This volume presents results of recent scientific investigations of the Izu–Bonin (Ogasawara) – Mariana (IBM) arc system. The IBM arc system extends 2500 km south from near Tokyo, Japan, to beyond Guam, USA (Fig. 1), and is an excellent example of an intra-oceanic convergent margin (IOCM). The IOCM are built on oceanic crust and contrast with arcs built on continental crust such as Japan or the Andes. Andean-type convergent margins are better studied than IOCM because the former are more visible and accessible, and are greater hazards to humanity. There are nevertheless numerous reasons why studying IOCM is essential for understanding how subduction zones work. Because anatexis of sialic crust is precluded at IOCM, studying their magmatic products removes one layer of complexity about how arc magmas form and fractionate. These igneous rocks are derived entirely from subducted materials and the mantle wedge, possibly accompanied by re-melting of juvenile arc crust, so their study allows a much more confident assessment of mantle-to-crust fluxes and processes to be made than is possible for arcs built on continental crust. The IOCM offer modern analogs for how continental crust forms, constraining models of process and rate. Because IOCM are far removed from continents they are not affected by the large volume of sediments carried to the sea by glaciers and rivers. The consequent thin sedimentary cover makes it much easier to study arc infrastructure and determine the mass and composition of subducted sediments. Active hydrothermal systems found on the submarine parts of IOCM give us a chance to study how many of Earth's important ore deposits formed. The IBM arc system is a particularly good example of an IOCM, for the following reasons. (1) The IBM arc system is an end-member convergent margin. The Marianas are the type example of a `decoupled' arc, characterized by a deep, steep subduction zone (Uyeda & Kanamori 1979). Seismic expression of the subduction zone extends down to the maximum depth on earth (700 km beneath the Marianas; Jarrard 1986). Convergence rates vary from 3 cm/year in the south to 7 cm/year in the north (Seno 1989), providing an opportunity to examine relationships between arc processes and convergence rate. According to the strain classification of Jarrard (1986), IBM ranges from class 1 in the south (active back-arc spreading) to class 2 (very slow back-arc spreading) in the north. These along-strike variations are partly due to the fact that the age of the sea floor being subducted varies from ca 125 Ma in the north to ca 170 Ma east of the central Marianas (Nakanishi et al. 1992). The increase in age and density of the subducted plate from north to south along the arc is reflected in the deep structure of the subduction zone, where seismic tomography reveals that the subducted slab beneath the Izu–Bonin arc does not penetrate the 670-km discontinuity, whereas the slab beneath the Marianas appears to descend vertically to at least 1200 km deep in the mantle (van der Hilst et al. 1991). This may reflect the magnitude of body forces associated with subduction of Earth's oldest oceanic lithosphere. Similarly, the uniquely great age of lithosphere being subducted means that the subduction zone is uniquely cold (Bird 1978) so that elemental transport from the subducted slab to the overlying mantle is more likely to be dominated by fluids as opposed to melts than any other subduction zone on Earth (Peacock 1990). Finally, Mariana-type subduction zones are used as the low-seismicity end-member in the continuum of types of subduction channel models (Cloos & Shreve 1996). (2) The IBM arc system is large and diverse. The IBM arc system is one of the largest convergent margins on earth, and its vastness allows for the expression of a wide range of tectonic and magmatic styles. The IBM arc system from Japan to Guam is ∼ 2500 km long, a length that is rivaled among intra-oceanic convergent margins only by its twin in the southwest Pacific, the Tonga–Kermadec arc system, and by the much more complex Melanesian arc system. If the festooning Yap and Palau segments are included in the IBM arc system (to which they clearly belonged before the mid-Tertiary opening of the Parece Vela Basin), the IBM arc system spans approximately a third of a hemisphere. The arc system is also broad, with a maximum width from the remnant arc of the Kyushu–Palau Ridge to the Mariana Trench of ∼ 1200 km. Although tectonic and magmatic activity is now limited to the eastern margin of this region, crust produced by the IBM arc system during its 45 Ma history covers an area slightly larger than the subcontinent of India. Different parts of the IBM arc system are characterized by different types of tectonic and igneous activity. In the north, the deeper parts of the Wadati–Benioff zone dip at ∼ 50° but steepen progressively southward, becoming vertical beneath the Marianas (Chiu et al. 1991). North of 24°N the arc and trench are straight, whereas the Marianas are strongly curved. Few seamounts are found near the trench on the Pacific plate north of 28°N, but these are much more common to the south, including the Ogasawara Plateau, the scattered seamounts of the Magellan chain, and the immense Caroline Ridge, the collision of which terminated subduction along the Yap Trench. The Izu–Bonin forearc is characterized by well-developed submarine canyons but lacks active serpentinite mud volcanoes, while the Mariana forearc has active serpentinite mud volcanoes but lacks submarine canyons. The frontal arc is developed discontinuously, for reasons that no one understands. The pedestal on which volcanoes of the active arc are built changes from lying a few hundred meters below sea-level in the north to lying more than 3 km deep in the vicinity of the Sofugan Tectonic Line and in the Marianas (Fig. 2). Active volcanoes rise above sea-level in the north, middle, and southern ends of the IBM arc system. Arc lava compositions also vary strongly along-strike of the IBM arc. These are compositionally bimodal in the north, where some of the most depleted arc tholeiites on earth are erupted along with subordinate low-K dacites; unusually enriched in the central portion, characterized by shoshonitic lavas; and transitional between calc-alkaline and tholeiitic in the Marianas. Arc volcano cross-chains characterize the arc north of the Sofugan Tectonic Line, are absent from the middle portion, and are occasionally found in the Mariana arc. The IBM arc is affected by intra-arc rifting in the north, flanks an actively spreading back-arc basin in the south, and is unrifted in the centre (Fig. 2). The geodynamic and magmatic diversity encountered across and along the IBM arc allows us many opportunities to try and better understand what is controlling these various manifestations of convergent margin tectonics and magmatism. Along-strike profiles of the Izu–Bonin – Mariana (IBM) arc system, from Japan (left) to Guam (right). The thick solid line shows the bathymetry and topography along the volcanic axis of the active arc, with the thin dashed horizontal line marking sea-level. The approximate locations of the principal island groups (Izu, Ogasawara (Bonin) –Volcano, and Mariana) are shown. Submarine volcanoes (and the Sofugan Tectonic Line, STL) are given as italicized abbreviations: Ku, Kurose; Ms, Myojin-sho; Do, Doyo; Kk, Kaikata; Kt, Kaitoku; F, Fukutoku-oka-no-ba; HC, Hiyoshi Volcanic Complex; Nk, Nikko; Fj, Fukujin; Ch, Chamorro; D, Diamante; R, Ruby; E, Esmeralda; T, Tracy. Subaerial volcanoes are given as normal abbreviations: O, Oshima; My, Miyakejima; Mi, Mikurajima; H, Hachijojima; A, Aogashima; Su, Sumisujima; T, Torishima; Sg, Sofugan; N, Nishinoshima; KIJ, Kita Iwo Jima; IJ, Iwo Jima; MIJ, Minami Iwo Jima; U, Uracas; M, Maug; As, Asuncion; Ag, Agrigan; P, Pagan; Al, Alamagan; G, Guguan; S, Sarigan; An, Anatahan. Locations of important zones of intra-arc and back-arc extension in the north (Bonin Arc Rifted Zone) and south (Mariana Trough Back-Arc Basin) are marked. The thick dashed line shows the maximum depth in the trench along its strike. Finally, the political limits of Japanese and US political jurisdiction are shown on the bottom. Frontal arc elements are not shown, but consist of the Bonin or Ogasawara Islands between 26° and 28°N and the Mariana frontal arc islands between 13° and 16 (3) Sedimentary sequences on the Pacific plate are simple, diagnostic, and completely subducted. Sediments on the soon-to-be-subducted Pacific plate comprise pelagic and ocean island basalt (OIB)-derived volcaniclastic sequences that are compositionally distinct from arc-derived sediments filling trenches at other convergent margins. Several multichannel reflection survey (MCS) profiles can be tied to an Ocean Drilling Program (ODP) hole penetrating to basement (Abrams et al. 1993), and the combined data show significant yet simple along-strike variations in the sedimentary sequence on the Pacific plate, most importantly in the abundance of off-ridge seamounts and associated volcaniclastic sequences in the south relative to the north. The systematic variation in the abundance of OIB volcaniclastics in subducted sediments allows for competing models advocating the source of arc elemental enrichments to be derived from subducted sediments (Elliott et al. 1997) or mantle (Stern et al. 1991) to be tested. With the exception of a few seamounts accreted to the inner wall of the IBM trench, nothing from the subducting plate is scraped off, greatly simplifying calculations of the volumes and compositions of input fluxes to the `subduction factory'. (4) Crustal structure is resolved. An ocean bottom seismograph (OBS) refraction profile across the northern part of the IBM arc (Suyehiro et al. 1996) provides the best image of crustal structure available for any intra-oceanic convergent margin, but we need more such profiles if we are to understand variations in crustal structure along this outstanding IOCM example. Thin sediment cover on the Pacific plate combined with a modest flux of sediment from the arc to the trench results in deep, sediment-starved trenches, including the Challenger Deep, the deepest hole on Earth's solid surface. There is little sediment to be scraped off, so the IBM arc system is characterized by the lack of an accretionary prism. The IBM arc system provides an outstanding example of what von Huene and Scholl (1991) call `type 2 trenches', where no accretionary prism forms. The inner trench walls mostly expose igneous basement of the forearc (Bloomer 1983; Ishii 1985). This lack of sedimentary cover means that the forearc basement can be studied, something that cannot be done at convergent margins with high sedimentation rates. This has allowed scientists to demonstrate that IBM forearc lithosphere formed at the time that subduction began, ∼ 45 million years ago or perhaps a little before, and provides important constraints for models of how subduction begins (Bloomer et al. 1995), including how the remarkable group of lavas known as boninites (after the type locality in the Bonin Islands) are generated (Pearce et al. 1992). Low sedimentation rates also allow us to study serpentinite mud volcanoes in the forearc and to sample slab-derived forearc fluids without having to worry about dewatering sediments in the accretionary prism (Fryer et al. 1995). An extremely important aspect of the IBM arc is that its northern end has been colliding with Japan for the past 15 million years. This has exhumed deep crustal sections in the Tanzawa Mountains and exposed arc supracrustal sequences on the Boso and Miura Peninsulas (Niitsuma 1989). These exposures can be correlated to the refraction profile and support the recent hypothesis that IBM middle crust is composed of a thick felsic layer (Suyehiro et al. 1996). The Tanzawa collision zone provides a natural laboratory where processes of terrane accretion can be studied, and the continuing collision poses a great seismic hazard for the greater Tokyo area. At its other end, the IBM arc is sheared off to form the greatest trench in the world, the Challenger Deep, providing another outstanding opportunity to examine deep crustal and upper mantle sections of this arc system. (5) The history of the arc is better known than that of any other convergent margin. This includes the beginning of subduction during the Eocene to the present. The location of especially the southern IBM arc at great distance and upwind from volcanoes of other arcs allows tephra recovered from drill holes and cores to be confidently assigned to this arc (Lee et al. 1995), while arc-derived volcaniclastic turbidites shed into proximal basins provide a record of how individual groups of volcanoes have evolved (Gill et␣al. 1994). Uplift of the forearc to form the frontal arc island chains of the Ogasawaras and southern Marianas provides opportunities to study Eocene arc basement and overlying sedimentary successions in detail. These opportunities combine to make the IBM arc system one of the best places on earth to track the magmatic evolution of a convergent margin through time. In addition to these purely scientific considerations, the IBM arc system lies in a region favored by logistics and international politics. It lies in low latitudes, where sea conditions generally favor marine investigations (although typhoons can cause serious if temporary problems). Its proximity to Japan means that it is easily reached by scientific cruises from that nation. The Mariana Trough contains the only active spreading ridge in the northwest Pacific, so East Asian and Japanese scientists wishing to study mid-ocean ridge processes are naturally attracted to the southern IBM. The United States (US) and Japan share political jurisdiction over the IBM arc system (Fig. 2), the only region where the two nations are geographical partners. These joint interests have deep historical roots: half a century before the US took control of Guam from Spain in 1898, American whalers occupied the Ogasawara Islands and used these to resupply the far-ranging US whaling fleet. The shared interests of the US and Japan favor future international scientific collaborative studies between the two nations, and provides an opportunity for scientists from other nations to participate. In spite of favorable scientific and political considerations, the IBM arc system has not been studied to the extent that it merits. There are many reasons why its study has been inhibited. Nearly all of the arc system is submerged, making it vastly more difficult to study than a similar-sized tract above sea-level. The size of the mostly submerged tract that needs to be studied is intimidating: the 2500 km from Guam to Tokyo is about the distance from San Diego to Vancouver, Dallas to New York, or London to Warsaw. The arc system lies partly in the realm of marine scientists and partly in the realm of land geologists, so the traditional division of earth scientists into marine and terrestrial groups has worked against scientific coordination, even among scientists from a single nation. This is compounded by political jurisdiction, which has focused studies at the ends: by Japanese scientists in the north and by US scientists in the south; with the result that the vast middle of the arc system has not been looked at as carefully. There is little communication between the small populations in the Izu, Ogasawara, and Mariana Islands; if a person wanted to travel from Saipan to Chichijima, he would have to fly to Tokyo and take the once-weekly ferry to the Ogasawara Islands. Efforts in the US sector are further hampered by the small population and land area of the Mariana Islands, which is split between two political entities: the Territory of Guam and the Commonwealth of the Northern Marianas, neither of which have a vote in the US Congress. Because of this and the fact that the Marianas are so far away from North America, US agencies such as National Oceanographic and Atmospheric Administration (NOAA) and the United States Geological Survey (USGS) have not carried out systematic surveys of US territorial waters and exclusive economic zones like those that have been done closer to the US. Marine investigations of the IBM arc system can be dangerous: the greatest peacetime loss of life ever to befall the oceanographic or earth science community occurred when the Myojin Volcano violently erupted on the night of 23 September 1952 and sunk the Japanese Maritime Safety Board survey vessel no. 5, the Kaiyo-maru, with a loss of 31 scientists and crew (Jaggar 1952). Finally, memories of the terrible war between the US and Japan may have inhibited studies in the past, because of the many thousands of casualties suffered during 1944–45 battles on Saipan, Guam, and Iwo Jima. Even an aircraft piloted by US President George Bush was shot down over the Ogasawara Islands! A workshop designed to stimulate studies of the IBM arc system was held from 27 July to 2␣August 1996 in Japan. Ten US and 75 Japanese scientists, as well as representatives from Germany, Britain, and Australia, attended the workshop, supported by the US National Science Foundation and the Japan Society for the Promotion of Science. The first part of the workshop was held at Shonan Village Center, Hayama, outside of Yokohama, and consisted of scientific presentations and discussions of future scientific efforts, along with a visit to facilities of the Japan Marine Science and Technology Center (JAMSTEC) in Yokosuka. The group enjoyed a field excursion to the Tanzawa Mountains in order to study exposures of the IBM arc crust exhumed by continuing collision between the northernmost IBM arc and Honshu. There were four working groups: Subduction Fluxes; Magmatic Evolution; Hydrothermal Activity and Mineralization; and Tectonics and Geophysics, each of which produced recommendations regarding research priorities. A summary of these recommendations has been published (Stern & Arima 1997), and further details can be found on the Worldwide Web at . An important result of the workshop was an invitation to participants to prepare their new findings for publication in a special volume of The Island Arc dedicated to the IBM arc system. This is that volume, and it is the first time that an international journal has been dedicated to this arc system. In the following pages the reader will find an impressive range of scientific reports on various aspects of the IBM arc system. The reader should note that this is not intended to be a general overview of the IBM arc system but a sampling of research now being carried out by the international community on this convergent margin. To make it easier for the reader to relate the different regions being studied, the approximate locations of the various studies are shown in Fig. 11. Location map for the Izu–Bonin – Mariana (IBM) arc system, modified after Taylor (1992). Numbered features are: 1, ridges associated with active and extinct arc segments, mostly shallower than 3 km (except for the Kyushu–Palau Ridge, which is much deeper); 2, active spreading ridge; 3, extinct spreading ridge; 4, convergent plate boundary or trench. Letters correspond to approximate study areas presented in this volume: A, Mohiuddin et al.; B, Lee & Ogawa; C, Soh; D, Kawate & Arima; E, Nakajima & Arima; F, Yamazaki & Yuasa; G, Takahashi et al.; H, Taira et al.; I, Ishizuka et␣al.; J, Usui & Glasby; K, Sun et al.; L, Meen et al.; M, Yamazaki & Murakami; N, Ikeda et␣al.; O, Peate & Pearce; P, Clift & Lee; Q, Stüben et al.; R, Stern & Smoot; S, Ohara & Ishii; T, Hawkins & Castillo; U, Cosca et al.; V, Fryer et al. It is the earnest hope of the editors and all participants in the workshop that this volume stimulates interest and further studies in the IBM arc system. There is a lot of work to be done before we understand all of the lessons that the IBM arc system has to teach us about how subduction zones operate. The reader is invited to join with us to help uncover these secrets!
An abstract is not available for this content so a preview has been provided. As you have access to this content, a full PDF is available via the ‘Save PDF’ action button.
Simultaneous measurements of ultrasonic compressional (V p ) and shear wave velocities (V s ) were conducted under pressures and temperatures representative of the lower crust on five mafic xenoliths derived from the northeast Japan arc crust. For comparison with the observed seismic heterogeneity of the lower crust of the northeast Japan arc, the velocity deviations (dV p and dV s ) of the xenoliths were calculated using the average reference V p (6.61 ± 0.1 km/s) and V s (3.76 ± 0.1 km/s). The dV p and dV s values of the xenoliths at 700°C and 0.8 GPa are 0.8% and −1.2% for hornblende gabbro, 4.8% and 4.0% for hornblende‐pyroxene gabbro, 0.9% and −1.5% for amphibolite, 2.5% and 1.7% for pyroxene amphibolite, and 0.9% and −5.6% for hornblendite, respectively. We compared the velocity perturbations in the lower crust determined by seismic tomography with dV p and dV s of the xenoliths using “V p −V s −V p /V s deviation diagrams.” On the basis of our measurements, we suggest that (1) the seismically high‐V p and V s regions beneath the Tobishima Basin consist of hornblende‐pyroxene gabbro, (2) hornblende gabbro is a predominant rock type beneath the Dewa Hills and Ou Backbone Range, (3) the low‐velocity anomalies beneath the active volcano areas may be caused by the existence of partial melts of hornblende gabbro, and (4) the low‐V p and high‐V s regions beneath the Kitakami Mountains consist of quartz‐plagioclase‐bearing rocks. Our data demonstrate that the seismic heterogeneity in the lower crust of the northeast Japan arc reflects variations in rock composition and temperature that are related to the regional geological history.
The join MgSiO3-MgAl2SiO6 was studied with the special reference to the solubility of Al2O3 in enstatite, above 1300°C. MgAl2SiO6 is incorporated in enstatite as much as 3.5wt.% (1.75wt. % Al2O3) at atmospheric pressure. At subsolidus temperatures the phase assemblage protoenstatitess+forsterite was confirmed in the composition ranging from 96.5 to 94.5wt. % MgSiO3. In the more aluminous region the subsolidus phase as-semblage is protoenstatitess+forsterite+cordieritess. The experimental result indicates that the composition of protoenstatite lies not only on the join MgSiO3-MgAl2SiO6 but also has a solid solution area on the MgO-Al2O3-SiO2 plane as shown in Fig. 4. The unit-cell dimensions of the protoenstatitess crystallized at 1300°C were determined. a, b, and V decrease with increasing Al2O3 contents in the protoenstatitess. The application of the results to meterorite as well as terrestrial rock is discussed. The Al2O3 contents of enstatite in meteorite may be used as an indicator of temperature scale.
The Miocene Tanzawa plutonic complex, consisting mainly of tonalite intrusions, is exposed at the northern end of the Izu–Bonin – Mariana (IBM) arc system as a consequence of collision with the Honshu Arc. The Tanzawa plutonic rocks belong to the calc‐alkaline series and exhibit a wide range of chemical variation, from 43 to 75 wt% SiO 2 . They are characterized by relatively high Ba/Rb and Ce/Nb ratios, and low abundances of K 2 O, LIL elements, and rare earth elements (REE). Their petrographic and geochemical features indicate derivation from an intermediate parental magma through crystal fractionation and accumulation processes, involving hornblende, plagioclase, and magnetite. The Tanzawa plutonic complex is interpreted to be the exposed middle crust of the IBM arc, which was uplifted during the collision. The mass balance calculations, combining data from melting experiments of hydrous basaltic compositions at lower‐to‐middle crustal levels, suggest that parental magma and ultramafic restite were generated by dehydration partial melting (∼ 45% melting) of amphibolite chemically similar to low‐K tholeiitic basalt. Partial melting of hydrated mafic lower crust might play an important role in felsic middle‐crust formation in the IBM arc.
Studies of inclusions in natural diamonds suggest a genetic link between diamond formation and volatile-and alkali-rich mantle melts or fluids, broadly similar to kimberlite composition (Navon et a., 1988).Arima et al. (1993) demonstrated that diamond crystallized and grew in a volatile-rich silicate melt of kimberlite composition in high-pressure and high-temperature experiments (1800-2200°C and 7-7.7 GPa).This paper reports detailed morphological features of newly crystallized diamonds in the experiments by Arima et al. (1993).In addition, resorption (dissolution) of diamond in kimberlitic melts was studied experimentally in the graphite stability field (1300-1500°C and 2.5 GPa) under the W-l condition.This paper also reports a variety of resorption features observed on the seed diamonds.Crystallization and growth of diamond at 7 -7.7 GPa
