Nature and growth rate of the Northern Izu–Bonin (Ogasawara) arc crust and their implications for continental crust formation

The bulk composition of the continental crust throughout geological history is thought by most previous workers to be andesitic. This assumption of an andesitic bulk composition led to an early hypothesis by 72 ) that the continental crust was created by arc magmatism. This hypothesis for the origin of continental crust was challenged by several authors because: (i) the mean rate of arc crust addition obtained by 50 ) is too small to account for some certain phases of rapid crustal growth; and (ii) the bulk composition of ocean island arcs, the main contributor to the Archean and early Proterozoic crust, is basaltic rather than andesitic ( 4 ; 49 ). New data from the Northern Izu–Bonin arc are presented here which support the 72 ) hypothesis for the origin of the continental crust by andesitic arc magma. A geological interpretation of P wave crustal structure obtained from the Northern Izu–Bonin arc by 66 ) indicates that the arc crust has four distinctive lithologic layers: from top to bottom: (i) a 0.5–2-km-thick layer of basic to intermediate volcaniclastic, lava and hemipelagite (layer A); (ii) a 2–5-km-thick basic to intermediate volcaniclastics, lavas and intrusive layer (layer B); (iii) a 2–7-km-thick layer of felsic (tonalitic) rocks (layer C); and (iv) a 4–7-km-thick layer of mafic igneous rocks (layer D). The chemical composition of the upper and middle part of the northern Izu–Bonin arc is estimated to be similar to the average continental crust by 73 ). The rate of igneous addition of the Northern Izu–Bonin arc since its initial 45-Ma magmatism was calculated as 80 km³/km per million years. This rate of addition is considered to be a reasonable estimate for all arcs in the western Pacific. Using this rate, the global rate of crustal growth is estimated to be 2.96 km³/year which exceeds the average rate of crustal growth since the formation of the Earth (1.76 km³/year). Based on this estimate of continental growth and the previously documented sediment subduction and tectonic erosion rate (1.8 km³/year, 24 ), several examples of growth curves of the continental crust are presented here. These growth curves suggest that at least 50% of the present volume of the continental crust can be explained by arc magmatism. This conclusion indicates that arc magmatism is the most important contributor to the formation of continental crust, especially at the upper crustal level.     

Seismic Activity Around and Under Krakatau Volcano,Sunda Arc: Constraints to the Source Region of Island Arc Volcanics

There is general agreement that calc-alkaline volcanic rocks at convergent plate margins are genetically related to the process of subduction (Ringwood, 1974; Maaloe and Petersen, 1981; Hawkesworth et al., 1997). However, opinions on the mode and site of generation of primary magma for island arc volcanism differ substantially. The site of generation of calc-alkaline magma is thought to be either in the mantle wedge (Plank and Langmuir, 1988; McCulloch and Gamble, 1991) or in the subducting slab (White and Dupré, 1986; Defant and Drummond, 1990; Edwards et al., 1993; Ryan and Langmuir, 1993). We present seismological evidence in favour of the latter concept. A distinctive seismicity pattern around and under the Krakatau volcano was identified during systematic studies of the SE Asian convergent plate margins by means of global seismological data. A column-like cluster of events, probably associated with the dynamics of the volcano, is clearly separated from the events in the Wadati-Benioff zone. The accuracy of hypocentral determinations of the events of the cluster does not differ from the accuracy of the events belonging to the subducting slab. The depths of the cluster events vary from very shallow to about 100 km without any apparent discontinuity. On the other hand, there is a pronounced aseismic gap in the Wadati-Benioff zone directly beneath the volcano at depths between 100-150 km. The Krakatau cluster connects this aseismic gap to the volcano at the surface. The pervasive occurrence of earthquakes in the continental wedge between the subducting slab and the Earth surface bears witness to the brittle character of the continental lithosphere and casts doubt on the existence of large-scale melting of mantle material. The aseismic gap (Hanu and Vank, 1985), interpreted by us as a partially melted domain occurring in subducted slabs in practically all active subduction zones that reach depths greater than 100 km, is here used as evidence for the location of the primary source region of island arc volcanics in the subducting plate.     

Early history of the Izu–Bonin – Mariana arc system: Evidence from Belau and the Palau Trench

Volcanic rocks of the Kyushu–Palau Ridge (KPR) from Deep Sea Drilling Project (DSDP) site 448 and from Belau comprise a low-to-medium-K arc tholeiitic series. Belau rocks include (probable) Mid-Eocene low-Ca type-3 boninite and pre-Early Oligocene–Early Miocene low-K arc tholeiitic basalt, basaltic andesite, andesite and dacite. Palau Trench samples include sparsely phyric high-Mg, -Cr and -Ni rocks which resemble the Belau boninite and Izu–Bonin – Mariana (IBM) system boninites. The high-Mg Palau Trench samples also resemble other primitive arc lavas (e.g. arc picrites). Their chemistry suggests an origin involving steep thermal gradients in multiply depleted mantle. Subduction of hot, young lithosphere under a young hot upper plate is postulated to explain this occurrence. The KPR is inferred to be the source of Eocene boninite and arc tholeiitic terranes presently in forearc regions of the IBM system. A model is presented here showing how many IBM boninites may have originated in a small area near Belau. These have migrated eastward by episodic back-arc opening accompanying eastward migration of arcs and trenches. Oldest known KPR rocks ( ca 47.5 Ma at DSDP site 296), and presumed KPR-derived exotic terranes of Guam ( ca 43.8 Ma), presage the postulated Eocene ( ca 42–43 Ma) change in Pacific plate motion invoked as the cause of subduction initiation at the KPR. The KPR has been rotated more than 40° clockwise since the Eocene, thus the age mismatch may indicate a different tectonic style, for example transtension or transpression, in earliest KPR history.     

Implications from the seismic crustal structure of the northern Izu–Bonin arc

The results of a controlled source seismic reflection–refraction experiment carried out in 1992 reveal the following characteristics of the northern Izu–Bonin (Ogasawara) oceanic island arc–trench system. (1) The crust rapidly thickens from the Shikoku back-arc basin to the arc, is thickest beneath the active rifts, and then gradually thins to the forearc. The thickness of the crust beneath the arc rift zone and the back-arc basin are ∼ 20 km and 8 km, respectively. (2) The Moho vanishes beneath the forearc. Velocities rapidly decrease eastwards beneath the inner trench wall. (3) The velocity of the lower crust of the arc and the back-arc basin is 7.1–7.3 km/s. This velocity is higher than the typical oceanic lower crust whose velocity is ∼ 6.7 km/s. (4) The velocity of the middle crust of the arc is ∼ 6 km/s. This layer does not exist beneath the back-arc basin. (5) A slight difference in the velocity gradient of the middle crust exists between the arc rift zone and the forearc. Based on these findings and previous studies, it is inferred that: (i) the middle crust is probably granitic rock and formed in more than two episodes; (ii) the lower crust formed by igneous underplating which may also have affected part of the back-arc basin; and (iii) the root of the serpentinite diapir on the inner trench wall is a low-velocity mantle wedge that was probably caused by large amounts of water released from the subducting Pacific plate at depths shallower than 30 km.     

Triple oxygen isotope constraints on the origin of ocean island basalts

Understanding the origin of ocean island basalts(OIB) has important bearings on Earth's deep mantle.Although it is widely accepted that subducted oceanic crust, as a consequence of plate tectonics, contributes material to OIB's formation, its exact fraction in OIB's mantle source remains ambiguous largely due to uncertainties associated with existing geochemical proxies. Here we show, through theoretical calculation, that unlike many known proxies, triple oxygen isotope compositions(i.e.D^(17 )O) in olivine samples are not affected by crystallization and partial melting. This unique feature, therefore, allows olivine D^(17 )O values to identify subducted oceanic crusts in OIB's mantle source. Furthermore, the fractions of subducted ocean sediments and hydrothermally altered oceanic crust in OIB's mantle source can be quantified using their characteristic D^(17 )O values. Based on published D^(17 )O data, we estimated the fraction of subducted oceanic crust to be as high as 22.3% in certain OIB, but the affected region in the respective mantle plume is likely to be limited.     

40Ar/ 39Ar and K–Ar geochronological age constraints for the inception and early evolution of the Izu–Bonin – Mariana arc system

K–Ar and ⁴⁰Ar/³⁹Ar dates are presented for locations in the Izu–Bonin – Mariana (IBM) forearc (Ocean Drilling Program (ODP) sites 786 & 782, Chichijima, Deep Sea Drilling Program (DSDP) sites 458 & 459, Saipan), and Palau on the remnant arc of the Kyushu–Palau Ridge. For a number of these locations, the ⁴⁰Ar/³⁹Ar plateau and ³⁶Ar/⁴⁰Ar versus ³⁹Ar/⁴⁰Ar isochrons give older ages than the K–Ar results. The most important results are: (i) at site 786, initial construction of the proto-IBM (now forearc) basement occurred at least by ca 47–45 Ma, consistent with the age of the immediately overlying sediments (middle Eocene nannofossil Zone CP13c); the younger pulse of construction dated at ca 35 Ma by K–Ar could not be confirmed by ⁴⁰Ar/³⁹Ar analysis; (ii) ⁴⁰Ar/³⁹Ar ages for the initial construction of the Mariana portion of the IBM system are as old as those of the Izu–Bonin portion, for example at site 458, initial construction commenced at least by ca 49 Ma and at ca 47 Ma at Saipan (Sankakayuma Formation); and (iii) a combination of K–Ar and ⁴⁰Ar/³⁹Ar ages indicate continued boninite magmatism in the Izu–Bonin forearc (and remnant arc at Palau) until ca 35 Ma. Subduction inception including boninite series rocks along most of the exposed length of the IBM system, clearly preceded by some 5 million years the Middle Eocene (ca 43.5 Ma) change in Pacific plate motion. Boninitic series magmatism persisted at locations now exposed in the forearc for ～ 15 million years after arc inception concurrently with low-K tholeiitic series eruptions from a subaerial arc system, established at ≥ 40 Ma, on the Kyushu–Palau Ridge. For the Mariana portion of the IBM system, reconstruction of the proto-arc places this activity adjacent to the concurrent but orthogonally spreading Central Basin Ridge of the West Philippine Basin. It is possible that a combination of subduction of a young North New Guinea Plate beneath newly created back-arc basin crust may account for some of the features of the Mariana system. It is clear, however, that the understanding of the processes of subduction initiation and early IBM arc development is incomplete.     

Initiation and evolution of intra-oceanic subduction in the Uralides: Geochemical and isotopic constraints from Devonian oceanic rocks of the Southern Urals, Russia

Abstract The Southern Uralides are a collisional orogen generated in the Late Devonian–Early Carboniferous by the collision of the Magnitogorsk island arc (MA) generated in the Early to Middle Devonian by intra‐oceanic convergence opposite to the continental margin, and the continental margin of the East European craton. A suture zone of the arc to the continental margin, the Main Uralian Fault (MUF), is marked by ophiolites and exhumed high‐pressure–low‐temperature metamorphic rocks of continental origin. The pre‐orogenic events of the Southern Urals and their geodynamic setting are traced by means of fluid‐immobile incompatible trace elements (rare earth elements and high field strength elements) and Sr–Nd–Pb isotope geochemistry of the MA suites, in particular the protoarc suite with boninites and probably ankaramites, and the mature arc comprised of island arc tholeiitic (IAT) suites, transitional IAT to calc‐alkaline (CA), and CA suites. The MA volcanics result in genetically distinct magmatic source components. In particular, depleted normal‐mid‐oceanic ridge basalt‐type mantle sources with various enrichments in a slab‐derived aqueous fluid component are evident. The enriched component is not involved in significant amounts, as testified by the rather radiogenic Nd isotopes and unradiogenic Pb isotopes. Further information on the pre‐orogenic events is provided by the Mindyak Massif metagabbros derived from diverse gabbroic protoliths that were affected by oceanic rodingitization, and subsequently by a high‐temperature (HT) metamorphism related to the development of a metamorphic sole. The HT metamorphism has the same age as the protoarc volcanism, and constrains the initiation of subduction at approximately 410 Ma. Consequently, the maximum timespan between initial intra‐oceanic convergence and final collision is approximately 31 my, a duration consistent with that of present‐day ongoing collisions in the western Pacific. The characteristics of early volcanism and the traces of a metamorphic sole provide useful criteria to attribute most MUF ophiolites to the Tethyan type with a complex pre‐orogenic evolution.     

Continuous real‐time analysis of the isotopic composition of precipitation during tropical rain events: Insights into tropical convection

To investigate stable isotopic variability of precipitation in Singapore, we continuously analysed the δ‐value of individual rain events from November 2014 to August 2017 using an online system composed of a diffusion sampler coupled to Cavity Ring‐Down Spectrometer. Over this period, the average value (δ¹⁸O_Avg), the lowest value (δ¹⁸O_Low), and the initial value (δ¹⁸O_Init) varied significantly, ranging from ?0.45 to ?15.54‰, ?0.9 to ?17.65‰, and 0 to ?13.13‰, respectively. All 3 values share similar variability, and events with low δ¹⁸O_Low and δ¹⁸O_Avg values have low δ¹⁸O_Init value. Individual events have limited intraevent variability in δ‐value (Δδ) with the majority having a Δδ below 4‰. Correlation of δ¹⁸O_Low and δ¹⁸O_Avg with δ¹⁸O_Init is much higher than that with Δδ, suggesting that convective activities prior to events have more control over δ‐value than on‐site convective activities. The d‐excess of events also varies considerably in response to the seasonal variation in moisture sources. A 2‐month running mean analysis of δ¹⁸O reveals clear seasonal and interannual variability. Seasonal variability is associated with the meridional movement of the Intertropical Convergence Zone and evolution of the Asian monsoon. El Niño–Southern Oscillation is a likely driver of interannual variability. During 2015–2016, the strongest El Niño year in recorded history, the majority of events have a δ¹⁸O value higher than the weighted average δ¹⁸O of daily precipitation. δ¹⁸O shows a positive correlation with outgoing longwave radiation in the western Pacific and the Asian monsoon region, and also with Oceanic Niño Index. During El Niño, the convection centre shifts eastward to the central/eastern Pacific, weakening convective activities in Southeast Asia. Our study shows that precipitation δ‐value contains information about El Niño–Southern Oscillation and the Intertropical Convergence Zone, which has a significant implication for the interpretation of water isotope data and understanding of hydrological processes in tropical regions.     

Estimating the thickness of a faulted mechanical layer based on fault spacing: An example from Miyako‐jima Island,southern Ryukyu Arc,Japan

The thickness of the mechanical layer that hosts a group of faults can be estimated from the spacing of saturated faults (i.e. the constant spacing between faults when the fault system is fully developed and has attained its final fault density). We measured fault spacing for a group of saturated active normal faults on Miyako‐jima Island (southern Ryukyu Arc, Japan) and estimated the thickness of the faulted mechanical layer. The measured fault spacing is 1.30 ±0.14 km, and the thickness of the mechanical layer is < 3 km, which is the upper limit of the seismogenic zone. This faulted mechanical layer corresponds to a sedimentary layer in which earthquakes cannot occur. Results indicate that the shallow (< 3 km depth) normal faults on Miyako‐jima Island do not have the potential to cause medium‐size earthquakes as individual faults. The origin of the shallow normal faults might be related to the presence of a larger‐scale, deeper fault. The results indicate that fault spacing provides important information on the potential magnitude of earthquakes associated with active faults.     

Geochemistry and Sr–Nd–Pb isotopic constraints on the petrogenesis of Cenozoic lavas from the Pali Aike and Morro Chico area (52°S), southern Patagonia,South America

Geochemical and isotopic analyses (Sr–Nd–Pb) of late Miocene to Quaternary plateau lavas from the Pali Aike and Morro Chico areas (52°S) were undertaken to constrain the melting processes and mantle sources that contributed to magma generation and the geodynamic evolution of southernmost Patagonia, South America. The Pali Aike and Morro Chico lavas are alkaline (Pali Aike, 45–49 wt.% SiO₂; 4.3–5.9 wt.% Na₂O+K₂O) and subalkaline (Morro Chico, 50.5–50.8 wt.% SiO₂; 4.0–4.4 wt.% Na₂O+K₂O), relatively primitive (Pali Aike, 9.5–13.7 wt.% MgO; Morro Chico, 7.6–8.8 wt.% MgO) mafic volcanic rocks that have typical intraplate ocean island basalt‐like signatures. Incompatible trace element ratios and isotopic ratios of the Pali Aike and Morro Chico lavas differ from those of the majority of Neogene southern Patagonian slab window lavas in showing more enriched characteristics and are similar to high‐μ (HIMU)‐like basalts. The rare earth element (REE) modeling to constrain mantle melting percentages suggests that these lavas were produced by low degrees of partial melting (1.0–2.0% for Pali Aike lavas and about 2.6–2.7% for Morro Chico lavas) of a garnet lherzolite mantle source. The major systematic variations of Sr–Nd–Pb isotopes in southern Patagonian lavas are related to geographic location. The Pali Aike and Morro Chico lavas from the southernmost part of Patagonia have lower ⁸⁷Sr/⁸⁶Sr and higher ¹⁴³Nd/¹⁴⁴Nd and ²⁰⁶Pb/²⁰⁴Pb ratios, relative to most of the southern Patagonian lavas erupted north of 49.5°S, pointing to a HIMU‐like signature. An isotopically depleted and HIMU‐like asthenospheric domain may have been the main source of magmas in the southernmost part of Patagonia (e.g. Pali Aike, Morro Chico, and Camusu Aike volcanic field), suggesting the presence of a major discontinuity in the isotopic composition of the asthenosphere in southern Patagonia. On the basis of geochemical and isotope data and the available geological and geotectonic reconstructions, a link between the HIMU asthenospheric mantle domain beneath southernmost Patagonia and the HIMU mega‐province of the southwestern Pacific Ocean is proposed.     

Evolution of West Rota Volcano, an extinct submarine volcano in the southern Mariana Arc: Evidence from sea floor morphology, remotely operated vehicle observations and 40Ar–39Ar geochronological studies

Abstract West Rota Volcano (WRV) is a recently discovered extinct submarine volcano in the southern Mariana Arc. It is large (25 km diameter base), shallow (up to 300 m below sealevel), and contains a large caldera (6 × 10 km, with up to 1 km relief). The WRV lies near the northern termination of a major NNE‐trending normal fault. This and a second, parallel fault just west of the volcano separate uplifted, thick frontal arc crust to the east from subsiding, thin back‐arc basin crust to the west. The WRV is distinct from other Mariana Arc volcanoes: (i) it consists of a lower, predominantly andesite section overlain by a bimodal rhyolite‐basalt layered sequence; (ii) andesitic rocks are locally intensely altered and mineralized; (iii) it has a large caldera; and (iv) WRV is built on a major fault. Submarine felsic calderas are common in the Izu and Kermadec Arcs but are otherwise unknown from the Marianas and other primitive, intraoceanic arcs. ⁴⁰Ar–³⁹Ar dating indicates that andesitic volcanism comprising the lower volcanic section occurred 0.33–0.55 my ago, whereas eruption of the upper rhyolites and basalts occurred 37–51 thousand years ago. Four sequences of rhyolite pyroclastics each are 20–75 m thick, unwelded and show reverse grading, indicating submarine eruption. The youngest unit consists of 1–2 m diameter spheroids of rhyolite pumice, interpreted as magmatic balloons, formed by relatively quiet effusion and inflation of rhyolite into the overlying seawater. Geochemical studies indicate that felsic magmas were generated by anatexis of amphibolite‐facies meta‐andesites, perhaps in the middle arc crust. The presence of a large felsic volcano and caldera in the southern Marianas might indicate interaction of large normal faults with a mid‐crustal magma body at depth, providing a way for viscous felsic melts to reach the surface.