The Punchbowl block is a fault-bounded crustal sliver in the eastern San Gabriel Mountains of southern California with important implications for conflicting reconstructions of the San Andreas fault system. Detailed mapping, determination of conglomerate-clast and sandstone compositions, and dating of detrital and igneous zircon of Oligocene–Miocene strata define two distinct subbasins and document initiation of extension and volcanism ca. 25–24 Ma, followed by local exhumation of the Pelona Schist, and transition from alluvial-fan to braided-fluvial deposition. Strata of the Punchbowl block correlate with those of other regions in southern California, confirming 40–50 km of dextral slip on the Punchbowl fault, and supporting reconstructions with 60–70 km of dextral slip on the San Gabriel/Canton fault and ∼240 km of dextral slip on the southern San Andreas fault. Provenance and probable correlations of Punchbowl-block strata argue against 80–110 km of dextral slip on the San Francisquito–Fenner–Clemens Well fault and limit the time interval during which such slip could have occurred. Synthesis of these findings with previous work produces paleogeographic reconstructions of the Punchbowl block and its probable correlatives through time.

The Punchbowl block, a crustal sliver between the Punchbowl and San Andreas faults (Fig. 1; e.g., Dibblee, 1987), contains Oligocene–Miocene strata, the older parts of which had not been thoroughly investigated prior to this study. Similar Oligocene–Miocene strata are present in the Tejon, Soledad, and Orocopia regions of southern California, which lie on different sides of the San Gabriel/Canton and San Andreas faults (Fig. 1; e.g., Crowell, 1975a). Some palinspastic reconstructions show the Tejon, Soledad, and Orocopia regions as correlated and originally adjacent to each other (e.g., Hill and Dibblee, 1953; Crowell, 1962, 1975a; Carman, 1964; Ehlig and Ehlert, 1972; Bohannon, 1975), whereas others do not correlate some of these regions (e.g., Powell, 1981, 1993; Spittler and Arthur, 1982; Frizzell et al., 1986). The central Punchbowl block is generally accepted as an offset equivalent of the Soledad region; therefore, its previously understudied strata provide important constraints on these palinspastic reconstructions. Furthermore, Oligocene–Miocene strata of the Punchbowl block straddle the Fenner fault, a component of a proposed early trace of the San Andreas fault, the existence of which is debated (e.g., Powell, 1981, 1993; Richard, 1993). For these reasons, we conducted a detailed study of these strata. Our findings support original alignment of the Tejon, Soledad, Punchbowl, and Orocopia regions, and the slip estimates implied thereby, and they argue against an early trace of the San Andreas fault system along the Fenner fault.

Oligocene–Miocene Basins

Punchbowl Block

The Punchbowl block contains the Punchbowl Formation (Noble, 1953, 1954), ∼1500 m of fluvial/alluvial conglomerate, sandstone, and minor mudstone, which accumulated during the middle-late Miocene (Fig. 2; Tedford and Downs, 1965; Woodburne and Golz, 1972; Woodburne, 1975; Dibblee, 1987; Liu, 1990). A distinct basal member is middle Miocene in age (Noble, 1954; Liu, 1990; personal commun. with Allen and Whistler inLiu, 1990). The older units documented in this study (the Paradise Springs and Vasquez formations; Fig. 2) have been interpreted in previous studies as either part of the basal Punchbowl Formation (e.g., Noble, 1954; Dibblee, 2002a, 2002b, 2002c), or deposits in a fault-bounded sliver along the Punchbowl fault that originated in a separate basin (Weldon et al., 1993).

Tejon Region

The oldest nonmarine strata of the Tejon region belong to the Plush Ranch Formation (Fig. 2; Carman, 1954, 1964), composed of alluvial and lacustrine conglomerate, sandstone, siltstone, shale, limestone, and evaporites (Carman, 1964; Cole and Stanley, 1995; Hendrix et al., 2010). Interbedded basalt has been dated by whole-rock and plagioclase K-Ar methods as ca. 26–23 Ma (Crowell, 1973; recalculated after Dalrymple, 1979; Frizzell and Weigand, 1993). Northwest of Plush Ranch basin, on the opposite side of Mount Pinos (which includes exposures of Pelona Schist), Oligocene–Miocene strata are generally mapped as Simmler Formation (Fig. 1; e.g., Kellogg and Miggins, 2002; Dibblee, 2005a, 2005b, 2006b), but are considered equivalent to the Plush Ranch Formation (personal commun. with Hill and Dibblee inCarman, 1964). These strata coarsen upward, from mostly sandstone at the base to coarse conglomerate at the top (Dibblee, 2005a, 2005b).

Atop the Plush Ranch Formation, in angular unconformity, there is the nonmarine Caliente Formation (named by T.W. Dibblee Jr. inStock, 1947; Schwade, 1954), which is composed of fluvial and lacustrine conglomerate, sandstone, and mudstone, with minor tuffaceous and limestone beds (Fig. 2; Carman, 1964; Ehlert, 2003).

Soledad Region

The oldest nonmarine strata of the Soledad region belong to the Vasquez Formation (Fig. 2; Sharp, 1935; Jahns and Muehlberger, 1954; Muehlberger, 1958), which is dominated by alluvial conglomerate and sandstone (Jahns and Muehlberger, 1954; Hendrix and Ingersoll, 1987). The Vasquez Formation is preserved in three subbasins (Fig. 1; Jahns and Muehlberger, 1954; Muehlberger, 1958; Hendrix and Ingersoll, 1987). The Vasquez Rocks and Texas Canyon subbasins, south of Sierra Pelona (where Pelona Schist is exposed; Fig. 1), are interpreted to have been depositionally distinct but kinematically linked for most of their history (Bohannon, 1976; Hendrix and Ingersoll, 1987; Hendrix, 1993; Hendrix et al., 2010). The Vasquez Rocks subbasin contains interbedded volcanic rocks, primarily basaltic andesite with some rhyodacite and rhyolite (Hendrix and Ingersoll, 1987; Frizzell and Weigand, 1993), dated by whole-rock and plagioclase K-Ar methods as ca. 26–23 Ma (Crowell, 1973; Spittler, 1974; Woodburne, 1975; recalculated after Dalrymple, 1979; Frizzell and Weigand, 1993). Strata of the Charlie Canyon subbasin, north of Sierra Pelona (Fig. 1), coarsen upward, from siltstone and fine sandstone at the base to coarse conglomerate near the top (Sams, 1964; Hendrix and Ingersoll, 1987).

Atop the Vasquez Formation, in angular unconformity, there are strata designated as the Tick Canyon Formation by Jahns (1939, 1940; see also Fig. 2 herein). They consist of alluvial, fluvial, and lacustrine conglomerate, sandstone, siltstone, and claystone (Jahns, 1940; Woodburne, 1975). These strata are overlain by, and were originally considered part of, the Mint Canyon Formation (Kew, 1923, 1924), but they were subsequently distinguished on the basis of an inferred disconformity (Jahns, 1940). Subsequent work discounted the presence of a disconformity (Ehlert, 1982, 2003; Lander, 1985; Bishop, 1990). In this study, we use the term “Tick Canyon strata” to distinguish these deposits from overlying strata of the Mint Canyon Formation and underlying strata of the Vasquez Formation.

Tick Canyon strata contain an unroofing sequence, culminating up section in clasts of Pelona Schist (Ehlert, 1982, 2003; Hendrix, 1993). The Tick Canyon strata also contain abundant volcanic clasts, most of which resemble volcanic rocks of the Vasquez Formation (Hendrix, 1993). The Charlie Canyon subbasin of the Soledad region also contains an unroofing sequence in the form of a Pelona Schist–bearing, poorly sorted breccia, stratigraphically above the Vasquez Formation but below the Mint Canyon Formation (e.g., Sams, 1964; Weber, 1994; Dibblee, 1997; Coffey, 2015). This breccia has previously been considered as the basal Mint Canyon Formation (Dibblee, 1997) and a separate formation (part of the San Francisquito Canyon breccia by Sams, 1964; the Powerhouse breccia-conglomerate by Weber, 1994). In this study, these strata are referred to as Tick Canyon strata.

The Mint Canyon Formation consists primarily of fluvial, alluvial, and lacustrine conglomerate, sandstone, and mudstone (Fig. 2; Kew, 1923, 1924; Ehlert, 1982, 2003). The Mint Canyon Formation is overlain by the dominantly marine Castaic Formation (Crowell, 1954), which consists of shale, sandstone, and minor conglomerate (Crowell, 1954; Ehlert, 1982). The contact between the Mint Canyon and Castaic Formations is an angular unconformity in some places, and it is apparently conformable and gradational in others (Fig. 2; Ehlert, 1982).

Orocopia Region

The only Oligocene–Miocene strata of the Orocopia region belong to the nonmarine Diligencia Formation (Fig. 2; Crowell, 1975b), which is composed of alluvial, fluvial, and lacustrine conglomerate, sandstone, siltstone, and limestone (Spittler and Arthur, 1982; Law et al., 2001; Ingersoll et al., 2014). It contains interbedded basalt and andesite flows, dated by whole-rock and plagioclase K-Ar methods as ca. 24–21 Ma (Crowell, 1973; Spittler, 1974; recalculated after Dalrymple, 1979; Frizzell and Weigand, 1993), and andesitic sills and dikes (Spittler and Arthur, 1982; Terres, 1984).

Potential Source Rocks

The oldest rocks of the San Gabriel and Punchbowl blocks are Paleoproterozoic gneisses, formed 1800–1660 Ma (Silver, 1968; Ehlig, 1981; Barth et al., 2001; Premo et al., 2007; Nourse and Premo, 2016), which were intruded by a complex of anorthosite, gabbro, syenite, and norite (e.g., Crowell, 1975a; Ehlig, 1981) ca. 1200 Ma (Barth et al., 1995, 2001). Perturbation of the Paleoproterozoic gneisses during intrusion produced discordant zircon with 207Pb/206Pb ages of 1760–1300 Ma (Silver et al., 1963; Barth et al., 1995, 2001). “Anorogenic” plutons (1460–1400 Ma) are present throughout southern California and surrounding regions (Anderson and Bender, 1989), including minor exposure in the eastern San Gabriel Mountains (Premo et al., 2007; personal commun. with J. Nourse, 2017 inHoyt et al., 2018).

The Proterozoic basement of southern California is intruded by numerous, overlapping Mesozoic plutons (e.g., Ehlig, 1981). In the San Gabriel Mountains, the oldest of these is the compositionally zoned Mount Lowe intrusion, emplaced 218–207 Ma (Barth and Wooden, 2006); similar-age plutons are present in the southern Mojave region (e.g., Barth et al., 1997). Most plutons in southern California are substantially younger than the Mount Lowe intrusion (e.g., Barth et al., 1997). In the San Gabriel Mountains, distinct magmatic episodes occurred 170–149 Ma and 90–75 Ma (e.g., Silver, 1971; May and Walker, 1989; Barth et al., 2008), producing quartz diorite to quartz monzonite (e.g., Ehlig, 1981).

The Paleogene San Francisquito Formation is exposed north of Blue Ridge in the Punchbowl block (Dibblee, 1967, 1987). It consists of almost entirely marine shale, mudstone, sandstone, conglomerate, and minor carbonate, both in the Punchbowl block and in its type area in the Soledad region (Dibblee, 1967; Kooser, 1982); coeval marine deposits are present in the Tejon (e.g., Kellogg et al., 2008) and Orocopia (Crowell and Susuki, 1959; Advocate et al., 1988) regions.

The gneissic, granitic, and sedimentary rocks of the San Gabriel Mountains lie within the upper plate of the Vincent thrust (e.g., Ehlig, 1981). The lower plate is composed of Pelona Schist, which is predominantly meta-arkose (Haxel and Dillon, 1978; Ehlig, 1981; Jacobson et al., 2011). This relationship is exposed in a structural window in the southern San Gabriel Mountains. Pelona Schist is also exposed in the core of an anticlinorium along Blue Ridge in the Punchbowl block (e.g., Dibblee, 1968), and in the core of anticlinoria in the Tejon and Soledad regions (e.g., Ehlig, 1968; Crowell, 1975a). The correlated Orocopia Schist is exposed in an anticlinorium in the Orocopia region (e.g., Crowell, 1962; Ehlig, 1968; Haxel and Dillon, 1978; Jacobson et al., 2007; Ingersoll et al., 2014).

Tectonic Reconstructions of the Southern San Andreas Fault System

The central Punchbowl block has been correlated with the Soledad region, implying 40–50 km of dextral slip on the Punchbowl fault (Dibblee, 1967, 1968; Ehlig, 1968, 1981; Powell, 1993). Correlation has also been suggested between the central Punchbowl block and either the northwestern Orocopia region (Ehlert and Ehlig, 1977; Ehlig and Joseph, 1977) or the northern Little San Bernardino Mountains (Ehlig and Joseph, 1977; Matti and Morton, 1993).

Correlation of the Tejon, Soledad, and Orocopia regions (Fig. 1) has been proposed and refined based on similarities in both basement rocks, and sedimentary and volcanic strata (e.g., Hill and Dibblee, 1953; Crowell, 1962, 1975a; Carman, 1964; Ehlig and Ehlert, 1972; Bohannon, 1975; Ehlert, 1982, 2003; Weigand, 1982; Frizzell and Weigand, 1993; Ingersoll et al., 2014; Hoyt et al., 2018). Such correlations imply 60–70 km and ∼240 km of dextral slip along the San Gabriel/Canton and San Andreas faults, respectively (e.g., Crowell, 1975a; Ingersoll et al., 2014). These correlations have been widely accepted, but also challenged by alternate reconstructions (e.g., Smith, 1977; Powell, 1981, 1993; Spittler and Arthur, 1982; Frizzell et al., 1986; Matti and Morton, 1993; Weldon et al., 1993), most of which suggest that 80–110 km of dextral slip occurred along the San Francisquito–Fenner–Clemens Well fault, and only 42 km and 160–185 km occurred along the San Gabriel/Canton and southern San Andreas faults, respectively (e.g., Powell, 1993). Some of the differences among these reconstructions may be reconcilable (Darin and Dorsey, 2013).

We mapped the study area in the central Punchbowl block at 1:12,000 scale, documented sediment composition via conglomerate-clast counts, sandstone point counts, and U-Pb dating of detrital zircon, and determined ages of igneous rocks via U-Pb dating of zircon.

Stratigraphy

Vasquez Formation

The stratigraphically lowest nonmarine strata in the study area are compositionally distinct from overlying strata and contain interbedded volcanic rocks. Accordingly, we consider them a separate formation, which we refer to as the Vasquez Formation based on probable correlation with the Vasquez Formation of the Soledad region, as discussed below. The Vasquez Formation of the study area is composed primarily of bright-red conglomerate and sandstone. The conglomerate exhibits low degrees of rounding and sorting, a muddy matrix, and, commonly, reverse grading. South of Blue Ridge, the Vasquez Formation nonconformably overlies granitoid, although the base of the section is excised by the Blue Ridge fault along much of its length (Fig. 3). Interbedded trachyandesite, previously undescribed, is present near the base of the section (Fig. 3), in places capped by thinly bedded tan limestone (too small to map). North of Blue Ridge, the Vasquez Formation also lies depositionally atop granitoid, possibly with some intervening San Francisquito Formation in places (Fig. 3). The Vasquez Formation north of Blue Ridge contains lenses of very poorly sorted, very angular, matrix-poor megabreccias (map unit PεNvg in Figs. 3 and 4) interbedded with Vasquez Formation conglomerate and sandstone. These deposits fit the definitions of “crackle breccia facies” and “jigsaw breccia facies” (i.e., Yarnold and Lombard, 1989, p. 12).

The sorting, rounding, grading, and matrix within Vasquez conglomerate suggest deposition by debris-flow and hyperconcentrated-flow mechanisms (e.g., Blackwelder, 1928; Fisher, 1971). We interpret the Vasquez Formation to represent primarily proximal alluvial-fan deposits. Using the criteria of Yarnold and Lombard (1989) and Yarnold (1993), we interpret lenses of granitoid crackle and jigsaw breccia in the Vasquez Formation north of Blue Ridge as rock-avalanche deposits. We interpret the thin interval of thinly bedded limestone atop the interbedded trachyandesite as lacustrine, likely the result of ponding against the volcanic flows.

The high relative abundance of trachyandesite and rhyolite clasts immediately up section of the trachyandesite flows (see below) suggests that the rhyolite clasts are from deposits closely related to, and thus broadly coeval with, the trachyandesite flows. If so, then the 24.4 ± 0.9 Ma age we determined for the rhyolite clasts (Table S1 in the Supplemental Files1) should approximate the age of the trachyandesite flows, and thus the age of the interbedded strata. Therefore, deposition of the Vasquez Formation in the Punchbowl block probably began ca. 25–24 Ma or earlier.

Paradise Springs Formation

Overlying the Vasquez Formation, there are compositionally distinct strata here informally dubbed the “Paradise Springs formation.” The Paradise Springs formation is texturally similar to the Vasquez Formation: It is composed of bright-red conglomerate and sandstone, with the conglomerate exhibiting low to moderate degrees of rounding and sorting, and a sandy to muddy matrix. The Paradise Springs formation is present both north and south of Blue Ridge; in both cases, basal strata and the contact with the Vasquez Formation are covered by Quaternary deposits (Fig. 3). South of Blue Ridge, along the Blue Ridge fault, there is a spatially restricted breccia (map unit Npss in Fig. 3). This breccia is very angular and very poorly sorted, with a deep-maroon sandy matrix.

We interpret the Paradise Springs formation to represent alluvial-fan deposits. The textural differences compared to the Vasquez Formation suggest less proximal deposition, except for the breccia unit.

Punchbowl Formation

North of Blue Ridge, the Paradise Springs formation is overlain by the Punchbowl Formation (Noble, 1953, 1954), with no discernible angular discordance (Fig. 3). The main member of the Punchbowl Formation is gray to white to tan conglomerate, sandstone, and minor mudrock. The conglomerate is typically well sorted and subrounded to rounded, with sandy matrix. Channels are abundant and exhibit normal grading. We recognized basal strata of the Punchbowl Formation as a distinct member (Fig. 3). This basal member is composed of pink to red to buff conglomerate and sandstone. It is texturally intermediate between the overlying main member of the Punchbowl Formation and the underlying Paradise Springs formation, though more similar to the former. The contact between the basal and main members of the Punchbowl Formation is transitional.

The Punchbowl Formation has been interpreted as dominantly fluvial (e.g., Dibblee, 1987). The basal Punchbowl Formation’s intermediate texture and color compared to the Paradise Springs and Punchbowl formations suggest that it represents the transition from alluvial to fluvial deposition.

Sediment Composition

Conglomerate-Clast Composition

Conglomerate-clast composition was determined at locations across the study area. A flexible grid was affixed to conglomeratic beds, and the lithology of the clast at each crosshair was determined until 100 counts were reached; grid spacing was varied between locations such that it always exceeded the average clast size.

Conglomerate-clast composition data are shown in Figure 5 (raw data are in Table S2 [see footnote 1]). Clasts of Pelona Schist are only present in the upper part of the Miocene strata (the Paradise Springs and Punchbowl formations). Throughout the study area, granitoid makes up the majority of the conglomerate-clast population of the Vasquez Formation. Adjacent to and immediately up section from the interbedded trachyandesite flows, trachyandesite and rhyolite clasts make up 30%–39% of the conglomerate clasts; the more stable rhyolite clasts are found in low abundance throughout most of the section. The Paradise Springs formation conglomerate is dominated by sandstone clasts derived from the San Francisquito Formation; south of Blue Ridge, granitoid clasts are also abundant. Pelona Schist clasts are absent in the lower part of the Paradise Springs formation, but they constitute up to 7% of the conglomerate clasts higher in the section; the spatially restricted breccia unit is composed entirely of Pelona Schist clasts. The Punchbowl Formation, including the basal member, is dominated by granitoid and gneiss clasts; clasts of sandstone from the San Francisquito Formation are present, but they are much less abundant than in the underlying Paradise Springs formation.

Sandstone Composition

Sandstone composition was determined for samples collected throughout the study area. Sandstone samples were mounted as standard 30-µm-thick thin sections by R.A. Petrographic and then etched with concentrated hydrofluoric acid (HF) and stained with a saturated solution of sodium hexanitrocobaltate (III) (Na3Co[NO2]6). Etching and staining distinguish quartz (unetched; unstained), potassium feldspar (stained with yellow dots), and plagioclase feldspar (heavily etched; unstained; Gabriel and Cox, 1929; Reeder and McAllister, 1957; Ingersoll and Cavazza, 1991). For each thin section, 400–500 points were analyzed using the Gazzi-Dickinson method of point counting (Gazzi, 1966; Dickinson, 1970; Ingersoll et al., 1984). For each sample, grid spacing was greater than the average grain size. Categories of framework grains (>0.0625 mm) are defined in Table 1; points interstitial to framework grains were also counted. Where diagenetic alteration had occurred, framework grains were counted as their original grain type. Point counts were also performed on samples collected and prepared by others in the same manner; details are given in Table S3 (see footnote 1).

Most of the Punchbowl block Oligocene–Miocene strata show considerable variation in sandstone composition (Fig. 6). Sandstone from the main member of the Punchbowl Formation, however, clusters very tightly on total quartz–total feldspar–total (nonquartzose) lithics (QFL) and monocrystalline quartz–potassium feldspar–plagioclase feldspar (QmFkFp) ternary diagrams; the large spread on the metamorphic-volcanic-sedimentary lithic (LmLvLs) ternary diagram is presumably caused by the extremely low abundance of lithic grains (grain categories are defined in Table 1; Hoyt et al., 2018). Sandstone from the basal Punchbowl Formation shows more variation than the overlying main Punchbowl Formation but less variation than the underlying Vasquez and Paradise Springs formations (Fig. 6). The Vasquez Formation, which is dominated by granitoid and volcanic conglomerate clasts, is dominated by feldspar, especially plagioclase, in the sandstone fraction. The Paradise Springs formation, which is dominated by sandstone conglomerate clasts from the San Francisquito Formation, contains significantly more quartz and sedimentary lithics in the sandstone fraction.

Detrital-Zircon Ages

Nine >1 kg sandstone samples were collected for detrital-zircon analysis (Fig. 3). Samples were crushed and sieved and then separated by density and magnetism using: (1) a Mineral Technologies MD Gemini shaking table at Pomona College or tetrabromoethane (ρ = 2.97 g/cm3), (2) a neodymium hand magnet and a model L-1 Frantz isodynamic magnetic separator, and (3) methylene iodide (ρ = 3.32 g/cm3). At the Arizona LaserChron Center at the University of Arizona, large splits of these separates were mounted, together with grains of zircon references, on epoxy plugs, which were polished to a depth of ∼20 μm to expose crystal interiors.

Detrital zircon was dated by U-Pb methods using laser-ablation–multicollector–inductively coupled plasma–mass spectrometry (LA-MC-ICP-MS) at the Arizona LaserChron Center, using protocols described by Gehrels et al. (2006, 2008). Material was ablated using a Photon Machines Analyte G2 excimer laser and then carried in helium into the plasma source of a Nu high-resolution ICP-MS. A 204Pb-based common Pb correction was performed, using the common Pb composition from Stacey and Kramers (1975). U/Pb, Th/U, and Pb-isotope relative sensitivities were determined using zircon reference Sri Lanka (563.5 ± 1.6 Ma; ∼518 ppm U and 68 ppm Th; Gehrels et al., 2008). The accuracy of final ages was verified using zircon reference R33 (419.3 ± 0.4 Ma; Black et al., 2004). Analytical data are reported in Table S4 (see footnote 1). Additional details of preparation and analysis were given in Coffey (2015).

Relative-probability distributions of detrital-zircon ages are given in Figure 7 (raw data are in Table S4 [footnote 1]). Distributions for all samples south of Blue Ridge (Fig. 7A) are dominated by peaks at ca. 1200 Ma and 1660–1800 Ma, and they exhibit low, broad peaks at ca. 1400 Ma. Distributions for all samples north of Blue Ridge (Fig. 7B) lack ca. 1200 Ma peaks and exhibit smaller peaks at ca. 1400 Ma and 1660–1800 Ma. Many samples contain peaks at ca. 150 Ma, and some samples contain peaks at ca. 220 Ma, ca. 75 Ma, and/or ca. 25 Ma.

Provenance Interpretations

All Oligocene–Miocene deposits south of Blue Ridge and the Fenner fault contain substantial ca. 1200 Ma zircon (Fig. 7A), whereas zircon of this age is entirely absent from all such deposits north of Blue Ridge and the Fenner fault (Fig. 7B). This is consistent with a drainage divide along Blue Ridge during the Miocene (Sadler, 1993; Hoyt et al., 2018), and it indicates that this divide was established prior to deposition of the lower Paradise Springs formation, and probably prior to deposition of the Vasquez Formation. Strata north and south of Blue Ridge would thus have formed in distinct basins, hereafter referred to as the “northern” and “southern” subbasins, respectively.

The granitoid depositionally beneath the Vasquez Formation (compositionally near the quadruple point of granite, quartz monzonite, granodiorite, and quartz monzodiorite on a QAPF diagram [Streckeisen, 1974] and dated as 146 ± 3 Ma 1σ; Table S1 [see footnote 1]) is presumably the primary proximal source of the Vasquez Formation. This interpretation accounts for the dominance of granitoid clasts in the conglomerate fraction (Fig. 5), the consistency of sandstone composition with that of both the granitoid and a granitoid conglomerate clast (Fig. 6), and the presence of ca. 150 Ma detrital-zircon ages (Fig. 7). South of Blue Ridge, these proximal alluvial-fan deposits likely interfingered with more distal deposits shed approximately northward from western and central parts of the San Gabriel block (presumably adjacent prior to Punchbowl fault slip), where the anorthosite-gabbro-syenite-norite complex and Mount Lowe intrusion are exposed. This would account for the ca. 1200 Ma and ca. 220 Ma detrital-zircon ages, respectively, in the Vasquez Formation samples south of Blue Ridge (Fig. 7A), as well as the variability in sandstone composition (Fig. 6).

The Paradise Springs formation, in contrast to the Vasquez Formation, was derived primarily from the San Francisquito Formation, as documented by the abundance of sandstone clasts within Paradise Springs conglomerate (Fig. 5), and the greater relative abundance of quartz and sedimentary lithic fragments and lower relative abundance of plagioclase feldspar within Paradise Springs sandstone (Fig. 6). Because present exposures of San Francisquito Formation are abundant in the vicinity of Paradise Springs formation exposures, but they are absent in most of the area of Vasquez Formation exposure (Fig. 3), this contrast in source rock may be more spatial than temporal. Prominent peaks in the detrital-zircon age distributions at ca. 75 Ma, ca. 150 Ma, and 1660–1800 Ma imply diverse ultimate source rocks. These are also the dominant zircon ages for the San Francisquito Formation of the Soledad region (Jacobson et al., 2011), and so most of this diversity probably results from recycling of San Francisquito zircon, rather than direct input from multiple sources.

North of Blue Ridge, the presence of Pelona Schist clasts in the upper part of the Paradise Springs formation, but their absence in the lower part and in the underlying Vasquez Formation, represents an unroofing sequence (Fig. 5; Ingersoll and Colasanti, 2004; Colasanti and Ingersoll, 2006). A similar unroofing sequence is present south of Blue Ridge (Fig. 5). The spatially restricted schist breccia within the Paradise Springs formation, interpreted as a proximal alluvial-fan deposit, is located along the Blue Ridge fault, which bounds the anticlinorium of Pelona Schist along Blue Ridge. This suggests that the Pelona Schist of Blue Ridge was the source of this breccia, and likely also the source of the Pelona Schist clasts within the unroofing sequence of the Paradise Springs formation.

Basal Punchbowl Formation sandstone is compositionally much more variable than that of the overlying Punchbowl Formation, but it is less variable than that of the underlying Paradise Springs and Vasquez formations (Fig. 6). The basal Punchbowl Formation contains a smaller proportion of gneissic clasts than overlying Punchbowl conglomerate, but a larger proportion than the Paradise Springs and Vasquez formations (Fig. 5). These observations support our interpretation of the basal Punchbowl Formation as documenting the shift in depositional environment from alluvial fans to an integrated braided-fluvial system. In the northern subbasin, a ca. 235 Ma peak is prominent in the upper Paradise Springs formation (samples KTC-14-dz7 and dz10; Fig. 7B) and in the basal Punchbowl Formation (sample KTC-14-dz8), but rare or absent down section (sample KTC-14-dz6) and in the San Francisquito Formation of the Soledad region and the Pelona-Orocopia schist (Jacobson et al., 2011). Therefore, the presence of this age peak may document the beginning of this change in drainage patterns. The detrital-zircon age distribution for the basal Punchbowl Formation sample is similar to those of the overlying main member of the Punchbowl Formation (Hoyt et al., 2018), but with more ca. 220 Ma zircon and less ca. 245 Ma zircon. This may indicate that the source of the ca. 245 Ma zircon, which Hoyt et al. (2018) suggested was likely Middle Triassic plutons of the southern Mojave region (e.g., Barth et al., 1997), was one of the last source areas to become integrated into the developing Punchbowl Formation drainage system. The provenance of the main member of the Punchbowl Formation was discussed by Hoyt et al. (2018).

Structure

Blue Ridge Fault

South of Blue Ridge, the contact between the Pelona Schist and the Vasquez and Paradise Springs formations, mapped as depositional by Dibblee (2002a, 2002c), is a fault, here termed the “Blue Ridge fault,” along its entire length. The fault is poorly exposed, but in places, slickensides and a narrow zone of hydrothermal alteration are present; slickenlines indicate almost pure dip slip (Fig. 3). The Blue Ridge fault terminates to the northwest against the Fenner fault. Accordingly, it predates the Fenner and Punchbowl faults, and it is likely a normal fault associated with the phase of extension that much of southern California underwent beginning ca. 25 Ma (e.g., Tennyson, 1989), in which case it would be coeval with deposition of the Vasquez Formation. Truncation of Paradise Springs strata indicates that the Blue Ridge fault either remained active until this time or was reactivated.

Fenner Fault

North of Blue Ridge, the contact between the Pelona Schist and the San Francisquito and Vasquez Formations is the Fenner fault. Noble (1954) and this study (Fig. 3) inferred continuation of the Fenner fault westward through the Paradise Springs formation to the Punchbowl fault, whereas other studies (e.g., Liu, 1990; Weldon et al., 1993; Dibblee, 2002a) have interpreted the Fenner fault as predating, and being overlain by, the Paradise Springs formation. Differing interpretations are possible because of poor, discontinuous exposure of the Fenner fault in this region.

Punchbowl Fault

The Punchbowl fault is a regionally significant strand of the San Andreas fault (e.g., Dibblee, 1967), with documented reverse-dextral slip (e.g., Chester and Chester, 1998). In the southeastern part of the study area, a single fault trace is present; to the northwest, the fault splits into two diverging subparallel traces (Fig. 3; Dibblee, 2002a, 2002c). The northeastern branch dips southwestward, and the dip at the surface is shallower farther northwest (Fig. 3; orientations were not measured on the southwestern branch). Uplift of the San Gabriel Mountains has generally been greater in the eastern half, as indicated by higher elevations and exposure of deeper structural levels (e.g., Bull, 1987). Accordingly, the present surface exposure of the Punchbowl fault within the study area is an oblique view, with progressively deeper structural levels exposed progressively southeastward. Splitting of the fault and shallowing of the dip of the northeastern branch to the northwest suggest a positive flower structure, in which a nearly vertical strike-slip fault at depth splits into two branches, which shallow and exhibit more reverse slip upward (i.e., Wilcox et al., 1973; Sylvester, 1988).

Smaller Faults and Folds

Faults and folds subparallel to the Punchbowl fault in the Punchbowl Formation in the northwestern part of the study area (in the region of A-A′ in Figs. 3 and 4) are presumably transpressional features associated with slip and shortening along the Punchbowl fault. The fault southeast of the Punchbowl syncline (near the middle of A-A′ and B-B′ in Figs. 3 and 4) does not terminate in the basal Punchbowl Formation, as indicated by previous studies (e.g., Dibblee, 2002a), but rather it cuts, and thus postdates, the lower part of the main member of the Punchbowl Formation. Additionally, shear-sense indicators along exposure of this fault zone in a road cut immediately west of B-B′ (Figs. 3 and 4) suggest oblique, reverse-dextral slip. Accordingly, this fault is here interpreted as a transpressional feature related to the Punchbowl fault, rather than as an extensional feature from earlier in the Miocene. This fault presumably originally dipped southwest (Fig. 8A) and underwent horizontal-axis rotation to achieve its present, steep northeast dip (Fig. 8B). This rotation was likely also responsible for the steeply southwest-dipping Punchbowl strata in this area (Fig. 3). This fault was likely kinematically linked with the Punchbowl fault and may splay off the Punchbowl fault at depth, as part of the positive flower structure proposed above (Figs. 4 and 8). Reverse motion on the Fenner fault could also be part of this flower structure.

Faults and folds within the San Francisquito Formation and granitoid basement are subparallel to both the Punchbowl fault and the (presumably older) Blue Ridge fault (Fig. 3), and they could be either high-angle reverse faults and associated folds related to transpression, potentially comprising part of the proposed flower structure, or listric normal faults and associated folds related to Oligocene–Miocene extension. It is also possible that these faults began as normal faults during Oligocene–Miocene extension, and they were subsequently reactivated as reverse faults because of their favorable orientation; such reactivation at a larger scale could potentially explain why the trace of the Punchbowl fault so closely follows that of the Blue Ridge fault. These possibilities predict different ages, geometries, and kinematics for these faults; additional detailed work might clarify these relations.

Vincent Thrust

Southwest of the southeastern part of the Punchbowl fault, the Vincent thrust separates Pelona Schist from mylonitic gneiss (Ehlig, 1981; Jacobson, 1983). Previous mapping (e.g., Dibblee, 2002c) inferred a fault subparallel to the Punchbowl fault along the southwestern edge of an intrusive body (map unit Pε?gd in Fig. 3) of probable late Oligocene age (May and Walker, 1989; Nourse, 2002), against which the Vincent thrust terminates. We did not find evidence for this fault; rather, our mapping suggests that the Vincent thrust was intruded by this intrusive body (Fig. 3). The Vincent thrust may have acted as a conduit, along which magma could more easily flow (H-H′ and I-I′ in Fig. 4); this would explain the anomalously large size of the body intruding the Vincent thrust compared to the surrounding, coeval sills and dikes intruding the Pelona Schist and mylonitic basement (Fig. 3; Dibblee, 2002c).

Basin Development

Mid-Cenozoic basin development likely began with initiation of normal faulting in the latest Oligocene. Distinct subbasins formed north and south of Blue Ridge, separated by an ancestral Blue Ridge topographic high, which acted as a drainage divide. Relief and erosion along this topographic high were sufficient to produce the Vasquez Formation alluvial-fan deposits that dominate the margins of the two subbasins. These proximal deposits interfingered with more distal, finer-grained alluvial deposits derived from the opposite margins of these subbasins. The Pelona Schist along Blue Ridge was entirely in the subsurface throughout deposition of the Vasquez Formation, presumably covered by the granitoid presently exposed on either side of Blue Ridge (Fig. 3), clasts of which dominate the conglomerate fraction of the Vasquez Formation in both subbasins. Extension via normal faulting was accompanied by bimodal volcanism, which produced the trachyandesite flows and the rhyolite and trachyandesite conglomerate clasts found in the Vasquez Formation of the southern subbasin (Figs. 3 and 5). The southern subbasin may have been primarily a half graben controlled by a normal fault on its southern margin, with the Blue Ridge fault forming later as an antithetic normal fault. This would explain both the substantial fine-grained sediment input from the south implied by detrital-zircon data and tilting of Vasquez strata away from, rather than toward, the Blue Ridge fault (Fig. 3).

The northern and southern subbasins persisted until (or were regenerated during) deposition of the Paradise Springs formation, separated as before by an ancestral Blue Ridge drainage divide. Alluvial-fan deposits in these subbasins were sourced from the San Francisquito Formation and, in the southern subbasin, granitoid and Vasquez volcanic rocks. During deposition of the Paradise Springs formation, Pelona Schist was first exposed along Blue Ridge and began contributing detritus to both subbasins.

Following deposition of the Paradise Springs formation, drainage patterns north of Blue Ridge gradually changed, resulting in transition from alluvial-fan to braided-fluvial deposition as the basal member of the Punchbowl Formation accumulated. The main member of the Punchbowl Formation accumulated in a well-integrated fluvial system with its headwaters outside the Punchbowl block (Hoyt et al., 2018). A similar transition may have occurred south of Blue Ridge (Hoyt et al., 2018), but no post–Paradise Springs strata are preserved in this part of the Punchbowl block.

The southern San Andreas fault became active ca. 5 Ma (Nicholson et al., 1994; Ingersoll and Rumelhart, 1999; Oskin et al., 2001; Crowell, 2003; Oskin and Stock, 2003), with the Punchbowl fault probably representing the initial main trace (Sharp and Silver, 1971). The San Andreas fault has been transpressional throughout this part of southern California as a result of a restraining double bend (term of McClay and Bonora, 2001) of regional scale, extending from San Gorgonio Pass (Fig. 1) in the southeast to the Tejon region just north of the Garlock fault in the northwest (e.g., Hill and Dibblee, 1953; Ingersoll and Coffey, 2017). This transpression shut down the Punchbowl drainage system and caused uplift, deformation, and erosion of Oligocene–Miocene strata.

Regional Stratigraphic Correlations

Vasquez Formation

The Vasquez Formation of the Punchbowl block generally resembles the Plush Ranch, Vasquez, and Diligencia Formations of the Tejon, Soledad, and Orocopia regions, respectively, in terms of age, lithology, depositional mechanisms, and relationships with surrounding units. The Vasquez Formation of the Punchbowl block partly overlaps in sandstone composition with the Vasquez Formation of the Soledad region (Fig. 9). In all four regions, lacustrine deposits are rare, but where they occur, they are immediately above volcanic strata. The basin geometry we propose for the southern subbasin is analogous to that described for the Plush Ranch basin of the Tejon region and the Texas Canyon subbasin of the Soledad region—a major fault along the southern margin that exposed Proterozoic basement and generated large alluvial-fan systems, the fine-grained, distal parts of which interfingered and mixed with proximal, coarse-grained deposits shed from granitoid along the northern margin of the basin, which was bounded by a smaller, antithetic fault (Hendrix, 1993; Cole and Stanley, 1995; Hendrix et al., 2010). In both the southern subbasin and the Texas Canyon subbasin, Vasquez Formation strata and underlying granitoid along the northern margin are in fault contact with an anticlinorium of Pelona Schist (e.g., Hendrix, 1993). Both the southern subbasin and the Vasquez Rocks subbasin contain interbedded, intermediate volcanic rocks and detritus of the anorthosite-gabbro-syenite-norite complex derived from the south (Fig. 10A; Hendrix and Ingersoll, 1987). In light of these similarities, the southern subbasin of the central Punchbowl block is likely a close equivalent of the Plush Ranch basin of the Tejon region and the Texas Canyon and Vasquez Rocks subbasins of the Soledad region.

The northern subbasin of the Punchbowl block lies north of the Fenner fault and Blue Ridge anticlinorium, a position analogous to that of the Charlie Canyon subbasin of the Soledad region (e.g., Dibblee, 1997), with which it is likely equivalent. Similar rockslide megabreccias near the top of the Vasquez Formation in each subbasin (Fig. 10B; Sams, 1964; Weber, 1994; Dibblee, 1997) support this correlation. These subbasins may also correlate with the Simmler Formation north of Pelona Schist exposures in the Tejon region. Consistent with this correlation, there is a general increase in clast size up section in both the Charlie Canyon subbasin (Sams, 1964; Hendrix and Ingersoll, 1987) and the Simmler Formation (Fig. 10B; Dibblee, 2005a, 2005b). The Diligencia Formation occupies an analogous position north of the Orocopia Mountains anticlinorium in the Orocopia region, and thus may correlate with these basins as well (Fig. 10B; Ingersoll et al., 2014).

Paradise Springs Formation

The Paradise Springs formation resembles the Tick Canyon strata of the Soledad region: Both represent primarily alluvial-fan deposits with highly variable and partly overlapping sandstone composition (Fig. 11), and both contain unroofing sequences documenting exhumation of Pelona Schist (Sams, 1964; Ehlert, 1982, 2003; Hendrix, 1993; Weber, 1994; Dibblee, 1997; Coffey, 2015). Additionally, both units overlie sandstone, conglomerate, and coeval volcanic strata of the Vasquez Formation, and both are overlain by Middle–Upper Miocene sandstone and conglomerate with no significant angular discordance (Fig. 10). The Paradise Springs formation of the southern subbasin presumably correlates with the type Tick Canyon strata, south of Sierra Pelona, whereas that of the northern subbasin presumably correlates with the Tick Canyon strata of Charlie Canyon subbasin, north of Sierra Pelona.

Punchbowl Formation

The Punchbowl Formation is largely coeval with the Caliente and Mint Canyon Formations of the Tejon and Soledad regions, respectively, but it represents a distinct drainage system (Hoyt et al., 2018). Strata mapped as Punchbowl Formation are present southeast of Sierra Pelona, at the eastern edge of the Soledad region (e.g., Dibblee, 2001). A sample of these strata analyzed by Coffey (2015) closely matched those of the Punchbowl Formation of the Punchbowl block (Hoyt et al., 2018) in both sandstone composition and detrital-zircon age distributions, suggesting that these strata are indeed part of the Punchbowl Formation. Additional sampling and study of these strata would be required to determine their paleogeographic significance.

Regional Tectonic and Paleogeographic Reconstructions

Punchbowl Fault

Previous correlation of the central Punchbowl block and the Soledad region has been based on correlation of (1) the anticlinoria of Pelona Schist along Blue Ridge and Sierra Pelona, (2) the Fenner and San Francisquito faults along the northern margins of these anticlinoria, and (3) the presence of San Francisquito Formation north of these anticlinoria (Dibblee, 1967, 1968; Ehlig, 1968, 1981; Powell, 1993). Our correlation of Oligocene–Miocene strata on either side of these anticlinoria confirms these correlations, and the 40–50 km of dextral slip on the Punchbowl fault that they imply.

San Gabriel/Canton and San Andreas Faults

Our findings support reconstructions of the San Gabriel/Canton and San Andreas faults that closely align crustal blocks of the Tejon, Soledad, and Orocopia regions (e.g., Hill and Dibblee, 1953; Crowell, 1962, 1975a; Carman, 1964; Ehlig and Ehlert, 1972; Bohannon, 1975; Ehlert, 1982, 2003; Weigand, 1982; Frizzell and Weigand, 1993; Ingersoll et al., 2014; Hoyt et al., 2018). In particular, the presence of Oligocene–Miocene strata on both flanks of the Blue Ridge anticlinorium helps link the Soledad region, where strata south of the Sierra Pelona anticlinorium have been given greater emphasis (e.g., Crowell, 1975a; Hendrix and Ingersoll, 1987), with the Orocopia region, where Oligocene–Miocene strata are presently preserved only on the north side of its anticlinorium (e.g., Crowell, 1975b). Correlating the Diligencia basin of the Orocopia region with the Charlie Canyon subbasin of the Soledad region (e.g., Bohannon, 1975; Ingersoll et al., 2014), using the northern subbasin of the central Punchbowl block as the link, eliminates problems with correlation of the Soledad and Orocopia regions (e.g., those discussed by Law et al., 2001).

San Francisquito–Fenner–Clemens Well Fault

The similarity of the stratigraphic sequences in the northern and southern subbasins of the central Punchbowl block argues against the proposed 80–110 km of dextral slip along the Fenner fault (and the correlated San Francisquito and Clemens Well faults; Powell, 1981, 1993), which separates the two subbasins. In particular, the presence of clasts of San Francisquito Formation sandstone in Paradise Springs formation conglomerate of the southern subbasin (Fig. 5) seems incompatible with this magnitude of slip. These clasts link the southern subbasin, which is south of the Fenner fault, with the San Francisquito Formation, which is north of the Fenner fault (Fig. 3). Even if deposition of the Paradise Springs formation occurred after this proposed Fenner fault slip, the presence of a drainage divide along Blue Ridge throughout deposition of the Paradise Springs formation, discussed above, would have prevented transport of these clasts across the Fenner fault into the southern subbasin. Rather, these clasts are presumably derived from San Francisquito Formation outcrops originally present south of the Fenner fault. Furthermore, whereas restoration of 60–70 km of dextral slip on the San Gabriel/Canton fault aligns the Simmler Formation of the Tejon region with its likely correlative, the Vasquez Formation of the Charlie Canyon subbasin of the Soledad region (e.g., Ingersoll et al., 2014), restoration of 80–110 km of dextral slip along the San Francisquito–Fenner–Clemens Well fault, as proposed by Powell (1981, 1993), would eliminate this cross-fault match.

If dextral slip of 80–110 km accumulated along the San Francisquito–Fenner–Clemens Well fault, then most or all of it would have done so in the few million years after deposition of the Vasquez and Diligencia Formations, and before deposition of the Paradise Springs formation and Tick Canyon strata (Fig. 2). Unroofing sequences document exhumation of the Pelona Schist in the northern subbasin of the Punchbowl block (Paradise Springs formation; Ingersoll and Colasanti, 2004; Colasanti and Ingersoll, 2006; this study) and in the Charlie Canyon subbasin of the Soledad region (Tick Canyon strata; Sams, 1964; Weber, 1994; Dibblee, 1997; Coffey, 2015). These unroofing sequences link these strata, which are north of the San Francisquito–Fenner–Clemens Well fault, with the Pelona Schist anticlinoria south of this fault. Any significant dextral slip must have postdated deposition of the Vasquez and Diligencia Formations, which are truncated by the San Francisquito–Fenner–Clemens Well fault (e.g., Jahns and Muehlberger, 1954; Muehlberger, 1958; Crowell, 1975b).

Whereas the San Francisquito–Fenner–Clemens Well fault is largely high angle and bears evidence of some dextral slip (e.g., Crowell, 1962; Stanley, 1966; Konigsberg, 1967; Spittler and Arthur, 1982; Terres, 1984; Ebert, 2004; Yan et al., 2005), it has been suggested that this represents minor (i.e., ≤10 km; Crowell, 1962; Spittler and Arthur, 1982; Terres, 1984; Ebert, 2004) reactivation of what was originally an Oligocene–Miocene normal fault (e.g., Spittler and Arthur, 1982; Hendrix and Ingersoll, 1987; Goodmacher et al., 1989; Robinson and Frost, 1996; Bunker and Bishop, 2001; Yan et al., 2005; Jacobson et al., 2007; Ingersoll et al., 2014). We suggest that down-to-north normal faulting along the Fenner fault likely contributed to subsidence of the northern subbasin of the central Punchbowl block.

Synthesis

Correlation of Oligocene–Miocene strata of the central Punchbowl block with those of the Tejon, Soledad, and Orocopia regions supports the hypothesis that all four crustal blocks were aligned prior to slip along the San Gabriel/Canton and San Andreas/Punchbowl faults (e.g., Hill and Dibblee, 1953; Crowell, 1962, 1975a; Carman, 1964; Dibblee, 1967, 1968; Ehlig, 1968, 1981; Ehlig and Ehlert, 1972; Bohannon, 1975; Ehlert, 1982, 2003; Weigand, 1982; Frizzell and Weigand, 1993; Ingersoll et al., 2014; Hoyt et al., 2018). Combining these correlations with paleomagnetic data (Terres, 1984; Terres and Luyendyk, 1985; Hornafius et al., 1986; Carter et al., 1987; Ellis et al., 1993), and stratigraphic and provenance data (see references above and in Fig. 12 caption), we propose the following sequence of paleogeographic reconstructions: (1) Normal faulting initiated ca. 25 Ma, forming two parallel belts of basins separated by a topographic high along Sierra Pelona (Hendrix and Ingersoll, 1987) and Blue Ridge. Alluvial-fan deposits, partly sourced from this topographic high, and bimodal volcanic flows accumulated in these basins (Fig. 12A). (2) The Canton fault (precursor to the San Gabriel fault; e.g., Crowell, 2003) initiated ca. 18 Ma. East of the Canton fault, Pelona-Orocopia schist was exhumed along the ancestral Sierra Pelona/Blue Ridge/Orocopia Mountains topographic high (e.g., Sams, 1964; Konigsberg, 1967) as these crustal blocks underwent clockwise vertical-axis rotation (Fig. 12B). Exposed Pelona Schist contributed detritus to alluvial-fan deposits on either side of this topographic high. (3) Beginning ca. 15 Ma, braided-fluvial systems began to develop on either side of the ancestral Sierra Pelona/Blue Ridge/Orocopia Mountains topographic high. By ca. 13 Ma, these fluvial systems were established, with the intervening topographic high continuing to act as a drainage divide (Fig. 12C). (4) Continued dextral slip along the Canton fault and its successor, the San Gabriel fault (e.g., Crowell, 2003), displaced the Tejon block from its eastern equivalents. This disrupted the Caliente/Mint Canyon drainage system and allowed a marine incursion into the western Soledad region, where the Castaic Formation began to accumulate (Fig. 12D; e.g., Crowell, 1954; Ehlert, 1982). The Punchbowl drainage system may have supplied sediment to the Castaic Formation and/or breached the ancestral Sierra Pelona/Blue Ridge drainage divide to deliver sediment to the eastern Soledad region. The northern Caliente Formation, which is compositionally distinct from the rest of the Caliente Formation (Hoyt et al., 2018), may have formed during this time as well. (5) Slip began along the southern San Andreas fault ca. 5 Ma, inducing transpression, which shut down the Punchbowl drainage system and deformed and uplifted Oligocene–Miocene strata. Slip along the Punchbowl fault, probably the initial trace of the southern San Andreas fault, separated the Soledad block from the Punchbowl block. Following abandonment of the Punchbowl fault in favor of the present trace of the southern San Andreas fault, the Punchbowl block was separated from the Orocopia block. Shortening between the southern Sierra Nevada and northern Peninsular Range batholiths likely induced additional vertical-axis rotations along the San Andreas fault system, bringing the crustal blocks in Figure 12D into their present orientations (e.g., Ingersoll and Coffey, 2017).

Oligocene–Miocene strata of the central Punchbowl block are composed of three distinct units: the Vasquez, Paradise Springs, and Punchbowl formations. These strata were deposited in distinct subbasins north and south of a drainage divide along what is now Blue Ridge. The alluvial-fan deposits and interbedded volcanic rocks of the Vasquez Formation document extension via normal faulting initiating ca. 25–24 Ma. The alluvial-fan deposits of the Paradise Springs formation contain an unroofing sequence that documents final exhumation of the Pelona Schist during the middle Miocene. A distinct basal member of the Punchbowl Formation records transition to the braided-fluvial deposits that comprise the main member of the Punchbowl Formation.

Correlation of the Vasquez and Paradise Springs formations with equivalent strata of the Soledad region confirms previous estimates of 40–50 km of dextral slip along the Punchbowl fault. Probable correlation with strata of the Tejon and Orocopia regions supports estimates of 60–70 km of dextral slip along the San Gabriel/Canton fault and ∼240 km of dextral slip along the southern San Andreas fault. Conversely, similarities between the northern and southern subbasins of the central Punchbowl block and probable correlations between strata of the Tejon and Soledad regions argue against dextral slip of 80–110 km for the San Francisquito–Fenner–Clemens Well fault.

We thank Clinton Colasanti for conducting preliminary research and collecting sandstone samples used in this study, Johanna F. Hoyt for help with collecting, and Dallon Stang for assistance during a reconnaissance field session. We thank Mark Pecha, Nicky Giesler, and Chelsi White of the University of Arizona LaserChron Center for assistance in analyzing detrital zircon and processing the resulting data, Juliet Ryan-Davis (now at U.S. Geological Survey), Jonathan Harris, and Jade Star Lackey of Pomona College for assistance with mineral separation and for the use of their water shaking table, and Winnie Wu (now at Long Beach City College) and Matthew Wielicki (now at University of Alabama) of the University of California, Los Angeles (UCLA) for help with sample preparation. We thank An Yin for helpful comments on the M.S. thesis upon which this paper is largely based, and Carl Jacobson for useful discussion. We thank Peter Haproff, Randon Flores, and Drew Gomberg for discussion of map relationships, and the camp at Paradise Springs for access to its property. We also thank Bryan Murray and Kim Bishop for helpful review of a previous version of this manuscript.

Fieldwork and microscope thin sections were paid by a Graduate Student Research Grant from the Geological Society of America awarded to Kevin T. Coffey. Detrital-zircon analyses were paid by a UCLA Academic Senate research grant awarded to Raymond V. Ingersoll. The ion-microprobe facility at UCLA is partly supported by a grant from the Instrumentation and Facilities Program, Division of Earth Sciences, National Science Foundation. The Arizona LaserChron Center is partly supported by National Science Foundation grant EAR-1032156.

1Supplemental Files. Table S1: Igneous-zircon data and methods. Table S2: Conglomerate-clast-count data and lithologic descriptions. Table S3: Raw and recalculated sandstone point-count data. Table S4: Detrital-zircon data. Please visit https://doi.org/10.1130/GES02025.S1 or access the full-text article on www.gsapubs.org to view the Supplemental Files.
Science Editor: Raymond M. Russo
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.