Search

Primary andesitic magmas could be an important component of arc magma genesis and might have played a key role in the advent of continents. Recent studies hypothesized that primary andesitic magmas occur in the oceanic arc, where the crust is thin. The Kermadec arc has the thinnest crust among all the studied oceanic arcs (<15 km in thickness); however, there are no studies that corroborate the formation of primary andesitic magmas in the arc. The aim of this study is to develop a better understanding of primary andesites in oceanic arcs through the petrology of the Kermadec arc. Here, we present the petrology of volcanic rocks dredged from the Kibblewhite Volcano in the Kermadec arc during the R/V SONNE SO-255 expedition in 2017. Magma types range from andesite to rhyolite at the Kibblewhite Volcano, but basalts dominate at the neighboring cones. This study focuses on magnesian andesites from the northeastern flank of this volcano. The magnesian andesites are nearly aphyric and plagioclase free but contain microphenocrysts of olivine (Fo84–86) and clinopyroxene (Mg# = 81–87). Using olivine addition models, the primary magmas were estimated to contain 55–56 wt % SiO2 and 10–12 wt % MgO, similar to the high-Mg andesites observed in other convergent plate margins, indicating the generation of primary andesitic magma beneath the Kibblewhite Volcano. The trace element and isotopic characteristics of the magnesian andesites are typical of volcanic rocks from the Kermadec arc. This indicates that the subduction of a young plate or melting of a pyroxenitic source is not necessary to produce magnesian andesites. Instead, we propose that the magnesian andesites were produced by the direct melting of the uppermost mantle of the Kermadec arc. The thin crust of the Kermadec arc should yield low-pressure conditions in the uppermost mantle, allowing the sub-arc mantle to generate primary andesitic melts. This study supports the hypothesis that primary andesitic magmas generate in the arc where the crust is thin and provides a new insight into the magma genesis of the Kermadec arc. [ABSTRACT FROM AUTHOR]

The relationship between plutonic and volcanic components of magmatic plumbing systems continues to be a question of intense debate. The Oki-Dōzen Islands, Sea of Japan, preserve outcrops of temporally associated plutonic, hypabyssal, and volcanic rocks. Post-intrusion uplift juxtaposed Miocene syenites in inferred faulted contact with volcanic trachytes that are cut by rhyolite hypabyssal dikes. This provides a window deep into the timing and origins of magma storage architecture and dynamics. Zircon is ubiquitous in all samples; our aim is to determine what its age and composition can reveal about the plutonic-volcanic connection. Here we show magma source characteristics are recorded in zircon Hf isotopes; source composition and assimilation of heterogeneous hydrothermally altered crust in zircon O isotopes; and extensive fractional crystallization in zircon trace elements. Combined with new U-Th-Pb SHRIMP zircon ages, 6.4–5.7 Ma, compositional data show pluton formation was by protracted amalgamation of discrete magma pulses. The rhyolite dike preserves an evolved fraction segregated from these discrete magmas. Synchronous with plutonism was a volcanic eruption of trachyte magma derived from the same source, which may have stalled at a relatively shallow depth prior to eruption. Stalling occurred at least above the amphibole stability zone because amphibole-compatible Sc and Ti were not depleted in the trachyte melt resulting in elevated values of these in volcanic compared to plutonic zircon. Identifying smaller episodic magma pulses in a larger magmatic complex places constraints on potential magma fluxes and eruptible volumes. High-flux, large volume, plume-related ocean island magmatic systems may have extensive vertically distributed multi-stage magmatic reservoirs and subduction-related systems transcrustal magma reservoirs. By contrast, Oki-Dōzen was a low-flux system with incremental pluton growth and small- to moderate-scale eruptions. [ABSTRACT FROM AUTHOR]

Ankaramites, which are clinopyroxene-rich basalts with primitive whole-rock compositions (Mg# >65), are common in oceanic arcs and are characterized by high whole-rock CaO/Al 2 O 3 (>1.0) ratios and olivine crystals with anomalously low nickel contents (<0.2 wt% NiO). These geochemical characteristics cannot be explained by the melting of ordinary mantle peridotite. However, their origin is critical for understanding the formation of primary magmas in oceanic arcs. Here, we investigated olivine-hosted melt inclusions (MIs) from ankaramites and magnesian andesites of the Kibblewhite Volcano in the Kermadec arc. The MIs from the ankaramites have similar major and trace element characteristics to the host rocks, indicating that the ankaramites did not result from an accumulation of mafic minerals but rather represent the primary magma in the Kibblewhite Volcano. The MIs from the magnesian andesites were hosted in forsteritic olivine xenocrysts with a wide range of NiO contents (Fo 90 – 92 ; 0.13–0.39 wt% NiO) and have similar major element compositions to the ankaramites but exhibit a wide range of CaO/Al 2 O 3 (0.85–1.54). The trace element characteristics of the MIs from the magnesian andesites do not match those of the host rocks, indicating that they are not primary melts of the magnesian andesites but primitive basaltic melts generated before the magnesian andesites formed. Interestingly, the CaO/Al 2 O 3 ratio of MIs from the magnesian andesites was negatively correlated with the NiO content of their host olivines. This correlation suggests that the composition of the primary basaltic magmas of the Kibblewhite Volcano changed continuously from peridotite-derived to ankaramitic. This correlation could not be explained by grain-scale process, crustal anatexis, or contribution of slab-derived carbonate-rich fluids. Instead, we propose that this correlation can be explained by the interaction of the ascending primary basaltic melts with the lithospheric mantle. During melt-mantle interaction, the assimilation of clinopyroxene and fractionation of olivine and orthopyroxene caused the CaO/Al 2 O 3 ratio to increase in the melt and the Ni content to decrease. Furthermore, because the magnesian andesites have low CaO/Al 2 O 3 ratios and could be derived from a clinopyroxene-poor mantle lithology, the interaction between the melt and mantle may also be closely related to the origin of the magnesian andesites at Kibblewhite Volcano. This interpretation provides a new perspective on the origin of the oceanic arc ankaramites and why primary andesitic and basaltic magmas coexist in the Kibblewhite Volcano. • Ankaramites are Ca-rich and Ni-poor porphyritic basalts that are common in oceanic arcs. • Melt inclusions from Kibblewhite Volcano show similar compositions to ankaramites. • Ankaramite is a primary magma component in oceanic arcs. • Interaction between melt and mantle can produce ankaramitic melts. • Harzburgite formed by melt-mantle interactions is the source of high-Mg andesites. [ABSTRACT FROM AUTHOR]

Pagan is one of the largest volcanoes along the Mariana arc volcanic front. It has a maximum elevation of 570 m (Mt. Pagan), but its submarine flanks descend to 2000–3000 m below sea level, and are unexplored. Bathymetric mapping and ROV Hyper-Dolphin dives (HPD1147 and HPD1148) on the submarine NE and SW flanks of Pagan were carried out during cruise NT10-12 of R.V. Natsushima in July 2010. There are no systematic compositional differences between subaerial lavas reported in the literature and differentiated submarine lavas collected in HPD1148, with <7 wt % MgO, suggesting they are derived from the same magmatic system. However, these differentiated lavas show complexities including magma mixing; thus we concentrate on magnesian submarine lavas (>7 wt % MgO). Twenty least-fractionated basalts (48·5–50 wt % SiO2) collected during HPD1147 extend to higher MgO (10–11 wt %) and Mg# (66–70) than the subaerial lavas. Olivine (up to Fo94) and spinel (Cr# up to 0·8) compositions suggest that these Pagan primitive magmas formed from high degrees of mantle melting. Two basalt types can be distinguished based on their geochemistry at similar (10–11 wt %) MgO; these erupted recently, 500 m apart. Both contain clinopyroxene and olivine phenocrysts and are referred to as COB1 and COB2. Lower TiO2, FeO, Na2O, K2O, incompatible trace element abundances, and Nb/Yb suggest that COB1 formed from higher degrees of mantle melting. In addition, light rare earth element (LREE) enrichment and higher Th/Nb in COB2 contrast with LREE depletion and lower Th/Nb in COB1. Higher Ba/Th and Ba/Nb and lower Th/Nb indicate that the main subduction addition in COB1 was dominated by hydrous fluid, whereas that in COB2 was dominated by sediment melt. Sr–Nd–Pb–Hf isotopes are also consistent with this interpretation. These observations suggest that the subduction component responsible for the greater degree of melting of the COB1 source was mostly hydrous fluid. The origin of such different metasomatic agents resulted in different primary magmas forming in the same volcano. Both hydrous fluid and sediment melt components may have unmixed from an originally homogeneous supercritical fluid in or above the subducting slab below the volcanic front. These may have been added separately to the mantle wedge peridotite (mantle diapir) and resulted in two neighboring but completely different primary magmas from the same diapir. Moreover, these primitive lavas suggest that even for intra-oceanic arcs assimilation–fractional crystallization is inevitable when these magmas evolve in the crust and, in addition, that phlogopite is present in their mantle residue and thus played an important role in their genesis. [ABSTRACT FROM PUBLISHER]

Quaternary volcanoes in the Northeast Japan arc can be grouped into 10 volcanic clusters striking transverse to the arc; these have an average width of 50 km, and are separated by parallel gaps 30–75 km wide. This clustering of volcanic centres, topographic profiles, low-velocity regions in the mantle wedge and local negative Bouguer gravity anomalies along the Japan Sea side of the volcanic arc are closely correlated. All these observations may be related to locally developed hot regions within the mantle wedge that have the form of inclined, 50 km wide fingers. Each of the 10 fingers recognised extends from deep mantle (>150 km) below the back-arc region towards the shallower mantle (∼50 km) beneath the volcanic front. Quaternary volcanoes are built immediately above the hot mantle fingers. The volcanic basement along the fingers has been uplifted by repeated injection of magmas into the crust, accompanied by Quaternary volcanic activity at the surface. Although volcanic activity is rare along the Japan Sea coast, tomographic results show that hot, low-velocity mantle fingers exist within the mantle wedge. The negative Bouguer anomalies at the rear of the volcanic arc could be caused by magmas supplied from the hot mantle fingers; these have not yet been erupted, but have accumulated at the Moho discontinuity. [Copyright &y& Elsevier]

During German R/V SONNE cruise (SO-255) mafic basaltic to magnesian andesitic lavaswere recovered from the Kermadec arc front volcano Kibblewhite (Hirai et al., EGU2018).Notably, these primitive lavas contain forsteritic olivines and olivine-hosted melt inclusions,which were re-homogenized using a heating stage. Here, we present the chemical variationsof the melt inclusions and their host olivines and discuss the interaction between mafic meltsand subarc lithospheric mantle. The olivines are divided into two groups based on their occurrence in different hostrock and NiO contents. Olivines (Fo90.3−92.4) found as phenocrysts in ankaramites(olivine-clinopyroxene rich basalts; 50.4–50.6 wt.% SiO2; 13.1–13.5 wt.% MgO; Mg# =71.8–72.1) show unusually low Ni contents (912–1462 ppm). Olivines (Fo91.0−92.2)occurring as xenocrysts in magnesian andesites (57.5–57.6 wt.% SiO2; 5.5–5.6 wt.%MgO; Mg# = 56.8–57.3) show higher Ni contents (1108–3057 ppm) than olivines inankaramites. All re-homogenized melt inclusions have primitive basalt compositions (49.6–52.0 wt.%SiO2; 12.7–15.9 wt.% MgO, 0.36–0.58 wt.% TiO2, 1.2–2.0 wt.% Na2O+K2O, 0.06–0.37wt.% SO3, and 0.05–0.20 wt.% Cl), which could represent parental melts for the ankaramites.However, melt inclusions hosted by olivines in ankaramites have lower Al2O3 andhigher CaO contents than those hosted by olivine xenocrysts in magnesian andesites;CaO/Al2O3 values are 1.2–1.6 and 0.9–1.3 respectively, indicating that there is a negativecorrelation between Ni contents in host olivines and CaO/Al2O3 values in their meltinclusions. The variable Ni and constant Fo contents in the host olivines could be explained by aninteraction of primary mafic melts with lithospheric mantle, which causes assimilation ofpyroxenes in the lithosphere and simultaneous fractional crystallization of olivines (AFC;Kelemen, 1990; Tamura et al., 2018). Moreover, the high CaO/Al2O3 values in the meltinclusions hosted by the ankaramite olivines cannot be produced by partial melting oflherzolite (e.g. as compilation by Médard et al., 2004). These melts have most likelyextensively interacted with the subarc lithospheric mantle. In addition, a negative correlationbetween Ni content in the host olivine and CaO/Al2O3 ratio in their melt inclusionssupports the melt-mantle interaction. Conversely, xenocrystic olivines in the andesiteswith high Ni contents must have been crystallized from more primitive basalticmelts within lithospheric mantle prior to significant melt-mantle interaction. Theseolivines were then trapped by primary magnesian andesite magmas as xenocrysts. [ABSTRACT FROM AUTHOR]

The process by which continental crust has formed is not well understood, though such crust mostly forms at convergent plate margins today. It is thus imperative to study modern intra-oceanic arcs, such as those common in the western Pacific Ocean. New seismic studies along the representative Izu-Bonin intra-oceanic arc provide unique along-strike images of arc crust and uppermost mantle to complement earlier, cross-arc lithospheric profiles. These reveal two scales (1000–10 km scale) of variations, one at the scale of the Izu versus Bonin (thick versus thin) arc crust, the other at the intervolcano (∼50 km) scale. These images show that: (1) the bulk composition of the Izu-Bonin arc crust is more mafic than typical continental crust, (2) the middle crust with seismic velocities similar to continental crust is predominantly beneath basaltic arc volcanoes, (3) the bulk composition beneath basaltic volcanoes changes little at thick and thin arc segments, and (4) a process to return lower crustal components to the mantle, such as delamination, is required for an arc crust to evolve into continental crust. Continued thickening of the Izu-Bonin crust, accompanied by delamination of lowermost crust, can yield velocity structure of typical continental crust. [ABSTRACT FROM AUTHOR]