IODP

Previous   |    Next

doi:10.2204/iodp.sp.350.2013

Scientific objectives

The primary objective of proposed Site IBM-3C is to test three pairs of alternative hypotheses about crustal genesis and mantle evolution:

  1. Geochemically asymmetric crust, which is most like “average continent” in the rear arc, is either (a) a fundamental trait of crust in oceanic arcs that is produced in a steady state throughout arc history from Paleogene inception or (b) a secondary trait that develops only after back-arc spreading (Fig. F15);

  2. Intracrustal differentiation amplifies this asymmetry (a) continuously as a steady-state process or (b) mostly during nonsteady-state events such as arc rifting; and

  3. After or near the cessation of the Shikoku back-arc basin opening, rear-arc magmatism either (a) started from the western end of the rear-arc seamount chains and migrated eastward (Fig. F7) or (b) started at the same time along the length of the rear-arc seamount chains but ended from west to east (Fig. F5).

Figure F15 illustrates the alternatives for Hypothesis 1 that can be tested by drilling. We call them the “from the beginning” and “from the middle” alternatives. Colors in Figure F15 simplify chemical differences between the volcanic front and rear arc that are predicted by these two hypotheses. During steady-state arc growth, crust develops that is “continental” in velocity structure and seismically similar beneath both the volcanic front and rear arc but is heterogeneous in chemical composition. Magmas at the volcanic front are rich in fluid-mobile recycled slab components (e.g., Sr, Pb, and U) that swamp the mantle, yet these magmas are so depleted in mantle-derived fluid-immobile elements (e.g., Nd, Hf, and Nb) that they are dissimilar to “average continental crust” in detail. This is less true in the rear arc where the less-depleted mantle, diminished slab fluid signature, possible addition of melt from subducted sediment, and lower degrees of mantle melting create crust that is more typical of the continents and allow the temporal history of the mantle source to be tracked more easily. Although the asymmetry is known in general in Izu from Neogene volcanic rocks obtained by dredging, the best way to assess its variability during the Neogene, and to learn how far back in arc history it extends, is to obtain a temporal record by drilling the volcaniclastic sediment in the rear arc. The alternative hypothesis is that the asymmetry is only true in the Neogene Izu arc and that magmatism was uniformly less depleted and/or uniformly rich in fluid-mobile recycled slab components (e.g., Sr, Pb, and U) during the Oligocene and Eocene. The latter would indicate that the subduction parameters that cause geochemical asymmetry differed in early arc history. These two hypotheses, from the beginning and from the middle (Fig. F15), and others can be tested only by recovery of the Eocene–Oligocene tephra and turbidites in the rear arc.

The second hypothesis is that nonsteady-state events play a major role in the evolution of arc crust. One alternative is that intracrustal recycling, which creates felsic magmas and possibly is forming the distinctive 6.0 km/s “tonalitic” middle crust, is heightened during periods of rifting preceding back-arc spreading (e.g., since 3 Ma) and that this recycling amplifies the across-arc chemical asymmetry. We know from ODP Legs 125 and 126 that the current phase of arc rifting produced a marked increase in felsic magmatism at the arc front, and we know from dredging that there are along-arc and across-arc differences in the chemical composition of tonalites and rhyolites, but only drilling can test this hypothesis by providing a stratigraphic record of felsic magmas across the arc, especially in the rear arc. The 7 Ma tonalites from Manji Seamount provide a comparison between extrusive and intrusive rear-arc felsic rocks.

The third hypothesis is that the origin of the Izu rear-arc seamount chains can be related to mantle convection patterns (hot fingers in Fig. F7) (e.g., Tamura et al., 2002; Honda et al., 2007). Numerical simulations of small-scale convection under island arcs (Honda and Yoshida, 2005) suggest that a roll (finger)-like pattern of hot and cold anomalies emerges in the mantle wedge starting from the back-arc side of the rolls. Thus the small-scale convection hypothesis predicts that rear-arc magmatism migrated from west to east.

Road map for testing hypotheses

Testing these hypotheses requires obtaining a temporal record of across-arc variation in magma composition from the Eocene to Neogene. This should enable (1) identification of temporal changes of basaltic magma chemistry and interpretation of the source processes and (2) identification of temporal variation of intermediate and felsic magmas and interpretation of crust-level differentiation processes. This information is in hand for the volcanic front but missing for the rear arc, which overlies the majority of crust that is “continent type” in composition. This rear-arc information is also needed in order to compare these characteristics to what is already known about these parameters for the volcanic front. Specifically, our objectives are to establish the temporal history of across-arc variations during five time periods that stand out in the rear-arc evolution:

  1. 3 Ma to the present: we will determine whether rear-arc and rift-type magmatism have overlapped since the onset of rifting at 3 Ma, and whether rift-type mafic and felsic magmatism changed during that time.

  2. 9–3 Ma: we will also establish whether rear-arc magmatism changed with time, how it compares with arc-front magmatism, and the role of felsic magmatism in producing middle crust in both settings.

  3. 17–9 Ma: if rear-arc magmatism migrated from west to east during this time frame, strata of this age should be missing in the proposed drilling site (Fig. F5).

  4. 25–17 Ma: we will determine whether volcanism stopped in the rear arc during opening of nearby Shikoku Basin, as it did at the volcanic front.

  5. >25 Ma: we will determine whether rear-arc magmatism changed with time; that is, whether Oligocene rear-arc and frontal-arc magmas differed in the Oligocene (during the initial stages of arc development), and especially whether felsic materials differed in their abundance, character, and mode of origin during arc evolution.

These determinations will be made using standard igneous geochemical tools applied to volcaniclastic materials (and any lavas encountered). These tools include bulk rock major, trace element, and Sr-Nd-Hf-Pb isotope chemistry and the same applied to glass shards, minerals, and their melt inclusions. Some of these tools (e.g., REE + HFSE trace elements and Nd-Hf isotopes, especially in minerals like pyroxenes) are not much affected by the level of alteration expected. Geochronology is essential and will be established using paleontology, paleomagnetism, and Ar-Ar and U/Pb dating of zircon in felsic materials. The provenance and mode of deposition of volcaniclastic sediment is also essential and will be established by examining the morphology of grains and the overall character of sedimentary units (e.g., Bednarz and Schmincke, 1994; McPhie and Allen, 2003).

These five objectives will establish the effects of a fundamental characteristic of island-arc magmatism (across-arc geochemical variations) on crustal production in that environment and will constrain the fundamental reasons for the variations themselves. This temporal record is also necessary to assess the evolution of the mantle wedge and slab, to evaluate processes of intracrustal differentiation, and to calculate mass balance and flux models of crustal growth.

Additional drilling discovery opportunities

Physical volcanology

As noted in the Introduction, most rocks in the upper crust of arcs are submarine volcaniclastics. Previous studies of the IBM arc system have revealed the importance of thick, pumice-rich pyroclastic units as a component of rift basins and arc-front volcano aprons (e.g., Nishimura et al., 1992; Tani et al., 2008). Pumice-rich pyroclastic units are also common in marine arc-related basin fills exposed on land (e.g., Busby, 2004; Busby et al., 2006), and their ultimate origin as the products of explosive eruptions is widely accepted. However, it is less clear how to distinguish eruption-fed products (e.g., Busby-Spera, 1984, 1986, 1988; Busby et al., 2003), strictly contemporaneous with an eruption, from those generated by resedimentation of temporarily stored pumiceous facies (e.g., Critelli et al., 2002). A further source is the collapse of volcaniclastic aprons, recently recognized as a major sediment source in the Miocene arcs of the North Island of New Zealand (Allen, 2004). Also unclear is how to distinguish the products of totally submerged explosive eruption plumes (Kokelaar and Busby, 1992; Busby, 2005) versus plumes that break the water/air interface or are totally subaerial (e.g., McPhie and Allen, 2003). Data from well-preserved examples where the context is well constrained, such as Izu, have the potential to greatly refine our currently primitive criteria and test some inferences based on outcrop examples. We will attempt to distinguish not simply the compositions of source volcanoes for rear-arc pyroclastic components but also their proximity, vent setting, and whether they were eruption fed or resedimented. We note that these methods led to the serendipitous discovery of a new type of deep seafloor pyroclastic eruption during ODP Leg 126 (Gill et al., 1990), and we believe that more rear-arc drilling will lead to more such discovery.

Drilling at proposed Site IBM-3C will test whether or not there is asymmetry in the physical volcanology of arcs as well as in magma compositions, how the differences evolve temporally, and how the differences can be applied to studies of paleo-arcs worldwide. Plausible cross-arc influences on the physical volcanology of arc volcanoes include

  • Magma composition: spatial and temporal gradients in magma compositions, especially SiO2 and volatiles, should be accompanied by variations in eruption styles and volcano types. Higher SiO2 and volatile magmas favor powerful explosive eruptions and the production of diverse, widely dispersed pyroclastic facies, as well as lavas and domes. On the other hand, lower SiO2 and volatile-poor magmas favor lavas or domes and subordinate, weakly explosive eruptions.
  • Vent environment: vents for rear-arc volcanoes are likely to be submerged; in contrast, vents for arc-front volcanoes can be either submerged (particularly in the early stages of arc evolution) or subaerial. The presence of water greatly alters the dynamics of eruptions and, hence, also the products. Deep water may suppress explosive activity, whereas shallow water may introduce the possibility of magma-water interaction; any water promotes quench fragmentation. Additionally, water depths can control styles of mineralization at vents (e.g., volcanic-hosted massive sulfides do not form in shallow water).
  • Presence or absence of wet sediment: ancient successions show that magmas intrude, rather than erupt, in submerged settings where wet sediment has accumulated, forming peperites (e.g., Busby-Spera and White, 1987; Skilling et al., 2002). The products are sill-sediment complexes and/or cryptodome complexes showing complex contact relations with the host sediment. Sill-sediment and/or cryptodome complexes are predicted to be a common feature of the rear arc in contrast to the arc front, where they may be present but largely limited to the earliest stages of arc evolution.

Microbiology

Proposed Site IBM-3C represents a potentially exciting opportunity to study the microbiology of the deep subseafloor, in the opinion of microbiologists including K. Edwards (University of Southern California, USA) and M. Schrenk (Department of Biology at East Carolina University, USA) with whom we have consulted. Of all IODP sites on the horizon, this one should have the most abundant vesicular basaltic glass. Such glass has extremely large amounts of reactive surface area and, based on what is known from dredged lavas, should have high levels of the oxidant Fe3+ and the nutrient P (certainly relative to mid-ocean-ridge basalt [MORB] lava). The site would also provide new pressure, temperature, and pore water conditions in which to explore for subseafloor microbial ecology and biogeochemistry (20–50 MPa, ≤100°C, and high Ca-Cl2 pore water). We predict relatively unaltered glass in at least the uppermost 600 m and quite altered glass below 1500 mbsf, with increasing alteration in between, based on what was found at ODP Sites 792 and 793 in the Izu fore arc. There, the lower depth corresponded to a marked change in pore water chemistry (increased Ca and inorganic C and decreased Si, SO4, and Mg) and decrease in porosity (Egeberg, 1992). It is uncertain whether the level of microbiological activity would be abnormally high (because of bioavailability of oxidants and nutrients) at proposed Site IBM-3C or low (because of decreased permeability and increased rock/water equilibration). However, we feel that this scientific objective should continue to be explored at this site.

Tonalite emplacement, mineralization, and exhumation

The proposed site lies downslope from the only known submarine example of porphyry copper mineralization in a rear arc, at Manji Seamount (Ishizuka et al., 2003c). Rounded cobbles of chalcopyrite-bearing quartz-magnetite stockwork and 7 Ma gabbroic to tonalitic plutonic rocks have been dredged from its flat-topped, subaerially eroded summit, which currently lies at 700 mbsl). The Shinkai 2000 diving survey discovered exposures of classic potassic and propylitic alteration, indicative of activity of hypersaline fluids, and closely associated plutonic rocks. Because rounded cobbles occur on the seamount’s flanks, clasts and heavy minerals (sulfides) also may be found at proposed Site IBM-3C. If so, we might discover a history of rear-arc tonalite intrusion, mineralization, exhumation, and submergence.

Top of page   |    Previous   |    Next