Geological Society of America Bulletin

Home , Alarcon Basin, Cerro Prieto, Salton Buttes, Salton Trough

Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Geological Society of America Bulletin

Oceanic magmatism in sedimentary basins of the northern Gulf of California rift

Axel K. Schmitt, Arturo Martín, Bodo Weber, Daniel F. Stockli, Haibo Zou and Chuan-Chou Shen

Geological Society of America Bulletin 2013;125, no. 11-12;1833-1850 doi: 10.1130/B30787.1

Email alerting services click www.gsapubs.org/cgi/alerts to receive free e-mail alerts when new articles cite this article Subscribe click www.gsapubs.org/subscriptions/ to subscribe to Geological Society of America Bulletin Permission request click http://www.geosociety.org/pubs/copyrt.htm#gsa to contact GSA Copyright not claimed on content prepared wholly by U.S. government employees within scope of their employment. Individual scientists are hereby granted permission, without fees or further requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in subsequent works and to make unlimited copies of items in GSA's journals for noncommercial use in classrooms to further education and science. This file may not be posted to any Web site, but authors may post the abstracts only of their articles on their own or their organization's Web site providing the posting includes a reference to the article's full citation. GSA provides this and other forums for the presentation of diverse opinions and positions by scientists worldwide, regardless of their race, citizenship, gender, religion, or political viewpoint. Opinions presented in this publication do not reflect official positions of the Society.

Notes

Oceanic magmatism in sedimentary basins of the northern Gulf of California rift

Axel K. Schmitt1,†, Arturo Martín2, Bodo Weber2, Daniel F. Stockli3, Haibo Zou4, and Chuan-Chou Shen5 1Department of Earth and Space Sciences, University of California–Los Angeles, Los Angeles, California 90095-1567, USA 2Departamento de Geología, Centro de Investigación Cientíﬁ ca y de Educación Superior de Ensenada (CICESE), Carretera Ensenada–Tijuana No. 3918, Zona Playitas, Ensenada, B.C., C.P. 22800, México 3Department of Geological Sciences, University of Texas at Austin, EPS RM 1.130, 1 University Station C9000, Austin, Texas 78712-0254, USA 4Department of Geology and Geography, Auburn University, 210 Petrie Hall, Auburn, Alabama 36849-5305, USA 5High-Precision Mass Spectrometry and Environment Change Laboratory (HISPEC), Department of Geosciences, National Taiwan University, Taipei 10617, Taiwan, R.O.C.

ABSTRACT value implies minor (<20%) assimilation of margins is due to decompression melting in the continental crustal rocks, which, however, is mantle, which ultimately produces composition- Rift-related magmatism in the northern- an upper limit because of crystal-scale evi- ally uniform, nearly anhydrous mid-ocean-ridge most Gulf of California and the adjacent sub- dence for magma contamination by uncon- basalts (MORB). First-order compositional dif- aerial Salton Trough and Cerro Prieto basins solidated sediment at the time of eruption. ferences often result from the transition between comprises intermediate to rhyolitic surfi cial Zircon crystals in felsic rocks (rhyolite lavas, these two magmatic regimes, where subduction and buried lava fl ows and domes, including intrusive microgranites, and granophyre xe- is associated with predominantly intermedi- their xenolith cargo. In addition, geothermal noliths) have trace-element and submantle ate compositions, whereas bimodal suites are a drill wells frequently penetrate subsurface δ18O compositions that are robust indicators hallmark of continental rifting (e.g., Bryan and gabbroic to granitic sills and dikes, which for a mafi c source that has exchanged oxygen Ernst, 2008). During incipient rifting, the prox- intruded into Colorado River delta fl uviatile by interacting with meteoric hydrothermal imity to sources of continent-derived detritus and lacustrine sediments. Combined single- fl uids. Collectively, these data imply that oce- shed into rift basins can also exert strong con- crystal U-Th-Pb and (U-Th)/He zircon ages anic rifting has initiated in the Salton Trough trol on the compositional diversity of mag- reveal late Pleistocene to Holocene eruption and Cerro Prieto basins. There, MORB-type mas and their representation in the geological ages for three volcanic centers in adjacent magmas formed mafi c intrusions within thick record. These controls include (1) density fi lter- rift basins (from N to S): Salton Buttes (erup- sedimentary basin fi ll, where they became ex- ing of magmas where negatively buoyant mafi c tion age: 2.48 ± 0.47 ka; 95% confi dence), posed to deep-reaching hydrothermal fl uids. magmas stall within sequences of low-density Cerro Prieto (maximum eruption age: 73 ± Diverse intermediate- to high-silica rhyolitic sediments so that the record of surfi cial vol- 7 ka), and Roca Consag (eruption age: 43 ± magmas that are prevalent at the surface canic rocks is not representative of magma fl ux 6 ka). U-Th zircon and allanite crystalliza- are produced by fractional crystallization of at depth (e.g., Fuis et al., 1984); (2) rapid sub- tion ages are close to the eruption ages, with mafi c parental magmas with minor assimila- sidence and burial, which conceal earlier phases the exception of Roca Consag lava, the zir- tion of sediments or pre-rift basement rocks, of magmatism (e.g., Herzig, 1990; Hurtado- con population of which is dominated by and by partial melting of hydrothermally al- Brito, 2012); (3) contamination of magmas dur- zircon with ca. 1 Ma crystallization ages, a tered mafi c intrusions. ing ascent through thick sequences of sedimen- population interpreted to be recycled from tary basin fi ll (through melting and assimilation, an unknown crustal source underlying the INTRODUCTION or mingling between magma and sediment; e.g., Wagner basin. Nd isotopic ratios for sub- Gibson et al., 1997); and (4) metamorphism and surface micro gabbros from Cerro Prieto The southwestern North American conti- hydrothermal alteration caused by fl uid circu-

(εNd = +8.9) overlap with values for mid- nental margin evolved from a convergent plate lation through porous sediments in magmati- oceanic-ridge basalts (MORB) from the East boundary into a series of rift basins and embry- cally active rift zones (e.g., Einsele et al., 1980; Paciﬁ c Rise, adjacent to the southern Gulf onic oceanic spreading centers interconnected McKibben et al., 1988). of California. Cerro Prieto microgranites by NW-SE–oriented transform fault systems In order to better constrain the origins of and Salton Sea basaltic xenoliths have simi- during the late Cenozoic (Lonsdale, 1989; Oskin rift-related magmatism and the interaction larly elevated εNd values. The lowest εNd value and Stock, 2003). This plate boundary reconﬁ g- between magma and sediment during conti- for late Pleistocene–Holo cene igneous rocks uration encompasses a transition between fun- nental breakup, we studied a comprehensive from the northern Gulf of California is for damentally different magma production mecha- suite of surface and subsurface magmatic rocks

Cerro Prieto dacitic lava (εNd = +0.6). This nisms: during subduction, melting is triggered from the northern Gulf of California, consist- by hydration of the mantle wedge, whereas ing of samples from two subaerial rift basins †E-mail: [email protected] magmatism in continental rifts and divergent (Salton Trough, Cerro Prieto) and the submarine

GSA Bulletin; November/December 2013; v. 125; no. 11/12; p. 1833–1850; doi: 10.1130/B30787.1; 14 ﬁ gures; 3 tables; Data Repository item 2013358.

Wagner basin (Fig. 1). In addition to whole- rock geochemical analysis, we also focused on single-crystal geochemical and geochrono- SAF logical analysis of zircon as a robust indicator 1 mineral. Single-crystal zircon geochronology IF permits us to constrain magmatic crystallization U.S.A. Mexico 2 and eruption ages at high temporal resolution, ...... CPF...... even for subsurface rocks highly altered by geo- ...... thermal activity. Alteration-resistant geochemi- ...... W...... cal indicators such as whole-rock Nd isotopes ...... 3...... P ...... agner...... a ...... and zircon oxygen isotopes indicate dominantly ...... c ...... MORB-type magma sources for parental mag- i ...... f ...... Delfín...... i ...... 4...... mas, with remelting of hydrothermally altered ...... N...... 30°...... c ...... O ...... juvenile maﬁ c crust as an important mechanism ...... VFZ...... c ...... to produce high-silica magma compositions. e ...... BF...... 6...... a ...... Thick sedimentary rift basin inﬁ ll controls n ...... 7...... 5...... Tiburo...... transform fault ...... magma ascent and biases extrusive magma- ...... n ...... (inferred) ...... tism toward silicic compositions, but melting of ...... sediment or prerift basement overall contributes ...... spreading center ...... 8...... negligibly to even the most silicic magma com- ...... 9...... Guay...... positions. Pleistocene-Holocene ...... m...... a...... s volcanic centers ...... W 115° ...... EPR MORB-affinity ...... GEOLOGICAL BACKGROUND ...... 1. Salton Buttes ...... N 25° ...... Mid-Miocene to Holocene 2. Cerro Prieto ...... Magmatic Evolution 3. Roca Consag ...... 4. Isla San Luis ...... Baja California and western Sonora are well- 5. Isla San Esteban ...... known examples of diverse, and in part unusual 6. Coronado postconvergent magma compositions (e.g., Till 7. San Borja Alarco et al., 2009; Calmus et al., 2011). These are 8. Tres Vírgenes n predated by an earlier phase of arc vol canism 9. Isla La Tortuga EPR related to the terminal phase of subduction of N Farallon plate fragments (20–12 Ma). Subduc- Oceanic crust (<3.5 Ma) 100 km tion-related rocks are widespread along the Other post-convergence volcanics (<12 Ma) entire eastern margin of the Baja California peninsula (Fig. 1). They are termed Comondú Puertecitos volcanics (6 - 3 Ma) ...... Comondu Group and

...... W 110° ...... volcanic group in the southern half of the penin- ...... northern correlatives (25 - 12 Ma) sula (e.g., Hausback, 1984; Sawlan, 1991; Umhoefer et al., 2001), with equivalent rocks Figure 1. Distribution of subduction and postsubduction volcanism in Baja California and also preserved in the northern half (Martín the Gulf of California (after Schmitt et al., 2006). Solid lines show major faults related to et al., 2000). After cessation of subduction, NW-SE translation between the Pacifi c and North American plates (after Lonsdale, 1989): postsubduction basaltic fi ssure eruptions (e.g., SAF—San Andreas fault, IF—Imperial fault; CPF—Cerro Prieto fault; VFZ—Volcanes Bellon et al., 2006; Benoit et al., 2002) and fault zone; BF—Ballenas fault. Locations of major rift basins are indicated. EPR—East peralkaline silicic ignimbrite volcanism initi- Pacifi c Rise; MORB—mid-ocean-ridge basalt. ated after ca. 12 Ma and lasted until ca. 8 Ma (Mora-Klepeis and McDowell, 2004; Oskin and Stock, 2003; Vidal-Solano et al., 2008). These associated vents are confi ned to the axes of en- Wagner basins. Numerous magmatic intrusives, silicic rocks are mostly exposed in northwestern echelon basins separated by NW-SE–trending volcanic edifi ces, and their pyroclastic deposits mainland Mexico, but remnants of pyroclastic right-lateral transform faults (e.g., Batiza, 1978; have been seismically imaged within the Upper fl ow deposits are locally preserved along the González-Escobar et al., 2009, 2010). Toward and Lower Delfi n basins, and several volcanic eastern margin of Baja California (Ferrari et al., the mouth of the Gulf of California, there is edifi ces exist along the Ballenas transform fault 1999; Oskin and Stock, 2003). Contemporane- clear evidence from magnetic striping for sea- and the Volcanes fault zone along the edge of ous with an initial phase of marine incursion fl oor spreading (Lonsdale, 1989), whereas the Baja California continental crust (Fig. 1; into the proto–Gulf of California, dominantly to the north the basins are blanketed by thick Persaud et al., 2003; Hurtado-Brito, 2012). Iso- bimodal volcanism persisted between ca. 8 and deposits of continental detritus, which prohibit lated late Pleistocene to recent volcanic centers 3 Ma, for example, in the Puertecitos volcanic geomagnetic verifi cation of oceanic spreading. are also scattered outside the basins along the fi eld (Fig. 1; Martín-Barajas et al., 1995). Seismic-refl ection data (Martín et al., 2013), eastern coast of Baja California: Tres Vírgenes Post–3 Ma magmas in the southern Gulf of however, indicate ~40–60-km-wide, newly (Capra et al., 1998; Schmitt et al., 2006, 2010); California are largely tholeiitic basalts, and their formed oceanic crust in the Tiburón, Delfín, and Isla San Luis (Paz-Moreno and Demant, 1999);

1834 Geological Society of America Bulletin, November/December 2013 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Northern Gulf of California oceanic magmatism and within the Ballenas Channel (Martín-Barajas A Mullet 117° 116° 115° et al., 2008). ea Island Of the three northernmost basins in the Gulf SAF Salton T nS of California, the Salton Trough and Cerro Colorado Red Prieto basins are subaerial due to their proximity A Island to the Colorado River delta. The origin of these rough Salton Salto basins has long been interpreted to be analo- Buttes Rock Hill gous to the fully oceanic basins farther south Obsidian in the Gulf of California (Elders et al., 1972). Peninsular Butte The Wagner and Consag basins adjacent to the IF 0 1 2 km south are submarine, albeit with average water Lake Cahuilla depths that are much shallower than in the cen- Ranges Gila River tral and southern Gulf of California. Although shoreline surfi cially emplaced basalt is absent in the sub- Cerro Prieto B aerial basins, mafi c intrusions are demonstrably Desert abundant at depth based on drill well penetra- B Evaporation tion, and their existence is also evident from CP pond Colorado F 32° xenolith populations, and from the geochemical M-205 characteristics of consanguineous felsic lavas M-194 River Cerro M-201 (see later herein). E-30 M-203 A particular puzzling phenomenon is the Prieto NL-1 persistence, and in some locations re-initiation, volcano of magmatism with subduction-type chemical Roca affi nities along strike of the Gulf of California GV-2 Consag WB rift (Bigioggero et al., 1995; Martín-Barajas 5 0 2.5 5 km 115° 114° et al., 1995; Capra et al., 1998; Negrete-Aranda and Canon-Tapia, 2008). Within the northern- Figure 2. Geological overview map of the northern Gulf of California with locations for most Gulf of California rift system (Fig. 2), such surface volcanoes. On-land depressions below sea level are indicated by gray shading; solid diverse compositional types of volcanism exist lines indicate major faults (see Fig. 1 for abbreviations). Insets show locations of lava domes side-by-side to the present day: In the Salton and sampled geothermal wells for the (A) Salton Sea and (B) Cerro Prieto geothermal fi elds. Trough, a prominent silicic gap exists between WB—Wagner basin. basaltic and rhyolitic end members, whereas silicic to intermediate magma compositions dominate the adjacent Cerro Prieto basin (Figs. 3 and 4). Multiple models for this wide spec- (Fig. 2). Early pyroclastic deposits overlying mafi c crust (Schmitt and Vazquez, 2006). This trum of postconvergent magmatism in the Gulf lacustrine sediments of Pleistocene–Holocene is supported by the presence of rhyolitic glass, of California have been proposed: subduction Lake Cahuilla are covered by lava fl ows and representing partial melt in the basaltic xeno- metasomatism enhancing mantle fertility com- domes up to 40 m thick, but their elevations are liths (Robinson et al., 1976). bined with thermal insulation of the mantle by entirely below sea level. Domes show wave-cut Cerro Prieto basin contains only one volcano thick sediments (Lizarralde et al., 2007), melt- benches, and rounded pumice rafts are present (Fig. 2), a composite lava dome that rises to 223 m ing of slab remnants (Aguillon-Robles et al., at paleoshorelines, suggesting eruption prior to above sea level and is surrounded by periph- 2001) or lower-crustal metabasites (Castillo, or during a highstand of Lake Cahuilla, the natu- eral autobrecciated lava injections into uncon- 2008), variable degrees of peridotite melting ral precursor of the Salton Sea. Because all fi ve solidated Colorado River deltaic sediments. The (Robinson et al., 1976), fractional crystalliza- domes are petrologically similar and occur on a dacitic lava is fi ne grained and crystal poor, with tion of basaltic magma (Herzig, 1990; Herzig single N-S–trending lineament, they are likely rare plagioclase phenocrysts present. Rounded and Jacobs, 1994), partial melting of granitic connected by a common feeder dike (Kelley xenoliths of indurated and baked sediments are basement (Reed et al., 1984), assimilation of and Soske, 1936; Robinson et al., 1976). Salton locally enclosed in the autobrecciated lava. continental crust concomitant with crystal frac- Buttes rhyolite lavas are crystal poor, with only Roca Consag (Fig. 2) is an isolated volcanic tionation (Martín-Barajas and Weber, 2003; minor plagioclase phenocrysts present. They plug located near the SW terminus of the sub- Vidal-Solano et al., 2008), and partial melting of contain, however, a diverse assemblage of marine Wagner basin (González-Escobar et al., hydrothermally altered basaltic crust (Schmitt xenoliths of basalt, granite, and metasediment 2010). It consists of microphyric low-K dacite, and Vazquez, 2006). (Kelley and Soske, 1936; Robinson et al., 1976; and likely subvolcanic intrusive rock. Its spatial Schmitt and Vazquez, 2006). Basaltic xenoliths extent is ~0.002 km2, and its elevation is ~40 m Recent Volcanism in the Salton Trough, have depleted trace-element signatures and above sea level. ε 143 144 Cerro Prieto, and Wagner Basins elevated Nd ( Nd/ Nd) coupled with low Previous radiometric dating for the Salton 87Sr/86Sr, characteristic of an East Pacifi c Rise Buttes and Cerro Prieto resulted in mostly K-Ar In the Salton Trough, the series of fi ve rhyo- mantle source (Herzig and Jacobs, 1994; Rob- ages. For Salton Buttes obsidian, ages range lite domes (from NE to SW: Mullet Island, two inson et al., 1976). The granitic xenoliths are from 33 ± 36 ka to <10 ka. Late Pleistocene domes of Red Island, Rock Hill, and Obsidian derived from juvenile crust rather than base- to Holocene U-Th zircon crystallization ages Butte) is collectively termed the Salton Buttes ment and originated from remelting of hydrated exist for Salton Buttes obsidian and xenoliths

Geological Society of America Bulletin, November/December 2013 1835 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Schmitt et al.

Salton Sea lava basaltic xenoliths granophyre xenoliths Cerro Prieto lava subsurface Roca Consag lava

ABhigh P(H2 O) intermediate

P(H2 O)

Figure 3. Major- and trace- low P(H2 O) element variation diagrams for northern Gulf of California igneous rocks. Panel B shows CD compositional ﬁ elds for melting of hydrous basaltic rocks under different water pressures (Thy et al., 1990). Thick dashed line in panel E is the limit of

K2O enrichment from closed- system mid-ocean-ridge basalt (MORB) fractionation (France et al., 2010). Solid and dashed EF arrows in panel H schematically indicate compositional trends approx. limit for closed-system remelting, and of MORB fractional crystallization cou- fractionation high pled with assimilation of continental crustal material (AFC), medium respectively. Data sources: this study, and compiled from low Robinson et al. (1976), Reed et al. (1984), Herzig and Elders GHmelting/fractionation (1988), and Herzig (1990).

AFC

(Schmitt and Vazquez, 2006). Obsidian hydra- Subsurface Magmatism For the Salton Trough, U-Pb zircon ages for tion methods yielded surface exposure age lava ﬂ ows and pyroclastic deposits (including estimates between ca. 8.4 ka and ca. 2.5 ka Earlier volcanic episodes in the northern Gulf distal Bishop Ash) that are present at depths (Friedman and Obradovich, 1981; cf. Anovitz of California rift basins are largely obscured of ~1.7–2.2 km range between ca. 420 ka and et al., 1999). Recently published (U-Th)/He zir- due to rapid subsidence and sedimentation. 760 ka, yielding subsidence and sedimenta- con ages for a granophyre xenolith indicate an However, geothermal drill wells have fre- tion rates of ~2–4 mm/yr (Schmitt and Hulen, eruption age for Red Island of 2.48 ± 0.47 ka quently penetrated igneous rocks at various 2008). At depth, these Pleistocene sediments are (Schmitt et al., 2013). Two groundmass K-Ar depths in the subsurface, suggesting long-lived presently undergoing prograde metamorphism, analyses for Cerro Prieto are comparatively and widespread magmatic activity. Subsurface with neoblastic biotite and garnet indicating imprecise at 100 ± 60 ka and 120 ± 70 ka (Reed igneous rocks are frequently hydrothermally temperatures >350 °C (McDowell and Elders, et al., 1984), with no radiometric ages available altered and resided at reservoir temperatures 1980). Seismically, this zone of metamorphosed for Roca Consag prior to this study. of up to 390 °C (Schmitt and Hulen, 2008). sediments is identiﬁ ed as an ~10-km-thick

1836 Geological Society of America Bulletin, November/December 2013 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Northern Gulf of California oceanic magmatism

Figure 4. Spidergrams for late and felsic intrusive rocks occur in close spatial Pleistocene mafi c and intermedi- Salton Sea (lava) Cerro Prieto (subsurface) proximity within an ~20 m depth interval in ate rocks from Cerro Prieto and Roca Consag (lava) some wells (e.g., between 3027 and 3048 m Roca Consag normalized to mid- Miocene arc volcanics in well E-30; Fig. 5). ocean-ridge basalt (N-MORB; Niu et al., 1999). Fluid-compati- METHODS ble trace elements are plotted on left. Miocene subduction- fluid Sampling related intermediate volcanic enrichment rocks from northeastern Baja or crustal contamination For accessory mineral geochronology and California (Martín et al., 2000) geochemistry (trace elements and oxygen iso- plotted for comparison show topes), lava from Cerro Prieto and Roca Consag similar enrichments, despite the was sampled to complement published data for absence of ongoing subduction. the Salton Buttes surface and subsurface rocks Note that subduction-related (Schmitt and Vazquez, 2006; Schmitt and Hulen, addition of incompatible ele- 2008; Schmitt et al., 2013). We also extracted ments and contamination with continental crustal rocks lead to largely indistinguishable zircon from Cerro Prieto well cuttings, but only trace-element patterns. Data sources: this study, except Salton Buttes lava data, which are two felsic intrusives yielded juvenile zircon. from Herzig and Elders (1988). Because of the requirement to process comparatively large amounts of samples (tens of grams) for heavy mineral separation, we acknowledge contamination of the cuttings by overlying wall high-velocity section that also could contain an crystalline groundmass (e.g., E-30 3048 m; rock in the wells. Whole-rock geochemical unknown volume of mafi c intrusive rocks (Fuis E-30 3069 m; M-201 3534 m; M-205 2579 m). data (including Th, Sr, and Nd isotopes) were et al., 1984). Gravity and magnetic anomalies in Intrusive intermediate rocks are dominated by obtained from samples where literature data the Salton Trough have also been interpreted as plagioclase with a trachytic texture (e.g., M-203 were unavailable or incomplete. For whole-rock recording the presence of shallow mafi c intru- 3921 m; M-203 3954 m). Felsic intrusive rocks geochemical analysis of cuttings, a gram-sized sions (Kasameyer and Hearst, 1988). are leucocratic microgranites with plagioclase aliquot of magmatic fragments was handpicked From the geothermal operator at Cerro Prieto microphenocrysts surrounded by granophyric under a binocular microscope, and thus contami- (Comisión Federal de Electricidad [CFE]), we intergrowths of quartz and alkali-feldspar (e.g., nation by wall-rock fragments can be ruled out. obtained well cuttings identifi ed by core loggers E-30 3027 m). Opaque minerals, and in some For comparison, published whole-rock compo- as igneous rocks that were penetrated in several cases sulfi des (sphalerite), zircon, and allanite, sitional and isotopic data were compiled (Rob- geothermal wells (Fig. 5). In total, ~100 cut- are present as accessory minerals (e.g., NL-1 inson et al., 1976; Herzig, 1990; Herzig and ting samples from seven wells (E-30, M-194, 3129 m). Remarkably, gabbroic, intermediate, Jacobs, 1994; Schmitt and Hulen, 2008). NL-1, GV-2, M-201, M-203, M-205), each sample consisting of ~50–100 g of millimeter- sized fragments sampled at 1 m intervals, were petrographically investigated using a binocu- % igneous lar microscope, petrographic microscope, and 0 100 0 100 0 100 0 100 0 100 0 100 0 100 0 scanning electron microscope (SEM). igneous (v = extrusive) dated sample In these wells, the shallow sections (to ~1.5 km 500 undifferentiated alluvium, mudstone, shale, and quartzite depth in the SW and ~2.5 km in the NE) are com- )m(ecafruswoleb posed of unconsolidated sediments, whereas 1000 deeper sections are dominantly shale and quartz- 1500 }v ite. Igneous samples comprise vol canic rocks at }v shallow levels (e.g., NL-1 1494 m; M-203 1608 felsic m) with thicknesses of ~20 m (Fig. 5). These 2000 intermediate basaltic volcanic rocks are aphyric, glassy, and vesicu- lated, and they represent buried pyroclastic fall- 2500

out deposits. At greater depth, igneous rocks htpe 3000 appear to be sill-like intrusions based on similar depths of intersection in neighboring wells (e.g., d 3500 microgranites at ~3000 m in E-30 and NL-1). These intrusions are up to ~40 m in thickness, 4000 although exact determination is compromised E-30 M-194 NL-1 M-203 M-201 M-205 GV-2 by mixing between cuttings derived from different depth intervals. Mafi c intrusives are fi ne- Figure 5. Simplifi ed well logs showing distribution of igneous rocks grained gabbroic rocks, with plagioclase and in Cerro Prieto geothermal wells. Percentages of igneous cuttings clinopyroxene as the dominant phenocrysts with per 3 m depth interval were visually estimated under a binocular ophitic texture and a brownish glassy to micro- microscope.

Geological Society of America Bulletin, November/December 2013 1837 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Schmitt et al.

Geochronology ded in epoxy, and polished to expose crystal ments are required. These were ground in an interiors. A 20 nA primary beam was focused agate mortar, and the majority of powder was U-Th and U-Pb Zircon onto a spot with ~15 μm diameter. Second- retained for Sr and Nd isotopic analysis. Only Zircon crystals for U-Th and U-Pb dating ary ion intensities were acquired in 10 magnet ~100 mg samples of rock powder were fused (Tables DR1 and DR21) were density separated cycles per analysis, resulting in a crater depth of into millimeter-sized glass beads using an elec- from crushed rocks using heavy liquids. Clear ~1 μm. Throughout the manuscript, we report trically heated tungsten crucible in an Ar atmo- and euhedral zircon crystals were subsequently and plot individual data points with 1σ errors, sphere. The glass beads typically contained ~1–4 handpicked, placed with a prism face onto a fl at but we state average age uncertainties at 95% wt% of WO3 from the crucible material, and all indium (In) metal surface, and pressed into the confi dence. Half-lives and isotopic ratios are data are reported normalized to 100% W-free. In using a fl at tungsten carbide metal anvil and an from Jaffey et al. (1971) and Cheng et al. (2000). XRF and glass bead EMPA major elements for arbor press. The In-mounted zircon crystals were Cerro Prieto sample CP05–2 agree within ±5%, ultrasonically cleaned in 1 N HCl and deionized (U-Th)/He Zircon except for FeO and Na2O, which are ~6% and water, and then coated with a gold layer, several The (U-Th)/He age determinations were car- ~10% lower in the fused beads compared to tens of nanometers thick, to provide a conduc- ried out at the University of Kansas using labora- XRF values. Trace-element analyses of glasses tive surface. Analysis spots on unsectioned crys- tory procedures described in Biswas et al. (2007). and zircons (Table DR3 [see footnote 1]) were tals are identifi ed as “rim,” whereas “interior” In cases where unsectioned crystals where ana- conducted by SIMS using energy-fi ltering to refers to crystals analyzed after sectioning and lyzed by SIMS, zircons were extracted from the suppress molecular interferences modifi ed from polishing to tens of micrometers using abra- mounts using a steel picking needle. Individual the procedure in Monteleone et al. (2007). NIST sives. For samples with too little material avail- crystals were wrapped in Pt foil, heated for 10 SRM610 glass was used as a primary standard able for heavy mineral separation (e.g., NL-1), min at 1290 °C, and reheated until >99% of the (Pearce et al., 1997), and accuracy was moni- thin sections of cuttings were screened by SEM, He was extracted from the crystal. All ages were tored by analysis of BHVO-2 glass and 91500 and the interiors of accessory zircon and allanite calculated using standard α-ejection corrections zircon. For most trace elements (in particular were analyzed in situ. using morphometric analyses (Table 1; Farley rare earth elements [REEs] + Y), SIMS abun- Isotopic analysis was conducted by second- et al., 1996). After laser heating, zircons were dances for BHVO-2 (whole rock) and 91500 ary ion mass spectrometry (SIMS) using the unwrapped from the Pt foil and dissolved using (zircon) agree within ±12% with published val- CAMECA IMS 1270 at the University of Cali- ues (Wilson, 1997; Liu et al., 2010). double-step HF-HNO3 and HCl pressure-vessel fornia–Los Angeles (UCLA; techniques modi- digestion procedures (Krogh, 1973). U, Th, and fi ed from Reid et al., 1997; Schmitt et al., 2006; Sm concentrations were determined by isotope Oxygen Isotopes in Zircon Vazquez and Reid, 2004). The main differences dilution–inductively coupled plasma–mass spec- Oxygen isotopes in zircon (Table DR4 [see to previously published methods are crystal rim trometry (ID-ICP-MS) analysis. The laboratory footnote 1]) were analyzed by SIMS using μ analysis to ~5 m depths at lateral beam diam- routinely analyzes zircon standards with indepen- multicollection dual Faraday cup analysis as μ eters of ~25–35 m, and the use of dual electron dently determined ages, and we report averages described in Trail et al. (2007). Unknowns multipliers separated by two atomic mass units for Fish Canyon tuff zircons of 27.8 ± 0.1 Ma were bracketed by analysis of AS3 standard δ18 in order to simultaneously collect background (1 relative standard deviation RSD% = 4%; n = zircon with O = 5.34‰ (Trail et al., 2007), 230 16 + intensities at mass/charge ~244 and Th O at 285). Reported age uncertainties refl ect the and accuracy was estimated by analyzing 91500 δ18 mass/charge ~246. Dual collection reduces the reproducibility of replicate analyses of these zircon as a secondary standard ( O = 9.98‰; magnet cycling time by ~20%. Primary beam laboratory standard samples (Farley et al., 2002), Wieden beck et al., 2004), which was mounted 16 – currents ( O ) were 40–60 nA, and total analy- but for much younger zircons such as the ones in the same geometry as the unknowns. The sis duration per spot was 25 min. Zircon stan- analyzed here, analytical uncertainties are sig- deviation from the published value for 91500 dard AS3 was used for Th/U relative sensitivity nifi cantly larger. We accounted for these by is ~0.2‰, commensurate with the SIMS repro- calibration and as an equilibrium zircon standard multiplying the average age uncertainty by the ducibility of the primary standard AS3. 230 238 for which ( Th)/( U) = 1.008 ± 0.007 (mean square-root of the MSWD. square of weighted deviates [MSWD] = 0.51; U-Th Isotopes in Whole Rocks n = 33) was obtained on analyses interspersed U and Th were separated using ion-exchange with the unknowns. Zircon elemental concen- Compositional and Isotopic Analysis column methods described in Shen et al. (2003). 238 16 + 90 16 + trations were calculated from U O / Zr2 O4 Isotopic measurements were conducted on a intensity ratios calibrated on zircon stan- Major and Trace Elements in Whole Thermo Electron Neptune multicollector (MC) dard 91500 with U = 81.2 ppm (Wiedenbeck Rocks and Zircon ICP-MS at the High-Precision Mass Spec- et al., 2004). Whole-rock compositional analysis of lava trometry and Environment Change Labora- U-Pb zircon analyses were conducted using samples was conducted using X-ray fl uores- tory (HISPEC ), Department of Geosciences, the CAMECA IMS 1270 at UCLA following cence (XRF) and ICP-MS analysis at University National Taiwan University (Shen et al., 2012). techniques outlined in Grove et al. (2003) and of Washington (Table 2). For cuttings from geo- A triple-spike, 229Th-233U-236U, isotope-dilution Schmitt et al. (2003). Zircons were handpicked thermal well samples, major and trace elements method was employed to correct mass bias and from heavy-liquid mineral separates, embed- were analyzed in bulk using electron microprobe determine uranium concentration (Shen et al., analysis (EMPA) and SIMS, respectively, by the 2002). Uncertainties in concentration and iso- 1GSA Data Repository item 2013358, Table DR1 fused-bead method (Table 2; Nicholls, 1974). topic data include corrections for blanks, instru- (U-Th zircon geochronology), Table DR2 (U-Pb Fused-bead analysis minimizes sample con- mental fractionation, multiplier dark noise, zircon geochronology), Table DR3 (zircon trace elements), and Table DR4 (zircon oxygen isotopes), is sumption, and aids in avoiding contamination by spectral interferences, and errors associated available at http://www.geosociety.org/pubs/ft2013 sedimentary rocks in the well cuttings because with quantifying the isotope composition in the .htm or by request to [email protected]. only small amounts of handpicked igneous frag- spike solution (Table 3).

1838 Geological Society of America Bulletin, November/December 2013 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Northern Gulf of California oceanic magmatism

TABLE 1. SUMMARY OF (U-Th)/He AND U-Th ZIRCON GEOCHRONOLOGY (Fig. 6), as well as U-Th zircon crystallization Age ± U Th 4He Mass ages from rhyolite lavas and the granophyre Sample-grain (ka) (ka) (ppm) (ppm) Th/U (nmol/g) (μg) Ft† xenolith (Fig. 7; data from Schmitt and Vazquez, Cerro Prieto lava (32°25′26″N, 115°18′3″W) CP09-04-1 14400 1200 196 279 1.42 16.4 10.7 0.81 2006; Schmitt et al., 2013). The Red Island CP09-04-2 98700 7900 201 137 0.683 97.7 6.9 0.78 (U-Th)/He zircon eruption age is 2.48 ± 0.47 ka CP09-04-4 87700 7000 322 227 0.705 132 3.9 0.74 (Schmitt et al., 2013). In the 230Th/232Th versus CP09-04-5 5900 500 613 1282 2.09 20.9 3.5 0.71 238 232 CP09-04-6 117100 9400 141 47 0.334 71.7 3.9 0.74 U/ Th activity diagram, Salton Buttes zir- CP09-04-9* 68.4 5.5 227 59 0.259 0.073 12.2 0.82 con compositions fall in a wedge-shaped fi eld CP09-04-10 105000 8400 497 213 0.429 235 4.5 0.75 CP09-04-11 22000 1800 288 174 0.605 27.3 2.2 0.70 (“sphenochron”), with the low U-Th apex of this CP09-04-13 172 14 224 84 0.375 0.171 5.3 0.76 wedge pointing toward the whole-rock compo- CP09-04-14 86600 6900 113 71 0.625 47.1 5.9 0.77 sition (Figs. 7A and 7B; Schmitt and Vazquez, CP09-04-M1 533600 42700 76.7 40.8 0.531 188 9.3 0.73 CP09-04-M2 415800 33300 10.8 4.7 0.434 23.6 73.9 0.86 2006; Schmitt et al., 2013). This indicates that Cerro Prieto sedimentary enclave (32°25′21″N, 115°18′45″W) the duration of zircon crystallization was pro- 230 CP0702D-1 901 72 207 85 0.410 0.818 4.2 0.74 tracted relative to the half-life of Th. U-Th CP0702D-2 26400 2100 344 107 0.311 39.4 4.5 0.75 zircon–whole rock model ages date back to ca. CP0702D-3 389 31 35.8 20.5 0.573 0.067 8.0 0.78 20 ka in the lavas, and ca. 40 ka in the grano- CP0702D-4 8400 700 543 409 0.753 20.9 3.4 0.72 CP0702D-5* 78.0 6.2 193 132 0.683 0.064 2.1 0.67 phyre xenolith, with the youngest crystals and CP0702Dr-1 5800 500 144 19 0.135 3.67 6.6 0.79 crystal populations yielding ages of 5.5 ± 1.2 ka CP0702Dr-2 3500 300 140 26 0.188 2.08 4.0 0.76 CP0702Dr-3 2200 200 258 32 0.123 2.38 3.5 0.74 and 2.9 ± 0.6 ka, respectively (Figs. 7A and 7B). CP0702Dr-4 3300 300 64.3 50.6 0.787 1.01 3.3 0.73 The onset of crystallization therefore must have CP0702Dr-5 15700 1300 136 41 0.297 9.32 4.4 0.76 preceded the eruption by at least several tens of Average age: 73 ± 7 ka (n = 2) thousands of years, but zircon likely continued ′ ″ ′ ″ Roca Consag lava (31°6 45 N, 114°29 17 W) to crystallize until shortly before eruption. RC01-24A-1 75.1 6.0 161 263 8.55 1.64 2.2 0.62 RC01-24A-2* 53.1 4.3 80.3 38.9 2.08 0.485 6.2 0.78 RC01-24A-3* 53.1 4.2 176 107 3.47 0.609 4.4 0.72 Cerro Prieto RC01-24A-4* 36.4 2.9 60.2 30.2 2.76 0.503 3.4 0.73 RC01-24A-5* 59.8 4.8 95.7 35.4 2.22 0.370 5.3 0.75 RC01-24A-6* 42.5 3.4 280 132 6.33 0.469 3.9 0.73 An extensive search for zircon in Cerro RC01-24A-7* 54.5 4.4 500 301 6.41 0.602 4.4 0.75 Prieto by processing ~20 kg of rock yielded RC01-24A-8* 30.8 2.5 233 46 0.990 0.198 8.3 0.79 RC01-24A-9* 50.2 4.0 71.7 29.8 1.82 0.415 5.8 0.77 only scarce (~20) mostly small, rounded, and RC01-24A-10* 51.7 4.1 185 54 2.47 0.290 7.6 0.78 brownish zircon crystals in the heavy mineral RC01-24A-11* 41.6 3.3 39.6 17.7 1.42 0.447 6.6 0.78 fraction. Only three clear, needle-shaped zircon RC01-24A-12* 38.0 3.0 208 103 1.33 0.495 4.4 0.76 Average age: 43 ± 6 ka (n = 11; mean square of weighted deviates [MSWD] = 7.1) crystals were found, but these were too small *Used for average. for (U-Th)/He analysis because long stopping †Ft—correction factor for 4He ejection errors 1σ. distances of alpha particles result in signifi cant loss of radiogenic 4He (Farley et al., 1996). U-Th and U-Pb dating confi rmed that most Sr-Nd Isotopes in Whole Rock spectrometer (TIMS) equipped with eight vari- brownish Cerro Prieto zircons are xenocrystic Sample preparation and element separation able collectors and one fi xed Faraday collector with ages between ca. 60 Ma and ca. 2 Ga (Fig. for Sr and Nd isotope analyses were carried out in static mode. Strontium isotope compositions 8A; Tables DR1 and DR2 [see footnote 1]), in PicoTrace® clean laboratory facilities at Cen- were measured either at LUGIS or at Scripps whereas the clear crystals (one crystal analyzed tro de Investigación Científi ca y de Educación Institution of Oceanography, La Jolla, Califor- with multiple spots) display signifi cant U-Th Superior de Ensenada (CICESE), Ensenada, nia, using a nine-collector, Micromass Sector disequilibrium. Combining all disequilibrium Mexico. Approximately 100 mg aliquots of 54 TIMS. Isotopic ratios were corrected for U-Th zircon analyses yields an isochron age of handpicked and powdered whole rock were mass fractionation by normalizing to 86Sr/88Sr = 86 +37/–33 ka (MSWD = 1.3; n = 8; Fig. 6B). heated at 165 °C for 15 h in a mixture of ~3–4 mL 0.1194 and 146Nd/144Nd = 0.7219. Neodymium An alternative treatment of the data is to com- double-distilled concentrated HF, ~1 mL double- isotopes are reported relative to 143Nd/144Nd = bine individual zircon spots with the whole-rock distilled concentrated HNO3, and a few drops of 0.51185 for the La Jolla Nd standard, and composition to calculate model crystallization

distilled concentrated HClO4 using a PicoTrace Sr isotopes are reported relative to NBS 987 ages (Fig. 6B). In this case, we obtained model DAS® pressure digestion system. After evapo- 87Sr/86Sr = 0.71025. Because of the young age crystallization ages between 19 +24/–22 ka and ration, the residues were dissolved in 2 N HCl. of the samples, no age corrections were applied, 121 +128/–80 ka, but whole-rock compositions Strontium and REEs were separated using quartz and elemental abundances are from XRF and for Cerro Prieto are likely affected by alteration glass columns ﬁ lled with Dowex AG50-WX8 SIMS whole-rock analyses (Table 2). (see following), and therefore this approach cation-exchange resin in HCl medium. Sm and may result in age bias. Nd were separated in quartz-glass columns ﬁ lled GEOCHRONOLOGY RESULTS Xenocrystic zircon can be used to date the with LN-Spec® resin in HCl medium. eruption age if 4He has degassed completely Neodymium isotope ratios were measured Salton Buttes during contact with the magma (Blondes et al., at Laboratorio Universitario de Geoquimica 2007). To test this approach, we analyzed indi- Isotopica (LUGIS), Instituto de Geofísica, Uni- The best evidence for the youthfulness of vidual xenocrysts extracted from the lava sam- versidad Nacional Autónoma de México, using the Salton Buttes is from published (U-Th)/He ple, as well as zircon separated from an ~10-cm- a Finnigan MAT 262 thermal ionization mass zircon ages for Red Island granophyre xenolith diameter baked mudstone enclave. This enclave

Geological Society of America Bulletin, November/December 2013 1839 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Schmitt et al. § ″ ″ 45 17 ′ ′ 31° 114° 06 29 RC01-24B § ″ ″ 45 17 0.20 0.29 0.5 0.6 1.0 1.4 1.2 1.6 2.0 2.0 0.2 0.3 7.8 8.0 7.9 8.4 4.1 3.4 ′ ′ 31° 59 66 10 10 114° 300 295 783 540 06 29 Roca Consag lava RC01-24A § ″ ″ 60 30 0.47 <0.1 0.11 0.46 0.05 0.11 0.90 0.20 0.25 1.2 1.6 5.6 5.4 0.2 0.6 ′ ′ 32° 00005 0.000023 0.000005 115° Cerro dome 24 18 0 . Prieto N CP07-2 † ″ ″ 5 ′ 25 6.0 6.8 1.90 – 0.15 0.03 – – – – – – – 25 – ′ 32° 26 23 25 27 115° 18 734 720 280 317 247 227 Cerro dome 25 Sr values for samples CP07-2 and RC01-24B Prieto S CP05-2 86 Sr/ 87 ″ 7–0 ′ ′ ″ 0 0 9.3 0.2 6.7 0 0 2957 115°9 32°25 4.567 0 M-205 41.608 . ′ ″ ″ ′ 3987 32°23 M-203 115°11 33.463 12.434 ′ ″ ″ 5–0 ′ 0 on State University. 0 8.9 5.7 8.2 7.2 5.3 6.6 5.8 – 0 0 3954 32°23 0 M-203 115°11 33.463 12.434 . nal Autónoma de México, except nal ′ ″ ″ 50 ′ 0 0 0 0 3921 32°23 0 M-203 115°11 33.463 12.434 . ′ ″ ″ 60 ′ 0 0 9.2 16 7.1 11 9.1 12 13 14 0.2 0.3 0.3 – 8.9 6 0 0 3534 32°24 0 M-201 115°10 34.768 23.248 . 0.5130937 0.5129439 0.5129346 – 0.512983 – 0.512671 0.512849 0.512811 0.704711 0.708185 0.7085760.000002 0.0000035 0.000004 – 0.704815 – 0.0000025 – – 0.7053749 0.703736 0.00001 0.7036664 0.0000085 0.000014 ′ ″ ″ ′ *sgnittucl – – 3513 32°24 M-201 115°10 34.768 23.248 ′ l ″ ″ ′ ewot – – 3080 32°24 M-201 115°10 34.768 23.248 e irP o ′ ″ ″ rreC ′ – – 2511 32°24 M-194 115°12 42.055 33.873 ′ ″ ″ ′ 8.3 6.2 9.0 15 – – E-30 3069 32°23 115°12 51.315 57.086 ′ ″ ″ ′ 7.8 7.4 5.7 7.0 13 E-30 3048 32°23 115°12 51.315 57.086 4––––– 5––––– ′ ″ ″ 8––––– 4–––––0 2 5 ′ TABLE 2. WHOLE-ROCK COMPOSITIONS OF LAVAS AND INTRUSIVE ROCKS FROM CERRO PRIETO AND ROCA CONSAG AND ROCA AND INTRUSIVE ROCKS FROM CERRO PRIETO 2. WHOLE-ROCK COMPOSITIONS OF LAVAS TABLE 7 0 9 0 0 6 0 0 3 8 0 0 1 0 0 0 E-30 3027 32°23 0 5 7 0 115°12 51.315 57.086 . . . . 30 50 ′ ″ ″ 8 50 50 1 ′ 2 8 0 2 0 7 0 9 0 4 0 2 0 0 0 1 NL-1 1494 32°24 0 7 0 5 115°12 42.055 33.873 . . . . 50 ′ ″ ″ 10 40 70 3 ′ 4 0 6 0 4 0 8 0 1.519.02 1.450.134.01 8.937.85 0.36 0.153.89 4.290.80 3.73 7.950.26 1.45 0.07 3.76 0.29 0.82 8.64 2.40 0.28 1.70 0.13 3.20 4.31 4.05 9.38 8.28 0.05 1.96 0.16 3.37 5.11 1.50 9.70 9.71 0.26 1.65 0.21 2.49 2.71 0.82 9.86 7.00 0.10 1.85 0.19 2.86 2.05 2.40 9.33 6.12 0.33 2.54 0.16 4.36 2.73 0.98 11.05 6.15 0.64 0.87 0.20 4.38 5.65 1.77 6.22 10.38 0.65 0.79 0.07 3.00 1.11 0.76 6.72 5.50 0.36 0.75 0.09 4.52 1.36 1.37 6.12 5.34 0.34 1.82 0.14 4.52 3.51 1.34 10.48 8.07 0.33 0.15 2.30 0.49 2.71 2.23 6.26 4.67 0.22 2.98 0.52 0.09 0.63 0.63 5.59 0.11 3.08 0.09 5.40 0.37 0.85 1.60 3.74 3.34 0.13 5.29 0.38 0.06 1.49 1.58 2.20 0.15 5.85 0.06 4.06 1.58 0.65 5.67 0.13 4.16 0.68 0.13 .3380.53.59.44.64.53 6.52.02.0 5.61.51.8 3.20.600.76 2.90.400.69 5.6 5.83.23.318.06.45.35.75.99.44.48.46.47.45.3– 3.83.7178.16.44.57.0126.49.87.45.86.0– 160.863.70.66 0.934.6 130.87 3.9 0.552.62.6125.55.04.35.29.85.68.25.52.74.5– 2.70.35 4.5 11 0.832.72.7145.55.64.15.5105.27.65.03.35.1– 18 3.10.36 0.39 1.0 21 4.2 11 0.38 1.6 1.8 1.3 11 2.1 10 2.14.5 1.40.3 0.84 0.68 1.9 1.0 5.5 18 1.1 0.76 0.6 1.5 1.4 1.1 8.5 0.57 1.4 0.2 2.1 1.8 – 0.55 0.71 2.5 – 1.7 3.1 0.80 1.2 1.4 1.8 1.4 1.9 0.78 2.3 1.5 0.72 1.2 3.1 1.4 1.0 1.2 0.78 1.8 0.91 1.5 0.87 0.44 1.3 1.2 0.27 0.70 1.6 0.77 – – ––––––––––––– 4 0 2 0 11 14 53 32 20 64 15 52 29 27 26 38 11 1227271465038474670306852664057531516 15 13 16 69 79 21 29 18 23 21 26 20 27 29 46 11 22 32 38 25 31 33 31 17 23 24241095548344784488048294231296.07.7 56.016.8 56.0 16.7 71.9 14.0 57.0 15.398.5 54.5 16.1 99.0 57.7 15.4 98.8 58.2 16.6 98.6 58.6 15.0 96.9 51.1 15.3 98.1 62.2 18.1 97.9 61.8 18.0 98.8 61.3 15.6 99.2 58.4 16.6 97.4 68.6 98.4 15.1 67.2 98.2 14.7 67.0 95.3 16.9 67.8 99.8 17.0 99.6 100.1 99.7 0 0 1 0 1608 239 240 1019 242 166 638 434 281 62 461 524 670 358 228130 226 132 326 699 206 337 182 270 255 249 347 265 173 470 162 207 456 357 370 310 280 220 226 244 32°23 7 0 5 0 M-203 115°11 12.434 33.463 . . . . 0 0 0 0 # # Nd total Sr 144 3 :e 3 86 5 2 O 2 O Major oxides and trace elements by X-ray ﬂuorescence (XRF); Pomona College. Major oxides by XRF; trace elements inductively coupled plasma–mass spectrometry (ICP-MS); GeoAnalytical Laboratory Washingt Sr-Nd isotopes analyzed at Laboratorio Universitario de Geoquimica Isotopica (LUGIS), Instituto Geofísica, Universidad Nacio 2 # O 2 p O O nducted at University of California–Los Angeles; last number in sample name *Major oxides by electron microprobe analysis; trace elements secondary ion mass spectrometry using glass beads; analyses co nducted at University of California–Los † § # Nd/ 2 2 2 Sr/ yT Nd Latitude (°N): TiO Al Fe MnO MgO CaO Na K P LOI (%) Total elements (ppm) Trace Sample: (Scripps). Nb Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu 1 s.e. 1 s.e. 1 s.e. indicates depth in m. SiO Rb Longitude (°W): Sr Y Zr Major oxide (wt%) 87 143 ε

1840 Geological Society of America Bulletin, November/December 2013 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Northern Gulf of California oceanic magmatism

30 than previously indicated based on compara-

U) A Salton Buttes tively imprecise K-Ar ages averaging ca. 100 ka 238 25 (Reed et al., 1984), although the exact eruption U)/(

234 20 age remains poorly constrained. Regardless, the presence of undegassed zircon xenocrysts has 15 important implications for the timing of con- U) ( tamination of Cerro Prieto magmas en route to 238 10 2.48±0.47 ka (n = 27) surface (see following). Th)/( 5 Seven intermediate to microgranitic Cerro 230 0 Prieto subsurface samples were processed, but only two microgranite samples in wells NL-1 B Cerro Prieto Th) ( 20 and E-30 yielded zircon and allanite of late 232 73±7 ka Pleistocene age (Table DR1 [see footnote 1]). U)/(

238 15 (n = 2) The other samples contained exclusively older zircon, most likely from contamination of the 10 cuttings by fragments from the overlying sedi- lava Th) ( ments (Table DR2 [see footnote 1]). For the 232

rank order sediment enclave 5 microgranites, the resulting U-Th isochron ages

Th)/( overlap within uncertainty (36 +15/–14 ka and 230 0 42 +15/–14 ka (Fig. 6D). The MSWD value of

C 1.2 and 1.8 for NL-1 and E-30, respectively, Roca Consag( ( suggests negligible nonanalytical scatter. Two- 10 point model isochrons can be calculated using either Cerro Prieto lava or NL-1 allanite under Th (ppm) ( 43±6 ka 232 the assumption that they were in isotopic equi- (n = 11) 5 librium with zircon at the time of crystallization. In both scenarios, there would be a signiﬁ cant portion of zero-age zircon in the E-30 popula- ka Ma 0 tion, but because of low U concentrations, the U (ppm) 0 1 precision of individual model ages is limited.

238 20 40 60 80 10 100 1000 In any case, the data for subsurface zircon are 3.482 0.0034 15.29 0.01 0.974 0.002 0.696 0.001 1.400(U-Th)/He 0.003 0.974 0.003 zircon age indicative of magma presence at shallow levels within the Cerro Prieto basin at ca. 40 ka W 0.597 0.0005 3.104 0.004 0.641 0.002 0.588 0.001 1.091 0.003 0.720 0.002 W 5.67 0.0025 18.30 0.02 0.990 0.002 0.947 0.001 1.046 0.002 0.996 0.002 W 5.99 0.0036 19.83 0.01 0.997 0.002 0.923 0.001 1.080 0.002 1.000 0.002 W ″ ″ ″ Figure 6. Individual (U-Th)/He ages and or younger. 05 08 03 44 ″ ′ ′ ′ ′ averages of (A) Salton Buttes (Schmitt

n et al., 2013); (B) Cerro Prieto, and (C) Roca Roca Consag o i t

a Consag. Note that zircon crystals in B are c oL N, 115°18 N, 115°38 N, 115°38 N, 115°36 ″ ″ xenocrystic (from a lava sample, and a sedi- Roca Consag zircons are overwhelmingly 15 ″ 15 25 ″ 51 ′ ′ ′ ′ mentary enclave), and that they have re- xenocrystic (Fig. 8B), with the youngest U-Pb tained pre-eruptive 4He. The youngest ages zircon ages at ca. 120 ka (n = 2). One of these 33°11 33°10 32°25 in B thus indicate a maximum eruption age. zircon crystals (RC z19) also displays U-series disequilibrium with (230Th)/(238U) = 0.85, attest- ing to its young age (Tables DR1 and DR2

TABLE 3. U-Th ISOTOPIC COMPOSITIONS (ACTIVITIES IN PARENTHESES) FOR WHOLE ROCKS COMPOSITIONS (ACTIVITIES IN PARENTHESES) 3. U-Th ISOTOPIC TABLE was collected in the periphery of the Cerro [see footnote 1]). An outstanding feature in the Prieto dome where autobrecciated lava inter- Roca Consag zircon population is a dominant ﬁ ngers with surrounding playa sediments. In ca. 1 Ma age peak, signiﬁ cantly predating the both cases, however, (U-Th)/He ages for xeno- younger population (Fig. 8B). This implies zir- crystic zircons scatter widely between ca. 68 ka con recycling from a previously undiscovered epytkc and ca. 550 Ma (Fig. 5B; Table 1). We attribute Pleistocene source in the Wagner basin. The this to incomplete degassing of zircons, sug- abundance of older crystals is likely due to a o

R gesting that the lava entrained the crystals (and contribution of Colorado River sediment to the of the mean. Activity ratios were calculated by the decay constants used in Cheng et al. (2000). of the mean. σ enclaves) only shortly before cooling and solidi- zircon population in Roca Consag lavas. The ﬁ cation. The two youngest (i.e., most degassed) (U-Th)/He zircon ages for Roca Consag were xenocrysts overlap within uncertainty and yield calculated without disequilibrium corrections an average age of 73 ± 7 ka (n = 2). This is simi- because their old crystallization ages imply sec- lar to the U-Th zircon crystallization age for ular equilibrium (Fig. 5C; Table 1). The average the few non-xenocrystic zircons in Cerro Prieto (U-Th)/He age is 43 ± 6 ka (n = 11; MSWD =

Analytical errors are 1 lava (a maximum estimate for the eruption 7.1 for individual 2σ errors of 16%). This aver- elpmaS age). Collectively, these data strongly suggest age excludes the smallest grain (zRC-1), with Note: I70-24 Red Island granophyre xenolith SB0402 Obsidian Butte lava 33°10 G2-112 Obsidian Butte lava (cuttings from geothermal well) CP05-2 Cerro Prieto lava that the eruption of Cerro Prieto occurred later its signiﬁ cant age uncertainty because of an

Geological Society of America Bulletin, November/December 2013 1841 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Schmitt et al.

A Salton Buttes lava B Salton Buttes 30 granophyre

e in 20 il 20 e n u i q il Figure 7. U-Th zircon and whole- e u 10 q rock data (activities indicated e 10 in parentheses) for (A) Salton 5 Buttes lava, (B) Salton Buttes 0 5 granophyre, (C) Cerro Prieto whole rock lava, and (D) Cerro Prieto sub- zircon rim 0 surface intrusives. Data in A zircon interior and B are from Schmitt et al. (2013); other data are from this study. Panels A and B show C Cerro Prieto lava D Cerro Prieto subsurface model isochron ages based on +37 86 -33 ka (n = 8) +15 pairing zircon with whole-rock allanite NL-1 36-14 ka (n = 15) data. Panels C and D show zir- zircon NL-1 con isochrons and error enve- zircon E-30 e n lopes from regression of the li e i in u il zircon data (hatched), in com- q u e 27 q parison to paired whole-rock e (allanite) and zircon model isochron ages. 0 0 +15 42 -14 ka (n = 51)

α unusually large -ejection correction. We adopt A kinked trend in Al2O3 versus SiO2 with a xenoliths, the presence of amphibole indicates this average of 43 ka as the eruption age of Roca peak at SiO2 = 62 wt% exists for Cerro Prieto higher degrees of hydration, compared to clino- Consag. The elevated MSWD is likely due to intermediate rocks, which also shows a subtle pyroxene-bearing gabbros in Cerro Prieto wells. unrecognized systematic uncertainties regard- break in slope for CaO and MgO versus SiO2. Among trace elements plotted in Figure 3, ing U and Th zonation of the crystals, and we This clearly rules out binary mixing as a mecha- Sr shows signiﬁ cant scatter and deviates from account for this by multiplying the age uncer- nism for generating intermediate compositions. the Ca trend. We attribute Sr mobility to hydro- tainty with the square-root of the MSWD. The trends in Figure 3 are broadly consistent thermal alteration, which limits the use of Sr with the increasing abundance of plagioclase in as a petrogenetic indicator (this also holds for CHEMICAL COMPOSITIONS the fractionating assemblage, but similar varia- 87Sr/86Sr; see following). Zr abundances gen-

tions can result from variable degrees of partial erally increase with SiO2, suggesting zircon- Whole-Rock Major and Trace Elements melting of hydrous basalt at low to intermediate undersaturated conditions throughout most of

H2O partial pressure P(H2O) (<170 MPa; Thy the magmatic differentiation path. For Cerro New and published whole-rock analyses for et al., 1990). Importantly, experimental glass Prieto, this is consistent with the scarcity or northern Gulf of California igneous rocks indi- compositions for partial melting of hydrated absence of zircon in intermediate rocks and their cate substantial compositional diversity rang- basalt under higher P(H2O) between 300 and exclusive presence in microgranites. ing from basalt to high-silica rhyolite (Fig. 3; 500 MPa (Thy et al., 1990) are higher in Al2O3 MORB-normalized trace elements for mostly Table 2). A characteristic difference between than observed for Cerro Prieto rocks (Fig. 3B). mafic to intermediate compositions (Cerro

Salton Buttes and Cerro Prieto is the strongly K2O broadly increases with increasing SiO2, but Prieto and Roca Consag; Fig. 3) are enriched bimodal character of the Salton Buttes lavas and most compositions plot above values permissive in fl uid-compatible elements—typically inter- xenoliths, whereas Cerro Prieto surface and sub- for closed-system fractional crystallization of preted as an indicator for fl uid addition to the surface rocks have a dominantly intermediate MORB. Although major-element compositional mantle during subduction. Comparison with population (Fig. 3). Despite these differences, trends are broadly similar for MORB fractional regional subduction-related volcanic rocks (San end-member compositions are almost indistin- crystallization and partial melting of MORB-type Luis Gonzaga volcanic fi eld; Martín et al., 2000) guishable: The least altered (based on low-K) lower crust (France et al., 2010), the presence of shows that similar enrichments exist for equiva- gabbros from Cerro Prieto are equivalent to those excess K2O (Fig. 3E) suggests that closed-system lent SiO2 (ranging between ~54 and ~67 wt%), at the Salton Buttes, and high-silica end-member fractional crystallization of a MORB parent can- including a prominent negative Nb anomaly. compositions (Cerro Prieto microgranite; Salton not be the sole process that produced felsic melts Positive Sr anomalies indicate the presence of Buttes rhyolite and granophyre) equally overlap. in the northern Gulf of California. For the basaltic cumulate plagioclase in some intrusive samples.

1842 Geological Society of America Bulletin, November/December 2013 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Northern Gulf of California oceanic magmatism

87 86 ε a trend with anticorrelated Sr/ Sr and Nd, as lava n = 16 age (Ma) expected for magmatic mixing and/or assimilation involving crustal sources such as Peninsu- wells n = 55 lar Ranges batholith or metasedimentary rocks (Fig. 9). Subsurface samples (including Cerro A Prieto microgabbros) are displaced from the Cerro Prieto depleted-mantle trend toward elevated 87Sr/86Sr (Fig. 9). This is also the case for the Salton 50 ka Buttes rhyolite lava and a granophyre xenolith

Pb 100 ka (Fig. 9). We interpret this as the result of Sr exchange with sedimentary wall rock during

Pb/ hydrothermal alteration in an active geothermal

207 206 150 ka ε ﬁ eld. Although no consistent trend in Nd versus

SiO2 exists over the entire data set (Fig. 3H), ε 200 ka there are two delineating end members: (1) Nd is nearly invariant over the entire range of SiO2 for Cerro Prieto gabbro and microgranite, which is indicative of closed-system magma differen- ε tiation; and (2) Nd systematically decreases with

age (Ma) increasing SiO2 for intermediate Cerro Prieto samples, which potentially indicates that fractional crystallization is coupled to assimilation- n = 54 fractional crystallization (AFC). B Roca Consag U-Th whole-rock isotopic compositions for Salton Buttes and Cerro Prieto plot to the left of the equiline (Fig. 6; Table 3). This is characteristic for magmas generated by decompression melting of the mantle, in the absence of a high-U ﬂ uid-mobile component. Their (238U)/(232Th)

Pb/ Pb overlaps with MORB compositions from the

207 206 150 ka East Pacifi c Rise at 21°N, with the exception of the extremely low U/Th for Cerro Prieto lava. This is likely due to secondary altera- 300 ka 200 ka 1000 ka 500 ka tion (causing U-loss), which is also indicated by strong (234U)/(238U) disequilibrium, and thus Cerro Prieto lava is not further considered here. 238U/ 206 Pb The comparatively low (230Th)/(238U) relative to MORB could indicate protracted differen- Figure 8. U-Pb zircon ages for (A) Cerro Prieto and (B) Roca tiation time scales (i.e., aging of the source over Consag. Large panels are 207Pb/206Pb vs. 238U/206Pb concordia dia- 200 k.y.). We, however, favor mixing with and/or grams showing data uncorrected for common Pb and isochrons for assimilation of crustal rocks in secular equilib- mixtures between anthropogenic common Pb (207Pb/206Pb = 0.8283) rium because of the isotopic heterogeneity in ε 143 144 87 86 and radiogenic Pb (ages in ka). Smaller panels are relative prob- Nd ( Nd/ Nd) and Sr/ Sr (Fig. 10) and the ε ability plots for zircon ages (ages in Ma) showing zircons from decrease in Nd with increasing SiO2 (Fig. 3H). (A) Cerro Prieto lava and well samples and (B) Roca Consag lava. Zircon in Cerro Prieto igneous well samples is likely derived from Zircon Trace Elements and Oxygen Isotopes wall rock and is interpreted to represent the detrital age distribution of Colorado River and local sediment sources. The dominant Because hydrothermal overprint affected sub- ca. 1 Ma peak in Roca Consag results from crystal recycling from surface samples, we used trace elements and an unidentifi ed source that is buried and submerged in the northern oxygen isotopic compositions of zircon as petro- Gulf of California. genetic indicators (Figs. 11–13; Tables DR3 and DR4 [see footnote 1]). Zircon’s imperviousness to hydrothermal alteration is demonstrated by Whole-Rock Isotopic Compositions with those of regional MORB lavas (East Pacifi c the preservation of highly variable oxygen iso- Rise, and Alarcon basin), and they imply that topic compositions in detrital zircon crystals ε Relative to bulk earth, Nd is elevated for all the mantle source underneath the Salton Trough from the same depth intervals as the magmatic magmatic compositions with the most positive and Cerro Prieto basin is asthenospheric mantle , zircons from dikes and sills encountered in the values in basaltic xenoliths from Salton Buttes similar to the mantle underneath oceanic spread- Cerro Prieto and Salton Sea wells. The hetero- ε δ18 and microgabbros from Cerro Prieto ( Nd = +8.9; ing systems (cf. Lizarralde et al., 2007). Cerro geneity in O for detrital zircon contrasts with Fig. 9; Table 2). These values closely overlap Prieto and Roca Consag lava samples fall on the oxygen isotopic homogeneity of young

Geological Society of America Bulletin, November/December 2013 1843 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Schmitt et al.

subsurface lavas and xenoliths Salton Buttes lava lite and dacite lavas, trace-element compositions Salton Buttes Salton Buttes basaltic Cerro Prieto lava straddle the lower limit determined for continen- Cerro Prieto Salton Buttes granophyre Roca Consag lava tal crustal zircon, whereas zircons from Salton Alarcon lavas Buttes basaltic xenoliths and Cerro Prieto microgranites plot below the continental crustal ﬁ eld. EPR hydrothermal Roca Consag zircon crystals in the ca. 1 Ma alteration age population, by contrast, have trace-element

d compositions that are consistent with continental

ε crustal sources. This is also the case for detrital Nd

144 bulk earth zircon crystals from Cerro Prieto wells. Zircon /N 0.2 crystallization temperatures of 700–800 °C

Nd mixing AFC were estimated from Ti-in-zircon thermometry 0.4 (Fig. 13; Ferry and Watson, 2007). These are

143 0.6 temperature minima because they were calcu- meta- 0.8 lated for titanium oxide activity a = 1 and PRB granodiorite sediments TiO2 silica activity aSiO2 = 1. The presence of quartz

in the microgranites indicates aSiO2 = 1, but aTiO2 is subunity because of the absence of rutile. If, 87Sr/ 86 Sr for example, zircon crystallization would occur at aTiO2 = 0.5 and aSiO2 = 1, this would elevate Figure 9. Nd vs. Sr isotopic compositions of northern Gulf of Cali- zircon crystallization temperatures by ~70 °C fornia igneous rocks in comparison with East Paciﬁ c Rise (EPR) (Ferry and Watson, 2007). One exception is zir- and Alarcon lavas (Castillo et al., 2002, and references therein). con in quartz-free basaltic xenoliths; in this case,

Model curves for mixing and assimilation-fractional crystalliza- subunity aSiO2 would be compensatory with aTiO2 tion (AFC) are indicated for Peninsular Ranges Batholith (PRB) or <1. Regardless of these uncertainties, zircon Ti metasedimentary rocks (Herzig and Jacobs, 1994). Samples plotting abundances and model crystallization tempera- to the right of these curves are presumably affected by hydrother- tures are similar to those of oceanic zircons (Fig. mal alteration caused by 87Sr/86Sr exchange with country-rocks due 13). This is an important constraint on the con- 143 144 ditions of zircon crystallization in basalt-derived to high ﬂ uid mobility of Sr, whereas Nd/ Nd (εNd) remains largely constant. Data sources: this study, Herzig and Jacobs (1994), and melts that requires temperatures signiﬁ cantly Schmitt and Hulen (2008). below basalt liquidus temperatures.

DISCUSSION magmatic zircons in individual samples, indi- the source magma for the ca. 1 Ma Roca Consag Timing of Synrift Magmatism in cating the absence of oxygen isotopic exchange zircon population. the Northern Gulf of California between zircon and fl uids in the modern geo- Zircon trace elements (including Ti and thermal system (Fig. 12). Magmatic zircon REEs; Table DR3 [see footnote 1]) for Salton New and recently published (U-Th)/He, δ18O values for Salton Buttes lavas and basaltic Buttes (Schmitt and Vazquez, 2006) and Cerro U-Th, and U-Pb ages constrain late Pleistocene xenoliths (Schmitt and Vazquez, 2006) fall into a Prieto overlap with values for oceanic crustal to Holocene pulses of magmatism within the broad range averaging ~5‰ (±1‰; Fig. 12). This zircon (Fig. 11). In the case of zircon from rhyo- northern Gulf of California rift basins: Surface is broadly comparable to two zircons in Cerro Prieto lavas that were suffi ciently large to be analyzed for δ18O (Fig. 12). Salton Sea subsurface s s s s lavas (Schmitt and Hulen, 2008) are also similar e e c xc 18 x e e δ e n in their zircon O values (Fig. 12), and overall, % ili Figure 10. U-Th whole-rock ac- % 0 u 0 2 q these values overlap with those of oceanic crustal tivity ratios (indicated by paren- 4 e zircon (Grimes et al., 2011). Cerro Prieto micro- theses) in comparison with East EPR δ18 ~21°N granite zircons are strongly depleted in O, with Pacifi c Rise and mid-ocean-ridge average values of ~2‰. These values fall below basalt (MORB) compositions. 232 MORB field mantle values (5.3‰ ± 0.3‰; Valley, 2003) and A schematic aging trend for require substantial exchange with meteoric water MORB with 40% excess 230Th in the source rocks for the microgranite lavas, is indicated, but a similar trend Th)/ Th) ageing (~200 ka) and/or

230 crustal contamination consistent with similarly low values for Salton can result from contamination (( Buttes granophyre zircons (Schmitt and Vazquez, with secular equilibrium rocks. Salton Buttes lava 2006). Roca Consag Quaternary zircon averages Salton Buttes granophyre Note that Table 3 lists Cerro Cerro Prieto lava δ18 O = 6.0‰ (±0.9‰; Fig. 12), the highest values Prieto lava with low (234U)/(238U) for non-xenocrystic zircon present in the north- suggestive of alteration. ern Gulf of California samples, indicating a com- 238 232 paratively high proportion of continental crust in ((U)/ Th)

1844 Geological Society of America Bulletin, November/December 2013 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Northern Gulf of California oceanic magmatism

A In previous studies, older pulses (ca. 400– Salton Buttes 500 ka) of volcanic and intrusive magmatism have been documented in the Salton Trough continental field subsurface in close proximity to the Salton Buttes (Schmitt and Hulen, 2008). These older lava pulses, however, did not contribute to the zircon basaltic xenolith ocean crustal field population in the younger lavas and xenoliths. This implies that silicic intrusions are small in B Cerro Prieto volume and segregated. Alternatively, preexist- Figure 11. Zircon U vs. Yb ing zircon could have been resorbed by heating abundances for northern Gulf and addition of mafi c magmas in subsequent of California igneous rocks. lava intrusive episodes. An intriguing observation in Fields for continental and oce- subsurface this context is the abundance of ca. 1 Ma zircon anic zircon are from Grimes detrital crystals with a continental crustal signature in et al. (2007). U zircon [ppm] Roca Consag lavas. Roca Consag xeno crystic C Roca Consag (~1 Ma) zircon crystals lack signifi cant radiogenic 4He despite their age. Thus, they must have degassed, potentially during magmatic assimilation. The source of the dominant ca. 1 Ma component in Roca Consag lavas remains enig- matic in the absence of any rocks of equivalent age onshore. Quartz-phyric volcanic rocks with a K-Ar feldspar age of 1.4 ± 0.5 Ma have Yb zircon [ppm] been encountered in the Altar basin (northwestern Sonora) in an exploratory well at ~3.8 km depth (Pacheco et al., 2006). At present, we can volcanism in the Salton Buttes occurred only that were entrained by air-drilling, and (2) in only speculate that the ca. 1 Ma Roca Consag ~2.5 k.y. ago (Schmitt et al., 2013), the erup- lava samples, there is residual radiogenic 4He zircons originated from felsic magmas, which, tion of Cerro Prieto is <80 ka as indicated by the in pre-Quaternary zircons, which indicates that based on high δ18O and trace-element patterns least radiogenic 4He in xenocrystic zircon, and these crystals were only briefl y in contact with for zircon, had a more continental affi nity than Roca Consag erupted at ca. 40 ka (this study). the melt, and were likely entrained during the younger zircon-crystallizing magmas in the These ages for volcanism in these rift basins eruption. Salton and Cerro Prieto basins. This permits us imply a very recent heat pulse for the associated geothermal reservoirs. This has been previously proposed on the basis of detrital K-feldspar mantle: 5.0–5.6‰ thermochronology for the Salton Sea geother- A lava and xenoliths mal reservoir, indicating present-day peak temperatures that could not have been maintained for more than a few thousand years (Heizler and Salton Buttes Harrison, 1991). Moreover, young U-Th zircon Figure 12. Zircon δ18O com- lava positions for northern Gulf of basaltic xenolith crystallization ages from subsurface samples Cerro Prieto lava indicate that intrusive magmatism was coeval California igneous rocks (this Roca Cosag lava with eruptive activity, both within the Salton study; Schmitt and Vazquez, B subsurface Trough and Cerro Prieto basins. The combined 2006; Schmitt and Hulen, 2008). evidence from U-Th zircon crystallization ages Mantle-derived zircons (via reveals protracted magmatic activity within closed-system fractionation of these rift basins throughout the late Pleistocene mantle melts) fall in a narrow into the Holocene. compositional range around Salton Buttes With the exception of Roca Consag (see fol- 5.3‰ (Valley, 2003); deviations rank order Cerro Prieto lowing), all rocks studied here have a signifi - to lower values require exchange cant abundance of juvenile zircons. In samples with meteoric water, whereas C detrital where late Pleistocene to Holocene zircons are higher values are indicative of continental crustal origins. present, they lack xenocrystic cores. Although exchange with continental VSMOW—Vienna standard abundant older crystals exist in subsurface meteoric water crust samples and lavas such as Cerro Prieto, there mean ocean water. are two indications that these are not truly magmatic xenocrysts: (1) In the case of cuttings from subsurface igneous rocks, the abundance of older crystals correlates with the (visually 18 estimated) abundance of sediment fragments O zircon [‰ VSMOW]

Geological Society of America Bulletin, November/December 2013 1845 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Schmitt et al.

result from shallow contamination of the lavas during emplacement onto unconsolidated sedi- 2 Salton Buttes basaltic xenolith ment. Regardless of the ambiguity whether Cerro Prieto subsurface Roca Consag mixing and/or assimilation took place via con- detrital tinental crustal anatexis or entrainment of soft Ti zircon [ppm] sediments, binary mixing calculations indicate ocean crustal field that even the most contaminated lavas contain at most ~20% of a continental crustal end member (Fig. 9). In addition to fractional crystallization of MORB-type magmas with minor assimilation of continental crustal rocks, oxygen isotopes in zircon require another, previously largely overlooked, process in the genesis of rift-related 594 magmas in the northern Gulf of California. The strong depletion of δ18O in microgranitic zircons

Ti-in-zircon temperature (°C, aTiOTi-in-zircon = 1) ε coupled with high Nd indicate remelting of a δ18O zircon [‰ VSMOW] hydrothermal MORB-type crust. Thus, juvenile crust of oceanic affi nity must have exchanged Figure 13. Zircon δ18O vs. Ti-in-zircon temperatures for northern with meteoric waters during episodes of hydro- Gulf of California igneous rocks. Oceanic fi eld is from Grimes et al. thermal alteration, resulting in δ18O depletion (2011). VSMOW—Vienna standard mean ocean water. relative to mantle values. Remelting of this hydrothermally altered oceanic crust through reinjection of fresh mafi c magma is a viable to bracket the onset of magmatism with oceanic solidated sediments: 87Sr/86Sr in particular is scenario to generate rhyolitic melts that directly δ18 ε affi nity in the northern Gulf of California vulnerable to contamination because Sr typi- inherit the low O and MORB-like Nd of their between ca. 1 Ma (“continental crustal” zircons cally decreases with differentiation, and thus source. Such low δ18O rhyolites are common in in Roca Consag) and ca. 0.5 Ma (“oceanic” zir- differentiated rocks are more easily affected oceanic rifts such as Iceland (e.g., Martin and cons in buried Salton Buttes lavas), although by entrainment of crustal xenoliths or sec- Sigmarsson, 2007), and some oceanic plagio- we acknowledge that this is highly tentative ondary alteration. Consequently, we interpret granites also display depleted δ18O values because of the very limited availability of igne- the strong displacement of subsurface rocks to (Grimes et al., 2011). ε δ18 ous rocks for this critical age range. the right of plausible magmatic mixing/assimi- The combination of high Nd and low O 87 86 ε lation trends in the Sr/ Sr versus Nd diagram clearly rules out anatexis of any plausible conti- MORB versus Continental Crustal Sources (Fig. 9; Schmitt and Hulen, 2008) as a result of nental crustal sources for Salton Sea and Cerro Sr-isotopic exchange between rock and hydro- Prieto felsic magmas (Figs. 10 and 12). This is Geophysical models remain ambiguous with thermal fl uid. Hence, for constraining magma further supported by zircon trace elements: Low regard to the state of continental rupture and sources, we restrict our discussion in the fol- U relative to Yb distinguishes most Cerro Prieto oceanic spreading in the northern Gulf of Cali- lowing to fl uid-insensitive indicators such as microgranite zircons and those from Salton fornia: Extrapolation of modern plate velocities whole-rock Nd isotopes, and trace elements Buttes basaltic xenoliths from typical continen- implies largely uniform amounts of extension and oxygen isotopes in magmatic zircons. tal crustal zircon (Fig. 11). We emphasize that ε δ18 along the entire rift zone, but direct evidence Salton Buttes and Cerro Prieto have Nd val- the zircon evidence ( O and trace elements) for Pliocene–Pleistocene oceanic crust for- ues in the least differentiated (basaltic) rocks is crucial for detecting this mechanism because mation such as magnetic striping is lacking that are strikingly similar to East Pacifi c Rise zircon records a primary magmatic signal that (Lonsdale, 1989). Early proponents of oceanic and Alarcon lavas, where seafl oor spreading is is not overprinted by alteration in the modern crustal formation in the northern Gulf of Cali- unambiguous (Fig. 9). This suggests that basal- hydrothermal system. By contrast, whole-rock fornia have reasoned, on the basis of composi- tic magmas along strike of the Gulf of California enrichment patterns in fl uid-mobile trace ele- tional and isotopic similarities between MORB share a common asthenospheric mantle source. ments are less reliable, and could be errone- and basaltic xenoliths in the Salton Buttes, that Even some of the most felsic rocks have Nd- ously interpreted as inherited from a hydrated rifting has proceeded there to a stage of oce- isotopic compositions that are identical to those mantle source, similar to that of subduction anic crust formation (Herzig and Jacobs, 1994; of basalts (Fig. 9). Although some of the inter- zones (Fig. 4). Based on whole-rock data alone, Robinson et al., 1976). The origins of interme- mediate and felsic rocks are displaced to lower it thus would be diffi cult to distinguish between ε diate to silicic magmas and their relationships Nd values relative to regional mantle-derived source enrichment (i.e., due to fl uid addi- to basaltic magmas, however, have remained rocks, they lack a clear signature for anatexis tion from a subducted slab) and hydrothermal ambiguous. Crustal melting has been previ- of continental crust as a signifi cant contributor alteration of basaltic crustal precursor rocks ously proposed (e.g., Reed et al., 1984), but to rift-related magmatism in the northern Gulf as the cause of the enrichment of fl uid-mobile this was based on geochemical indicators that of California (Fig. 9). Because xenocrystic zir- trace elements (Fig. 4). By contrast, zircon geo- subsequently proved to be unreliable, espe- cons in these lavas frequently retain pre-eruptive chemistry is indicative of a MORB-like source, cially for rocks for which the composition may 4He, caution has to be exercised in interpreting which is characterized by lower U/Yb com- ε have been altered by hydrothermal activity or shifts to lower Nd as true magmatic mixing or pared to arc-like and continental crustal rocks contaminated during emplacement in uncon- assimilation trends. Instead, this trend could (Fig. 11).

1846 Geological Society of America Bulletin, November/December 2013 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Northern Gulf of California oceanic magmatism

Felsic Magmas in Incipient Oceanic sheared continental Spreading Centers with Thick crust, metasediments and mafic intrusions Evolved surface volcanism Sedimentary Cover WEand geothermal activity 0 Having established a dominantly MORB- Meteoric water Sedimentation and Peninsular type source for maﬁ c magmas in the northern metamorphism Cratonic Ranges Gulf of California from Nd-isotopic evidence, Crust 10 Batholith we now discuss the styles of differentiation in Hydrothermal the context of other occurrences of differenti- alteration & remelting ated magmas in mid-ocean-ridge environments. There, basaltic magmas are known to undergo 20 signiﬁ cant differentiation, which is recorded by MORB-type evolved intrusive rocks in oceanic crustal seg- magma ascent and intrusion ments (plagiogranites). Only occasionally and

Depth (km) in specifi c tectonic settings, such as propagat- 30 ing rift tips or rift-transform intersections, have felsic mid-ocean-ridge lavas been documented intermediate to rhyolite (e.g., Schmitt et al., 2011; Wanless et al., 2010). 40 basalt In the Gulf of California province, oceanic spreading occurs in short (tens of kilometers) Asthenospheric mantle upwelling segments separated by transform faults, which vertical exaggeration ~2:1 is a broadly similar setting to rift-transform 50 intersections where felsic mid-ocean-ridge 0100Distance (km) lavas have erupted (e.g., Juan de Fuca; Schmitt et al., 2011). Fractional crystallization of paren- Figure 14. Schematic cross section illustrating the magmatic processes tal basalt is a potential mechanism to generate in northern Gulf of California rift basins. Active magmatism involves mid-ocean-ridge basalt (MORB)–type parental magmas derived SiO2-rich magmas, but this is inconsistent with from partial melting of decompressing asthenosphere, with inter- geochemical data such as the observed K2O enrichments, and oxygen isotopic depletion, mediate and rhyolitic magmas generated by fractional crystalli- which cannot be explained by closed-system zation and remelting of hydrothermally altered mafi c intrusions. differentiation of pristine MORB. The presence Only late Pleistocene (ca. 1 Ma) zircons from Roca Consag indicate of rhyolite melt pockets in basaltic xenoliths minor melting and/or assimilation of continental crust. Underlying from the Salton Buttes is direct petrographic model of crustal structure from gravity variations is modifi ed from evidence that remelting of hydrated (amphi- Parsons and McCarthy (1996). bole-bearing) mafi c rocks is viable for producing felsic melts (Robinson et al., 1976). Felsic melt pockets are also highly amenable to zircon on average lower than mantle. Unequivocally 3H), which can be explained by AFC involv- crystallization (Schmitt and Vazquez, 2006). low-δ18O zircons, such as crystals in Cerro Prieto ing minor amounts of continental crustal rocks. A comparison with glass compositions experi- microgranites and Salton Buttes granophyres, Binary mixing between high-silica rhyolites mentally produced for partial melting of basalt however, require a low-δ18O source compared to and basalts can be ruled out because the inter- in the presence of H2O indicates that remelting mantle-derived magmas. Isotopically light oxy- mediate compositions deviate from isotopic and

occurred under fairly low H2O (and total) pres- gen typical for near-surface waters must have trace-element mixing trajectories (Fig. 3). The ε sures, likely <170 MPa (Fig. 3B; Thy et al., infi ltrated a mafi c source (inferred from high Nd values for intermediate rocks from Cerro ε 1990). This places the region of basaltic intru- Nd) prior to remelting and differentiation. Prieto and Salton Sea overlap, suggesting that sion and remelting at a depth shallower than Thus far, we have identifi ed two key similari- the overall amount of assimilation is similar in ~7 km, close to the base of low-density and low- ties for magmas in the Salton Trough and Cerro both systems, although Cerro Prieto lavas have ε 4 seismic-velocity basin fi ll (Fig. 14; Fuis et al., Prieto (Roca Consag lavas appear to be distinct, lower Nd. The presence of only partially He- 1984; Parsons and McCarthy, 1996). but because of limited exposure and sampling, degassed detrital zircon crystals in Cerro Prieto Given the limited range in δ18O of pristine this is a tentative conclusion): Mafi c end mem- lavas, however, indicates that at least some MORB magmas, fractional crystallization is bers are similar and MORB-like, and melting continental crustal material was incorporated expected to produce intermediate to rhyolitic of a MORB-type crust has occurred subsequent late, and not via anatectic assimilation at depth magmas that are only moderately elevated to hydrothermal alteration, producing high- (Fig. 5). This urges caution in using whole-rock (by ~1‰–2‰) in δ18O relative to the parental silica rhyolite magmas. As noted previously, analyses, which may have become contami- magma compositions, translating into a narrow a conspicuous distinction between both mag- nated upon ascent and emplacement at shallow range of mantle-derived zircon (Valley, 2003) matic systems is that magma compositions are depth, in particular during magma interaction when accounting for oxygen isotopic fraction- strongly bimodal in the Salton Trough, whereas with unconsolidated sediment. ation between zircon and melt (~1‰ at 800 °C; intermediate intrusive and extrusive rocks exist There is strong evidence based on oxygen Trail et al., 2009). Magmatic compositions of at Cerro Prieto. Cerro Prieto intermediate rocks and Nd isotopes for remelting of young MORB- Cerro Prieto and Salton Buttes lavas marginally (barring E-30 3027 m microgranite) show a type crust as a source for felsic magmas, as δ18 ε ε overlap with this range, but their O values are trend to lower Nd with increasing SiO2 (Fig. indicated by the horizontal trend in Nd versus

Geological Society of America Bulletin, November/December 2013 1847 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Schmitt et al.

SiO (Fig. 3). Conclusive evidence for continen- tion and the solidus of mafi c rocks, which read- Benoit, M., Aguillon-Robles, A., Calmus, T., Maury, R.C., 2 Bellon, H., Cotten, J., Bourgois, J., and Michaud, F., tal crustal anatexis as a signifi cant source for ily exchange oxygen with fl uids. Remelting of 2002, Geochemical diversity of Miocene volcanism magmatism in the northern Gulf of California is newly formed oceanic-type crust underlying in southern Baja California, Mexico: Implication absent, despite minor degrees of crustal assimi- the sediments by intrusion of mantle-derived of mantle and crustal sources during the opening of an asthenospheric window: The Journal of Geol- lation in some of the felsic lavas. This conclu- basalts is the dominant origin of felsic mag- ogy, v. 110, p. 627–648, doi:10.1086/342735. sion holds for post–1 Ma magmas, whereas ca. mas in this rift. Continental crustal sources, or Bigioggero, B., Chiesa, S., Zanchi, A., Montrasio, A., and Vez- 1 Ma xenocrystic zircons in Roca Consag lavas subduction-metasomatized mantle, thus contrib- zoli, L., 1995, The Cerro Mencenares volcanic center, Baja California Sur; source and tectonic control on post- display some continental affi nity. Even where ute only marginally, if any, to melt generation. subduction magmatism within the Gulf Rift: Geologi- ε cal Society of America Bulletin, v. 107, p. 1108–1122, minor assimilation is indicated by lower Nd In this respect, the lower crust in the rift basins doi:10.1130/0016-7606(1995)107<1108:TCMVCB>2.3 values, this would at most amount to an ~20% of the northern Gulf of California—albeit dis- .CO;2. crustal contribution (Fig. 9). Major elemental tinct from typical oceanic crustal sequences by Biswas, S., Coutand, I., Grujic, D., Hager, C., Stockli, D., compositional differences between Salton Sea a thick overburden of sediments—has become and Grasemann, B., 2007, Exhumation and uplift of the Shillong Plateau and its infl uence on the eastern and Cerro Prieto exist, but they are of secondary fundamentally oceanic in nature. This study Himalayas: New constraints from apatite and zircon relevance with regard to the main mechanisms emphasizes the importance of single-crystal zir- (U-Th-[Sm])/He and apatite fi ssion track analyses: of magma production in the mantle, and remelt- con analysis to see through the effects of near- Tectonics, v. 26, TC6013, doi:10.1029/2007TC002125. Blondes, M.S., Reiners, P.W., Edwards, B.R., and Biscontini, ing of hydrothermally altered mafi c crust. surface contamination by soft sediment–magma A., 2007, Dating young basalt eruptions by (U-Th)/He The deeper parts of the basin crust in the interaction, and pervasive hydrothermal altera- on xenolithic zircons: Geology, v. 35, p. 17–20, doi: 10.1130/G22956A.1. northern Gulf of California are postulated to tion, which are typical for rift basins fi lled by Bryan, S.E., and Ernst, R.E., 2008, Revised defi nition of be dominated by basaltic intrusions, yet unlike continent-derived detritus. large igneous provinces (LIPs): Earth-Science Re- basins in the southern Gulf of California, sur- views, v. 86, p. 175–202, doi:10.1016/j.earscirev.2007 ACKNOWLEDGMENTS .08.008. fi cial basalts are absent. Thick sedimentary Calmus, T., Pallares, C., Maury, R.C., Aguillon-Robles, A., blanketing thus plays an important role for We thank Julio Alvarez Rosales and staff at Bellon, H., Benoit, M., and Michaud, F., 2011, Vol- magmatic differentiation in rifts (Fig. 14), but Comisión Federal de Electricidad for their support in canic markers of the post-subduction evolution of Baja obtaining well cuttings from Cerro Prieto. Michael California and Sonora, Mexico: Slab tearing versus not for basaltic magma generation, which is Huh is thanked for assistance with petrographic log- lithospheric rupture of the Gulf of California: Pure and compositionally homogeneous along the entire ging of well cuttings and fused bead preparation. We Applied Geophysics, v. 168, no. 8–9, p. 1303–1330. extent of the Gulf of California rift zone (cf. thank Gabriel Rendón for zircon mineral separation Capra, L., Macias, J.L., Espindola, J.M., and Siebe, C., and Pat Castillo for whole-rock Sr-isotope analysis of 1998, Holocene Plinian eruption of La Virgen Volcano, Lizarralde et al., 2007). A major effect of the Baja California, Mexico: Journal of Volcanology and Cerro Prieto and Roca Consag lavas. We also thank sedimentary basin infi ll is that it controls the Geothermal Research, v. 80, p. 239–266, doi:10.1016 Peter Schaaf and Gabriela Solís Pichardo, both at /S0377-0273(97)00049-8. neutral buoyancy level of magmas. Felsic mag- LUGIS , UNAM, for assistance with running the Finni- Castillo, P.R., 2008, Origin of the adakite–high-Nb basalt mas have lower density and tend to ascend to gan MAT262. We are also thankful for comments by asso ciation and its implications for postsubduction shallower levels, whereas basaltic melts reach reviewers Andy Barth, Teresa Orozco-Esquivel, and magmatism in Baja California, Mexico: Geological Society of America Bulletin, v. 120, p. 451–462, doi: neutral buoyancy at deeper levels within the Associate Editor Brian McConnell, as well as Sci- ence Editor Nancy Riggs. This work was supported 10.1130/B26166.1. thick sedimentary pile. This is consistent with by grants through UC-MEXUS CN 07-60 and Na- Castillo, P.R., Hawkins, J.W., Lonsdale, P.F., Hilton, D.R., the presence of intermediate to felsic compo- Shaw, A.M., and Glascock, M.D., 2002, Petrology of tional Science Foundation MARGINS (grant EAR- Alarcon Rise lavas, Gulf of California: Nascent intra- sitions of samples collected both in subaerial 0948162). The ion microprobe facility at the University continental ocean crust: Journal of Geophysical Re- and submarine volcanoes in the northern Gulf of California–Los Angeles is partly supported by a search–Solid Earth, v. 107, no. B10, 2222, doi:10.1029 grant from the Instrumentation and Facilities Program, /2001JB000666. of California. By contrast, basaltic to andesitic Division of Earth Sciences, National Science Founda- Cheng, H., Edwards, R.L., Hoff, J., Gallup, C.D., Rich- eruptions have only been reported in the Balle- tion. U-Th isotopic determinations were supported by ards, D.A., and Asmerom, Y., 2000, The half-lives nas Channel and the lower Delfi n basin, where Taiwan ROC National Science Council grants (NSC of uranium-234 and thorium-230: Chemical Geol- 100-2116-M-002-009, 101-2116-M-002-009, and ogy, v. 169, p. 17–33, doi:10.1016/S0009-2541 sedimentary deposits are thin (Martín et al., (99)00157-6. NTU 102R7625 to Chuan-Chou Shen). 2013). As a second effect of high sedimenta- Einsele, G., Gieskes, J.M., Curray, J., Moore, D.M., Aguayo, tion rates, we propose that thermal insulation REFERENCES CITED E., Aubry, M.-P., Fornari, D., Guerrero, J., Kastner, M., Kelts, K., Lyle, M., Matoba, Y., Molina-Cruz, by sediments facilitates subsequent remelting of Aguillon-Robles, A., Calmus, T., Benoit, M., Bellon, H., A., Niemitz , J., Rueda, J., Saunders, A., Schrader, H., basaltic intrusions, which is aided by hydration Maury, R.O., Cotten, J., Bourgois, J., and Michaud, F., Simoneit , B., and Vacquier, V., 1980, Intrusion of ba- during low-temperature water-rock interactions, 2001, Late Miocene adakites and Nb-enriched basalts saltic sills into highly porous sediments, and resulting from Vizcaino Peninsula, Mexico: Indicators of East hydrothermal activity: Nature, v. 283, no. 5746, p. 441– δ18 also lowering the O of the rocks. Pacifi c Rise subduction below southern Baja Califor- 445, doi:10.1038/283441a0. nia?: Geology, v. 29, p. 531–534, doi:10.1130/0091 Elders, W.A., Rex, R.W., Meidav, T., Robinson, P.T., and CONCLUSIONS -7613(2001)029<0531:LMAANE>2.0.CO;2. Biehler, S., 1972, Crustal spreading in Southern Anovitz, L.M., Elam, J.M., Riciputi, L.R., and Cole, California: Science, v. 178, no. 4056, p. 15–24, doi: D.R., 1999, The failure of obsidian hydration dating: 10.1126/science.178.4056.15. Late Pleistocene to Holocene magmatism in Sources, implications, and new directions: Journal Farley, K.A., Wolf, R.A., and Silver, L.T., 1996, The effects of Archaeological Science, v. 26, p. 735–752, doi: of long alpha-stopping distances on (U-Th)/He ages: the northern Gulf of California rift zone occurs 10.1006/jasc.1998.0342. Geochimica et Cosmochimica Acta, v. 60, p. 4223– in an environment characterized by high deposi- Batiza, R., 1978, Geology, petrology, and geochemistry 4229, doi:10.1016/S0016-7037(96)00193-7. tion rates of continent-derived detritus in narrow of Isla Tortuga, a recently formed tholeiitic island in Farley, K.A., Kohn, B.P., and Pillans, B., 2002, The effects Gulf of California: Geological Society of America of secular disequilibrium on (U-Th)/He systematics rift basins. Intrusion of this sedimentary pack- Bulletin, v. 89, p. 1309–1324, doi:10.1130/0016-7606 and dating of Quaternary volcanic zircon and apatite: age by mafi c to silicic magmas causes thermal (1978)89<1309:GPAGOI>2.0.CO;2. Earth and Planetary Science Letters, v. 201, p. 117– overprint and metamorphism, but falls short of Bellon, H., Aguillon-Robles, A., Calmus, T., Maury, R.C., 125, doi:10.1016/S0012-821X(02)00659-3. Bourgois, J., and Cotten, J., 2006, La Purisirna volcanic Ferrari, L., Lopez-Martinez, M., Aguirre-Diaz, G., and generating voluminous anatexis of sediments. fi eld, Baja California Sur (Mexico): Miocene to Qua- Carrasco-Nunez, G., 1999, Space-time patterns of Hydration of juvenile mafi c crust via deep ternary volcanism related to subduction and opening Ceno zoic arc volcanism in central Mexico: From the of an asthenospheric window: Journal of Volcanology Sierra Madre Occidental to the Mexican volcanic belt: circulation of meteoric fl uids occurs in these and Geothermal Research, v. 152, no. 3–4, p. 253–272, Geology, v. 27, p. 303–306, doi:10.1130/0091-7613 basins, lowering the oxygen isotopic composi- doi:10.1016/j.jvolgeores.2005.10.005. (1999)027<0303:STPOCA>2.3.CO;2.

1848 Geological Society of America Bulletin, November/December 2013 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Northern Gulf of California oceanic magmatism

Ferry, J.M., and Watson, E.B., 2007, New thermodynamic ogy, v. 22, p. 991–994, doi:10.1130/0091-7613 phosed deltaic lacustrine sediments; the Salton Sea models and revised calibrations for the Ti-in-zircon (1994)022<0991:CVATSE>2.3.CO;2. geothermal system, California: Economic Geol ogy and Zr-in-rutile thermometers: Contributions to Min- Hurtado-Brito, J.C., 2012, El Registro Volcánico en las and the Bulletin of the Society of Economic Geolo- eralogy and Petrology, v. 154, p. 429–437, doi:10.1007 Cuencas Rift del Norte del Golfo de California a Partir gists, v. 83, no. 3, p. 511–523, doi:10.2113/gsecongeo /s00410-007-0201-0. de Sísmica de Refl exión [Master’s thesis]: Ensenada, .83.3.511. France, L., Koepke, J., Ildefonse, B., Cichy, S.B., and De- Mexico, CICESE, 100 p. Monteleone, B.D., Baldwin, S.L., Webb, L.E., Fitzger- schamps, F., 2010, Hydrous partial melting in the Jaffey, A.H., Flynn, K.F., Glendenin, L.E., Bentley, W.C., ald, P.G., Grove, M., and Schmitt, A.K., 2007, Late sheeted dike complex at fast spreading ridges: Ex- and Essling, A.M., 1971, Precision measurement of Miocene–Pliocene eclogite facies metamorphism, perimental and natural observations: Contributions to half-lives and specifi c activities of 235U and 238U: Physi- D’Entrecasteaux Islands, SE Papua New Guinea: Mineralogy and Petrology, v. 160, p. 683–704, doi: cal Review C: Nuclear Physics, v. 4, p. 1889–1906, Journal of Metamorphic Geology, v. 25, p. 245–265, 10.1007/s00410-010-0502-6. doi:10.1103/PhysRevC.4.1889. doi:10.1111/j.1525-1314.2006.00685.x. Friedman, I., and Obradovich, J., 1981, Obsidian hydration dat- Kasameyer, P.W., and Hearst, J.R., 1988, Borehole grav- Mora-Klepeis, G., and McDowell, F.W., 2004, Late Miocene ing of volcanic events: Quaternary Research, v. 16, p. 37– ity measurements in the Salton Sea Scientifi c Drill- calc-alkalic volcanism in northwestern Mexico: An 47, doi:10.1016/0033-5894(81)90126-5. ing Project well state 2–14: Journal of Geophysical expression of rift or subduction-related magmatism?: Fuis, G.S., Mooney, W.D., Healy, J.H., McMechan, G.A., Research–Solid Earth and Planets, v. 93, p. 13,037– Journal of South American Earth Sciences, v. 17, p. 297– and Lutter, W.J., 1984, A seismic refraction survey 13,045, doi:10.1029/JB093iB11p13037. 310, doi:10.1016/j.jsames.2004.08.001. of the Imperial Valley region, California: Journal of Kelley, V.C., and Soske, J.L., 1936, Origin of the Salton Negrete-Aranda, R., and Canon-Tapia, E., 2008, Post- Geophysical Research, ser. B, v. 89, p. 1165–1189, volcanic domes, Salton Sea, California: The Journal of subduction volcanism in the Baja California Penin- doi:10.1029/JB089iB02p01165. Geology, v. 44, p. 496–509, doi:10.1086/624444. sula, Mexico: The effects of tectonic reconfi guration Gibson, S.A., Thompson, R.N., Weska, R.K., Dickin, A.P., Krogh, T.E., 1973, A low-contamination method for hydro- in volcanic systems: Lithos, v. 102, p. 392–414, doi: and Leonardos, O.H., 1997, Late Cretaceous rift-related thermal decomposition of zircon and extraction of U 10.1016/j.lithos.2007.08.013. upwelling and melting of the Trindade starting mantle and Pb for isotopic age determinations: Geochimica et Nicholls, I.A., 1974, Direct fusion method of preparing sili- plume head beneath western Brazil: Contributions to Cosmochimica Acta, v. 37, p. 485–494. cate rock glasses for energy-dispersive electron-micro- Mineralogy and Petrology, v. 126, no. 3, p. 303–314, Liu, Y.S., Hu, Z.C., Zong, K.Q., Gao, C.G., Gao, S., Xu, probe analysis: Chemical Geology, v. 14, p. 151–157, doi:10.1007/s004100050252. J.A., and Chen, H.H., 2010, Reappraisement and doi:10.1016/0009-2541(74)90124-7. González-Escobar, M., Aguilar-Campos, C., Suarez- refinement of zircon U-Pb isotope and trace ele- Niu, Y.L., Collerson, K.D., Batiza, R., Wendt, J.I., and Vidal, F., and Martin-Barajas, A., 2009, Geom- ment analyses by LA-ICP-MS: Chinese Science Regelous, M., 1999, Origin of enriched-type mid- etry of the Wagner basin, upper Gulf of California Bulletin, v. 55, no. 15, p. 1535–1546, doi:10.1007/ ocean ridge basalt at ridges far from mantle plumes: based on seismic reflections: International Geol- s11434-010-3052-4. The East Pacifi c Rise at 11 degrees 20′N: Journal of ogy Review, v. 51, no. 2, p. 133–144, doi:10.1080 Lizarralde, D., Axen, G.J., Brown, H.E., Fletcher, J.M., Gon- Geophysical Research–Solid Earth, v. 104, p. 7067– /00206810802615124. zalez-Fernandez, A., Harding, A.J., Holbrook, W.S., 7087, doi:10.1029/1998JB900037. Gonzalez-Escobar, M., Suarez-Vidal, F., Hernandez-Perez, Kent, G.M., Paramo, P., Sutherland, F., and Umhoefer, Oskin, M., and Stock, J., 2003, Pacifi c–North America plate J.A., and Martin-Barajas, A., 2010, Seismic refl ection– P.J., 2007, Variation in styles of rifting in the Gulf of motion and opening of the Upper Delfi n basin, north- based evidence of a transfer zone between the Wagner California: Nature, v. 448, p. 466–469, doi:10.1038 ern Gulf of California, Mexico: Geological Society and Consag basins: Implications for defi ning the struc- /nature06035. of America Bulletin, v. 115, no. 10, p. 1173–1190, tural geometry of the northern Gulf of California: Geo- Lonsdale, P.F., 1989, Geology and tectonic history of doi:10.1130/B25154.1. Marine Letters, v. 30, no. 6, p. 575–584, doi:10.1007 the Gulf of California, in Winterer, E.L., Hussong, Pacheco, M., Martín-Barajas, A., Elders, W., Espinosa- /s00367-010-0204-0. D.M., and Decker, R.W., eds., The Eastern Pacifi c Cardeña, J.M., Helenes, J., and Segura, A., 2006, Grimes, C.B., John, B.E., Kelemen, P.B., Mazdab, F.K., Ocean and Hawaii: Boulder, Colorado, Geological Stratigraphy and structure of the Altar basin of NW Wooden, J.L., Cheadle, M.J., Hanghoj, K., and Society of America, The Geology of North Amer- Sonora: Implications for the history of the Colorado Schwartz, J.J., 2007, Trace element chemistry of zircons ica, v. N, p. 499–521. River delta and the Salton Trough: Revista Mexicana from oceanic crust; a method for distinguishing detrital Martín, A., Fletcher, J.M., Lopez-Martinez, M., and Men- de Ciencias Geológicas, v. 23, no. 1, p. 1–22. zircon provenance: Geology, v. 35, no. 7, p. 643–646, doza-Borunda, R., 2000, Waning Miocene subduction Parsons, T., and McCarthy, J., 1996, Crustal and upper doi:10.1130/G23603A.1. and arc volcanism in Baja California: The San Luis mantle velocity structure of the Salton Trough, south- Grimes, C.B., Ushikubo, T., John, B.E., and Valley, J.W., Gonzaga volcanic fi eld: Tectonophysics, v. 318, p. 27– east California: Tectonics, v. 15, no. 2, p. 456–471, 2011, Uniformly mantle-like δ18O in zircons from oce- 51, doi:10.1016/S0040-1951(99)00305-4. doi:10.1029/95TC02616. anic plagiogranites and gabbros: Contributions to Min- Martín, A., Hurtado, J.C., Cañon, E., Weber, B., and Schmitt, Paz-Moreno, F.A., and Demant, A., 1999, The Recent Isla eralogy and Petrology, v. 161, p. 13–33, doi:10.1007 A.K., 2013, Extent and composition of the new crust San Luis volcanic centre: Petrology of a rift-related /s00410-010-0519-x. beneath rift basins in the northern Gulf of California: volcanic suite in the northern Gulf of California, Grove, M., Jacobson, C.E., Barth, A.P., and Vucic, A., 2003, Geological Society of America Abstracts with Pro- Mexico: Journal of Volcanology and Geothermal Temporal and spatial trends of Late Cretaceous–early grams, v. 45, no. 6, p. 22. Research, v. 93, p. 31–52, doi:10.1016/S0377-0273 Tertiary underplating of Pelona and related schist be- Martin, E., and Sigmarsson, O., 2007, Crustal thermal (99)00083-9. neath Southern California and southwestern Arizona, state and origin of silicic magma in Iceland: The Pearce, N.J.G., Perkins, W.T., Westgate, J.A., Gorton, M.P., in Johnson, S.E., Paterson, S.R., Fletcher, J.M., Girty, case of Torfajokull, Ljosufjoll and Snaefellsjokull Jackson, S.E., Neal, C.R., and Chenery, S.P., 1997, G.H., Kimbrough, D.L., and Martín-Barajas, A., eds., volcanoes: Contributions to Mineralogy and Petrol- A compilation of new and published major and trace Tectonic Evolution of Northwestern México and the ogy, v. 153, p. 593–605, doi:10.1007/s00410-006 element data for NIST SRM 610 and NIST SRM Southwestern USA: Geological Society of America -0165-5. 612 glass reference materials: Geostandards News- Special Paper 374, p. 381–406. Martín-Barajas, A., and Weber, B., 2003, Petrology of Plio- letter—The Journal of Geostandards and Geoanaly- Hausback, B.P., 1984, Cenozoic volcanic and tectonic evolu- Pleistocene volcanic rocks in the northern Gulf of sis, v. 21, no. 1, p. 115–144, doi:10.1111/j.1751-908X tion of Baja California Sur, in Frizzell, V., ed., Geology California rift system; insights into crustal composi- .1997.tb00538.x. of the Baja California Peninsula: Society of Economic tion: Geological Society of America Abstracts with Persaud, P., Stock, J.M., Steckler, M.S., Martín-Barajas, A., Paleontologists and Mineralogists (SEPM), Pacific Programs, v. 35, no. 4, p. 28–29. Diebold, J.B., González-Fernández, A., and Moun- Section, Bakersfi eld, California, p. 219–236. Martín-Barajas, A., Stock, J.M., Layer, P., Hausback, B., tain, G.S., 2003, Active deformation and shallow Heizler, M.T., and Harrison, T.M., 1991, The heating du- Renne, P., and Lopez-Martinez, M., 1995, Arc-rift structure of the Wagner, Consag, and Delfín Basins, ration and provenance age of rocks in the Salton Sea transition volcanism in the Puertecitos Volcanic Prov- northern Gulf of California, Mexico: Journal of Geo- geothermal fi eld, Southern California: Journal of Vol- ince, northeastern Baja-California, Mexico: Geological physical Research: Solid Earth, v. 108, no. B7, 2355, canology and Geothermal Research, v. 46, p. 73–97, Society of America Bulletin, v. 107, p. 407–424, doi: doi:10.1029/2002JB001937. doi:10.1016/0377-0273(91)90077-D. 10.1130/0016-7606(1995)107<0407:ARTVIT>2.3 Reed, M.J., Wright, R., and Richards, R.G., 1984, Relation- Herzig, C.T., 1990, Geochemistry of igneous rocks from .CO;2. ship between volcanism and hydrothermal activity at the Cerro Prieto geothermal fi eld, northern Baja Cali- Martín-Barajas, A., Weber, B., Schmitt, A.K., and Lons- Cerro Prieto, Mexico: Transactions, Geothermal Re- fornia, Mexico: Journal of Volcanology and Geother- dale, P., 2008, Recent rift volcanism in the northern sources Council, v. 8, p. 217–221. mal Research, v. 42, p. 261–271, doi:10.1016/0377 Gulf of California and the Salton Trough: Why a Reid, M.R., Coath, C.D., Harrison, T.M., and McKeegan, -0273(90)90003-X. preponderance of evolved magmas?: San Francisco, K.D., 1997, Prolonged residence times for the young- Herzig, C.T., and Elders, W.A., 1988, Nature and signifi cance American Geophysical Union, Fall Meeting, abstract est rhyolites associated with Long Valley Caldera; of igneous rocks cored in the state 2–14 research bore- T11A–1841. 230Th-238U ion microprobe dating of young zircons: hole, Salton Sea Scientifi c Drilling Project, California: McDowell, S.D., and Elders, W.A., 1980, Authigenic layer Earth and Planetary Science Letters, v. 150, p. 27–39, Journal of Geophysical Research–Solid Earth and silicate minerals in borehole Elmore-1, Salton Sea geo- doi:10.1016/S0012-821X(97)00077-0. Planets, v. 93, no. B11, p. 13,069–13,080, doi:10.1029 thermal fi eld, California, USA: Contributions to Min- Robinson, P.T., Elders, W.A., and Muffler, L.J.P., 1976, /JB093iB11p13069. eralogy and Petrology, v. 74, p. 293–310, doi:10.1007 Quaternary volcanism in Salton Sea geothermal fi eld, Herzig, C.T., and Jacobs, D.C., 1994, Cenozoic vol canism /BF00371699. Imperial Valley, California: Geological Society of and two-stage extension in the Salton Trough, South- McKibben, M.A., Andes, J.P., and Williams, A.E., 1988, America Bulletin, v. 87, p. 347–360, doi:10.1130 ern California and northern Baja California: Geol- Active ore formation at a brine interface in metamor- /0016-7606(1976)87<347:QVITSS>2.0.CO;2.

Geological Society of America Bulletin, November/December 2013 1849 Downloaded from gsabulletin.gsapubs.org on January 15, 2014 Schmitt et al.

Sawlan, M.G., 1991, Magmatic evolution of the Gulf of Salton Trough (California): Geology, v. 41, p. 7–10, Mexico: Sedimentary Geology, v. 144, p. 125–147, Cali fornia rift, in Dauphin, J.P., and Simoneit, B.R., doi:10.1130/G33634.1. doi:10.1016/S0037-0738(01)00138-5. eds., The Gulf and Peninsular Province of the Califor- Shen, C.-C., Lawrence Edwards, R., Cheng, H., Dorale, Valley, J.W., 2003, Oxygen isotopes in zircon: Reviews in nias: American Association of Petroleum Geologists J.A., Thomas, R.B., Bradley Moran, S., Weinstein, Mineralogy and Geochemistry, v. 53, p. 343–385, Memoir 47, p. 301–369. S.E., and Edmonds, H.N., 2002, Uranium and thorium doi:10.2113/0530343. Schmitt, A.K., and Hulen, J.B., 2008, Buried rhyolites isotopic and concentration measurements by magnetic Vazquez, J.A., and Reid, M.R., 2004, Probing the accumu- within the active, high-temperature Salton Sea geo- sector inductively coupled plasma mass spectrometry: lation history of the voluminous Toba magma: Sci- thermal system: Journal of Volcanology and Geo- Chemical Geology, v. 185, p. 165–178, doi:10.1016 ence, v. 305, p. 991–994, doi:10.1126/science.1096994. thermal Research, v. 178, p. 708–718, doi:10.1016 /S0009-2541(01)00404-1. Vidal-Solano, J.R., Demant, A., Moreno, F.A.P., Lapierre, /j.jvolgeores.2008.09.001. Shen, C.-C., Cheng, H., Edwards, R.L., Moran, S.B., Ed- H., Ortega-Rivera, M.A., and Lee, J.K.W., 2008, In- Schmitt, A.K., and Vazquez, J.A., 2006, Alteration and re- monds, H.N., Hoff, J.A., and Thomas, R.B., 2003, Mea- sights into the tectonomagmatic evolution of NW Mex- melting of nascent oceanic crust during continental surement of attogram quantities of 231Pa in dissolved ico: Geochronology and geochemistry of the Miocene rupture: Evidence from zircon geochemistry of rhyo- and particulate fractions of seawater by isotope dilution volcanic rocks from the Pinacate area, Sonora: Geolog- lites and xenoliths from the Salton Trough, California: thermal ionization mass spectroscopy: Analytical Chem- ical Society of America Bulletin, v. 120, p. 691–708, Earth and Planetary Science Letters, v. 252, p. 260– istry, v. 75, p. 1075–1079, doi:10.1021/ac026247r. doi:10.1130/B26053.1. 274, doi:10.1016/j.epsl.2006.09.041. Shen, C.-C., Wu, C.-C., Cheng, H., Edwards, R.L., Hsieh, Wanless, V.D., Perﬁ t, M.R., Ridley, W.I., and Klein, E., Schmitt, A.K., Grove, M., Harrison, T.M., Lovera, O., Y.-T., Gallet, S., Chang, C.-C., Li, T.-Y., Lam, D.D., 2010, Dacite petrogenesis on mid-ocean ridges: Evi- Hulen , J., and Walters, M., 2003, The Geysers–Cobb Kano, A., Hori, M., and Spötl, C., 2012, High-precision dence for oceanic crustal melting and assimilation: Mountain magma system, California (Part 1): U-Pb zir- and high-resolution carbonate 230Th dating by MC-ICP- Journal of Petrology, v. 51, p. 2377–2410, doi:10.1093 con ages of volcanic rocks, conditions of zircon crys- MS with SEM protocols: Geochimica et Cosmochimica /petrology/egq056. tallization and magma residence times: Geochimica et Acta, v. 99, p. 71–86, doi:10.1016/j.gca.2012.09.018. Wiedenbeck, M., Hanchar, J.M., Peck, W.H., Sylvester, P., Cosmochimica Acta, v. 67, p. 3423–3442, doi:10.1016 Thy, P., Beard, J.S., and Lofgren, G.E., 1990, Experi- Valley, J., Whitehouse, M., Kronz, A., Morishita, Y., /S0016-7037(03)00140-6. mental constraints on the origin of Icelandic rhyo- Nasdala, L., Fiebig, J., Franchi, I., Girard, J.P., Green- Schmitt, A.K., Stockli, D.F., and Hausback, B.P., 2006, lites: The Journal of Geology, v. 98, p. 417–421, wood, R.C., Hinton, R., Kita, N., Mason, P.R.D., Nor- Eruption and magma crystallization ages of Las doi:10.1086/629413. man, M., Ogasawara, M., Piccoli, P.M., Rhede, D., Tres Virgenes (Baja California) constrained by Till, C.B., Gans, P.B., Spera, F.J., MacMillan, I., and Blair, Satoh, H., Schulz-Dobrick, B., Skar, O., Spicuzza, combined Th-230/U-238 and (U-Th)/He dating of K.D., 2009, Perils of petrotectonic modeling: A view M.J., Terada, K., Tindle, A., Togashi, S., Vennemann, zircon: Journal of Volcanology and Geothermal Re- from southern Sonora, Mexico: Journal of Volcanology T., Xie, Q., and Zheng, Y.F., 2004, Further characteri- search, v. 158, p. 281–295, doi:10.1016/j.jvolgeores and Geothermal Research, v. 186, no. 3–4, p. 160–168, sation of the 91500 zircon crystal: Geostandards and .2006.07.005. doi:10.1016/j.jvolgeores.2009.06.014. Geoanalytical Research, v. 28, p. 9–39, doi:10.1111 Schmitt, A.K., Stockli, D.F., Niedermann, S., Lovera, Trail, D., Mojzsis, S.J., Harrison, T.M., Schmitt, A.K., /j.1751-908X.2004.tb01041.x. O.M., and Hausback, B.P., 2010, Eruption ages of Watson, E.B., and Young, E.D., 2007, Constraints Wilson, S.A., 1997, Data Compilation for USGS Reference Las Tres Vírgenes volcano (Baja California): A tale on Hadean zircon protoliths from oxygen isotopes, Material BHVO-2, Hawaiian Basalt: U.S. Geologi- of two helium isotopes: Quaternary Geochronol- Ti-thermometry, and rare earth elements: Geochem- cal Survey Open-File Report: http://crustal.usgs.gov ogy, v. 5, no. 5, p. 503–511. istry Geophysics Geosystems, v. 8, Q06014, doi: /geochemical_reference_standards/basaltbhvo2.html Schmitt, A.K., Perﬁ t, M.R., Rubin, K.H., Stockli, D.F., 10.1029/2006GC001449. (accessed 28 July 2012). Smith, M.C., Cotsonika, L.A., Zellmer, G.F., Ridley, Trail, D., Bindeman, I.N., Watson, E.B., and Schmitt, W.I., and Lovera, O.M., 2011, Rapid cooling rates at an A.K., 2009, Experimental calibration of oxygen iso- SCIENCE EDITOR: NANCY RIGGS active mid-ocean ridge from zircon thermochronology: tope fractionation between quartz and zircon: Geo- ASSOCIATE EDITOR: BRIAN MCCONNELL Earth and Planetary Science Letters, v. 302, p. 349– chimica et Cosmochimica Acta, v. 73, p. 7110–7126, MANUSCRIPT RECEIVED 28 AUGUST 2012 358, doi:10.1016/j.epsl.2010.12.022. doi:10.1016/j.gca.2009.08.024. REVISED MANUSCRIPT RECEIVED 25 MAY 2013 Schmitt, A.K., Martín, A., Stockli, D.F., Farley, K.A., and Umhoefer, P.J., Dorsey, R.J., Willsey, S., Mayer, L., and MANUSCRIPT ACCEPTED 7 AUGUST 2013 Lovera, O.M., 2013, (U-Th)/He zircon and archaeo- Renne, P.R., 2001, Stratigraphy and geochronology of logical ages for a late prehistoric eruption in the the Comondú Group near Loreto, Baja California Sur, Printed in the USA

1850 Geological Society of America Bulletin, November/December 2013