John W. Shervais

Pre-4.2 AE mare-basalt volcanism in the lunar highlands

The concept that the plutonism of the lunar highlands and the mare-type volcanism are two separate problems in both time (> 4.4 AE versus < 3.95 AE) and space is seriously questioned by the discovery of a 4.23-AE low-Ti mare basalt from Fra Mauro Formation. Apollo 14 breccia 14305 contains a clast (,122) which is an olivine gabbronorite that is texturally and mineralogically similar to several Apollo 12 basalts (e.g., 12005, 12035, 12040). It consists of cumulus olivine (40 modal %; Fo 62–70) and Ti-chromite (2.5 modal %); post-cumulus phases include low-Ca pyroxene (29 modal %; Wo 7–13 En 68–75), augite (10 modal %; Wo 31–40 En 47–50), plagioclase (15 modal %, An 82–93), and ilmenite (4 modal %, 5–7 MgO). The TiO2 content of this rock = 4.3%; CaO/Al2O3 ⋍ 1.0, CaO = 5.1%; MgO/FeO ⋍ 1.0, MgO = 21.9%. The REE pattern, normalized to chondritic abundances, is approximately 30 × Ch and “hump-shaped” with a pronounced Eu depletion and a non-KREEPy signature. A four-point Rb-Sr isochron reveals an age of 4.23 ± 0.05 AE. The sample has a low initial 87Sr/86Sr= 0.69911 ± 3. The data presented here show that non-KREEPy, mare-type volcanism commenced at least as early as 4.2 AE in the Fra Mauro region and probably across much of the lunar surface. Massive bombardment during the “terminal cataclysm” and the subsequent veneer of younger mare basalts has obliturated most of the evidence for these ancient volcanic events. These old, mare-type volcanics may be related to basin-forming events such as made Procellarum (i.e., impact-triggered igneous activity).

Geochemical evidence for the tectonic setting of the Coast Range ophiolite: A composite island arc–oceanic crust terrane in western California

The Middle to Late Jurassic age Coast Range ophiolite (CRO) of California contains two geochemically distinct volcanic rock associations that formed in different tectonic settings. Volcanic rocks from the southern CRO (Point Sal, Cuesta Ridge, Stanley Mountain, Llanada, Quinto Creek, and Del Puerto) and parts of the northern CRO (Healdsburg, Elder Creek) are similar to low-K tholeiites and calc-alkaline rocks of the island-arc suite. The thin volcanic sections of these ophiolite remnants suggest formation by intra-arc rifting. In contrast, volcanic rocks from Stonyford seamount and Paskenta in the northern CRO are transitional subalkaline metabasalts with geochemical characteristics similar to enriched mid-ocean ridge basalts or ocean-island tholeiites. These rocks are associated with Tithonian radiolarian cherts and may be part of the Franciscan Complex. Alternatively, they may represent a change in tectonic setting within the CRO during the Late Jurassic. Regardless, the CRO as currently conceived cannot be considered a single terrane with one mode of origin.

Early Proterozoic oceanic crust and the evolution of subcontinental mantle: Eclogites and related rocks from southern Africa

Seven eclogitic nodules from kimberlites in southern Africa have been studied in detail for whole-rock, major- and trace-element geochemistry, petrography, and mineral chemistry by electron microprobe; high-purity mineral separates from six of these samples have been analyzed for trace elements and for the isotopic composition of Sr, Nd, and oxygen. Three eclogite groups are recognized: group A eclogites have very high Mg/Fe, low Na in pyroxene, moderate δ18O (+4.7 to +5.3), and low 87Sr/86Sr and 143/144Nd. Olivine and enstatite may also be present as accessory phases in this group. Group B eclogites have moderate to high Mg/Fe ratios, high Na in pyroxene, low δ18O (+3.0 to +3.4), high 87Sr/86Sr and 143Nd/144Nd ratios, and extremely LREE-depleted/ HREE-enriched garnets with no Eu anomalies. Group C eclogites have low Mg/Fe ratios, high Na in pyroxene, and variable Sr-, Nd-, and O-isotope compositions. Accessory feldspar is present in one sample that may be of crustal origin. Mineral separates of garnet and pyroxene have positive Eu anomalies. The group A eclogites are too refractory to represent magma compositions and must have formed as cumulate dike rocks in the upper mantle that contain a minor trapped liquid component. This is supported by the presence of accessory olivine and enstatite, and by their KSm and KNd, which are similar to empirical pyroxene/garnet partition coefficients. The Group B eclogites are extremely depleted in incompatible elements and have ϵNdvalues 10x to 20x MORB. The high 87Sr/86Sr of these eclogites is not consistent with their strongly depleted LREE and high ϵNd, and cannot be primary—it must have been imposed on the protolith. The low δ18O of these eclogites cannot form by mantle fractionation processes and must also be inherited from the protolith. Both the Sr- and oxygen-isotope data are consistent with high- temperature hydrothermal alteration of a basaltic protolith, followed by partial melting to form the refractory compositions observed now. The hydrothermal fluid may have been sea water, but secondary enrichment of the protolith in Rb is required to generate the observed high Sr ratios. The high Na content of the pyroxenes supports a spilitic alteration event. The major- and trace-element characteristics of the group C eclogites are consistent with recrystallization of a cumulate gabbro protolith. One group C rock is probably a garnet granulite derived from the lower crust. The other may represent the plutonic portion of oceanic crust. A reconstructed whole-rock isochron for the three Bellsbank eclogites yields an age of 2.1 ± 0.1 b.y. This implies that plate-tectonic processes involving the generation and subduction of oceanic crust have been active since the early Proterozoic. The early Proterozoic ocean crust consisted of two components: a plutonic section of gabbro cumulates (group C eclogites) and a volcanic section of basalt (group B eclogites). The volcanic component has undergone substantial hydro-thermal alteration and subsequent partial melting to form a refractory residue; the plutonic component is less modified.

Multi-Stage Origin of the Coast Range Ophiolite, California: Implications for the Life Cycle of Supra-Subduction Zone Ophiolites

The Coast Range ophiolite of California is one of the most extensive ophiolite terranes in North America, extending over 700 km from the northernmost Sacramento Valley to the southern Transverse Ranges in central California. This ophiolite, and other ophiolite remnants with similar mid-Jurassic ages, represent a major but short-lived episode of oceanic crust formation that affected much of western North America. The history of this ophiolite is important for models of the tectonic evolution of western North America during the Mesozoic, and a range of conflicting interpretations have arisen. Current petrologic, geochemical, stratigraphic, and radiometric age data all favor the interpretation that the Coast Range ophiolite formed to a large extent by rapid extension in the fore-arc region of a nascent subduction zone. Closer inspection of these data, however, along with detailed studies of field relationships at several locales, show that formation of the ophiolite was more complex, and requires several stages of formation. Our work shows that exposures of the Coast Range ophiolite preserve evidence for four stages of magmatic development. The first three stages represent formation of the ophiolite above a nascent subduction zone. Rocks associated with the first stage include ophiolite layered gabbros, a sheeted complex, and volcanic rocks with arc tholeiitic or (more rarely) low-K calc-alkaline affinities. The second stage is characterized by intrusive wehrlite-clinopyroxenite complexes, intrusive gabbros, Cr-rich diorites, and volcanic rocks with high-Ca boninitic or tholeiitic ankaramite affinities. The third stage includes diorite and quartz diorite plutons, felsic dike and sill complexes, and calc-alkaline volcanic rocks. The first three stages of ophiolite formation were terminated by the intrusion of mid-ocean ridge basalt dikes, and the eruption of mid-ocean ridge basalt or ocean-island basalt volcanic suites. We interpret this final magmatic event (MORB dikes) to represent the collision of an active spreading ridge. Subsequent reorganization of relative plate motions led to sinistral transpression, along with renewed subduction and accretion of the Franciscan Complex. The latter event resulted in uplift and exhumation of the ophiolite by the process of accretionary uplift.

Layered mafic sill complex beneath the eastern Snake River Plain: Evidence from cyclic geochemical variations in basalt

The eastern Snake River Plain in southern Idaho, western United States, is characterized by 1–2 km of Pleistocene to late Pliocene basalt overlying rhyolite caldera complexes. Cyclic variations in the chemical composition of basalts from 1136 m of scientific drill core show that the parent magmas of these lavas evolved by crystal fractionation at shallow to intermediate crustal depths, punctuated by episodic recharge with more primitive compositions and assimilation of adjacent wall rock. We have identified 10 upward fractionation cycles and four reversed cycles; assimilation of sialic crust was limited and mainly affects the oldest basalts, which directly overlie rhyolites. We infer that the crystal fractionation and/or recharge cycles took place in a series of sill-like intrusions at intermediate crustal depths that now form a layered mafic intrusion that underlies the eastern Snake River Plain at depth. This layered sill complex is represented by the ∼10-km-thick “basaltic sill” that has been imaged seismically at ∼12–22 km depth. The association of this mid-crustal sill complex with geochemical fractionation cycles in basalt supports the concept that exposed layered mafic intrusions may be linked to overlying basalt provinces that have since been removed by erosion.

Yellowstone plume-continental lithosphere interaction beneath the Snake River Plain

The Snake River Plain represents 17 m.y. of volcanic activity that took place as the North American continent migrated over a relatively fixed magma source, or hotspot. The identification of a clear seismic image of a plume beneath Yellowstone is compelling evidence that the Miocene to recent volcanism associated with the Columbia Plateau, Oregon High Lava Plains, Snake River Plain, Northern Nevada Rift and Yellowstone Plateau represents a single magmatic system related to a mantle plume. A remaining enigma is, why do radiogenic isotope signatures from basalts erupted over the Mesozoic–Paleozoic accreted terrains suggest a plume source while basalts erupted across the Proterozoic–Archean craton margin indicate an ancient subcontinental mantle lithosphere source? We show that ancient cratonic lithosphere like that of the Wyoming province superimposes its inherent isotopic composition on sublithospheric plume and/or asthenospheric melts. The results show that Yellowstone plume could have a radiogenic isotope composition similar to the mantle source of the early Columbia River Basalt Group and that the plume source composition has persisted to the present day.

Radioisotopic and biostratigraphic age relations in the Coast Range Ophiolite, northern California: Implications for the tectonic evolution of the Western Cordillera

The Coast Range ophiolite (CRO) in northern California includes two distinct remnants. The Elder Creek ophiolite is a classic suprasubduction zone ophiolite with three sequential plutonic suites (layered gabbro, wehrlite-pyroxenite, quartz diorite), a mafi c to felsic dike complex, and mafi c-felsic volcanic rocks; the entire suite is cut by late mid-oceanic-ridge basalt (MORB) dikes and overlain by ophiolitic breccia. The Stonyford volcanic complex (SFVC) comprises three volcanic series with intercalated chert horizons that form a submarine volcano enclosed in sheared serpentinite. Structurally below this seamount are melange blocks of CRO similar to Elder Creek. U/Pb zircon ages from plagiogranite and quartz diorites at Elder Creek range in age from 165 Ma to 172 Ma. U/Pb zircon ages obtained from CRO melange blocks below the SFVC are similar (166‐172 Ma). 40 Ar- 39 Ar ages of alkali basalt glass in the upper SFVC are all younger at ≈164 Ma. Radiolarians extracted from chert lenses intercalated with basalt in the SFVC indicate that the sedimentary strata range in age from Bathonian (Unitary Association Zone 6‐6 of Baumgartner et al., 1995a) near the base of the complex to late Callovian to early Kimmeridgian (Unitary Association Zones 8‐10) in the upper part. The SFVC sedimentary record preserves evidence of a major faunal change wherein relatively small-sized, polytaxic radiolarian faunas were replaced by very robust, oligotaxic, nassellarian-dominated faunas that included Praeparvicingula spp. We suggest that CRO formation began after the early Middle Jurassic (172‐180 Ma) collision of an exotic or fringing arc with North America and initiation of a new or reconfi gured east-dipping subduction zone. The data show that the CRO formed prior to the Late Jurassic Nevadan orogeny, probably by rapid forearc extension above a nascent subduction zone. We infer that CRO spreading ended with the collision of an oceanic spreading center ca. 164 Ma, coincident with the oldest high-grade blocks in the structurally underlying Franciscan assemblage. We further suggest that the “classic” Nevadan orogeny represents a response to spreading center collision, with shallow subduction of young lithosphere causing the initial compressional deformation and with a subsequent change in North American plate motion to rapid northward drift (J2 cusp) causing sinistral transpression and transtension in the Sierra foothills. These data are not consistent with models for Late Jurassic arc collision in the Sierra foothills or a back-arc origin for the CRO.

Growth of subcontinental lithosphere: Evidence from repeated dike injections in the Balmuccia lherzolite massif, Italian Alps

Abstract The Balmuccia alpine lherzolite massif is a fragment of subcontinental lithospheric mantle emplaced into the lower crust 251 Ma ago during the final, extensional phase of the Hercynian orogeny. The Balmuccia massif consists largely of lherzolite, with subordinate harzburgite and dunite, and an array of dike rocks formed in the mantle before crustal emplacement. Dike rocks include websterite and bronzitite of the Cr-diopside suite, spinel clinopyroxenite and spinel-poor websterite of the Al-augite suite, gabbro and gabbronorite of the late gabbro suite, and hornblendite of the hydrous vein suite. The dike rocks display consistent intrusive relationships with one another, such that Cr-diopside suite dikes are always older than dikes and veins of the Al-augite suite, followed by dikes of the late gabbro suite and veins of the hydrous vein suite. Phlogopite (phl) veinlets that formed during interaction with the adjacent crust are the youngest event. There are at least three generations of Cr-diopside suite dikes, as shown by crosscutting relations. Dikes of the Al-augite suite form a polybaric fractionation series from spinel clinopyroxenite to websterite and feldspathic websterite, which crystallized from aluminous alkaline magmas at relatively high pressures. The late gabbro suite of dikes intruded at lower pressures, where plagioclase saturation occurred before significant mafic phase fractionation. Hornblendite veins have distinct compositional and isotopic characteristics, which show that they are not related to either the Al-augite suite or to the late gabbro dike suite. Cr-diopside suite dikes have Nd and Sr isotopic compositions similar to those of the host lherzolite and within the range of compositions defined by ocean–island basalts. The Al-augite dikes and the hornblendite veins have Sr and Nd isotopic compositions similar to those of Cr-diopside suite lherzolite and websterite. The late gabbro dikes have Nd and Sr isotopic compositions similar to mid-ocean ridge basalt (MORB) asthenosphere. Lead isotopic compositions for all of the samples fall in the present-day MORB field on the 208 Pb / 204 Pb vs. 206 Pb / 204 Pb diagram but are displaced above this field on the 207 Pb / 204 Pb vs. 206 Pb / 204 Pb diagram. There is overlap in the data between the Cr-diopside suite and the Al-augite and hydrous vein suites, with the exception that the Cr-diopside websterite dikes have more radiogenic Pb than any of the other samples. In Pb–Pb space as well, the late gabbro suite has the least radiogenic isotopic compositions, reflecting a change in magma source region during uplift. These data show that tectonic thinning of subcontinental lithospheric mantle during extension caused a change in the source regions of mantle-derived magmas from an ocean island basalt (OIB)-like lithosphere to the underlying MORB asthenosphere. They also demonstrate that the upper mantle acquires its heterogeneous isotopic character through several different processes, including in situ radiogenic growth, addition of asthenospheric melts, dike-wall rock ionic exchange, redistribution of the lithospheric dike and vein materials by melting, and in the late stages of emplacement, assimilation of crustal materials.

A Field and Chemical Study of Serpentinization—Stonyford, California: Chemical Flux and Mass Balance

Serpentinized harzburgites and dunites in the Coast Range ophiolite near Stonyford, California, form massive, decameter- to kilometer-scale blocks in serpentinite schist; together these form serpentinite broken formation that grades into mélange where exotic blocks have been incorporated into the serpeninite schist. Whole rock geochemical data and modal reconstruction of protolith compositions show that serpentinization here proceeded essentially isochemically for Si, Mg, and Fe, whereas other elements (Ca, Al, Cr) were lost to an aqueous flux. Mass balance calculations based on actual primary and secondary mineral compositions show that significant Fe, Al, and Cr may be accommodated in serpentine. The transformation of orthopyroxene to serpentine (bastite) releases significant amounts of silica, which forms additional serpentine when it reacts with MgO released by the serpentinization olivine; this reaction suppressed brucite formation and accounts in part for the conservation of silica and magnesia documented by whole rock geochemistry. For normal harzburgites (20-25% modal orthopyroxene), approximately half of the potential brucite was suppressed. Volume expansion was considerable: 25-30% for pseudomorphic replacement of orthopyroxene, 50-60% for replacement of olivine. The increase in volume resulted primarily from the addition of water to hydrate the primary silicate assemblage; loss of Al and Ca to aqueous solutions results in a slight loss of mass. Expansion is accommodated by orthogonal fractures at both the microscopic and macroscopic scales. Subsequent movement along the macroscopic serpentinized fractures led to the formation of serpentinite broken formation, with a matrix of sheared and foliated serpentinite, and relict blocks of massive, less serpentinized peridotite. This movement may have resulted in part from the volume expansion of the peridotite, as rigid, less serpentinized blocks were forced to adjust to increased volumes in their totally serpentinized selvages by differential movements that forced the blocks to move in the direction of least principal stress.

Ion and electron microprobe study of troctolites, norite, and anorthosites from Apollo 14: evidence for urKREEP assimilation during petrogenesis of Apollo 14 Mg-suite rocks

Abstract Most of the Moon’s highland crust formed during the period 4.65–4.45 Ga ago from a vast magma ocean up to 800 km deep (Hess and Parmentier, 1995) . This early lunar crust comprises Fe-rich anorthosites with calcic plagioclase compositions. Subsequent evolution of the highland crust was dominated by troctolites, anorthosites, and norites of the Mg-suite. This plutonic series is characterized by calcic plagioclase, and mafic minerals with high mg# (=100∗Mg/[Mg + Fe]). These rocks evidently formed by partial melting of ultramafic rocks of the lunar mantle, but their bulk rock incompatible element characteristics are too enriched to represent such a primitive source. Previous studies have suggested that this enrichment in incompatible trace elements is the result of metasomatism of the crust by fluids rich in REE and P. The products of this suggested metasomatic event are REE-rich phosphates (typically whitlockite) deposited interstitially. Alternatively, the incompatible element-rich nature of these plutonic rocks may represent a characteristic of their parent magma, acquired prior to crystallization of the plutons. In an effort to distinguish the origin of this important lunar rock series, we have analyzed the REE content of primary cumulus phases in ten Mg-suite cumulates using SIMS, along with their major and minor element compositions by electron microprobe analysis. Nine of these samples have high mg#s, consistent with their formation from the most primitive parent melts of the Mg-suite. The data presented here show that Mg-suite troctolites and anorthosites preserve major and trace element characteristics acquired during their formation as igneous cumulate rocks and that these characteristics can be used to reconstruct related aspects of the parent magma composition. Our data show that primitive cumulates of the Mg-suite crystallized from magmas with REE contents similar to high-K KREEP in both concentration and relative abundance. The highly enriched nature of this parent magma contrasts with its primitive major element characteristics, as pointed out by previous workers. This enigma is best explained by the mixing of residual magma ocean urKREEP melts with ultramagnesian komatiitic partial melts from the deep lunar interior. The data do not support earlier models that invoke crustal metasomatism to enrich the Mg-suite cumulates after formation, or models which call for a superKREEP parent for the troctolites and anorthosites.