miller, 2002; Collins et al., 2006), and these processes, too, should result in clear vertical geochemical variations, especially in trace elements.

Gravity-driven crystal-liquid separation within a magma chamber predicts increasing SiO2, K2O, and Al2O3 and decreasing Fe2O3t, MgO, CaO, and Na2O with increasing elevation. Accumulation of feldspar should enrich cumulate rocks in several elements (Ba, Sr, and Eu), and separated liquid should be enriched in incompatible elements such as Th and La. Our data indicate that there are no such resolvable vertical compositional patterns in either the El Capitan or the Taft Granites (Figs. 11–14) and no vertical trends in the relative abundance and mean crystal size of feldspars and ferromagnesian minerals within the El Capitan Granite (Fig. 15). The heterogeneity in whole-rock composition and lack of regularity in texture indicate that the El Capitan Granite experienced a more complex history during emplacement than simple crystal-liquid segregation models would indicate. We cannot rule out the possibility that repeated intrusion and mixing during construction of the granites has masked any vertical variations that once existed.

Recent studies suggest that compositional heterogeneity within plutonic suites is largely inherited from the source of a magma rather than produced by in situ fractionation (Eichelberger et al., 2000; Clemens et al., 2010; Tappa et al., 2011; Coleman et al., 2012), although crystal-liquid separation can produce local variability. The vertical chemical and textural profiles within units on El Capitan are consistent with these studies. Because evidence of large-scale differentiation is lacking in the vertical profiles of the rocks of El Capitan, we suggest that the compositional heterogeneity of the ISYV is inherited from deeper levels rather than created by processes operating at the observed level. Such variation could originate at the magma’s source or during staging at a shallower level below the emplacement level.

Geochemistry and the Morphology of El Capitan

The differences in morphology between the southeast and southwest faces of El Capitan indicate that the distribution of Cretaceous granitoids may have played a role in shaping the monolith. The southeast face locally overhangs up to ∼35 m with a typical slope of ∼87° compared with the southwest face, which has a typical slope of ∼80°. This different morphology may, in part, be due to near-absence of mafic dikes on the southwest face in contrast to their ubiquitous presence on the southeast face. Mafic dikes have a greater abundance of weaker minerals, particularly biotite, and finer grain size, resulting in a lower rock-mass strength than the host granites. Fracturing is more intense in the diorite of North America than in the surrounding granite, resulting in smaller mean joint spacing. This may lead to a higher rockfall frequency in the diorite, compared to the granite. Rockfall volumes are usually smaller in the diorite, but more frequent, and thus erosion is greater in the mafic areas of El Capitan. The profile of Figure 20 highlights the concave shape of the cliff where the diorite of North America dike crops out (Matasci et al., 2014). Preferential weathering of weaker rocks can provide surfaces from which joints can propagate (Schmidt and Montgomery, 1995; Korup, 2008). We suggest that preferential weathering of mafic dikes may be the source of El Capitan’s distinctive shape and perhaps even contributed to localizing the original incision of Yosemite Valley.

CONCLUSIONS

New decimeter-scale geologic mapping of El Capitan in Yosemite Valley, California, reveals much about the complex emplacement history of this portion of the Sierra Nevada batholith: (1) Eight distinct intrusive episodes of granitoids from the intrusive suites of Yosemite Valley and Buena Vista Crest were mapped and correlated using crosscutting relations and geochronology, expanding and refining the relative and absolute chronologies of intrusion. These episodes reflect an intrusive history that spans ∼2.9 m.y. and occurred under a variety of thermal and rheologic conditions. (2) Brittle faulting is absent from the face of El Capitan. (3) Aplite dikes display geochemistry suggestive of local origin and were likely not injected laterally from the Tuolumne Intrusive Suite. (4) Mafic dikes are dominantly distributed on the eastern portion of the face and are hypothesized to be one of the reasons for El Capitan’s distinctive shape. (5) Evidence of large-scale in situ crystal-liquid fractionation is conspicuously absent from the whole-rock compositions and petrographic textures of the dominant silicic units, suggesting that deeper crustal processes were the cause of compositional heterogeneity displayed on the face of El Capitan.

Research was supported by National Geographic/Waitt grant W217-12, National Science Foundation grants EAR-0336070 and EAR-0538129, the Geological Society of America, Sigma Xi, the American Alpine Club, and the University of North Carolina’s Martin Fund. Further support was provided by Maptek, Bluewater Ropes, Metolius Climbing, and Patagonia. Partial funding for geochronology was provided by a gift from University of North Carolina alum Jesse Davis, and gracious support from emeritus professor John J.W. Rogers helped keep geochemical facilities running. Many thanks to Battista Matasci from the University of Lausanne, Greg Downing and Eric Hansen of www.xRez.com, and to photographers Tom Evans and Derek Ferguson. A thorough review of an earlier version of the manuscript by Jade Star Lackey was most helpful. The manuscript also benefitted from comments and reviews by Editor Shanaka de Silva, Associate Editor Rita Economos, reviewer Josh Schwartz, and an anonymous reviewer. Thanks to Greg Stock and the National Park Service for logistical support. Many of the ideas herein were developed in concert with Bryan Law, who was the first person to recognize Leaning Tower Granite on El Capitan. Much gratitude is owed to the 29 climbers who helped map the face of El Capitan.

1Supplemental Figure. Sample photographs. Please visit http://dx.doi.org/10.1130/GES01133.S1 or the full-text article on www.gsapubs.org to view the Supplemental File.
2Supplemental Table 1. Major element data determined by XRF. Please visit http://dx.doi.org/10.1130/GES01133.S2 or the full-text article on www.gsapubs.org to view Supplemental Table 1.
3Supplemental Table 2. Trace element data determined by XRF. Please visit http://dx.doi.org/10.1130/GES01133.S3 or the full-text article on www.gsapubs.org to view Supplemental Table 2.
4Supplemental Table 3. ICPMS data. Please visit http://dx.doi.org/10.1130/GES01133.S4 or the full-text article on www.gsapubs.org to view Supplemental Table 3.
5Supplemental Table 4. LAICP data. Please visit http://dx.doi.org/10.1130/GES01133.S5 or the full-text article on www.gsapubs.org to view Supplemental Table 4.
6Supplemental Table 5. Zircon TIMS age data from the Leaning Tower granite. Please visit http://dx.doi.org/10.1130/GES01133.S6 or the full-text article on www.gsapubs.org to view Supplemental Table 5.