Formation of pluton roofs, floors, and walls by crack opening at Split Mountain, Sierra Nevada, California

The mechanisms by which plutons are emplaced have been debated for more than a century (e.g., Wright, 1907; Buddington, 1959; Pitcher, 1993). Plutons with discordant contacts and a subcircular outline have commonly been interpreted to reflect diapirism, stoping, or both (e.g., Compton, 1955; Pitcher and Berger, 1972). However, there are several reasons to doubt the efficacy of diapirism and stoping as emplacement mechanisms, especially in the shallow crust. These include the mechanical, thermal, and waste disposal problems posed by stoping (Glazner and Bartley, 2006; Glazner, 2007) and the thermal and strain requirements of upper crustal diapirism (e.g., Mahon et al., 1988; Clemens and Mawer, 1992). An alternative explanation for plutons with sharp, discordant contacts is the opening of fractures to admit sills and/or dikes (e.g., Cruden and McCaffrey, 2001; Glazner and Bartley, 2006; Bartley et al., 2008). The Split Mountain area is an excellent place to evaluate these alternative emplacement models owing to the exceptional exposures of pluton–wall-rock contact relations.

The plutonic geology of the Split Mountain area is among the most spectacular and most photographed in the world (Fig. 1) because dramatic color contrasts between rock units render the gross spatial relations readily visible from the floor of Owens Valley to the east (Shelton, 1966; Coleman et al., 2005). The area is particularly notable for the thin screen of pre-batholithic rocks that traces contacts between several plutons and forms the roof, wall, or floor of various plutons (Plate 1; Figs. 1 and 2; Moore, 1963; Bateman, 1965). Such interpluton screens are common in batholiths but their interpretation is debated (John and Blundy, 1993; Bartley and Glazner, 1998; Coleman et al., 2005).

The U.S. Geological Survey Mount Pinchot and Big Pine geologic quadrangle maps (scale 1:62,500; Moore, 1963; Bateman, 1965) accurately depict the general geologic relations in the study area, but provide insufficient detail to permit precise interpretations of contacts and structures. The study area straddles the boundary between Kings Canyon National Park and the John Muir Wilderness, and resulting restrictions on access mean that close geologic examination requires ascent on foot from trailheads more than 2 km below the range crest. High altitude, rugged topography, abundant cliff exposures, and difficult access thus have led most geologists to interpret much of the geology from distant views and from below. Closer examination reveals that such distant views can be misleading owing to castellated weathering. Our work included both detailed mapping on the ground and examination of contacts from a low-flying airplane. A number of key relations either are located in narrow chutes not visible from a distance or are evident only when viewed from the same elevation or above.

Conventional geologic mapping was completed on topographic base maps (scale 1:10,000). Owing to steep topography, base maps were locally inaccurate and contacts were adjusted during compilation using color infrared orthophotoquads, which are subject to significant distortion in such steep terrain. Matching the topographic base to orthophotoquads and to various digital elevation models was only marginally successful in places, and the geologic map (Plate 1) represents our best attempt to marry these geographic bases to our mapping and to that of Moore (1963) and Bateman (1965). In places this required moving contacts from older mapping significant distances (>100 m). Ground-based mapping was supplemented with oblique aerial photography taken during two low overflights. Side-looking stereo pairs made from these photos proved especially useful for filling in mapping in areas too steep to climb.

The exceptionally steep topography around Split Mountain makes conventional geologic maps difficult to interpret. For this reason, we present our map both as a conventional geologic map on a shaded relief background (Plate 1) and as a draped overlay for viewing with Google Earth (Fig. 3; Supplemental File 1¹).

Correlation of metasedimentary units was aided by isotopic fingerprinting. Nd isotope analyses were performed using standard isotope-dilution thermal-ionization mass spectrometry at the University of North Carolina (methods in Glazner et al., 2008).

Pre-batholithic wall rocks in the study area include impure quartzite, schist, marble, and calc-silicate hornfels (Moore, 1963). Our data indicate that the rocks correlate with Cambrian formations defined to the northeast in the White and Inyo Mountains (Nelson, 1962).

The most areally extensive pre-batholithic rock unit comprises impure quartzite and (garnet) biotite schist with locally interlayered garnet-diopside–bearing calc-silicate hornfels. The quartzite and schist are dark brown to black, moderately to highly fissile, and lithologically layered on scales of 1–30 cm. The observed mineral assemblage is quartz + biotite + plagioclase + magnetite ± either garnet or white mica, but not both. These assemblages are stable over a broad range of pressure-temperature (P-T) conditions and, because garnet and white mica were not found to coexist, no thermobarometer reaction was identified. Therefore, petrography suggests only that metamorphism probably took place in the amphibolite facies, although Sisson et al. (1996) reported a garnet-plagioclase-biotite-aluminosilicate-quartz P-T estimate of 740 °C, 0.21 GPa for a sample from Taboose Creek. The schist commonly grades into migmatite within 1–5 m of its contact with the underlying Red Mountain Creek pluton. Elsewhere, the schist and quartzite are strongly folded and foliated but lack evidence for partial melting.

Correlation of the quartzite and schist with the Early Cambrian Campito Formation is based on the petrography of the unit and on its Nd isotopic composition, which we use as a geochemical fingerprint. The only dark brown to black quartzite known from the regional stratigraphy is the Andrews Mountain Member of the Campito Formation (Nelson, 1962). As in the Campito Formation (Mount, 1980; Ernst, 1996), the dark color of the quartzite reflects abundant magnetite. The dark quartzite locally contains well-preserved ripple cross-bedding (Fig. 4) that resembles cross-bedding seen in Campito quartzite in the White Mountains. The garnet- or white mica–bearing schist may represent the finer grained Montenegro Member of the Campito Formation. The quartzite and schist were not mapped separately because the two are intimately intermixed, possibly owing to tectonic imbrication or isoclinal folding.

The quartzite-schist unit has a distinctive whole-rock Nd isotopic composition that is characteristic of the Campito Formation. Farmer and Ball (1997) found that the Andrews Mountain Member belongs to a regionally extensive Early Cambrian stratigraphic interval of the Cordilleran miogeocline that has a distinctive, relatively primitive present-day Nd isotopic composition (ɛ_Nd = –4 to –6). Split Mountain quartzite and schist samples share this Nd isotopic signature (Fig. 5), which is quite distinct from other metasedimentary units of the Cordilleran miogeocline (–18; Farmer and Ball, 1997). Although the Nd isotopic composition of the quartzite-schist unit resembles that of the nearby Lamarck Granodiorite (Coleman et al., 1995), the Sm/Nd ratios of quartzite and schist samples are like those of the Campito Formation and are significantly greater than those of the Lamarck Granodiorite. Changing ɛ_Nd from –18 to –5 would require nearly all of the Nd in the metasedimentary rock to have been replaced by Nd from the granodiorite. This is unlikely given the geochemical immobility of Nd, and we therefore interpret the Nd isotopic composition of the schist and quartzite to be primary and to confirm correlation of the rock unit with the Campito Formation.

The Campito Formation is locally overlain (Plate 1; Fig 6A) by calcite marble that is variably converted to calc-silicate hornfels that contain garnet, diopside, and epidote. The scale and rhythmic character of lithologic layering in this unit where it is relatively intact (Fig. 6B) closely resembles that of marble 15 km north at Big Pine Creek that is correlated with the Poleta Formation (Moore and Foster, 1980), which is stratigraphically above the Campito Formation. This observation, coupled with generally upright cross-bedding in the Campito Formation (e.g., Fig. 4), indicates that the units are generally in their normal stratigraphic sequence.

The Red Mountain Creek pluton (hereafter the RMC pluton; corresponds to the Red Mountain Creek alaskite of Moore, 1963) is dominated by homogeneous, equigranular, coarse-grained leucogranite with a color index generally <2. The coarse-grained leucogranite grades upward into a fine-grained aplitic cap, 100–200 m thick, near the roof of the pluton. On the northeastern headwall of Cardinal Lake cirque, the leucogranite is commingled with several masses of epidotized diorite and mafic granodiorite that range to a few hundred meters across.

The gently dipping roof of the RMC pluton is familiar to many geologists owing to Shelton’s (1966) often-reproduced low-angle aerial photographs of the east faces of Split and Cardinal Mountains and the arête that joins them (Figs. 1 and 7). The roof is also subhorizontal and dramatically exposed on the northern side of Red Mountain Creek, in the headwall of Cardinal Lake cirque, on the south flank of Cardinal Mountain, and near the summit of Goodale Mountain (Plate 1; Figs. 7 and 8).

Westward from the summits of Cardinal and Goodale Mountains, the roof contact bends down to a subvertical dip (Plate 1; Fig. 9). We interpret the bend not to be a primary feature of the contact, but rather a product of later tectonic shearing (see following). The roof can also be traced northward, then eastward, from Split Mountain to the south flank of Mount Tinemaha where the contact descends eastward to the range front. The dip of the primary contact does not increase to the northeast in this area as previously reported (Zak and Paterson, 2006); rather, the dip of the contact remains gentle but the contact steps down across a series of high-angle faults (Plate 1; Figs. 8A, 8C; see following).

Zak and Paterson (2006) stated that many large blocks stoped from the roof are enclosed in the RMC pluton, but xenoliths are actually extremely rare (<<1 area %). We performed several traverses on the west side of the east-facing arête shown in Figure 7, as well as in the Taboose Pass and Red Lake areas, and xenoliths of any size are essentially absent. As is clear from Figure 7, xenoliths are at most very uncommon in the east-side cliffs.

As in many plutons, wall-rock inclusions of any size are found only near wall-rock contacts. However, many bodies that appear to be inclusions are probably in situ pieces of the roof. Overall, the roof contact is horizontal and discordant to wall-rock foliation, which generally dips moderately to the southwest (Figs. 10–12). In detail, the contact follows wall-rock foliation for a few meters to a few hundred meters, then cuts across the foliation to follow a different foliation plane. The resulting sawtooth irregularities, combined with the rugged topography, can produce the appearance of detached blocks when viewed from some directions. However, when viewed from more than one angle, such apparently isolated blocks commonly prove to be attached to the roof (Fig. 8D). Moreover, the topographic surface in much of the Red Mountain Creek area is near the elevation of the pluton roof (Fig. 3). The combination of rugged topography and the stepped pluton roof results in preservation of isolated but in situ pieces of the roof where there is slightly less erosion or minor downdropping by faults. This geometrical relationship is especially evident from the air.

The RMC pluton has not yet yielded a reliable U-Pb zircon date owing to a combination of inherited zircon and Pb loss (D.S. Coleman, 2004, personal commun.). A Late Jurassic age is inferred from field relations and comparison to nearby similar plutons. The RMC pluton clearly is older than 148 Ma because it is abundantly intruded by the Independence dike swarm (Moore, 1963; Chen and Moore, 1979; Carl and Glazner, 2002). The age of the RMC pluton relative to the Tinemaha pluton (described in the following) is unclear. The plutons are separated nearly everywhere by the Split Mountain screen and thus are rarely in contact. At the few discontinuities in the wall-rock screen, the plutons are separated by a few tens of meters of fine-grained biotite granite that has indistinct gradational contacts with both plutons. The RMC pluton is probably similar in age to the nearby Twin Lakes and Diamond leucogranite plutons, which are located to the southwest of the study area (Moore, 1963) and yielded U-Pb zircon ages of 164.5 ± 2.3 Ma (Mahan et al., 2003) and 161 ± 1 Ma (Carl, 2000), respectively.

The Tinemaha Granodiorite is composed of mafic hornblende granodiorite that commonly contains sparse centimeter-scale alkali feldspar phenocrysts (Bateman, 1965, 1992) and yielded a 165 Ma U-Pb zircon date (Chen and Moore, 1982). It is one among numerous Middle to Late Jurassic granodiorite plutons found in the Sierra Nevada and in nearby ranges to the east (e.g., Chen and Moore, 1982; Bateman, 1992; Whitmarsh, 1998; Coleman et al., 2003; Ernst et al., 2003; Bartley et al., 2007).

Exposures at Split Mountain are at the southern margin of the ∼75 km² Tinemaha pluton (Bateman, 1965) and are important because the contact of the Tinemaha pluton dips gently under the pluton and thus represents its floor. This relation is clearly apparent from the outcrop pattern on the east face of Split Mountain, on the northern wall of Red Mountain Creek canyon, and on Stecker Flat (Plate 1; Figs. 3, 8A, 8C, and 8D). In all of these locations, the Tinemaha pluton overlies the Split Mountain screen along a gently dipping contact oriented subparallel to the roof of the underlying RMC pluton.

The southwesternmost known outcrop of Tinemaha Granodiorite is perched on the west ridge of Cardinal Mountain, where the granodiorite overlies Cambrian metasedimentary strata and is intruded by the eastern contact of the Lamarck Granodiorite (Plate 1). This outcrop suggests that the floor of the Tinemaha pluton continued at least that far to the southwest. The Tinemaha and RMC plutons thus appear to have been vertically stacked over an area of at least 15 km² that encompasses most of the exposed area of the RMC pluton.

The floor of the Tinemaha pluton, like the roof of the underlying RMC pluton, crosscuts tectonic structures in metasedimentary wall rocks (Fig. 13). The Tinemaha Granodiorite is largely devoid of xenoliths except near its floor, where it encloses several large (∼10–200 m) bodies of marble and calc-silicate rock that appear to correlate with nearby Poleta Formation of the interpluton screen (Figs. 8A and 13). The ends of the Poleta blocks are steep, planar, and bounded by tabular bodies of granodiorite that may be dikes, but cliff exposures make direct observation difficult.

All pre-Cretaceous rocks in the study area were intruded by the Independence dike swarm (Moore and Hopson, 1961; Moore, 1963). The overwhelming majority of the dikes intruded at 148 ± 2 Ma (Carl and Glazner, 2002), although dikes of other ages are found locally (e.g., Coleman et al., 2000). The dike swarm is intensely developed in the Split Mountain area (Plate 1; Fig. 11) and locally as much as 25% of the outcrop area is occupied by dikes (Bartley et al., 2007). The dikes range widely in composition (e.g., Glazner et al., 2008), but >90% of those in the study area are mafic and range from diabase to hornblende diorite. Individual dikes on Split Mountain, Mount Tinemaha, and nearby Mount Prater are as much as 15 m thick, although all such dikes are highly composite (Coleman et al., 2005; Bartley et al., 2007; Glazner et al., 2008).

The Lamarck Granodiorite (Bateman, 1992) is a steep-walled, narrow, northwest-striking body ∼60 km long. Only a small part of the southeastern end is in the study area (Plate 1). It is mainly composed of texturally diverse hornblende granodiorite but ranges from gabbro to granite (Frost and Mahood, 1987). Geologic mapping indicates that the pluton is highly composite (Hathaway, 2002; Gracely, 2007), and U-Pb geochronology indicates that its growth duration was at least 2–3 m.y. (91.9–94.5 Ma; Coleman et al., 1995; Davis et al., 2012). Magmas that span the full compositional range were added to the pluton through the full duration of its assembly (Gracely, 2007), and at least parts of the pluton were assembled by amalgamation of steeply dipping intrusive sheets (Hathaway, 2002; Gracely, 2007; Clemons and Bartley, 2008). The anatomy of the Lamarck Granodiorite thus resembles the smaller and closely related McDoogle pluton (Fig. 2; Moore, 1963; Mahan et al., 2003), which is slightly older (95 ± 0.1 Ma; Davis et al., 2012) and is located immediately along strike to the southeast.

Individual intrusive sheets in the Lamarck pluton commonly strike parallel to the length of the pluton as a whole, but broadly east-west–striking sheets that cut the predominant structural grain are also common (Hathaway, 2002). This is exemplified by west-northwest–striking gabbroic dikes west of Taboose Pass (Plate 1), which have sharp planar margins but locally pass along strike into tabular zones of commingled mafic and felsic magma. The mafic dikes therefore represent late additions to the pluton that were emplaced before the host granodiorite was fully consolidated (Frost and Mahood, 1987).

West of Taboose Pass, the Lamarck Granodiorite concordantly intrudes subvertical Campito schist and quartzite of the Split Mountain screen (Fig. 9). Discrete bodies of marble and calc-silicate rocks derived from the Poleta Formation are locally present along the intrusive contact. Diffuse foliation in the Lamarck Granodiorite is broadly concordant with the contact but there is no evidence of solid-state deformation. Intense foliation in the wall rocks reflects deformation within the Sawmill Lake shear zone (see following). This geometry abruptly changes immediately north of the western ridge crest of Cardinal Mountain. The intrusive contact of the Lamarck pluton bends sharply to a 40°–50°E dip and truncates the wall-rock screen, the foliation that it contains, and the intrusive contact of the RMC pluton with the screen (Plate 1; Fig. 14). The intrusive contact maintains this moderate eastward dip to the base of the cirque headwall east of Cardinal Lake (Plate 1), where the contact is knife sharp against the RMC pluton and truncates Late Jurassic dikes. Rocks on both sides of the contact are unfoliated. Ascending the northern cirque wall, the wall-rock screen reappears between the Lamarck and RMC plutons and the intrusive contact returns to a subvertical dip.

Parts of several other smaller Cretaceous plutons are found in the southern part of the area of Plate 1 (Moore, 1963; Bradford, 1995). These include the Striped pluton, which is composed of fine-grained biotite granodiorite, and diverse rocks including the ca. 92 Ma Aberdeen mafic complex (Bradford, 1995) and the Goodale granodiorite and Siberia leucogranite, which have commingled contacts with the mafic complex and are therefore interpreted to be coeval with it.

Four aspects of intrusive contacts in the study area merit particular emphasis.

1. The contacts are sharp, with two possible exceptions related to the RMC pluton. First, its contact with granite of Taboose Creek is difficult to locate precisely in the field. This may reflect the lithologic similarities of leucocratic granites as much as gradation from one to the other. Second, Campito Formation schist in the roof of the RMC pluton is migmatitic for as much as 5 m above the contact. However, the fine-grained white leucosome in the migmatite does not resemble the coarse leucogranite of the adjacent pluton, and the leucosome appears mainly as irregular centimeter-scale segregations rather than as lit-par-lit intrusive sheets. Therefore, the migmatite appears to reflect local in situ partial melting of the wall rock, probably fluxed by fluids evolved from the underlying pluton.

2. The intrusive contacts are discordant to wall-rock structure (Figs. 10 and 11), except for the western contact of the RMC pluton near Taboose Pass and the adjacent eastern contact of the Lamarck Granodiorite (Fig. 9). Both concordant contacts reflect effects of the Taboose Pass shear zone (see following), although the role of the shear zone differs between the two contacts. The concordance is strictly local; each contact can be traced in outcrop to locations where it is discordant (Plate 1).

3. Primary intrusive contacts of the RMC, Tinemaha, and Striped plutons either currently dip gently or have been tectonically reoriented. Exposures of the floor of the Striped pluton are poor owing to frost-heaving and talus cover, but the subhorizontal orientations of the roof of the RMC pluton and the floor of the Tinemaha pluton are spectacularly displayed in prominent cliffs over an area of ∼15 km² (Plate 1; Figs. 1, 3, 7, 8A, 8C, 8D, 10, and 11).

4. The Jurassic RMC pluton is bounded on the south by the small Cretaceous intrusions mentioned herein, mainly the Goodale granodiorite (Plate 1; Fig. 8B). The southern primary intrusive contact of the RMC pluton thus is not preserved and its original southward extent and geometric form are unknown.

We address here the tectonic deformation history recorded by rocks in the Split Mountain area. The map pattern shown in Plate 1 mainly reflects the intrusive relationships discussed herein, but shear zones and faults described in this section are locally important. Outcrop-scale evidence summarized in the following indicates that wall rocks that compose the Split Mountain screen were ductilely deformed at least twice, before intrusion of the Jurassic plutons that bound the screen, and again after intrusion of the Independence dike swarm but before emplacement of the Cretaceous plutons. The latter period of deformation also sporadically affected Jurassic intrusive rocks, principally the Independence dikes and the Tinemaha Granodiorite. Rock fabrics formed in each ductile deformation phase locally had significant effects on magmatic intrusion. Brittle faults that offset contacts by mappable amounts are largely restricted to the northern side of the Red Mountain Creek drainage and near the eastern range front.

The oldest deformation feature recognized in the study area is a pervasive schistosity in the metasedimentary rocks, here termed S₁, which is consistently truncated by Jurassic intrusions (Plate 1; Figs. 10–12). Except locally in the limbs of mesoscopic to macroscopic folds (e.g., Fig. 13) and where reoriented by the Sawmill Lake shear zone (Plate 1; Fig. 9; see following), S₁ dips moderately toward the southwest. It commonly contains a lineation, L₁, that is defined by elongate mineral grains and plunges gently to moderately southwest.

The S₁ foliation is commonly folded by outcrop-scale, open to tight, sharp-hinged folds that, like the foliation that they deform, are consistently cut by the Jurassic intrusions (Fig. 13). Fold hinges that trend both northwest and northeast are observed, but northwest-trending folds are more common and dominate the pattern of foliation poles in stereographic projection (Fig. 12). Axial surfaces of northwest-trending folds consistently dip to the southwest and suggest that the deformation included a significant component of top-to-the-northeast shear, i.e., in a direction roughly concordant with L₁. All ductile structures older than the Jurassic plutons suggest northeast-directed contractional shear.

Discrete ductile deformation zones affect the Jurassic plutons and dikes as well as pre-batholithic wall rocks. The deformation zones are most readily observed and interpreted where they affect intrusive rocks, and therefore these zones are described first. However, the largest and most regionally significant deformation zone, the Sawmill Lake shear zone, mainly affects metasedimentary rocks within the field area.

Uniformly subvertical Independence dikes are abundant throughout the study area and most are ductilely deformed. Foliation in the dikes has a sigmoid pattern in plan view in which foliation in the center is oblique to the walls but intensifies and curves toward parallelism with the walls as the walls are approached (also see Moore, 1963; Carl et al., 1998; Glazner et al., 1999; Carl, 2000). The absence of the fabric in the wall rocks indicates that Independence dikes were less competent than the Jurassic plutons that they intruded. Each dike thus functioned as a ductile shear zone (Ramsay, 1980), the orientation of which was preset by the dike.

The obliquity of the foliation to the dike indicates the sense of shear, which varies with dike orientation. The majority of Independence dikes strike between west and northwest (Moore and Hopson, 1961; Carl and Glazner, 2002; Bartley et al., 2007), and these dikes consistently record sinistral shear (Moore, 1963; Carl et al., 1998). A small number of north-striking dikes are lithologically identical to the west- and northwest-striking dikes, have the same crosscutting relations, and locally can be seen to branch from west-striking dikes. The north-striking Independence dikes consistently record dextral shear (Fig. 12). Sheared Independence dikes thus define a conjugate pattern that accommodated bulk northeast-southwest shortening and northwest-southeast elongation (Fig. 15). The predominance of west- to northwest-striking dikes results in a net component of sinistral shear in the bulk strain field.

Phyllonitic shear zones that range from 1 cm to several meters thick are sporadic but common in the Tinemaha Granodiorite, and are also present but much less common in the RMC pluton. Phyllonite forms by hydration of framework and chain silicates to form weaker phyllosilicates, particularly chlorite and white mica, which then localize ductile deformation (e.g., White et al., 1980). Phyllonites observed in the Tinemaha Granodiorite range from joint surfaces coated with a thin veneer of ductilely sheared material, to millimeter-thick zones that are coplanar with unsheared joints, to fully developed phyllonitic shear zones as much as several meters thick. The phyllonite zones thus appear to have grown at preexisting joints where the hydrating fluid was introduced.

The diverse orientations of the phyllonite zones thus appear to be inherited from preexisting joint sets. As a result, the kinematics of the phyllonite zones, like the sheared Independence dikes, vary with orientation and also are consistent with bulk northeast-southwest horizontal contractional strain (Fig. 12). Reverse-sense shear is observed across phyllonite zones that strike northwest and dip either northeast or southwest. A component of sinistral shear is present across more westward-striking shear zones and a component of dextral shear appears across zones that strike more northward.

Kinematics of the phyllonite zones and the sheared Independence dikes thus combine to indicate a bulk flattening strain with northeast-southwest shortening and both northwest-southeast and vertical extension (Fig. 15). The absolute magnitude of displacement represented by sheared Independence dikes and phyllonite zones in the Tinemaha Granodiorite is difficult to quantify owing to the sporadic spatial distribution of the shear zones, but it is unlikely to be large enough to be regionally significant.

A broad, north-striking subvertical zone of intense ductile strain is exposed ∼1 km west of Taboose Pass, and we correlate it with the Sawmill Lake shear zone of Mahan et al. (2003) (Fig. 2). The zone is at least 300 m thick and probably reflects regionally significant displacement. Based on similar orientations and crosscutting relations, the Sawmill Lake shear zone probably formed close in time to and in the same regional strain field as the smaller shear zones. However, the Sawmill Lake shear zone is more subtle in the field than the phyllonite zones in the Jurassic plutons because, in this location, it is largely confined to metasedimentary rocks that were already strongly foliated.

The most obvious expression of the Sawmill Lake shear zone is macroscopic deformation of the Split Mountain interpluton screen. Downward bending of the screen westward from the summits of Cardinal and Goodale Mountains (Plate 1; Fig. 9) is accompanied by the appearance of an intense subvertical schistosity in the Campito Formation that is younger than S₁ based on relations to Independence dikes. Ductile shearing of Independence dikes is common throughout the study area, but their original intrusive geometry is scarcely affected and they remain visibly discordant to the S₁ wall-rock foliation (Fig. 12). In contrast, in the steep part of the interpluton screen west of Taboose Pass, the dikes are intensely disrupted and now are represented only by scattered boudins in a matrix of vertically foliated schist (Fig. 16A). The resulting mélange-like rock mass also contains granodiorite boudins probably derived from the Tinemaha pluton. These observations indicate that the Sawmill Lake shear zone is younger than any of the Jurassic intrusive rocks and therefore is younger than and distinct from S₁.

Although lineation in the Sawmill Lake shear zone plunges steeply, such lineation is not strongly developed, and consistent microstructural shear-sense indicators have not been observed in the study area. These characteristics suggest that the bulk strain field includes a large component of coaxial flattening. However, (1) the macroscopic deflection of contacts seen in Figure 9 indicates a significant component of west-side-down shear, consistent with kinematic data from the shear zone farther south (Mahan et al., 2003), and (2) during reconnaissance mapping of the shear zone on Mount Bolton Brown 5 km north of the study area, it was observed that Independence dikes had undergone intense dextral asymmetric boudinage (Fig. 16).

Rocks of the Sawmill Lake shear zone are bounded on the west by the intrusive contact of the Lamarck Granodiorite, which is concordant with the shear zone in the area west of Taboose Pass. However, as noted earlier, the Lamarck pluton completely cuts out the shear zone at Cardinal Lake and therefore the shear zone is older than the Lamarck Granodiorite. The preserved thickness of the Sawmill Lake shear zone thus is a minimum. Crosscutting and overprinting relations bracket the age of the shear zone between intrusion of the Independence dikes at 148 Ma and intrusion of the Lamarck Granodiorite at 92–94 Ma.

The Sawmill Lake shear zone has been traced 20 km northwestward to highly sheared rocks that are found on both side of, and locally within, the Lamarck pluton (Fig. 2; Clemons and Bartley, 2008). As in the present study area, intrusive contacts of the Lamarck pluton are commonly concordant to the shear zone fabric (Hathaway, 2002; Gracely, 2007) but locally sharply truncate it. Farther to the northwest, the Sawmill Lake shear zone may link to the Rosy Finch shear zone (Fig. 2; Tikoff and de St. Blanquat, 1997), but evidence is insufficient to evaluate this hypothesis.

Several high-angle, west- to northwest- to north-northeast–striking faults are exposed on Mount Tinemaha and on the northern flank of Split Mountain. Gently dipping contacts are consistently dropped down on the north and east sides of the faults (Plate 1; Fig. 8C). Steeply dipping markers have not been matched across the faults and therefore it is unknown if the faults also accommodated significant strike slip.

Bateman (1965) and Moore (1963) mapped numerous north- to northwest-striking faults near the eastern foot of the Sierra Nevada that are expressed as topographic and vegetation lineaments. Several of the faults, some of which are antithetic, cross Stecker Flat and Shingle Mill Bench (Plate 1) and produce minor offsets of bedrock contacts. The faults appear to be related to the frontal fault system of the Sierra Nevada and probably are the youngest structures in the study area.

Contact and fabric relations of the Red Mountain Creek, Tinemaha, and Lamarck plutons indicate that all of their intrusive contacts originated as fractures and therefore the plutons were emplaced either by stoping or by diking. The evidence reviewed below favors diking over stoping for the emplacement of all three plutons.

Flat and steep contacts. Everywhere that the original intrusive contact of the RMC pluton is preserved, it either dips gently or has been reoriented by later deformation (Plate 1; Figs. 7–11). All exposures of the contact, although now geometrically more complex owing to overprinting geologic events, collectively represent a portion of the original roof of the pluton. The overall plan-view shape and areal extent of the pluton thus are unknown.

This conclusion contrasts with the report that the northern contact of the RMC pluton is an intact steep wall that, at the northwest corner of the pluton, passes continuously into the roof (Zak and Paterson, 2006; e.g., see their fig. 1). We closely examined this area both on the ground (Plate 1) and from a low-flying aircraft (Figs. 8A, 8C), and the primary intrusive contact of the RMC everywhere dips gently. However, the contact is dissected by the high-angle Tinemaha fault zone, which does not appear on the maps of Bateman (1965) or Zak and Paterson (2006). The fault zone drops the contact down to the range front in a series of steps, most of which are concealed in couloirs and therefore can be recognized only through close-up inspection. If one is unaware of the fault zone, the overall map pattern naturally leads to the inference that the intrusive contact is steep. However, this clearly is not the case, and no roof-to-wall transition is exposed.

In contrast, the western contact of the RMC pluton is steep where it intersects the Sawmill Lake shear zone. However, as described herein, this clearly reflects reorientation of the contact in the shear zone. The original orientation of the portion of the contact that is within the shear zone cannot be determined from available evidence, but no evidence suggests that the primary dip in this location differed significantly from the gentle dip observed in all other exposures.

Stepped roof. The roof of the RMC steps irregularly, as emphasized by Zak and Paterson (2006), who interpreted the steps as evidence of stoping. However, in our view, the most significant aspect of the roof contact is not its stepped character, but rather the strikingly horizontal orientation and roughly planar overall shape of the roof in spite of its sharp discordance to wall-rock foliation (Figs. 1 and 7–11; cf. Glazner and Bartley, 2006). We find it difficult to envision a reason that piecemeal stoping would abruptly stop at a subhorizontal plane that cuts across foliation in the middle of a wall-rock unit. It appears more likely that this contact geometry was produced by propagation of a subhorizontal crack system that opened subvertically to admit magma. Because the crack system propagated through wall rocks that were mechanically anisotropic owing to an older foliation, the fractures locally exploited the foliation and produced the stepped geometry.

Xenolith abundance. Xenoliths are rare in the RMC pluton. Even if all of the dark rocks mapped inside the RMC pluton were stoped blocks, they compose a tiny fraction of the outcrop area of the pluton. Point-counting our map (Plate 1) and Figure 1 of Zak and Paterson (2006) yields area fractions of 0.2% and 0.3%, respectively.

However, these numbers overstate the abundance of xenoliths because at least some wall-rock bodies mapped in the RMC pluton are in situ pieces of the roof. Spectacular cliff exposures clearly show that the Tinemaha Granodiorite overlies the Split Mountain screen and the RMC pluton along a gently dipping floor contact (discussed in the following). A mass of Tinemaha Granodiorite also is perched on top of Stecker Flat, where it is surrounded by underlying RMC leucogranite. This body of Tinemaha is at least locally separated from underlying Red Mountain Creek leucogranite by Campito schist and quartzite, which are at much the same elevation as equivalent outcrops across the canyon to the north (Plate 1; Fig. 3). The contacts are cut by high-angle faults that are probably related to the Tinemaha fault zone and that similarly step the intrusive contacts mainly down toward the northeast. We therefore interpret Stecker Flat to provide in situ exposures of the RMC pluton roof (and of the Tinemaha pluton floor), preserved in part by high-angle faulting and isolated from other roof exposures by incision of the Red Mountain Creek drainage. Two blocks of Campito schist and quartzite that are mapped entirely within the RMC pluton on Stecker Flat, one to the north of the Tinemaha body and the other to the southeast, are at elevations similar to correlative rocks at the mapped intrusive contacts. These bodies are most likely in situ outliers of the roof that were preserved by high-angle faults and isolated by erosion (Plate 1).

The abundance of wall-rock xenoliths in the RMC pluton cannot have been reduced by assimilation. RMC leucogranite is a high-silica, minimum-melt granite (73–77 wt% SiO₂; M. Pardue, 2008, personal commun.) composed almost entirely of quartz and feldspars. Assimilation of a significant amount of magnetite- and carbonate-rich wall rocks would cause obvious petrographic changes that are not observed, such as growth of wollastonite, diopside, and Fe-bearing phases. More fundamentally, minimum-melt granite magma is at the bottom of a thermal and energetic minimum in composition space and thus lacks any excess thermal energy to melt wall rocks (Bowen, 1928; Glazner, 2007).

Lit-par-lit dike injection is common at the margins of plutons that intrude previously foliated rocks, and the RMC pluton presents spectacular examples of the phenomenon (e.g., Fig. 7); Zak and Paterson (2006) interpreted these to record incipient stoping. However, in at least some instances, and possibly most, the wall rocks remained in situ and, even where now isolated inside the pluton, the wall rocks never were detached or engulfed in magma (also see Bartley et al., 2008). We have seen no conclusive evidence in the RMC pluton that injection of dikes into the pluton roof progressed to detachment of blocks from the roof into a magma chamber. As with many plutons worldwide, the presence of xenoliths almost exclusively at pluton margins is consistent with in situ isolation by successive injections of magma (Glazner and Bartley, 2006; Bartley et al., 2008).

Shape. The present outline of the RMC pluton (Plate 1) does not reflect either its original shape or extent. The mapped limits of the RMC pluton are defined by geologic features that are both younger than and unrelated to the pluton. On its eastern side, the pluton is truncated by the Cenozoic range-front fault zone and unconformably overlapped by Quaternary sediment; on the south, the pluton is cut off by intrusion of several younger plutons; the pluton’s western limit is governed by the Sawmill Lake ductile shear zone; and the outcrop pattern of the northern contact is greatly modified by the high-angle Tinemaha fault zone. Therefore, although Zak and Paterson (2006) considered the equant mapped shape of the RMC pluton to indicate emplacement by a combination of diapirism and stoping, this shape actually is the composite result of several later geologic events, and neither reflects the original pluton shape nor provides evidence relevant to pluton emplacement mechanisms.

A key test of the hypothesis that stoping is a significant process in pluton emplacement is whether large accumulations of stoped blocks are observed on pluton floors (e.g., Buddington, 1959). The floor of the Tinemaha pluton is spectacularly exposed in the study area and thus provides an excellent test case. A few blocks of wall rock are enclosed in the Tinemaha pluton at or near its floor (Plate 1; Figs. 8A and 13). The blocks are composed exclusively of marble and calc-silicate rocks derived from the Poleta Formation, which elsewhere in the study area depositionally overlies the Campito Formation (e.g., Fig. 6A). Structures inside the blocks, including foliation and folds, are concordant with comparable structures in the wall-rock screen between the Tinemaha and RMC plutons. It therefore appears likely that, rather than representing blocks stoped from the roof of the Tinemaha pluton, these bodies were isolated in situ by dike injection (cf. Mahan et al., 2003; Bartley et al., 2008). However, even if the marble bodies are stoped blocks, they represent a minuscule volume fraction of the Tinemaha pluton here and elsewhere (Sawka et al., 1990). The sparseness of such blocks indicates that stoping was volumetrically insignificant during emplacement of the Tinemaha pluton (also see Glazner and Bartley, 2006, 2008).

Contact relations of Lamarck Granodiorite in the study area resemble those of the McDoogle pluton (Mahan et al., 2003), which is directly along the southward projection of the Lamarck pluton and could be regarded as an early mafic phase of it (e.g., Coleman et al., 2005). Four observations indicate that Lamarck Granodiorite in the study area was emplaced largely, and probably entirely, by opening of fractures after displacement across the Sawmill Lake shear zone: (1) the abrupt truncation in Cardinal Lake cirque of the shear zone fabric and of the intrusive contact of the RMC pluton; (2) the absence of a solid-state fabric in the Lamarck Granodiorite; (3) the undeformed condition of the crosscutting mafic dike complex that is commingled with Lamarck Granodiorite; and (4) where the Lamarck was intruded against the RMC pluton without intervening metasedimentary rocks, the absence of evidence for solid-state flow in any of the rocks. These characteristics favor intrusion of the Lamarck by opening of fractures and suggest that subvertical, concordant sections of the contact reflect preferential fracturing of wall rocks parallel to the subvertical wall-rock foliation (cf. Mahan et al., 2003).

Only a few xenoliths were observed in the Lamarck pluton in the study area, and all are within a few tens of meters of the Split Mountain screen from which they were derived. Xenoliths are practically absent (<<0.01% area) from the rest of the Lamarck Granodiorite, with the exception of the early, volumetrically minor Treasure Lakes phase, which crops out 20 km northwest of the study area (Hathaway, 2002; Gracely, 2007; Clemons and Bartley, 2008). Structurally, the Treasure Lakes phase of the Lamarck pluton closely resembles the McDoogle pluton (Clemons and Bartley, 2008), and we therefore infer that the xenoliths in that phase of the Lamarck also were isolated in situ by intrusion of a plexus of dikes (cf. Mahan et al., 2003; Bartley et al., 2008).

Key features of the Split Mountain interpluton screen that indicate its origin include the following. First, rocks in the screen were ductilely deformed twice, but neither deformation phase corresponds in time to emplacement of any pluton that bounds the screen. Structures produced by the earlier period of deformation (Triassic–Jurassic?; cf. Stevens and Greene, 2000) are discordant to and are truncated by all intrusions in the study area. Structures produced by the second (latest Jurassic–Early Cretaceous) period of deformation overprint all of the Jurassic intrusive rocks and are cut by all of the Cretaceous intrusions. Therefore, the thin tabular shape of the screen was not produced by ductile stretching of an originally more equidimensional rock body during intrusion of one or more of the bounding plutons. Second, the intrusive contacts are sharp and are commonly accompanied by intrusion of dikes into wall rocks. Therefore, tensile fracture occurred at the pluton margins during intrusion, and opening of fractures to admit magma probably accounts for all of the sharp intrusive contacts. Third, the northern part of the screen is subhorizontal and bounded above and below by Jurassic plutons, whereas the southern part of the screen has steep contacts and is bounded on either side by Cretaceous plutons. The transition between the two segments is located on the western ridge of Cardinal Mountain where the screen enters the Sawmill Lake shear zone, which sheared the screen into a subvertical orientation.

Based on these observations, we interpret the thin, tabular shape of the Split Mountain interpluton screen to be the product of tensile splitting of wall rocks and injection of magma along subparallel fractures. The screen initially formed in the Jurassic in a subhorizontal orientation, as now observed on Split Mountain where the screen separates the Tinemaha and RMC plutons. The Jurassic plutons appear to be broadly laccolithic, having grown by subvertical opening of subhorizontal crack systems. Available evidence is insufficient to determine if space was made mainly by roof lifting, in which case the plutons are laccoliths in the strict sense, or by floor subsidence, in which case the plutons would be better termed lopoliths (e.g., Cruden and McCaffrey, 2001).

After Late Jurassic (ca. 148 Ma) emplacement of the Independence dike swarm but before emplacement of the Lamarck pluton (ca. 92–94 Ma), the Sawmill Lake shear zone sheared part of the screen into a subvertical orientation. Much of the Sierran arc underwent deformation during this time period (e.g., Saleeby et al., 1990; Tikoff and Teyssier, 1992; Tobisch et al., 1995; Mahan et al., 2003). Therefore, the Sawmill Lake shear zone appears to be part of a regional tectonic system rather than being a product of local magmatic processes.

The final stage in formation of the Split Mountain screen was intrusion by Cretaceous plutons, including the Lamarck Granodiorite, which was guided by the Sawmill Lake shear zone. We interpret the Lamarck Granodiorite, like the closely related McDoogle pluton, to have been emplaced by magmatic crack-seal (Bartley et al., 2008) in which fractures that opened to admit magma were guided by the steep shear zone foliation. The southern steep segment of the Split Mountain screen thus appears to reflect emplacement processes that were similar to the northern subhorizontal part, but intrusive contacts that bound the screen are steep rather than subhorizontal; fractures were steep because they were guided by mechanical anisotropy imparted by wall-rock foliation.

Given the limited evidence, any correlation and regional significance assigned to pre–Middle Jurassic deformation in the Split Mountain screen is speculative. The probable context for the deformation is the east Sierran thrust system (Dunne, 1986), a long-lived system of east-vergent contractional structures localized at the eastern margin of the Sierran arc throughout much of its history. However, it is unclear where the Cambrian metasedimentary rocks of the Split Mountain screen would fit into the tectonostratigraphy of this system. Restoration of 65 km of post–83 Ma dextral shear across Owens Valley (Kylander-Clark et al., 2005; Bartley et al., 2007) places the Split Mountain area adjacent to the southern Inyo Mountains, where folds and thrusts of the east Sierran thrust system are well known, but rocks at the surface are late Paleozoic and Mesozoic in age. The Cambrian rocks at Split Mountain therefore might represent unroofing of a deeper level that formed downdip of the east Sierran thrust structures now exposed in the Inyo Mountains, or they might belong to a higher thrust sheet that once overlay the rocks now exposed in the southern Inyo Mountains.

The orientations, timing relations, and kinematics of the Sawmill Lake shear zone and of smaller shear zones throughout the study area match the Sawmill Lake shear zone along strike to the south into which the McDoogle pluton intruded late synkinematically (Mahan et al., 2003). Mahan et al. (2003) noted that similar mylonitic deformation of pre-Cretaceous plutonic and metasedimentary rocks also is found along strike to the north near Long Lake (Hathaway, 2002; Clemons and Bartley, 2008). The oldest Cretaceous intrusions near Long Lake, the Inconsolable pluton and the Treasure Lakes phase of the Lamarck Granodiorite (95.4 and 94.5 Ma, respectively; Davis et al., 2012), are very close in age to the McDoogle pluton and, like the McDoogle, the Inconsolable and Treasure Lakes intrusions locally contain steep contractional mylonite zones (Clemons and Bartley, 2008).

The evidence thus indicates the presence in the eastern Sierra Nevada of a regionally extensive system of Early to early-Late Cretaceous mylonite zones that accommodated broadly east-west contraction and east-side-up shear. This kinematic regime clearly was active until ca. 95 Ma and only minor displacements occurred after this time. It remains unclear how much earlier the shear zones began to develop and how strain accumulation was distributed in time between 148 and 95 Ma. Glazner et al. (1999) suggested that sinistral shearing began during intrusion of Independence dike swarm at 148 Ma, whereas U-Pb geochronology and finite strain patterns in the Oak Creek pendant led Saleeby et al. (1990) to infer that shearing in that area after emplacement of the Independence dikes did not begin until after 110 Ma (see Mahan et al., 2003, for further discussion).

The predominance of northeast-southwest contraction (Fig. 15) suggests that, like the pre–Middle Jurassic deformation, Cretaceous shear zones in the study area are related to the east Sierran thrust system. These structures may represent internal wedge thickening behind the east-vergent thrusts exposed in the Inyo Mountains.

However, the steep dip of the shear zone and the predominance of east-side-up shear that persists for at least 50 km along the eastern edge of the Cretaceous Sierran arc suggest an alternative origin. Glazner et al. (2003) noted that both structural observations and isostatic requirements indicate that the Sierra Nevada batholith and its volcanic cover sank isostatically with respect to their surroundings. The age, kinematics, and location of the Sawmill Lake shear zone are consistent with the shear zone having formed to accommodate isostatic sinking of the pre–95 Ma Cretaceous magmatic arc.

Our mapping of the Split Mountain area was undertaken to understand the origin of the thin wall-rock screens that are commonly found between plutons in large batholiths such as the Sierra Nevada, and to examine the implications for pluton emplacement mechanisms. The Split Mountain interpluton screen originally formed in the Jurassic as a subhorizontal sheet between the vertically stacked, broadly laccolithic Tinemaha and RMC plutons. The screen thus was defined by subhorizontal fractures that opened to admit the magma from which the bounding plutons formed. The broadly horizontal but stepped boundaries of the screen are generally discordant to, but locally follow, the preexisting moderately dipping tectonic foliation in the screen. We interpret this pattern to reflect local exploitation of the preexisting foliation by the subhorizontally propagating crack system that guided emplacement of the plutons.

Late Jurassic to Early Cretaceous tectonic contraction and subvertical shear overprinted all Jurassic and older rocks in the area. Most of this deformation was focused into the regionally extensive Sawmill Lake shear zone, which sheared part of the Split Mountain screen and its bounding intrusive contacts into subvertical orientations. Much like the closely related McDoogle pluton (Mahan et al., 2003), the dike-like Lamarck Granodiorite then was emplaced into steep fractures that were guided by the shear zone foliation.

We thus interpret the geometry of the Split Mountain screen and its contact relations with the several plutons that bound it to indicate that the bounding plutons grew largely by opening of fracture systems to admit intruding sheets of magma. This interpretation is consistent with the hypothesis that plutons bounded by sharp, discordant contacts were not emplaced by stoping (Glazner and Bartley, 2006, 2008), but rather reflect incremental growth by magmatic crack-seal (Bartley et al., 2008).

Our work on the plutonic and wall-rock geology of the eastern Sierra Nevada was supported by National Science Foundation grants EAR-9814787, EAR- 0337351, and EAR-0538094 to Bartley, and EAR-9526803, EAR-9814789, EAR-0336070, and EAR-0538129 to Glazner. Ryan Taylor and Brent Miller helped generate the Nd isotopic data, and Michael Strane and Michael Stearns helped with drafting the geologic map. We are grateful to the late John Shelton for granting us permission to use his spectacular photo of Split Mountain as Figure 1. Our close collaborator in plutonic research, Drew Coleman, was not directly involved in work reported here (although he would have been if the U-Pb systematics of the Red Mountain Creek pluton were more tractable), but his contributions to our thinking about intrusive processes are implicit throughout this paper. Finally, association with the White Mountain Research Station and its capable staff over the years benefited this work greatly.