Scientific objectives

Understanding how subduction zones are initiated and continental crust forms in intraoceanic arcs requires knowledge of the inception and evolution of a representative intraoceanic arc, such as the IBM system. An intraoceanic setting is mandatory to avoid the obscuring geochemical, geophysical, and structural veils of preexisting continental crust and the practicality of depth of recovery of basement sections by drilling. The IBM satisfies these criteria. Understanding the evolution of the IBM system, particularly in the more recent half of its 50 m.y. history, has improved considerably over the past three decades, not the least from studies of ash and pyroclast materials recovered by ocean drilling. However, we have poorer records for the nature of arc development in the first half of the history of the system and very limited understanding of how this (or any other) arc was initiated. Expedition 351 will target, in particular, evidence for the earliest evolution of the IBM system following inception.

Expedition 351 is complementary, yet distinct, to IODP Expeditions 350 and 352. Expedition 350 will examine the history of rear arc volcanism adjacent to the currently active volcanic front of the IBM arc, whereas Expedition 352 will explore the sedimentary and igneous history preserved in the current IBM fore arc, potentially spanning the entire history of the system from inception onward. To some extent, the recovery objectives of the magmatic history of the arc during all these expeditions directly overlap but will also gain the vitally important perspective of across-arc strike recovery and identification of compositional variability of eruptive and potentially intrusive magmatic products.

Our drilling plan for Expedition 351 is conceptually straightforward, targeting a single site involving penetration of a thick section of sediment overlying oceanic crust of normal thickness; nevertheless, the water depth (4720 m), sediment thickness (1300 m), and consequent depth to basement are technically challenging.

In detail, there are several primary and some secondary objectives of Expedition 351 involving both sediment (oceanic crust Layer 1; see description in “Seismic studies/Site survey data”) and igneous basement (Layer 2) (Table T1). Overall, the geochemical and geophysical properties of the initial basement (pre-middle Eocene Layers 1 and 2 of the ASB) likely underlie the entire IBM system; these properties are therefore essential for a refined interpretation of the seismic structure obtained for the system.

Primary objectives

1. Determine the nature of the original crust and mantle that existed in the region prior to the beginning of subduction in the middle Eocene (targets: Layers 2 and 1).

An essential boundary condition for understanding the evolution of island arcs is to know the composition, structure, and age of the crust and mantle that existed before subduction began. For example, unlike the case for calculations of the crust formation rate at mid-ocean ridges, estimates of intraoceanic arc fluxes commonly subtract a “standard” 7 km thickness of oceanic crust within the total arc thickness as a probable pre-arc constituent (Reymer and Schubert, 1984; Fliedner and Klemperer, 2000). In addition, the simple assumption is made that Layer 2 of this crust is equivalent to normal mid-ocean-ridge basalt (MORB), despite the evidence for significant variability both compositionally and structurally worldwide of this layer (Cannat et al., 2006). Depending on the mode of arc growth, preexisting nonarc crustal components should contribute geochemically through assimilation and partial melting processes triggered during the passage of later arc magmas and could make up an important part of the lower arc crust. We know that contamination-assimilation processes are commonly established during development of arc magmas, typically only clearly recognizable when strong geochemical contrasts exist between potential contaminant-assimilant and the original magma (Arculus et al., 1983). In general, the presence of basement crust is assumed because recovery of samples from submature arc depths of 15–20 km is impossible. In the case of the proto-IBM system, we have an opportunity to define the geochemical nature of the preexisting basement; absolute trace element abundances together with ratios of these and various isotopic indicators of the basement rocks can then be utilized to quantify models of interaction between ascending IBM arc magmas and the basement.

In the northern IBM case, remarkably, the preexisting oceanic crust exists under 1–1.5 km of sediment in the ASB adjacent to the KPR remnant arc and perhaps also crops out on the lower fore-arc slope of the Bonin Trench (DeBari et al., 1999; Ishizuka et al., 2011a), making possible access to samples of the pre-arc oceanic crustal basement upon which the arc was constructed. We know the age of IBM inception was at ~52–50 Ma (Cosca et al., 1998; Ishizuka et al., 2011a). The ages of initial lithosphere foundering and the change to downdip subduction are consistent with geochronology of the Hawaiian-Emperor Seamount Chain putatively recording the change in Pacific plate motion. Recently published geochronology suggests that the bend in the seamount chain started at ~50 Ma and occurred over a period of ~8 m.y. (Sharp and Clague, 2006).

All of the back-arc basins of the Philippine Sea plate are underlain by asthenosphere of Indian Ocean character, geochemically distinct from mantle sources beneath the Pacific plate now being subducted along its eastern margin (Hickey-Vargas, 1998). In terms of a variety of radiogenic isotopic characteristics (e.g., Pb-Sr-Nd-Hf), the Pacific and Indian Ocean MORB and OIB are distinct (Hofmann, 1997). Their mantle sources, as a consequence, must have contrasting geochemical compositions, at least in terms of the ratios of parent/daughter pairs of isotopic systems. Although the origins and development of these characteristics are controversial (e.g., Mahoney et al., 1998, and references therein), it appears breakup and dispersal of the Gondwanan continental lithospheric fragments during the Paleozoic through to present have exposed asthenospheric mantle beneath the Indian Ocean that has been significantly modified geochemically compared with that of the sub-Pacific mantle. Juxtaposition of the Indian and Pacific MORB-source mantle types across a single one of a dense network of anomalously deep transform faults at the Australian-Antarctic Discordance (AAD) was explored, for example, during ODP Leg 187 (Kempton et al., 2002). The extent of Indian-type mantle migration eastward into the Pacific domain is not fully constrained, but the present locus of the IBM and former Vitiaz arc systems appear to mark a present-day limit. Possibly, the initiation of these arcs was in some way controlled by the juxtaposition of Pacific- and Indian-type mantles (e.g., localized by the type of transform fault cluster exposed at the AAD).

It is nevertheless clear the initial construction of the IBM arc, rather than developing solely upon oceanic crust, transected the present-day series of Cretaceous–Paleocene ridges (e.g., Amami Plateau and Daito and Oki-Daito Ridges) and intervening basins that formed, at least in part, an arc–back-arc system (Taylor and Goodliffe, 2004; Hickey-Vargas, 2005; Ishizuka et al., 2011b) (Fig. F2). Recent isotopic results for the Amami Plateau indicate the Philippine Sea plate also contains Pacific Ocean-type lithosphere, and the nature of the lithosphere on which the proto-IBM arc was built was likely diverse in character (Hickey-Vargas et al., 2008). It may be that decoding the nature of the magma source in the upper mantle that existed immediately before the IBM arc inception is a key to understanding the cause of the initiation of subduction zones and intraoceanic arc formation. It is also possible, of course, that the basement of the ASB is of back-arc character, generated from upper mantle that was contaminated by Cretaceous subduction processes. These questions can only be resolved by recovering and studying samples of the ASB basement.

2. Identify and model the process of subduction initiation and initial arc crust formation (target: Layer 1).

According to the model of forced subduction initiation, the nucleating margin will first undergo compression and localized uplift; the Macquarie Ridge complex (MRC) south of New Zealand is a present-day example. Some segments of the margin may be forced above sea level (such as at Macquarie Island along the MRC). Models show that the magnitude and horizontal wavelength of uplift are dependent on the age of the overriding plate, and knowledge of this age (see primary Objective 1 basement and also from ages of deepest sediment overlying the basement at Site IBM-1, this objective) is an important geodynamic input (Gurnis, et al., 2004). The self-nucleating model does not predict such a phase of uplift but predicts early extension. Understanding the response of the overriding plate during the initial stages in formation of the new subduction zone is thus essential for testing first-order competing proposals for subduction initiation.

The sedimentary sequence developed during subduction initiation to be recovered from the ASB at Site IBM-1 is crucial with respect to understanding the tectonic, petrologic, and geochemical consequences of the critical stages of arc inception. Depending on the wavelength of structural disturbance to the overriding plate, we anticipate erosion of crustal materials derived from the ridges and plateaus around the western and southern margins of the ASB. Sufficient geochemical and geochronological data have been collected (Ishizuka et al., 2011b, in press; Tani et al., 2012) for these crustal sources that the materials shed into the ASB at this time will be recognizable in recovered cores at Site IBM-1. Similarly, uplift along the present trace of the KPR at the northeastern margin of the ASB may have reworked preinception sediment, and this together with juvenile arc-inception volcaniclastic materials will also be recognizable in cores. Any paleontological monitors of seafloor depth changes through subduction initiation and later periods of time will be critical for analyses of bathymetric changes to the ASB seafloor (e.g., changes in the CCD as well as the nature and preservation of benthic foraminifers).

The character of the earliest juvenile magmatic outputs of the nascent IBM arc recovered from the sedimentary sequence at Site IBM-1 is of first-order importance for (1) constraining the composition of the mantle wedge at this stage of arc development, (2) determining the nature of subducted inputs, and (3) comparing with the magmatic sequences recovered so far by drilling and dredging in the current fore arc (e.g., Reagan et al., 2010) to be sought during Expedition 352. We have an opportunity to explore possible across-strike variations in the nature of juvenile arc magmatic outputs and to establish whether the sequence recognized for the fore arc from earliest MORB-like “fore-arc basalt” through boninite to the temporally more enduring and apparently stable output of island arc tholeiitic types, which also existed at the western margin of the earliest IBM arc.

In addition to chemical compositions and ages of fragmental materials deposited in the ASB concurrent with arc initiation, direct structural analysis of the cores will be a first-order requirement providing evidence for prevailing tension (normal faults), compression (thrust faults), or potentially tranquil tectonic environments. Evidence for temporal variations in these conditions will also be sought in order to model the state of stress in the overriding plate during arc initiation.

3. Determine the compositional evolution during the Paleogene of the IBM arc (target: Layer 1).

The complete tephra record of fore-arc/arc/back-arc volcanism subsequent to arc initiation in the middle Eocene to the isolation of the KPR as a remnant arc accompanying inception of the Shikoku back-arc basin at ~25 Ma (and possibly sporadically thereafter) will be obtained at Site IBM-1. Comprehensive analytical data exist for Neogene ash and pyroclastic materials recovered from the IBM fore arc (e.g., Bryant et al., 1999; Straub, 2003; Straub et al., 2004) and indicate remarkable stability (a function of subducted slab inputs, mantle wedge replenishment, and overriding plate inputs) of the northern part of this system, but the Paleogene record is sparse. In combination with the known Eocene–Recent lava/plutonic products (Ishizuka et al., 2011b), the ash and pyroclast record during this interval will allow us to determine the output variation through time along a transect of the northern IBM arc compared to Pacific plate inputs.

4. Establish geophysical properties of the ASB (targets: Layers 1 and 2).

The basement of the IBM arc comprises sedimentary and underlying igneous rock types that will become accessible for direct geophysical measurements through study of recovered cores and downhole logging at Site IBM-1 during Expedition 351. Following publication of the seminal seismic cross section obtained by Suyehiro et al. (1996), several other across- and along-strike seismic surveys have expanded our knowledge of the velocity structure of the IBM system (e.g., Kodaira et al., 2007). Recovery and analysis of core samples, together with comprehensive downhole logging of physico-chemical-structural properties of the pre-arc basement and overlying Layer 1, will clearly advance our understanding of the overall nature of the arc structure, permitting reliable estimates of age-composition characteristics of crustal units and enabling refined calculation of the critically important question of arc crustal growth rates.

There has been a long-standing debate concerning the initial geographic orientation and locations of the IBM system during subduction initiation (argued to be ~90) and its subsequent postulated 90 clockwise rotation to a present-day north–south strike during 50 m.y. (Hall et al., 1995; Yamazaki et al., 2010, and references therein). Among the primary objectives of Expedition 351 will be establishing the magnetic declination and its putative change as a function of time for both the ASB basement and overlying sediment of Layer 1. Although these measurements are technically challenging, there is a particular advantage offered at Site IBM-1 in attempting magnetic orientation measurements because of the apparently relatively undisturbed character of the strata forming Layer 1. The initial orientation of the IBM system and any subsequent rotations are important with respect to establishing the boundary conditions for subduction initiation and later plate interactions.

Secondary objectives

1. Recover sedimentary records of paleoceanographic conditions from Pliocene–Pleistocene to early Tertiary and possibly Late Cretaceous in the eastern Tethys–western Pacific (target: Layer 1).

The paleographic position of the ASB adjacent to Cretaceous-aged arcs and back-arc basins potentially provides an excellent opportunity to recover records of paleoceanographic conditions in the far eastern Tethys–western Pacific in the Late Cretaceous (and possibly older) through to the Tertiary. Among the exciting opportunities offered by penetration of the full thickness of oceanic crustal Layer 1 at Site IBM-1 are the transection of the Cretaceous/Tertiary boundary, the organic-rich shales of OAEs, the Paleocene/Eocene Thermal Maximum, and Eocene hyperthermals. If found, these records would be valuable, as there are very few sections recovered from this region. For example, Cretaceous OAEs would provide important constraints on the prevalence of anoxia at an extreme eastern meridional location compared with the classic Tethyan sites.

2. Recover sedimentary records of onset and persistence of the East Asian Monsoon and other climate-modulated land-sea correlations (target: Layer 1).

Integrated Ocean Drilling Program Expedition 346 will conduct two latitudinal transects (Japan Sea and northern East China Sea) to test hypotheses concerning tectonic linkages with the onset of the Asian Monsoon. Site IBM-1 is located to the southeast of the scheduled drilling sites in the East China Sea and may be complementary to those of Expedition 346. Although our site is distal from the Yangtze River discharge, a record of fluvial input is possible. Furthermore, potential aeolian inputs to the ASB include loess from east Asia. Thus, high-quality sediment recovered by advanced piston corer (APC) coring at Site IBM-1 should provide additional important information regarding timing and geographic distribution of terrestrial material in the western Pacific marine record.

3. Recover an ash record of the evolution of the Ryukyu-Kyushu arc (target: Layer 1).

Subduction of the Philippine Sea plate, on which the ASB now is situated, takes place in part along the Ryukyu-Kyushu arc to the northwest of Site IBM-1 (Fig. F2). We anticipate an ash record from this arc to be preserved within the sediment at Site IBM-1. Some particularly significant explosive events from the Ryukyu-Kyushu arc produced extensive ash deposits that are used as stratigraphic marker horizons in the Japanese islands. In addition to this feature, the temporal evolution of the Ryukyu-Kyushu arc located on the east Asian continental margin will provide important complementary evidence of arc evolution with the IBM system.