RESEARCH to Understand Subduction Initiation, Study Forearc Crust

RESEARCH

To understand subduction initiation, study forearc crust: To understand forearc crust, study ophiolites

R.J. Stern1*, M. Reagan2*, O. Ishizuka3*, Y. Ohara4*, and S. Whattam5* 1GEOSCIENCES DEPARTMENT, UNIVERSITY OF TEXAS AT DALLAS, RICHARDSON, TEXAS 75083-0688, USA 2DEPARTMENT OF GEOSCIENCE, UNIVERSITY OF IOWA, IOWA CITY, IOWA 52242, USA 3INSTITUTE OF GEOSCIENCE AND GEOINFORMATION, GEOLOGICAL SURVEY OF JAPAN/AIST, CENTRAL 7, 1-1-1, HIGASHI, TSUKUBA, IBARAKI 305-8567, JAPAN 4HYDROGRAPHIC AND OCEANOGRAPHIC DEPARTMENT OF JAPAN, TOKYO 104-0045, JAPAN 5DEPARTMENT OF EARTH AND ENVIRONMENTAL SCIENCES, KOREA UNIVERSITY, SEOUL 136-701, REPUBLIC OF KOREA

ABSTRACT

Articulating a comprehensive plate-tectonic theory requires understanding how new subduction zones form (subduction initiation). Because subduction initiation is a tectonomagmatic singularity with few active examples, reconstructing subduction initiation is challenging. The lithosphere of many intra-oceanic forearcs preserves a high-fi delity magmatic and stratigraphic record of subduction initiation. We have heretofore been remarkably ignorant of this record, because the “naked forearcs” that expose subduction initiation crustal sections are dis- tant from continents and lie in the deep trenches, and it is diffi cult and expensive to study and sample this record via dredging, diving, and drilling. Studies of the Izu-Bonin-Mariana convergent margin indicate that subduction initiation there was accompanied by seafl oor spreading in what ultimately became the forearc of the new convergent margin. Izu-Bonin-Mariana subduction initiation encompassed ~7 m.y. for the complete transition from initial seafl oor spreading and eruption of voluminous mid-ocean-ridge basalts (forearc basalts) to normal arc volcanism, perhaps consistent with how long it might take for slowly subsiding lithosphere to sink ~100 km deep and for mantle motions to evolve from upwelling beneath the infant arc to downwelling beneath the magmatic front. Many ophiolites have chemical features that indicate formation above a convergent plate margin, and most of those formed in forearcs, where they were well positioned to be tectonically emplaced on land when buoyant crust jammed the associated subduction zone. We propose a strategy to better understand forearcs and thus subduction initiation by studying ophiolites, which preserve the magmatic stratigraphy, as seen in the Izu-Bonin-Mariana forearc; we call these “subduction initiation rule” ophiolites. This understanding opens the door for on-land geologists to contribute fundamentally to understanding subduction initiation.

LITHOSPHERE; v. 4; no. 6; p. 469–483 | Published online 16 May 2012 doi: 10.1130/L183.1

INTRODUCTION there are few active examples, and (2) nearly in Eocene time (Ishizuka et al., 2011). Major all of the evidence for tectonic, magmatic, and subduction initiation episodes are hemispheric A better understanding of the mechanisms sedimentary responses to subduction initia- in scale and necessarily reorganize upper-man- by which new subduction zones form is criti- tion is preserved in forearcs, which are deeply tle fl ow, and in many cases are accompanied by cal for advancing the solid Earth sciences. Until submerged and buried beneath sediments. We widespread and voluminous igneous activity. we can reconstruct how and why this happens, would prefer to study subduction initiation in Here, we outline a strategy that promises to we cannot pretend to understand a wide range progress, but there are few places to do this. accelerate our understanding of processes asso- of important Earth processes and properties, One such active region however, is the Puysegur ciated with major subduction initiation episodes including lithospheric strength, composition, subduction zone off the coast of southern New by considering both the subduction initiation and density, and the driving force behind plate Zealand (LeBrun et al., 2003; Sutherland et al., record preserved in forearcs and insights from motions. In spite of this, our understanding of 2006). However, as only a narrow segment of studying well-preserved ophiolites. The record the subduction initiation process has advanced the Australia-Pacifi c transform plate margin is of subduction initiation is preserved in igneous slowly, for two important reasons: (1) Sub- affected, studies of Puysegur cannot capture all crust and upper-mantle residues and the associ- duction initiation is an ephemeral process, so of the processes that accompany major subduc- ated sediments on the overriding plate next to the tion initiation events, i.e., those that change the trench. These collectively comprise the forearc lithospheric force balance suffi ciently to cause (Dickinson and Sealey, 1979) and provide the changes in plate motion and stimulate volumi- best record of subduction initiation. Signifi cant *E-mails: [email protected]; mark-reagan@uiowa .edu; [email protected]; [email protected]; nous magmatism, as discussed herein. Such parts of forearcs may be lost by tectonic erosion [email protected]. episodes shaped the western margin of North (Scholl and von Huene, 2009); nevertheless, Editor’s note: This article is part of a special issue ti- America in Mesozoic time (Dickinson, 2004), whatever remains contains the best record of tled “Initiation and Termination of Subduction: Rock Re- established a convergent margin along SW the processes that accompanied subduction ini- cord, Geodynamic Models, Modern Plate Boundaries,” edited by John Shervais and John Wakabayashi. The Eurasia in Late Cretaceous time (Moghadem tiation of that particular convergent margin. We full issue can be found at http://lithosphere.gsapubs and Stern, 2011), and engendered most of the explore why this record has been overlooked and .org/content/4/6.toc. active subduction zones of the western Pacifi c summarize recent studies of forearc crust and

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 STERN ET AL.

upper mantle, and what the results reveal about of a convergent plate margin is provided by the some is scraped off to form an accretionary subduction initiation. The expense and diffi - Late Mesozoic of California, with the Francis- prism. Such situations of forearc thickening and culty of directly studying forearc igneous rock can mélange representing exhumed subduction- widening are globally unusual, because most exposures are huge obstacles to our progress, so zone material, the Great Valley Group repre- forearcs lose upper-plate crust to the subduc- we explore the potential of some ophiolites for senting the forearc basin, and the Sierra Nevada tion zone due to tectonic erosion, as a result of illuminating forearc composition and magmatic Batholith representing the roots of the magmatic normal faulting, oversteepening, and basal frac- stratigraphy. Ophiolites are exposed on land and arc. This is indeed an excellent example of a turing and abrasion along the plate interface. so are vastly easier and cheaper to study than sediment-rich convergent margin, but empha- Another misconception (due to bias toward forearcs. We conclude that those ophiolites that sis on California and other sediment-dominated studying sediment-rich convergent margins) is formed in a forearc provide important opportu- forearc examples has inhibited appreciation of that all inner-trench slopes have very low slopes nities for advancing our understanding of sub- forearc crust itself. (<3°), when, in fact, erosive margins, especially duction initiation. The strategy of comparative Many—but not all—continental forearcs those exposing igneous basement, are much study of igneous forearc crust and ophiolites, are excellent examples of convergent margins steeper, typically with slopes of 3°–7° (Clift and coupled with geodynamic modeling, promises affected by high sediment fl ux. Some continental Vannucchi, 2004). Estimates of the proportion of to lead to major advancements in our under- convergent margins—such as Peru-Chile and NE accretionary versus erosive convergent margins standing of subduction initiation processes. Japan—do not have high sediment fl ux, but these vary. According to Clift and Vannucchi (2004), have not been textbook examples because the 57% of the cumulative length of trenches is ero- FOREARCS interesting outcrops are in very deep water, and sive and 43% is accretionary, whereas Scholl thus are diffi cult and expensive to study. In con- and von Huene (2007) estimated that 74% and Forearcs comprise the bulk of any arc-trench trast, forearcs away from continents are mostly 26% are erosive and accretionary, respectively. system, occupying the ~150–200-km-wide sediment starved (Clift and Vannucchi, 2004). Thickness of sediment on the downgoing plate region above the subducted plate between the Such naked forearcs expose crust and upper is the single most important control on whether trench and the magmatic arc. Forearcs are rela- mantle, which are readily accessed by drilling a margin is erosive or accretionary. A sediment tively stable and low standing—intra-oceanic through thin sediment cover, as was done during thickness of ~500 m divides the two types of forearcs lie entirely below sea level—and are Deep Sea Drilling Project (DSDP) Leg 60 and margins. Other factors favoring tectonic erosion morphologically unimpressive compared to Ocean Drilling Program (ODP) Legs 125 and include collision of large bathymetric features spectacular volcanoes of the fl anking magmatic 126 in the Izu-Bonin-Mariana arc and ODP Leg such as seamounts (Clift and Vannucchi, 2004) arc and the tremendous gash of the trench. For 135 in the Tonga forearc (Bloomer et al., 1995). and presence of rasping grabens on the downgo- these reasons, it is understandable that forearcs ing plate (Hilde, 1983). were either overlooked or misunderstood when EROSIVE VERSUS ACCRETIONARY Although most forearcs are erosive, they are the geologic implications of plate tectonics were FOREARCS more poorly known than accretionary forearcs fi rst being explored in the late 1960s and 1970s. because they are harder to study, and a smaller During this time, thinking about forearcs was Subordinate proportions of forearcs are research community has been interested in them. dominated by examples on or near continents, accretionary, growing by deposition of large Erosive forearcs are exclusively submarine, so such as California, Japan, Alaska, and Indone- sediment loads from a fl anking continent, which studying them requires research vessels with sia (for an account of early thinking about what bury forearc crust beneath forearc basins and technology to examine and sample the bottom would come to be called forearcs, see Dickin- then overfl ow to the trench, where these sedi- (Fig. 1A). Compared to accretionary forearcs, son, 2001). Even today, the textbook example ments briefl y ride on the subducting plate before erosive forearcs lie in deeper water, farther from

Figure 1. Photographs of on-land (left) and submerged (right) forearc exposures. Left photo shows how easy it is for geoscientists to examine lithologies and structures on land. Right photo is taken from Shinkai 6500 YK1012, Dive 1231, ~6000 m deep in the southern Mariana Trench. Only one or two scientists at a time can go down to examine and sample rocks. Costs of an on-land ﬁ eld trip are a miniscule fraction of the expense of a submarine ﬁ eld trip. Photo on the left is Franciscan radiolarian chert exposed on the Marin headlands, California (photo by S. Graham). Field of view on right is ~7 m.

470 www.gsapubs.org | Volume 4 | Number 6 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 Subduction initiation, forearcs, and ophiolites | RESEARCH

the continents, making them more diffi cult and margins contain potential economic deposits cation. Seismic-refl ection profi ling over naked expensive to study. In contrast, accretionary of hydrocarbons, which attract the attention of forearcs yields much less interesting seismic forearcs lie in shallower water, and parts rise oil companies; erosive margins do not. Finally, refl ection images, and they are more diffi cult to above sea level (e.g., Kodiak Island, Alaska; accretionary margins attract more geophysical interpret (Fig. 2B). It is so much easier and more Shimanto Belt, SW Japan; and Nias and Men- interest than do erosive margins. Accretionary rewarding to study accretionary margins that it tawai, Sumatra), where they are relatively easy margins are characterized by progressive defor- is a wonder that erosive margins are studied as to study (Fig. 1B). Accretionary margins are mation of sedimentary layering, which spatially much as they are. found near continents and so are usually in some and temporally changes from fl at lying and In spite of the challenges of studying erosive nation’s territorial waters or exclusive economic unconsolidated at the trench to steeply dipping forearcs, it is important that we do so. Naked zone, which attracts study by the scientists of and lithifi ed arcward (Fig. 2A). Such lithostruc- forearcs expose igneous infrastructure in the that nation. Most of the great accretionary mar- tural variations reward seismic-refl ection profi l- inner trench wall and bury this crust under thin gins lie in the Northern Hemisphere, facing the ing with spectacular images, on which structural sediments of the forearc itself, providing the best largest continents, biggest rivers, and greatest interpretations, publications, and proposals can opportunities for in situ investigations of this sediment fl ux. The Northern Hemisphere is also be based. In contrast, intra-oceanic forearcs are lithosphere by drilling, dredging, and diving. where the richest nations are—the ones most generally sediment starved because they are far likely to support large-scale geoscientifi c efforts from continents, and thus large sediment fl uxes. FOREARC CRUST AND UPPER MANTLE needed for marine tectonic studies. In contrast, These comprise a subclass of erosive margins erosive margins often lie far away from conti- known as “naked forearcs” (Stern, 2002), which An important characteristic of many naked nents, rich nations, and scientists. Accretionary lack thick sedimentary cover and obvious imbri- forearcs is the presence of peridotite—exposed

3 V.E. ~1.5x Sedimented Forearc (Indonesia)

- - back thrust out-of-sequence thrust (backstop thrust) branchbranchch thrustththrrust reduced coherent

Depth (km) imaging décdédécollementolllleememem nt

Sediment-flooded A trench

2 5 km Naked forearc (Mariana) V.E. ~2x

B 21 Figure 2. Comparison of seismic-refl ection profi les of (A) sedimented and (B) naked (unsedimented) forearcs. (A) Prestack depth-migrated section of multichannel refl ection profi le off the Sunda Strait, from Kopp and Kukowski (2003) and interpreted by them. An arcward increase in material strength results in a segmentation of the margin. Faint seaward- and landward-dipping faults cut the trench fi ll in the protothrust zone of segment I, indicating the fi rst stages of faulting. The deformation front marks the onset of faulting in conjugate pairs of fore-thrusts and back thrusts. The frontal active accretionary prism (segment II) is composed of tilted thrust slices separated by regularly spaced thrust faults. The transition to the fossil accretionary prism of the outer high is marked by a prominent out-of-sequence thrust. Segment III forms the backstop to the frontally accreted material and dis- plays much reduced tectonic activity, mainly manifested in the occasional reactivation of previous thrusts, which helps adjust the taper. (B) Mariana forearc, trench, and part of incoming Pacifi c plate shown as depth section (MCS Line 22–23, from Oakley et al., 2008). Circles locate the points along the plate refl ection where depths were recorded. Normal faults on the incoming plate and the Mariana forearc are interpreted. The toe of the forearc is uplifted as the fl ank of a Pacifi c plate seamount subducts. M—seafl oor multiple; V.E.—vertical exaggeration.

LITHOSPHERE | Volume 4 | Number 6 | www.gsapubs.org 471

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 STERN ET AL.

0° 60°E 120°E 180° 120°W 60°W Figure 3. Known locations of peridotite exposures in the lower parts of inner-trench walls, shown in 60° bold: (1) South Sandwich (Pearce N et al., 2000); (2) Tonga (Fisher and Aegean Aleutians Engel, 1969; Bloomer and Fisher, Japan Cascades Makran Kuriles 1987); (3) Mariana (Bloomer, 1983; Himalaya Lesser Ohara and Ishii, 1998; Michibayashi 30° Philippines Izu-Bonin Antilles et al., 2007, 2009); (4) Izu-Bonin Mariana (Okamura et al., 2006; Ishizuka et Central America al., 2011). Mariana and Izu-Bonin New Britain forearcs also bring up peridotite 0° Indonesia Vanuatu in serpentine mud volcanoes (Fryer et al., 1995; Parkinson and Tonga Pearce, 1998). Peridotites repre- New

30° Hebrides Andes sent the bases of in situ ophiolites Kermadec South that make up the crust and upper Sandwich mantle of intra-oceanic forearcs. Other inner-trench walls may also expose peridotite, but most have 60° not been sampled as extensively S as these four.

upper mantle—in the lower trench wall. Trench as “bulk silicate earth” (BSE). Several studies an idealized upper-mantle composition, but one peridotite exposures demonstrate that the Moho have estimated PM compositions, including which acknowledges the extraction of the conti- is also exposed, along with a complete crustal Mg# (100Mg/[Mg + Fe] = 89–90), CaO (2.8– nental crust. “Pyrolite” is another idealized com-

section at shallower depths, and indicate the 3.7 wt%), and Al2O3 (3.5–4.5 wt%; see table 2 of position (Green and Falloon, 1998) that is very thickness of the crust. Lyubetskaya and Korenaga, 2007). These esti- similar to FMM. FMM and pyrolite approximate Peridotite is exposed in the inner walls of mates constrain minimum Mg# and maximum the composition of upper mantle that partially

intra-oceanic trenches at depths >8 km in the CaO and Al2O3 contents of the upper-mantle melts to generate oceanic crust beneath diver- Tonga Trench (Bloomer and Fisher, 1987) but source region of most basalts. Because Earth has gent plate boundaries (spreading ridges) and to can be found as shallow as 5800 m in the south- been recycling surface materials and melting to produce arc melts beneath convergent margins ern Mariana Trench (Michibayashi et al., 2009). make basalt for several billion years, signifi cant (note that other components such as pyroxenite Such depths are mostly beyond the reach of tracts of primitive upper mantle are unlikely to exist in the mantle, but these melt almost com- manned submersibles, which currently cannot exist. Instead, the concept of “fertile mid-ocean- pletely, leaving no identifi able residue). descend below 6500 m, so forearc peridotites ridge basalt (MORB)–type mantle” (FMM; If PM, BSE, FMM, and pyrolite were rocks are rarely sampled except by dredging. Still, we Pearce et al., 2000) is more useful. FMM is also instead of ideas, they would be classifi ed as know about intra-oceanic peridotite exposures in four trenches: Izu-Bonin, Mariana, Tonga, and South Sandwich (Figs. 3 and 4). In addition, mantle peridotite is brought up by serpentine mud volcanoes, which are common in the Mariana forearc and are also known from the Izu forearc (Fryer, 2002). Volcanics Because of its importance for understanding the nature and origin of intra-oceanic forearcs, some basic concepts about mantle peridotite in general and forearc peridotite in particular are Gabbro presented here. These are residues after partial melting and complement magmatic rocks such as lavas—especially basalts—which are more RB MO

common subjects of marine petrologic study. (km) sea level Depth below Because Earth has had mantle since shortly after it formed, but this has been modiﬁ ed by melt extraction and mixing with subducted materials, idealized compositions are useful for Peridotite this discussion. For example, “primitive” mantle (PM) refers to an idealized chemical composi- Figure 4. Interpretive section of a typical intra-oceanic forearc as reconstructed from dredging of the tion after the core segregated but before the con- Tonga Trench. Figure is from Bloomer and Fisher (1987), reproduced with permission of Journal of Geol- tinental crust was extracted. PM is also known ogy. N- and E- MORB refer to normal and enriched mid-ocean ridge basalt. V.E.—vertical exaggeration.

472 www.gsapubs.org | Volume 4 | Number 6 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 Subduction initiation, forearcs, and ophiolites | RESEARCH

lherzolite, an olivine-rich ultramaﬁ c rock that Ol FMM contains >5% clinopyroxene (cpx), the remain- ABDunite der being orthopyroxene and an aluminous 10 90 phase (spinel or garnet; Fig. 5A). Changes in 5 20 peridotite composition due to partial melting n 80 Harzburgite letio are simple and clear. Melt depletion dimin- p 30 e d 70 t ishes abundances of cpx, CaO, and Al O and l 2 3 e

increases Cr# (Cr/Cr + Al) in the residual spi- 40 15 FMM 60 Abyssal peridotite nel (Fig. 5). This is because basalts, which are rich in CaO and Al O (~12 and 16 wt%, 50 50 2 3 Lherzolite respectively), are generated by partial melting 25 of lherzolite, which is much poorer in CaO and Opx Cpx Mariana forearc peridotite Al O (~3 and ~4 wt%, respectively for FMM; 2 3 Figure 5. What happens to peridotites when Fig. 5B). Clinopyroxene contains nearly all they melt: (A) Changes in modal mineralogy the CaO in peridotite, whereas spinel in FMM (and thus rock type) with progressive melting, contains ~35% Al O (a few percent Al O is plotted on International Union of Geological 2 3 2 3 C also dissolved in pyroxene). Melt depletion Sciences (IUGS) classifi cation for peridotites Boninite decreases the proportion of cpx, so that the (LeBas and Streckeisen, 1991). Fertile mid- residue progressively changes from lherzolite to ocean-ridge basalt (MORB)–type mantle (FMM) is mostly peridotite, consisting of olivine (Ol), harzburgite (<5% cpx); extreme melt depletion orthopyroxene (Opx), and clinopyroxene (Cpx), Forearc peridotite MD yields dunite (>90% olivine; Fig. 5A). comprising lherzolite. Melting to produce basal- Even though most forearc peridotite expo- tic melts depletes peridotite in clinopyroxene, Cr# sures are serpentinized, there are robust mineral so that residual peridotite after ~20% melting and whole-rock compositional characteristics is harzburgite, with <5% clinopyroxene. Melting that are remarkably unaffected by such altera- of harzburgite yields boninite melts and resid- Backarc ual dunite. (B) Bulk-rock abundances of Al O basin Abyssal peridotite tion. These include changes in the proportions 2 3 peridotite versus CaO (volatile free, normalized to 100% of minerals (Fig. 5A), major-element bulk total), showing how melting depletes perido- FMM chemistry (Fig. 5B), and spinel compositions tites in these elements. The compositions of (Fig. 5C). All of these refl ect the amount of melt FMM and mantle residue after 5%, 15%, and Mg# depletion, as discussed already. 25% partial melting of FMM are from Pearce These approaches allow us to compare the and Parkinson (1993). Fields for abyssal and forearc peridotites are from Pearce et al. (1992). (C) +2 “refractoriness” of peridotites from various tec- Composition of spinels in peridotite, plotted on Cr# (Cr/Cr + Al) versus Mg# (Mg/Mg + Fe ) diagram. Fields are after Dick and Bullen (1984), modifi ed to show the composition of backarc basin tonic settings, i.e., how much melt depletion peridotites from the Mariana Trough (Ohara et al., 2002). Melting preferentially extracts Al from they have experienced. This reveals that forearc spinel, increasing Cr# as melting progresses. Composition of FMM spinels and approximate trend peridotites are the most depleted ultramafi c of melt depletion (MD) are also shown. rocks from any modern tectonic environment, with the highest Cr# spinels and lowest proportion of cpx and whole-rock abundances of CaO ger exist. Such transitory conditions are linked strength elements (HFSEs; elements with high

and Al2O3 (Bonatti and Michael, 1989). Not all to subduction initiation in the next section. valence and small ionic radius). High LILE/ forearc peridotites are so depleted; for example, Igneous rocks of exposed forearc crust above HFSE ratios in boninites reﬂ ect metasomatism some from the South Sandwich forearc include peridotites (Fig. 4) are only now becoming the of the source mantle by hydrous ﬂ uid released

lherzolites with up to 3.7% Al2O3 and 4.4% focus of geoscrutiny. At one time, forearc crust from subducted crust and sediments (Gill, 1981; CaO, along with spinels with Cr# as low as ~0.4 was thought to comprise oceanic crust that was Stern et al., 1991; Pearce et al., 1992). This fl uid (Pearce et al., 2000). Morishita et al. (2011) trapped when subduction began (e.g., Dickin- lowers peridotite melting temperature at the documented two populations of spinel in Izu- son and Sealey, 1979), so the origin of forearc same time that it re-enriches it in fl uid-mobile Bonin forearc dunites, one group with moder- igneous crust has not, until recently, been much elements, including LILEs and LREEs. ate Cr# (0.4–0.6), and the other with high Cr# studied. This is changing, partly because of what Not all naked forearcs have boninite, but (>0.8). Variations in spinel compositions not- is now recognized as the unusually strong deple- at least one of them—the Izu-Bonin-Mariana withstanding, forearc peridotites are dominated tion of forearc peridotites and because of inter- arc system—does (Fig. 6A; Stern et al., 1991; by ultradepleted compositions rarely found in est in boninites. Boninites are lavas with unusual Macpherson and Hall, 2001). Probably the best other tectonic environments. combinations of high silica and magnesium exposure of boninite in the world is found in The unusually depleted nature of forearc coupled with low calcium and aluminum abun- the Bonin (Ogasawara) islands (Taylor et al., peridotites requires unusual melting conditions: dances. These compositional features refl ect 1994). These boninites erupted on the seafl oor abnormally high temperature, volatile fl ux, or low-pressure melting of harzburgitic mantle ca. 46–48 Ma, shortly after subduction began both. Whatever the cause, these depletions are (e.g., Falloon and Danyushevsky, 2000), which ca. 52 Ma (Ishizuka et al., 2006, 2011). Recent all the more noteworthy because forearcs asso- is not otherwise observed on modern Earth. In studies of Izu-Bonin-Mariana forearc crust ciated with mature arcs have unusually low heat addition, boninites are enriched in large ion exposed in the inner-trench wall (Fig. 6B) reveal fl ow (Stein, 2003) and rarely are volcanically lithophile elements (LILEs; elements with low that boninite may be the uppermost component active. Whatever conditions caused the unusu- valence and large ionic radius) and light rare of forearc crust, underlain by thicker, slightly ally extensive melting beneath forearcs no lon- earth elements (LREEs) relative to high fi eld older tholeiitic basalts, which Reagan et al.

LITHOSPHERE | Volume 4 | Number 6 | www.gsapubs.org 473

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 STERN ET AL.

Mariana forearc, and we prefer to interpret this Lithology Age Approximate A (Ma) depth (m) as another exposure of basalts that formed when Izu-Bonin-Mariana subduction began. Arc tholeiites To conclude this section, it is clear that our and calc- 37–44 understanding of the composition of forearc Pacific alkaline plate rocks crust is incomplete, largely due to the diffi culty Japan of accessing and studying this material. Much Tokyo ∗ High-Mg Subaerially 44–45 of what we think we know is based upon studies . andesite exposed .. of the Izu-Bonin-Mariana forearc. We anticipate . . Boninite that Izu-Bonin-Mariana forearc crust is repre- Izu-Bonin-Mariana. 3 (and related sentative of igneous rocks that form when sub- . 44–48 30° differentiates) duction begins, but we cannot be sure until we have studied the igneous rocks of other forearcs . * 5500–4760 B* 2 Pillow lava in similar detail. . * hyaloclastite . Basalt (FAB) 50–52 . Philippine Sheeted dike SUBDUCTION INITIATION 6300 Sea plate arc system . Gabbro/ In this section, we discuss igneous activity .. Gabbro 20° N . Mesozoic 50–52 . associated with the formation of a new sub- . basalt . ? duction zone. As presented already, much of . Basalt . * ? 6780 what we understand about the igneous crust of * * forearcs comes from studies of the Izu-Bonin- 500 km * * 1 Peridotite Mariana system. The fundamental question we B address for this forearc, and one that is pertinent 130° E 140° 150° for many others is: Does subduction initiation generate this broad swath of crust, which ulti- Figure 6. (A) Schematic map of the Izu-Bonin-Mariana arc system, showing principal tectonic features and forearc crustal sections discussed in text. 1—study of Reagan et al. (2010); 2—study of mately forms the forearc (Fig. 8)? Also, what Ishizuka et al. (2011); 3—study of DeBari et al. (1999). Asterisk marks location of Deep Sea Drill- happens when the subduction zone itself forms? ing Project (DSDP) Site 458; B is location of Bonin (Ogasawara) Islands. White region is seafl oor For the Izu-Bonin-Mariana arc system, we have <2500 m deep. Dashed line shows location of profi le in Figure 8. (B) Schematic columnar section of documented the progression of igneous activity crust exposed in the Bonin inner-trench wall, from Ishizuka et al. (2011), used with permission by in the forearc about the time that the Pacifi c plate Elsevier. FAB—forearc basalt. changed its motion, and we have concluded that this activity resulted from the dynamic response of the crust and upper mantle to subduction initi- (2010) called “forearc basalts.” Forearc basalt richer in silica). In spite of these differences, ation. The basic idea is: at ca. 52 Ma, old, dense lavas and related dikes have chemical compo- most forearc basalts also have trace-element lithosphere of the Pacifi c plate began to sink, sitions similar to MORB. Forearc basalts were (Figs. 7B and 7C) and isotopic compositions perhaps due to differential subsidence across a fi rst recognized in the southern Mariana forearc (Reagan et al., 2010) that are MORB-like, lithospheric weakness, such as an old fracture SE of Guam, where they crop out trenchward whereas younger forearc lavas have composi- zone (Fig. 9A; Stern and Bloomer, 1992). We of thin boninites and younger arc rocks (Rea- tions suggesting that subducted fl uids were conclude that subduction initiation at this time gan et al., 2010). This outcrop pattern, as well involved in their genesis. For example, lavas was hemispheric in scale: much of the western as the volcanic stratigraphy drilled in the Mari- generated by melting in the presence of a fl uid Pacifi c, extending south from Izu-Bonin-Mari- ana forearc at DSDP Site 458 (Fig. 6B) indi- from a subducting plate (e.g., Mariana arc lavas ana to Fiji and the Tonga-Kermadec convergent cates that forearc basalts here are older than and boninites) typically have elevated Th/Yb margin, formed new subduction zones about this the boninites and were likely the fi rst lavas to compared to MORB. On a plot of Th/Yb versus time. This was accompanied by voluminous gen- erupt when the Izu-Bonin-Mariana subduction Nb/Yb, most Izu-Bonin-Mariana forearc basalts eration of forearc basalts, boninite, and related zone formed (Reagan et al., 2010; Ishizuka et plot with MORB along the unmodifi ed “mantle igneous rocks, much of which is now preserved al., 2011). Below the forearc basalts, there are array,” whereas younger, subduction-infl uenced in these forearcs. We are not sure whether these gabbroic rocks, and then mantle peridotites, as lavas trend toward Th/Yb typical of Mariana arc new subduction zones were caused by, or were summarized in Figure 6B (Ishizuka et al., 2011). lavas (Fig. 7E). Note that Figure 7 also plots the the cause of the change in the Pacifi c plate abso- Forearc basalts have major-element compo- chemostratigraphies of ophiolitic lavas, a point lute motion, from NNW to WNW at ca. 50 Ma, sitions that are broadly MORB-like, although which is discussed further below. as refl ected by the bend in the Emperor-Hawaii signifi cant differences exist, at least for Izu- DeBari et al. (1999) studied the Izu-Bonin- seamount chain (Sharp and Clague, 2006), but Bonin-Mariana forearc basalts. Forearc basalts Mariana inner-trench wall (6100–6500 m deep) both events happened about the same time.

are generally not as rich in TiO2 as is typical near 32°N (Fig. 6A) and documented the pres- There are many challenges to this summary of

MORB, which often contains >1.4 wt% TiO2. ence of MORB-like basalts. They interpreted Izu-Bonin-Mariana subduction initiation. These

Forearc basalts can also be more rich in SiO2 these to represent older oceanic crust that was include hypotheses that: (1) interaction with a than typical MORB (most forearc basalts con- trapped when subduction began. However, the mantle plume was responsible for boninite for- tain <51% but forearc basalt–boninite tran- composition of these lavas is identical to those of mation (Macpherson and Hall, 2001); (2) extru- sitional lavas at DSDP Sites 458 and 459 are forearc basalts from elsewhere in the Izu-Bonin- sion of Indian Ocean–Asian asthenosphere due

474 www.gsapubs.org | Volume 4 | Number 6 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 Subduction initiation, forearcs, and ophiolites | RESEARCH

A B

C D

E F

Figure 7. Whole-rock geochemical data of stratigraphically constrained (i.e., lower versus upper) subalkaline lavas and latest- stage dikes of Izu-Bonin-Mariana forearc lavas (left) and ophiolites of the Eastern Mediterranean–Persian Gulf region (right). Forearc chemical data are from Reagan et al. (2010) and Ishizuka et al. (2011). Ophiolite data are from Albania (Mirdita; data from Dilek et al., 2008); Greece (Pindos; data from Saccani and Photiades, 2004); Cyprus (Troodos; data from Flower and Levine, 1987); and Oman (Semail; data from Godard et al., 2003) as compiled by Whattam and Stern (2011). (A) Izu-Bonin-Mariana t (IBM) forearc and (B) Tethyan ophiolites on the SiO2 versus FeO /MgO subalkaline afﬁ nity discrimination diagram (Miyashiro, 1974). (C) Izu-Bonin-Mariana forearc and (D) Tethyan ophiolite lavas on chondrite-normalized rare earth element (REE) plots (REE concentrations from Nakamura, 1974). (E) Izu-Bonin-Mariana forearc and (F) Tethyan ophiolite lavas on the Nb/Yb versus Th/Yb plot (Pearce, 2008). Compositional data have been ﬁ ltered to include samples with reported major-oxide totals of t 98%–102% and loss on ignition (LOI) <7%. In A, total iron is expressed as FeO (= Fe2O3 × 0.89), and oxide concentrations were recalculated and normalized to 100% on an anhydrous (volatile-free) basis. In E and F, abbreviations are N- and E-M—normal and enriched mid-ocean-ridge basalt (MORB). Note that in B, the only “lower lava” samples that plot as calc-alkaline (n = 4) are from Troodos. In addition, Troodos is not represented on F because the data set of Flower and Levine (1987) does not include concentrations of Th, Nb, and Yb. Modern Mariana arc lava data (blue dots) in E are from Peate and Pearce (1998). FAB—forearc basalt; FC—fractional crystallization.

LITHOSPHERE | Volume 4 | Number 6 | www.gsapubs.org 475

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 STERN ET AL.

Volcanic front Inner forearc Outer forearc Trench Distance from trench (km) Figure 8. Simplifi ed P-wave veloc- 180 150 100 50 0 0 ity structure beneath Izu forearc (approximately E-W line at Vp ~1.9 sediments 30°50′E), modifi ed after Kamimura 3.3 Volcanics et al. (2002), with interpreted 5 4.3 4.8 Altered basalt and diabase 4.5 3.2 lithologies. Note that forearc 4.5 6.3 crustal thickness decreases from 5.8 6.3 Gabbro ~11 km near the volcanic front 10 6.7 6.8 6.5 to 5 km or less near the trench. 6.4 7.0 Moho Note also signifi cantly lower ~6.8 7.3 interface rust Vp P-wave seismic velocity (Vp) in nic c 15 cea uppermost mantle beneath the Depth (km) Mantle r o pe outer forearc (6.4–6.8 km/s vs. Vp ~8.0 Up Subduction 7.0 7.3–8.0 km/s), probably refl ecting st Vp ~ Subducted Moho 20 cru greater serpentinization beneath nic cea the outer forearc. r o Mantle Vp ~8.0 we Lo V.E. ~4x 25

to Tethys closure was a major cause of western Young Oceanic crust Old A Pacifi c arc rollback and basin opening (Flower LM LM et al., 2001); (3) boninite formation refl ected the intersection of an active spreading ridge and TF/FZ AM AM a subduction-transform fault transition (Des- champs and Lallemand, 2003); (4) compression Early protoforearc spreading (FAB) across a preexisting fracture zone was required B (Hall et al., 2003); (5) lateral compositional buoyancy contrast within oceanic lithosphere Upwelling fertile controls subduction initiation (Niu et al., 2003); asthenosphere; (6) the new Izu-Bonin-Mariana convergent mar- Sinking slab; rapid trench rollback no interaction with gin cut across, rather than followed, preexist- slab-derived fluid ing lithospheric fabric, such as remnant arcs, C Late protoforearc spreading (VAB/BON) fracture zones, and spreading ridges (Taylor and Goodliffe, 2004); and (7) subduction of the v v v v v v v v v v v Pacifi c-Izanagi spreading ridge triggered a chain Depleted mantle stagnates; reaction of tectonic plate reorganizations that strong interaction led to Izu-Bonin-Mariana and Tonga-Kermadec with slab-derived fluid subduction initiation (Whittaker et al., 2007). There is clearly a lot of uncertainty about D Localized magmatic arc Forearc crust what caused the Izu-Bonin-Mariana subduc- (volcanic front) Fig. 11 tion initiation and attendant boninite volca- v v v v v vv v v v v nism, but there is no disagreement that it was accompanied by voluminous igneous activity. True subduction; This formed Izu-Bonin-Mariana forearc crust trench rollback slows (Figs. 6 and 8) as well as crustal tracts well to the west of the present volcanic front, including the West Mariana Ridge and Kyushu-Palau Figure 9. Subduction initiation, formation of the forearc, and evolution of magmatic systems, modi- Ridge. Stern and Bloomer (1992) conserva- fi ed after Metcalf and Shervais (2008). (A) Older, thicker, colder, and denser lithosphere (right) is jux- tively (i.e., assuming generation of 6-km-thick taposed with young, thinner, hotter, and more buoyant lithosphere across a zone of weakness (e.g., 3 transform fault or fracture zone [TF/FZ]). (B) Subsidence of old lithosphere allows asthenosphere to crust) estimated that 1200–1800 km of crust fl ood over it. Upwelling asthenosphere melts due to decompression, generating mid-ocean-ridge were produced per kilometer of arc during Izu- basalt (MORB)–like basalt (forearc basalts of Reagan et al., 2010) accompanied by seafl oor spread- Bonin-Mariana subduction initiation, and that ing. (C) Continued lithospheric subsidence or beginning of downdip motion of slab is accompanied this episode lasted 10 m.y., for a crustal growth by penetration of slab-derived fl uids into upwelled mantle, causing melting of depleted harzbur- rate of 120–180 km3/km. This is equivalent to gite. (D) Downdip motion of lithosphere signals start of true subduction, which terminates rapid the volume of crust produced at a mid-ocean- trench rollback and protoforearc spreading. Forearc mantle cools, and igneous activity retreats ridge spreading at 2–3 cm/yr. In fact, Izu-Bonin- ~200 km to what becomes the magmatic arc. Izu-Bonin-Mariana subduction initiation encompassed ~7 m.y. for the complete transition from initial seafl oor spreading to normal arc volcanism. Mariana mean crustal thickness produced dur- BON—boninite; FAB—forearc basalt; VAB—volcanic arc basalts; LM—lithospheric mantle (gray); ing this episode was probably greater than 6 km, AM—asthenospheric mantle (white). Note tectonic setting of lava sequence shown in Figure 11. as shown in Figure 8 for the Izu forearc. For the

476 www.gsapubs.org | Volume 4 | Number 6 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 Subduction initiation, forearcs, and ophiolites | RESEARCH

Mariana forearc, Calvert et al. (2008) concluded leads logically to the idea that hinged subsid- the Ontong-Java Plateau on the north side of that 15-km-thick crust just east of the magmatic ence of the older, thicker, and denser Pacifi c the Solomon arc in Miocene time caused a new front formed during a brief magmatic episode plate allowed asthenosphere to well up and fi ll subduction zone to form to the south (Mann and early in the arc’s history. Furthermore, Izu- the widening chasm, accompanied by extensive Taira, 2004), in what may be the best actualistic Bonin-Mariana subduction initiation probably decompression melting (Fig. 8B). The sequence example of subduction polarity reversal. Stern lasted less than 10 m.y.; Ishizuka et al. (2011) of events summarized in Figure 9 encapsulates (2004) characterized this type of subduction estimated that it took ~7 m.y. before a stable the idea of spontaneous subduction initiation initiation, which includes the Puysegur trench magmatic arc was established near the position (Stern, 2004), including early extension and example, as refl ecting “induced” nucleation of that it still occupies. Finally, subduction erosion seafl oor spreading. On the other hand, we can- a subduction zone. Induced nucleation of a sub- has removed a signifi cant amount of forearc; a not rule out the possibility that induced (forced) duction zone conceptually contrasts with “spon- width of 1 km removed per million years trans- subduction initiation might also have been asso- taneous” nucleation of a subduction zone” such lates into 240 km3/km of lost crust. All these ciated with early voluminous igneous activity, as Izu-Bonin-Mariana may be. Stern (2004) fur- considerations indicate that Izu-Bonin-Mariana due to the likelihood of lithospheric collapse ther suggested that induced nucleation of a sub- crustal growth related to subduction initiation and asthenospheric upwelling, as modeled by duction zone may not be associated with volu- was signifi cantly greater than 120–180 km3/km. Hall et al. (2003). minous forearc igneous activity, such as that Such high crustal growth rates and the Does current knowledge about Izu-Bonin- forming the Izu-Bonin-Mariana forearc. If so, absence of large, central volcanoes of Eocene Mariana subduction initiation serve as a useful forearcs formed by induced nucleation of a sub- age in the Izu-Bonin-Mariana forearc imply model for reconstructing how other forearcs duction zone may have fundamentally different that crust was formed by seafl oor spreading, form? We cannot be sure because we know so crustal structures and origins than those formed at least during the early, forearc basalts–domi- little about their crust, but Izu-Bonin-Mariana during spontaneous nucleation of a subduc- nated episode. This is consistent with the pres- serves as a useful analogue for the Tonga-Ker- tion zone. The generalization that spontaneous ence of dense diabase dike swarms at the base madec forearc. This crust is exposed only on the nucleation of a subduction zone results in broad of the forearc basalts sequence in the Izu-Bonin- island of ‘Eua and was drilled near the trench at forearc magmatism, whereas induced nucleation Mariana forearc (Reagan et al., 2010; Ishizuka ODP Site 841. Arc tholeiite, gabbro, and harz- of a subduction zone does not, may not be true. et al., 2011). Furthermore, the composition of burgitic peridotite have been dredged from the At least one example of induced nucleation of forearc basalts, which is similar to MORB, inner-trench wall, suggesting to Bloomer et a subduction zone—Aleutian subduction initia- implies a similar origin by decompression al. (1995) that the Tonga forearc is fl oored by tion—may be an example of induced nucleation melting beneath a spreading ridge (Plank and crust similar in age and composition to that of of a subduction zone with attendant forearc Langmuir, 1992). Small central volcanoes may Izu-Bonin-Mariana, including Eocene boninite igneous activity. Scholl (2007) and Minyuk and have formed during later, boninitic volcanism, (Crawford et al., 2003). The oldest known rocks Stone (2009) suggested that Aleutian subduction for example, at ODP Site 786B (Lagabrielle et in the Tonga-Kermadec forearc are 46–40 Ma initiation exploited SSW-trending strike-slip al., 1992), but the thicker, older forearc basalt arc-type lavas occurring below Upper Middle faults related to the “tectonic escape” of Alaska, succession seems to have been emplaced by Eocene limestones on ‘Eua (Ewart and Bryan, associated with motion along the North Pacifi c tectonomagmatic processes akin to seafl oor 1972; Duncan et al., 1985; Tappin and Balance, Rim Orogenic Stream (Redfi eld et al., 2007). A spreading. There is no clear evidence about the 1994). ODP drilling at Site 841 in the Tongan curved system of long strike-slip faults propa- arrangement of spreading ridges that might have forearc recovered a thick sequence of low-K gated southwestward across the North Pacifi c existed; Stern (2004) inferred short segments arc tholeiitic rhyolites (Bloomer et al., 1994), Rim throughout Cenozoic time, disrupting older aligned oblique to the evolving convergent plate dated by McDougall (1994) at 44 ± 2 Ma. Far- subduction zones along the Bering Sea shelf boundary, although no good magnetic anomaly ther north in the Tongan forearc, true low-Ca edge. Recently obtained ca. 46 Ma 40Ar/39Ar patterns have yet been identifi ed in any forearc boninitic rocks and associated backarc basin– ages from Aleutian forearc igneous rocks (Jicha that could be interpreted as spreading fabric. type basalts of probable Eocene age have been et al., 2006; Minyuk and Stone, 2009) provide Given that seafl oor spreading best explains dredged at depths in excess of 4 km (Falloon minimum age constraints for the timing of Aleu- the tectonic environment for Izu-Bonin-Mariana et al., 1987) and have yielded ages between 45 tian subduction initiation, although Aleutian forearc crust formation, the logical conclusion is and 35 Ma (Bloomer et al., 1998). It is not clear subduction initiation generated middle Eocene that a strongly extensional environment existed whether or not these boninitic lavas were gen- magmatic rocks as early as ca. 50 Ma (Scholl, at that time. There is no evidence that early erated in the same ca. 52 Ma subduction initia- 2007). The effects of tectonic erosion at con- Izu-Bonin-Mariana subduction was accompa- tion event as that responsible for the subduction vergent margins must be considered for any nied by compression, as would be expected if initiation boninite–refractory-forearc mantle thoughtful subduction initiation analysis. Naked the new subduction zone was caused by one package of the Izu-Bonin-Mariana arc system forearcs are likely to be trimmed back by sub- plate being forced beneath the other (induced to the north. duction erosion, at rates that can vary from a few subduction initiation of Stern, 2004), although We emphasize that not all subduction zones to several kilometers per million years (Clift and such evidence (uplift-related unconformity, form by processes outlined here. As outlined by Vannucchi, 2004; Scholl and von Huene, 2007). thrust faulting) might have been obliterated by Stern (2004), some subduction zones form as Subduction erosion thus can remove all of the Eocene igneous activity. Such evidence of initial a result of plate-boundary reconfi gurations, for evidence for a magmatic forearc in several tens compression without forearc volcanism charac- example, as a result of terrane accretion or conti- of millions of years, as may be the case for the terizes the Puysegur mini-subduction initiation nental collision. Collision of India with Eurasia Andean forearc. episode, which serves as an excellent example is a good example of this, although a new sub- Keeping such caveats and complications of induced subduction initiation. The conclusion duction zone has not yet formed behind (south in mind, it seems reasonable to conclude that that a strongly extensional environment accom- of) India, refl ecting the great strength of Indian many forearcs form as a result of voluminous panied Izu-Bonin-Mariana subduction initiation Ocean lithosphere (Stern, 2004). Collision of yet ephemeral igneous activity accompanying

LITHOSPHERE | Volume 4 | Number 6 | www.gsapubs.org 477

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 STERN ET AL.

subduction initiation. This makes it worthwhile forearc crust (and accreted sediments) is a key important part of modern geologic thought since to reconsider the origins of forearc igneous component of orogens. These tracts of obducted the 1960s, but there is a lot of confusion about rocks that formed about the same time as sub- forearc lithosphere are known as ophiolites. the tectonic environment in which they formed. duction initiation. For example, the ca. 55 Ma Much of this misunderstanding results from Siletzia terrane of the Oregon and Washington OPHIOLITES a lack of appreciation of the disparate lines of Coast Ranges formed about the same time as evidence needed to reconstruct ophiolites, espe- Cascadia subduction initiation and is variously Ophiolites are fragments of oceanic litho- cially structural geology, igneous geochemistry, interpreted as an accreted oceanic plateau (Dun- sphere that have been tectonically emplaced on and marine geology. Ophiolites were originally can, 1982), produced by the Yellowstone plume land. A complete “Penrose” ophiolite includes thought to form at mid-ocean ridges, but we now head (Pyle et al., 2009), or due to a tear in the tectonized peridotite, gabbro, sheeted dikes, understand that most sediments and all crust on subducting slab (Humphreys, 2008). The pos- and pillow basalt (Anonymous, 1972), but this the downgoing plate are subducted. If the down- sibility that Siletzia formed in situ as a forearc idealization is rarely seen because ophiolites are going plate includes very thick (>1 km; Clift and magmatic response to Cascadia subduction ini- faulted and fragmented during emplacement or Vannucchi, 2004) sediments, some of this may tiation should also be entertained. because one, or more, unit was never generated. be scraped off, and sometimes seamounts may Finally, we should consider the possibil- Nevertheless, ophiolites are key petrotectonic be transferred from the subducting to the over- ity of subduction “re-initiation,” i.e., the case indicators, perhaps the single most impor- riding plate, but normal oceanic crust itself has where a subduction zone once existed, then tant indicator of ancient plate-tectonic activity never been demonstrated to be transferred from was extinguished, and then resumed at about (Stern, 2005). Ophiolites mark tectonic sutures, downgoing to overriding plate (i.e., by seismic- the same location, because of either induced or indicating both the location of ancient oceans refl ection profi ling) at any modern convergent spontaneous nucleation of a subduction zone. and convergent plate boundaries where buoyant margin (Fig. 10C). Changes in plate motion The Cretaceous and younger evolution of SW lithosphere was partially subducted, also known might trap mid-ocean-ridge crust in what may Japan serves as an example of this. Subduc- as collision zones (Dilek, 2003). As a result, ultimately become a forearc (e.g., Macquarie tion of Pacifi c seafl oor beneath NE Japan has ophiolites are often highly altered and faulted, Island; Varne et al., 2000), but this is likely a been continuous, but subduction of the Philip- and much effort and imagination are needed very unusual tectonic scenario. pine Sea beneath SW Japan has been episodic. to reconstruct the original crust and uppermost In the 1970s, geoscientists (beginning with During Cretaceous time, a subduction zone mantle section. Ophiolites are unequivocal evi- Miyashiro, 1973) began to recognize the simi- dipped beneath what is now SW Japan, asso- dence of seafl oor spreading and have been an larity of some ophiolite lava compositions to ciated with a robust magmatic arc. Subduction there ceased and was replaced by a transform fault (shear margin) during Paleogene time. This Forearc ophiolite: Easy to emplace Backarc ophiolite: Difficult to emplace lithospheric weakness was converted into a new AB subduction zone beginning in latest Oligocene Forearc Backarc basin time, with an attendant fl are-up of forearc igne- lithosphere lithosphere ous activity (Kimura et al., 2005). Other likely examples of subduction re-initiation include the Ophiolite Paleogene Cascadia system and the Late Juras- Ophiolite sic of California. To conclude this section, the igneous crust and uppermost mantle of forearcs do not generally represent trapped oceanic lithosphere but in fact typically form during upper-plate spreading associated with subduction initiation (Shervais, C MORB ophiolite: Almost impossible to emplace 2001; Stern, 2004). There is no doubt that we MORB need to better understand the composition and lithosphere mode of formation of forearc crust, not only for Buoyant crust its own sake but also to better reconstruct events accompanying subduction initiation. Such stud- Subducting plate ies require studying naked forearcs, with all the challenges this entails. In the next section, we Suture propose a complementary strategy that takes advantage of the fact that forearc lithosphere is commonly emplaced (obducted) when buoyant lithosphere—particularly continental crust—on Figure 10. Tectonic cartoon illustrating the relative feasibility of emplacing oceanic lithospheres the downgoing plate enters and clogs a sub- created in forearc, backarc basin, and mid-ocean-ridge settings. (A) It is relatively easy to emplace duction zone (Wakabayashi and Dilek, 2003). forearc lithosphere. Subduction of buoyant material commonly leads to failure of subduction zone, Introduction of buoyant lithosphere disrupts the and isostatic rebound of buoyant material emplaces ophiolite. (B) It is diffi cult to emplace backarc basin oceanic lithosphere. Compression and shortening across the arc will lead to uplift of the arc. normal operation of a subduction zone, so that (C) It is almost impossible to emplace true mid-ocean-ridge basalt (MORB) crust at a convergent subducted materials are partially regurgitated, plate boundary. Sediments and fragments of seamounts may be scraped off the downgoing plate, and overlying forearc lithosphere is lifted above but subducting MORB-type lithosphere is nowhere known to be transferred from downgoing to sea level (Glodny et al., 2005). Consequently, overriding plate (modifi ed after Stern, 2004).

478 www.gsapubs.org | Volume 4 | Number 6 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 Subduction initiation, forearcs, and ophiolites | RESEARCH

those of convergent margins, leading to the we propose an alternative defi nition that bet- tion progress from early decompression melts of concept of the “suprasubduction zone” ophio- ter captures the important elements of what are unmodifi ed fertile (lherzolitic) mantle to yield lite (Pearce et al., 1984). This created tension called ophiolites in the literature. An ophiolite forearc basalts to younger hydrous fl ux melts between conclusions based on structural stud- must consist of a signifi cant proportion of harz- of depleted (harzburgitic) mantle that has been ies, which indicated ophiolites form by seafl oor burgitic peridotite (depleted mantle) and pil- strongly modifi ed by subduction-related fl uids spreading, and those based on geochemistry, lowed basalt. Gabbro and sheeted dikes may be to yield late high-Mg andesites and boninitic which indicated ophiolites form at convergent missing, but there should be associated deep- lavas. If the subduction initiation process con- margins. This tension was temporarily rec- sea sediments. These units should be exposed tinues long enough to generate steady-state sub- onciled by the idea that suprasubduction zone above sea level. This defi nition is more consis- duction with downwelling mantle overlying the ophiolites formed in backarc basins (Dewey, tent with common usage of the term ophiolite sinking plate, then normal arc volcanics will cap 1976; Pearce, 2003). This reconciled structural than is the Penrose defi nition. the subduction initiation sequence as the locus evidence of seafl oor spreading with geochemi- Based on emplacement mechanisms, Waka- of magmatism retreats from the trench. This cal evidence for convergent margin setting, but it bayashi and Dilek (2003) distinguished four types magmagenetic evolution is portrayed in Fig- is diffi cult to see how a backarc basin, ~200 km of ophiolites: (1) “Tethyan” ophiolites, emplaced ures 7 and 11, which encapsulate the subduction from the convergent plate boundary, would be over passive continental margins; (2) “Cordille- initiation tectonic evolution shown in Figure 9. emplaced (Fig. 10B). This logic train is further ran” ophiolites, emplaced over subduction com- Whattam and Stern (2011) outlined the argu- derailed for some ophiolites interpreted as fossil plexes; (3) “ridge-trench intersection” ophiolites ments and implications of the subduction initia- backarc basins because the associated arc and emplaced through complex processes resulting tion rule and proposed that ophiolites showing forearc, which are typically ~200 km wide and from the interaction between a spreading ridge this progression be termed “subduction initia- ~30 km thick, are usually not identifi ed (e.g., and a subduction zone; and (4) the unique Mac- tion rule ophiolites.” Semail ophiolite interpretation of Godard et al., quarie Island ophiolite, which was subaerially It has long been recognized that some 2003). Furthermore, there are no known Ceno- exposed as a result of a change in plate-boundary ophiolites contain igneous rocks with strong zoic examples of backarc basin closure (Stern, confi guration along a mid-ocean-ridge system. chemical affi nities to both mid-ocean-ridge 2004), which is required to emplace a backarc The fi rst two types represent the overwhelming and arc basalts. Distinguishing compositions

basin ophiolite. majority of ophiolites, with the second two types of MORB-like lavas include >1% TiO2, vari- A simple and elegant solution to the ophio- comprising miniscule proportions. Tethyan- and able depletion in LREEs, absence of HFSEs lite emplacement problem is that whatever Cordilleran-type ophiolites reﬂ ect the fundamen- (e.g., Nb, Ta), depletions on spider (primitive crust comprises the forearc is most likely to be tally different nature of the subducted oceanic mantle– or N-MORB–normalized) diagrams, emplaced during plate collision. This process is realms, with Tethyan ophiolites subducting rela- La/Nb <1, etc.; we regard these early MORB- often referred to as “obduction,” although this tively narrow oceans before colliding with conti- like sequences as forearc basalts. In contrast,

term originally described how oceanic crust on nental fragments and Cordilleran ophiolites gen- volcanic arc–like basalts have lower TiO2, are the downgoing plate was thrust over the con- erally subducting Pacifi c seafl oor, which contains enriched in fl uid-mobile elements (e.g., LILEs vergent margin (Coleman, 1971). Forearc litho- no continental fragments. and LREEs), and have strong HFSE depletions sphere is optimally situated for obduction (Wak- Given that most ophiolites originate in relative to LREE (e.g., La/Nb >1, etc.; Pearce, abayashi and Dilek, 2003). It is straightforward forearcs, is it possible to independently deter- 2003). Recognition of both arc-like and MORB- to envisage emplacement of forearc lithosphere mine whether or not a given ophiolite formed like compositions in some ophiolites has above the same subduction system in which it during subduction initiation or as a result of encouraged some workers to infer formation in was generated due to isostatic uplift following some other process, as outlined herein? Such a backarc basin (e.g., Beccaluva et al., 2004) or a partial subduction of buoyant crust that jams the an assessment would be very useful because of complex, multistage tectonic history, for exam- subduction zone (Fig. 10A). This process—akin the diffi culties involved in studying submerged ple, eruption of the two suites in discrete tec- to sliding a spatula under a pancake or fried egg forearc igneous rocks and reconstructing sub- tonic environments (e.g., Saccani and Photiades, to lift it—is also most likely to yield the least- duction initiation processes. If ophiolites could 2004; Godard et al., 2003). The reasons against disrupted ophiolites, and those most likely to be so linked, it could pay huge dividends in formation as a backarc basin are outlined in the approximate the Penrose ophiolite ideal, such advancing our understanding of the fundamen- previous section. The conclusion here that the as Troodos (Cyprus) and Semail (Oman). Dur- tal tectonic province of forearc and processes of basaltic sections in ophiolites are forearc basalts ing the early days following the plate-tectonic subduction initiation. The next section summa- (Reagan et al., 2010) is supported by the obser- revolution, when it was thought that forearc crust rizes a chemostratigraphic approach for evalu- vation that most Izu-Bonin-Mariana forearc was relict, trapped oceanic crust, this provided ating whether or not a given ophiolite formed basalts have lower Ti/V ratios than MORB, an attractive way to emplace MORB-type ophio- during subduction initiation. which probably results from the enhanced melt- lites. As discussed previously herein, there is lit- ing that commonly occurs in nascent subduction tle evidence to support the idea that forearcs are THE SUBDUCTION INITIATION RULE settings. The presence of transitional lavas with composed of trapped oceanic crust that existed compositions that progress from forearc basalts in the region prior to the formation of a new sub- The subduction initiation rule (Whattam and to boninite with time at DSDP Sites 458 and 459 duction zone. It is increasingly clear from the Stern, 2011) predicts that ophiolites that form also supports this contention. Some ophiolites simple perspective of plausible emplacement as a result of subduction initiation processes may refl ect complex tectonic histories, but this mechanisms that forearcs are the most likely consist of a sequence of igneous rocks formed cannot serve as the general explanation for sub- source of ophiolites (Casey and Dewey, 1984; by a magma source that changed progressively duction initiation rule ophiolites. Furthermore, Milson, 2003; Metcalf and Shervais, 2008). in composition by the combined effects of melt any inference of complex tectonic histories is Because most ophiolites are very incom- depletion and subduction-related metasoma- inconsistent with the absence of unconformi- plete and rarely satisfy the Penrose defi nition, tism. Magmas erupted during subduction initia- ties or signifi cant sedimentary horizons between

LITHOSPHERE | Volume 4 | Number 6 | www.gsapubs.org 479

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 STERN ET AL.

The four Tethyan ophiolites mentioned here The Subduction Initiation Rule show subduction initiation rule relationships, with lower lavas being more MORB-like and Lava Mantle Age upper lavas being more arc-like, as Whattam and Stern (2011) showed. Troodos and Semail composition composition are at either end of the 3000-km-long ophiolite Youngest belt that marks the Late Cretaceous forearc of Calc-alkaline SW Asia, including the inner and outer ophio- Sometimes boninite Harzburgitic mantle source Mostly LREE-depleted lite belts of Zagros, Iran (Moghadam and Stern, Modified by fluids from Lower ΣREE 2011). The magmatic chemostratigraphies of the subducted crust /sediments Higher TiO2 , Y, Zr many Zagros ophiolites have not been worked Lower Ti/V Higher Cr/Y out, but these ophiolites in many cases have strong compositional afﬁ nities in some respect to arc magmas, for example, with respect to La/ Nb (arc lavas usually have La/Nb >1.4 according to Condie, 1999). It is very likely that we will be able to work out magmatic stratigraphies

1–3 km in more ophiolites around the world, but this will require thoughtful integrated structural and petrologic studies.

DISCUSSION AND CONCLUSIONS Tholeiitic LREE-depleted Subduction initiation rule ophiolites provide Higher ΣREE Higher TiO2 , Y, Zr Lherzolitic mantle source wonderful opportunities to better understand the Higher Ti/V Unaffected by fluids from crust and upper mantle of magmatic forearcs Lower Cr/Y subducted crust/sediments Oldest and the ways in which new subduction zones form. Identifying and studying subduction ini- Figure 11. The subduction initiation rule, simplifi ed after Whattam and Stern (2011). Ophiolites tiation rule ophiolites will provide easy access (and forearc crust) that form as a result of subduction initiation preserve systematic variations in basalt compositions (left), from mid-ocean-ridge basalt (MORB)–like at the base to volcanic arc)– to forearc lithosphere samples and structures, like basalt (VAB), even boninitic, at the top. This refl ects changes in the source mantle as a new allowing many scientifi c perspectives to be subduction zone starts, beginning with adiabatic upwelling of asthenosphere that is not infl u- engaged cheaply and easily. Where ophiolites enced by subduction-related fl uids (Fig. 9B) to form MORB-like basalts (forearc basalts) by seafl oor can be fi rmly linked to forearcs through applica- spreading. Fluids from the sinking lithosphere become increasingly important with time, eventu- tion of the subduction initiation rule and other ally reaching and metasomatically re-enriching the increasingly depleted mantle source of melts approaches, geodynamic models for subduc- (Fig. 9C). Asthenospheric upwelling diminishes in importance with time and is ultimately replaced by induced convection as sinking lithosphere begins downdip motion, and true subduction begins tion initiation can be more easily developed (Fig. 9D). This isolates the forearc mantle wedge, leading to establishment of localized magmatic and tested. Because subduction initiation rule arc behind a cold, dead forearc. LREE—light rare earth element; ∑REE— total rare earth element ophiolites are fossil forearcs, they often can be concentrations. traced across strike for tens of kilometers and along strike for hundreds of kilometers. Ero- sion of deformed subduction initiation rule ophiolite lavas with MORB-like versus volca- ingly interpreted to be products of magmatism ophiolites exposes various levels, allowing four- nic arc–like basalt compositions. Such inter- in a single (suprasubduction zone) tectonic set- dimensional reconstructions of timing as well ruptions might be diffi cult to identify because ting, albeit one that changed rapidly. as vertical, across-strike, and along-strike mag- of ophiolite deformation, but such an important It is worth emphasizing that it is generally matic variations. Sampling for geochronology is and distinctive feature should sometimes be diffi cult to recognize a magmatic stratigraphy in easier than for in situ forearcs because of this recognized if this interpretation is generally cor- ophiolites, because they are so often jumbled by exposure, and relations of such dated samples to rect. Conversely, the interpretation of a single, faulting, and because lavas and associated intru- the surrounding rocks and fabrics provide con- rapidly evolving magmagenetic environment is sions with different chemistries appear similar text for interpreting the ages. The multiple per- favored because such a hiatus is generally not in the fi eld. We recognize four ophiolites that spectives and levels of detail allowed by these observed. In addition, lavas with compositions have been studied in suffi cient detail to recon- approaches mean that many aspects of forearc that are transitional between forearc basalts and struct their magmatic stratigraphies: Mirdita, crust structure and subduction initiation can be boninites have been observed in some ophiolites Pindos, Troodos, and Semail (as summarized understood much better by indirect study of sub- (e.g., Dilek and Furnes, 2009). As a result, ophi- by Whattam and Stern, 2011). There are other duction initiation rule ophiolites than by study- olite compositional variability is increasingly examples of subduction initiation rule ophiolites ing forearcs directly, as can be seen by mentally ascribed to progressive depletion and metaso- that could also be considered, for example, the comparing the experiences captured in Figure 1. matism of the mantle source as the ophiolite 485–489 Ma ophiolite belt that can be traced Even as we recognize that studies of sub- magmatic crust forms (e.g., Shervais, 2001; >1000 km from Newfoundland down into the duction initiation rule ophiolites are essential Beccaluva at al., 2005; Dilek and Furnes, 2009). Taconic suture of New York (Bédard et al., for understanding forearcs, studies of in situ Ophiolite lava compositions thus are increas- 1998; Schroetter et al., 2003; Pagé et al., 2009). forearc crust need to move forward. We are only

480 www.gsapubs.org | Volume 4 | Number 6 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 Subduction initiation, forearcs, and ophiolites | RESEARCH

beginning to understand the igneous crust of a ferent than that seen for Izu-Bonin-Mariana, for Bloomer, S.H., Falloon, T., Petcovic, H., Nielsen, R.L., and Dun- can, R.A., 1998, Petrology and geochemistry of volcanic single forearc in any detail—that of Izu-Bonin- example, by having upper alkalic lavas instead and plutonic rocks from the Tonga forearc, southwest Mariana—and this example seems to be unusual of volcanic arc–like basalt overlying tholeiites. Pacifi c (abstract): Eos (Transactions, American Geophysi- in terms of the abundance of early-arc boninites. Above all, combined studies of subduction cal Union), v. 79, no. 45, Fall Meeting supplement, p. F420. Bonatti, E., and Michael, P.J., 1989, Mantle peridotites from We need to continue to test and refi ne what we initiation rule ophiolites and forearc crust need continental rifts to ocean basins to subduction zones: know about forearc magmatic evolution. It is to be integrated with geodynamic modeling to Earth and Planetary Science Letters, v. 91, p. 297–311, especially important that we drill and sample learn about the ways in which new subduction doi:10.1016/0012-821X(89)90005-8. Calvert, A.J., Klemperer, S.L., Takahashi, N., and Kerr, B.C., through the magmatic stratigraphy of an igneous zones form. Geodynamic modeling of subduc- 2008, Three-dimensional crustal structure of the Mari- forearc in order to better understand the sequence tion initiation needs to be fi rmly tethered in real- ana island arc from seismic tomography: Journal of and timing of magmas at one location, allowing Geophysical Research–Solid Earth, v. 113, doi:10.1029 ity, and the more that such models attempt to /2007JB004939. us to answer questions such as: What are the explain the rocks making up a real forearc and Casey, J.F., and Dewey, J.F., 1984, Initiation of subduction relative proportions of forearc basalts versus vol- ophiolite, the more rapidly our understanding of zones along transform and accreting plate boundaries, triple-junction evolution, and forearc spread- canic arc–like basalt/boninitic lavas? How does this fundamental Earth processes will advance. ing centres—Implications for ophiolitic geology and forearc basalt magmatism transition to volcanic obduction in Gass, G., Lippard, S.J., and Shelton, A.W., arc–like basalt magmatism? Is it gradational or ACKNOWLEDGMENTS eds., Ophiolites and Oceanic Lithosphere: Geological Society of London Special Publication 13, p. 269–290, abrupt? How long does forearc volcanism last? doi:10.1144/GSL.SP.1984.013.01.22. Answering such questions for an in situ forearc We thank Dave Scholl for edifying comments on Clift, P.D., and Vannucchi, P., 2004, Controls on tectonic accre- and comparing these answers with those for sub- tion versus erosion in subduction zones: Implications for accretionary prisms and Steve Graham for the the origin and recycling of the continental crust: Reviews duction initiation rule ophiolites will be key tests photo of a Franciscan chert exposure. Thought- of Geophysics, v. 42, RG2001, doi:10.1029/2003RG000127. of the ideas presented in this paper. Furthermore, ful reviews by Rod Metcalf, Jean Bédard, and Coleman, R.G., 1971, Plate tectonic emplacement of upper mantle peridotites along continental edges: Journal of there are aspects of forearc structure and evolu- editors John Shervais and John Goodge are Geophysical Research, v. 76, p. 1212–1222, doi:10.1029 tion that cannot be understood without studying much appreciated. This manuscript was written /JB076i005p01212. extant forearcs. For example, active serpentine while Stern enjoyed a Blaustein Fellowship at Condie, K.C., 1999, Mafi c crustal xenoliths and the origin of the lower continental crust: Lithos, v. 46, p. 95–101, mud volcanoes in the Mariana forearc serve as Stanford University. Stern’s research on intra- doi:10.1016/S0024-4937(98)00056-5. actualistic models for sedimentary serpentinite oceanic arcs has been supported by the National Crawford, A.J., Meffre, S., and Symonds, P.A., 2003, Chapter deposits associated with some forearcs (Fryer 25—120 to 0 Ma tectonic evolution of the southwest Science Foundation, most recently by grant Pacifi c and analogous geological evolution of the 600 et al., 1995; Fryer, 2002). Knowing that these OCE-0961352. Reagan acknowledges research to 220 Ma Tasman fold belt system, in Hillis, R.R., and mud volcanoes exist has stimulated the search support from National Science Foundation grant Müller, R.D., eds., Evolution and Dynamics of the Aus- tralian Plate: Geological Society of Australia Special for ancient serpentine mudfl ows (e.g., Teklay, EAR-0840862. This is University of Texas at Publication 22, p. 377–397. 2006). Another example is the problem of tec- Dallas Geosciences contribution 1228. DeBari, S.M., Taylor, B., Spencer, K., and Fujioka, K., 1999, tonic erosion, which can remove a few kilome- A trapped Philippine Sea plate origin for MORB from the inner slope of the Izu-Bonin Trench: Earth and Plan- ters of forearc crust in a few million years. Stud- REFERENCES CITED etary Science Letters, v. 174, p. 183–197, doi:10.1016 ies of active systems are generally aware that /S0012-821X(99)00252-6. some crust is missing, whereas studies of ophiol- Anonymous, 1972, Penrose fi eld conference on ophiolites: Deschamps, A., and Lallemand, S., 2003, Geodynamic set- Geotimes, v. 17, p. 24–25. ting of Izu-Bonin-Mariana boninites, in Larter, R.D., ites rarely consider this complication. Beccaluva, L., Coltori, M., Giunta, G., and Siena, F., 2004, and Leat, P.T., eds., Intra-Oceanic Subduction Systems: Another point worth emphasizing is that Tethyan vs. Cordilleran ophiolites: A reappraisal of dis- Tectonic and Magmatic Processes: Geological Society there may be a wide range of subduction ini- tinctive tectono-magmatic features of supra-subduction of London Special Publication 219, p. 163–185. complexes in relation to the subduction mode: Tectono- Dewey, J.F., 1976, Ophiolite obduction: Tectonophysics, tiation magmatic products. Our models now physics, v. 393, p. 163–174, doi:10.1016/j.tecto.2004.07.034. v. 31, p. 93–120, doi:10.1016/0040-1951(76)90169-4. are heavily biased toward the Izu-Bonin-Mar- Beccaluva, L., Coltorti, M., Saccani, E., and Siena, F., 2005, Dick, H.J.B., and Bullen, T., 1984, Chromian spinel as a pet- Magma generation and crustal accretion as evidenced rogenetic indicator in abyssal and Alpine-type perido- iana convergent margin, which has abundant by supra-subduction ophiolites of the Albanide- tites and spatially associated lavas: Contributions to boninitic lavas and serpentinite mud volcanoes. Hellenide Subpelagonian zone: The Island Arc, v. 14, Mineralogy and Petrology, v. 86, p. 54–76, doi:10.1007 The Izu-Bonin-Mariana forearc may serve as p. 551–563, doi:10.1111/j.1440-1738.2005.00483.x. /BF00373711. Bédard, J.H., Lauzière, K., Tremblay, A., and Sangster, A., Dickinson, W.R., 2001, Ch. 16: The coming of plate tectonics a useful example of intra-oceanic subduction 1998, Evidence from Betts Cove ophiolite boninites to the Pacifi c Rim, in Oreskes, N., ed., Plate Tectonics: initiation, but it could be less useful as a model for forearc seafl oor-spreading: Tectonophysics, v. 284, An Insider’s History of the Modern Theory of the Earth: for continental margin subduction initiation. p. 233–245, doi:10.1016/S0040-1951(97)00182-0. Boulder, Colorado, Westview Press, p. 264–287. Bloomer, S.H., 1983, Distribution and origin of igneous rocks Dickinson, W.R., 2004, Evolution of the North American Subcontinental mantle lithosphere is compo- from the landward slopes of the Mariana Trench: Jour- Cordillera: Annual Review of Earth and Planetary Sci- sitionally distinct from that beneath the ocean nal of Geophysical Research, v. 88, no. B9, p. 7411–7428, ences, v. 32, p. 13–45, doi:10.1146/annurev.earth.32 doi:10.1029/JB088iB09p07411. .101802.120257. basins (Bonatti and Michael, 1989), and conti- Bloomer, S.H., and Fisher, R.L., 1987, Petrology and geo- Dickinson, W.R., and Sealey, D.R., 1979, Structure and stra- nental and oceanic mantle may melt to different chemistry of igneous rocks from the Tonga Trench—A tigraphy of forearc regions: The American Association extents during subduction initiation. Because non-accreting plate boundary: The Journal of Geology, of Petroleum Geologists Bulletin, v. 63, p. 2–31. v. 95, p. 469–495, doi:10.1086/629144. Dilek, Y., 2003, Ophiolite concept and its evolution, in Dilek, there are likely to be signifi cant differences in Bloomer, S.H., Ewart, W., Hergt, J.M., and Bryant, W.T., 1994, Y., and Newcomb, S., eds., Ophiolite Concept and the magmatic products from that shown by Izu- Geochemistry of igneous rocks from ODP Site 841, Evolution of Geological Thought: Geological Society of Bonin-Mariana and ophiolites, we should keep Tonga forearc, in Hawkins, J.W., Parsons, L., Allan, America Special Paper 373, p. 1–16. J., et al., Proceedings of the Ocean Drilling Program, Dilek, Y., and Furnes, H., 2009, Structure and geochemis- an open mind when considering the origin of Scientifi c Results Volume 135: College Station, Texas, try of Tethyan ophiolites and their petrogenesis in any magmatic forearc built about the time that Ocean Drilling Program, p. 625–646. subduction rollback systems: Lithos, v. 113, p. 1–20, subduction began at a given convergent mar- Bloomer, S.H., Taylor, B., MacLeod, C.J., Stern, R.J., Fryer, doi:10.1016/j.lithos.2009.04.022. P., Hawkins, J.W., and Johnson, L., 1995, Early arc Dilek, Y., Furnes, H., and Shallo, M., 2008, Geochemistry of the gin. The Siletzia terrane of coastal Oregon and volcanism and the ophiolite problem: A perspective Jurassic Mirdita ophiolite (Albania) and the MORB to Washington (Duncan, 1982) could be an exam- from drilling in the western Pacifi c, in Taylor, B., and SSZ evolution of a marginal basin oceanic crust: Lithos, Natland, J., eds., Active Margins and Marginal Basins v. 100, p. 174–209, doi:10.1016/j.lithos.2007.06.026. ple of continental margin subduction initiation of the Western Pacifi c: American Geophysical Union Duncan, R.A., 1982, A captured island chain in the Coast where the magmatic sequence is somewhat dif- Geophysical Monograph 88, p. 1–30. Range of Oregon and Washington: Journal of Geophysi-

LITHOSPHERE | Volume 4 | Number 6 | www.gsapubs.org 481

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 STERN ET AL.

cal Research, v. 87, no. B13, p. 10,827–10,837, doi:10.1029 Ishizuka, O., Kimura, J.-I., Li, Y.B., Stern, R.J., Reagan, M., Tay- Minyuk, P.S., and Stone, D.B., 2009, Paleomagnetic deter- /JB087iB13p10827. lor, R.N., Ohara, Y., Bloomer, S.H., Ishii, T., Hargrove, U.S., mination of paleolatitude and rotation of Bering Island Duncan, R.A., Vallier, T.L., and Falvey, D.A., 1985, Volcanic III, and Haraguchi, S., 2006, Early stages in the evolution (Komandorsky Islands) Russia: Comparison with rota- episodes at ‘Eua, Tonga islands, in Scholl, D.W., and of Izu-Bonin arc volcanism: New age, chemical, and iso- tions in the Aleutian Islands and Kamchatka: European Vallier, T.L., eds., Geology and Offshore Resources topic constraints: Earth and Planetary Science Letters, Geophysical Union Stephan Mueller Special Publica- of the Paciﬁ c Island Arcs–Tonga Region: Houston, v. 250, p. 385–401, doi:10.1016/j.epsl.2006.08.007. tion Series, v. 4, p. 329–348, doi:10.5194/smsps-4-329 Texas, Circum-Paciﬁ c Council for Energy and Mineral Ishizuka, O., Tani, K., Reagan, M.K., Kanayama, K., Umino, -2009. Resources, and American Association of Petroleum S., Harigane, Y., Sakamoto, I., Miyajima, Y., Yuasa, M., Miyashiro, A., 1973, The Troodos ophiolitic complex was Geologists, Earth Science Series, v. 2, p. 281–290. and Dunkley, D.J., 2011, The timescales of subduction probably formed in an island arc: Earth and Planetary Ewart, A., and Bryan, W.B., 1972, Petrography and geochemis- initiation and subsequent evolution of an oceanic Science Letters, v. 19, p. 218–224, doi:10.1016/0012-821X try of the igneous rocks from E’ua, Tongan Islands: Geo- island arc: Earth and Planetary Science Letters, v. 306, (73)90118-0. logical Society of America Bulletin, v. 83, p. 3281–3298, p. 229–240, doi:10.1016/j.epsl.2011.04.006. Miyashiro, A., 1974, Volcanic rock series in island arcs and doi:10.1130/0016-7606(1972)83[3281:PAGOTI]2.0.CO;2. Jicha, B.R., Scholl, D.W., Singer, B.S., Yogodzhinsky, G.M., active continental margins: American Journal of Sci- Falloon, T.J., and Danyushevsky, L.V., 2000, Melting of refrac- and Kay, S.M., 2006, Revised age of Aleutian Island ence, v. 274, p. 321–355, doi:10.2475/ajs.274.4.321. tory mantle at 1.5, 2 and 2.5 GPa under anhydrous and arc formation implies high rate of magma production: Moghadam, H.S., and Stern, R.J., 2011, Geodynamic evolu-

H2O-undersaturated conditions: Implications for the Geology, v. 34, p. 661–664, doi:10.1130/G22433.1. tion of late Cretaceous Zagros ophiolites: Formation of petrogenesis of high-Ca boninites and the infl uence of Kamimura, A., Kasahara, J., Shinohara, M., Hino, R., Shio- oceanic lithosphere above a nascent subduction zone: subduction components on mantle melting: Journal bara, H., Fujie, G., and Kanazawa, T., 2002, Crustal Geological Magazine, v. 148, p. 762–801, doi:10.1017 of Petrology, v. 41, p. 257–283, doi:10.1093/petrology structure study at the Izu-Bonin subduction zone /S0016756811000410. /41.2.257. around 31°N: Implications of serpentinized materi- Morishita, T., Tani, K., Shukono, H., Harigane, Y., Tamura, A., Falloon, T.J., Green, D.H., and Crawford, A., 1987, Dredged als along the subduction plate boundary: Physics of Kumagai, H., and Hellebrand, E., 2011, Diversity of igneous rocks from the northern termination of the the Earth and Planetary Interiors, v. 132, p. 105–129, melt conduits in the Izu-Bonin-Mariana forearc mantle: Tofua magmatic arc, Tonga and adjacent Lau Basin: doi:10.1016/S0031-9201(02)00047-X. Implications for the earliest stage of arc magmatism: Australian Journal of Earth Sciences, v. 34, p. 487–506, Kimura, J.-I., Stern, R.J., and Yoshida, T., 2005, Re-initiation Geology, v. 39, p. 411–414, doi:10.1130/G31706.1. doi:10.1080/08120098708729428. of subduction and magmatic responses in SW Japan Nakamura, N., 1974, Determination of REE, Ba, Fe, Mg, Fisher, R.L., and Engel, C.G., 1969, Ultramafi c and basaltic during Neogene time: Geological Society of America Na and K in carbonaceous and ordinary meteorites: rocks dredged from the nearshore fl ank of the Tonga Bulletin, v. 117, p. 969–986, doi:10.1130/B25565.1. Geochimica et Cosmochimica Acta, v. 38, p. 757–775, Trench: Geological Society of America Bulletin, v. 80, Kopp, H., and Kukowski, N., 2003, Backstop geometry and doi:10.1016/0016-7037(74)90149-5. p. 1373–1378, doi:10.1130/0016-7606(1969)80[1373:UAB accretionary mechanics of the Sunda margin: Tecton- Niu, Y., O’Hara, M.J., and Pearce, J.A., 2003, Initiation of RDF]2.0.CO;2. ics, v. 22, no. 6, p. 1072, doi:10.1029/2002TC001420. subduction zones as a consequence of lateral com- Flower, M.F.J., and Levine, H.M., 1987, Petrogenesis of a Lagabrielle, Y., Sizun, J.-P., and Arculus, R.J., 1992, The con- positional buoyancy contrast within the lithosphere, tholeiite-boninite sequence from Ayios Mamas, Troo- structional and deformational history of the igneous a petrological perspective: Journal of Petrology, v. 44, dos ophiolite: Evidence for splitting of a volcanic arc?: basement penetrated at Site 786, in Fryer, P., Pearce, p. 851–866, doi:10.1093/petrology/44.5.851. Contributions to Mineralogy and Petrology, v. 97, J.A., Stokking, L.B., et al., Proceedings of the Ocean Oakley, A.J., Taylor, B., and Moore, G.F., 2008, Pacifi c plate p. 509–524, doi:10.1007/BF00375328. Drilling Program, Scientifi c Results, Volume 125: Col- subduction beneath the central Mariana and Izu-Bonin Flower, M.F.J., Russo, R.M., Tamaki, K., and Hoang, N., 2001, lege Station, Texas, Ocean Drilling Program, p. 263–276. forearcs: New insights from an old margin: Geochem- Mantle contamination and the Izu-Bonin-Mariana LeBas, M.J., and Streckeisen, A.L., 1991, The IUGS sys- istry Geophysics Geosystems, v. 9, no. 6, Q06003, (IBM) ‘high-tide mark’: Evidence for mantle extrusion tematics of igneous rocks: Journal of the Geological doi:10.1029/2007GC001820. by Tethyan closure: Tectonophysics, v. 333, p. 9–34, Society of London, v. 148, p. 825–833, doi:10.1144 Ohara, Y., and Ishii, T., 1998, Peridotites from southern Mari- doi:10.1016/S0040-1951(00)00264-X. /gsjgs.148.5.0825. ana forearc: Heterogeneous fl uid supply in mantle Fryer, P., 2002, Recent studies of serpentinite occurrences LeBrun, J.-F., Lamarche, G., and Collot, J.-Y., 2003, Subduc- wedge: The Island Arc, v. 7, p. 541–552. in the oceans: Mantle-ocean interactions in the plate tion initiation at a strike-slip boundary: The Cenozoic Ohara, Y., Stern, R.J., Ishii, T., Yurimoto, H., and Yamazaki, T., tectonic cycle: Chemie der Erde, v. 62, p. 257–302, Pacifi c-Australian plate boundary, south of New Zea- 2002, Peridotites from the Mariana Trough: First look doi:10.1078/0009-2819-00020. land: Journal of Geophysical Research, v. 108, no. B9, at the mantle beneath an active back-arc basin: Con- Fryer, P., Mottl, M., Johnson, L., Haggerty, J., Phipps, S., and 2453, doi:10.1029/2002JB002041. tributions to Mineralogy and Petrology, v. 143, p. 1–18, Maekawa, H., 1995, Serpentine bodies in the forearcs of Lyubetskaya, T., and Korenaga, J., 2007, Chemical com- doi:10.1007/s00410-001-0329-2. western Pacifi c convergent margins: Origin and asso- position of Earth’s primitive mantle and its variance: Okamura, H., Sarai, S., and Kim, Y.U., 2006, Petrology of ciated fl uids, in Taylor, B., and Natland, J., eds., Active 1. Methods and results: Journal of Geophysical forearc peridotite from the Hahajima Seamount, the Margins and Marginal Basins of the Western Pacifi c: Research, v. 112, B03211, doi:1029/2005JB004223. Izu-Bonin arc, with special reference to chemical char- American Geophysical Union Geophysical Monograph Macpherson, C.G., and Hall, R., 2001, Tectonic setting of acteristics of chromian spinel: Mineralogical Maga- 88, p. 259–279. Eocene boninite magmatism in the Izu-Bonin-Mariana zine, v. 70, p. 15–26, doi:10.1180/0026461067010310. Gill, J.B., 1981, Orogenic Andesites and Plate Tectonics: New forearc: Earth and Planetary Science Letters, v. 186, Pagé, P., Bédard, J.H., and Tremblay, A., 2009, Geochemical York, Springer-Verlag, 390 p. p. 215–230, doi:10.1016/S0012-821X(01)00248-5. variations in a depleted fore-arc mantle: The Ordovi- Glodny, J., Lohrmnann, J., Echtler, H., Gräfe, K., Seifert, W., Mann, P., and Taira, A., 2004, Global tectonic signifi cance cian Thetford Mines ophiolite: Lithos, v. 113, p. 21–47, Collao, S., and Figueroa, O., 2005, Internal dynamics of the Solomon Islands and Ontong Java Plateau doi:10.1016/j.lithos.2009.03.030. of a paleoaccretionary wedge: Insights from combined convergent zone: Tectonophysics, v. 389, p. 137–190, Parkinson, I.J., and Pearce, J.A., 1998, Peridotites from the isotope tectonochronology and sandbox modeling of doi:10.1016/j.tecto.2003.10.024. Izu-Bonin-Mariana forearc (ODP Leg 125): Evidence the south-central Chilean forearc: Earth and Planetary McDougall, I., 1994, Dating of rhyolitic glass in the Tonga for mantle melting interaction in a supra-subduction Science Letters, v. 231, p. 23–39, doi:10.1016/j.epsl forearc (Hole 841B), in Hawkins, J.W., Parsons, L., and zone setting: Journal of Petrology, v. 39, p. 1577–1618, .2004.12.014. Allan, J., et al., Proceedings of the Ocean Drilling Proj- doi:10.1093/petrology/39.9.1577. Godard, M., Dautria, J.-M., and Perrin, M., 2003, Geochemi- ect, Scientifi c Results Leg 135: College Station, Texas, Pearce, J.A., 2003, Supra-subduction zone ophiolites: The cal variability of the Oman ophiolite lavas: Relation- Ocean Drilling Program, p. 923–924. search for modern analogues, in Dilek, Y., and New- ship with spatial distribution and paleomagnetic direc- Metcalf, R.V., and Shervais, J.W., 2008, Supra-subduction comb, S., eds., Ophiolite Concept and the Evolution tions: Geochemistry Geophysics Geosystems, v. 4, zone ophiolites: Is there really an ophiolite conun- of Geological Thought: Geological Society of America no. 6, 8609, doi:10.1029/2002GC000452. drum?, in Wright, J.E., and Shervais, J.W., eds., Ophio- Special Paper 373, p. 269–293. Green, D.H., and Falloon, T.J., 1998, Pyrolite: A Ringwood lites, Arcs, and Batholiths: Geological Society of Amer- Pearce, J.A., 2008, Geochemical fi ngerprinting of oceanic concept and its current expression, in Jackson, I., ed., ica Special Paper 438, p. 191–222. basalts with applications to ophiolite classifi cation and The Earth’s Mantle: New York, Cambridge University Michibayashi, K., Tasaka, M., Ohara, Y., Ishii, T., Okamoto, A., the search for Archean oceanic crust: Lithos, v. 100, Press, p. 311–378. and Fryer, P., 2007, Variable microstructure of peridotite p. 14–48, doi:10.1016/j.lithos.2007.06.016. Hall, C.E., Gurnis, M., Sdrolias, M., Lavier, L.L., and Mül- samples from the southern Mariana Trench: Evidence Pearce, J.A., and Parkinson, I.J., 1993, Trace element models ler, R.D., 2003, Catastrophic initiation of subduction of a complex tectonic evolution: Tectonophysics, for mantle melting: Application to volcanic arc petro- following forced convergence across fracture zones: v. 444, p. 111–118, doi:10.1016/j.tecto.2007.08.010. genesis, in Prichard, H.M., Alabaster, T., Harris, N.B., Earth and Planetary Science Letters, v. 212, p. 15–30, Michibayashi, K., Ohara, Y., Stern, R.J., Fryer, P., Kimura, J.-I., and Neary, C.R., eds., Magmatic Processes and Plate doi:10.1016/S0012-821X(03)00242-5. Tasaka, M., Harigane, Y., and Ishii, T., 2009, Peridotites Tectonics: The Geological Society of London Special Hilde, T.W.C., 1983, Sediment subduction versus accretion derived from a ductile shear zone within backarc litho- Publication 76, p. 373–403. around the Pacifi c: Tectonophysics, v. 99, p. 381–397, spheric mantle, southern Mariana Trench: Results of Pearce, J.A., Lippard, S.J., and Roberts, S., 1984, Character- doi:10.1016/0040-1951(83)90114-2. a Shinkai 6500 dive: Geophysics Geochemistry Geo- istics and tectonic signifi cance of supra-subduction Humphreys, E., 2008, Cenozoic slab windows beneath the systems, v. 10, doi:10.1029/2008GC002197. zone ophiolites, in Kokelaar, B.P., and Howells, M.F., western United States, in Circum-Pacifi c Tectonics, Geo- Milson, J., 2003, Forearc ophiolites: A view from the west- eds., Marginal Basin Geology: Volcanic and Associated logic Evolution, and Ore Deposits Symposium—Pro- ern Pacifi c, in Dilek, Y., and Robinson, P.T., eds., Ophiol- Sedimentary and Tectonic Processes in Modern and ceedings Volume: Arizona Geological Society Digest, ites in Earth History: The Geological Society of London Ancient Marginal Basins: Geological Society of Lon- p. 389–398. Special Publications 218, p. 507–515. don Special Publication 16, p. 77–94.

482 www.gsapubs.org | Volume 4 | Number 6 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021 Subduction initiation, forearcs, and ophiolites | RESEARCH

Pearce, J.A., Vander Laan, S.R., Arculus, R.J., Murton, B.J., struction, in Hatcher, R.D., Jr., Carlson, M.P., McBride, gin, southwest New Zealand: New Zealand Journal of Ishii, T., Peate, D.W., and Parkinson, I.J., 1992, Boninite J.H., and Martínez Catalán, J.R., eds., 4-D Framework Geology and Geophysics, v. 49, p. 131–149. and harzburgite from Leg 125 (Bonin-Mariana forearc): of Continental Crust: Geological Society of America Tappin, D.R., and Balance, P.F., 1994, Contributions to the A case study of magma genesis during the initial Memoir 200, p. 9–32. sedimentary geology of ‘Eua Island, Kingdom of stages of subduction, in Fryer, P., et al., Proceeding Scholl, D.W., and von Huene, R., 2009, Implications of esti- Tonga: Reworking in an oceanic forearc, in Stevenson, of the Ocean Drilling Program, Scientifi c Results, Site mated magmatic additions and recycling losses at A.J., Herzer, R.H., and Balance, P.F., eds., Geology and 778–786, Bonin-Mariana Region: College Station, the subduction zones of accretionary (non-collisional) Submarine Resources of the Tonga–Lau–Fiji Region: Texas, Ocean Drilling Program, p. 623–659. and collisional (suturing) orogens, in Cawood, P., and South Pacifi c Applied Geoscience Commission Pearce, J.A., Barker, P.F., Edwards, S.J., Parkinson, I.J., Kröner, A., eds., Accretionary Orogens in Space and (SOPAC), Suva, Technical Bulletin 8, p. 1–20. and Leat, P.T., 2000, Geochemistry and tectonic sig- Time: The Geological Society of London Special Publi- Taylor, B., and Goodliffe, A.M., 2004, The West Philip- nifi cance of peridotites from the South Sandwich cation 318, p. 105–125. pine Basin and the initiation of subduction, revis- arc-basin system, South Atlantic: Contributions to Schroetter, J.-M., Pagé, P., Bédard, J.H., Tremblay, A., and ited: Geophysical Research Letters, v. 31, doi:10.1029 Mineralogy and Petrology, v. 139, p. 36–53, doi:10.1007 Bécu, V., 2003, Forearc extension and seafl oor spread- /2004GL020136. /s004100050572. ing in the Thetford Mines ophiolite complex, in Dilek, Taylor, R.N., Nesbitt, R.W., Vidal, P., Harmon, R.S., Auvray, Peate, D., and Pearce, J., 1998, Causes of spatial composi- Y., and Robinson, P.T., eds., Ophiolites in Earth History: B., and Croudace, I.W., 1994, Mineralogy, chemistry, tional variations in Mariana arc lavas: Trace element Geological Society of London Special Publication 218, and genesis of the Boninite series volcanics, Chichi- evidence: The Island Arc, v. 7, p. 479–495. p. 231–251. jima, Bonin Islands, Japan: Journal of Petrology, v. 35, Plank, T., and Langmuir, C.H., 1992, Effects of the melting Sharp, W.D., and Clague, D.A., 2006, 50-Ma initiation of p. 577–617, doi:10.1093/petrology/35.3.577. regime on the composition of the oceanic crust: Jour- Hawaiian-Emperor Bend records major change in Teklay, M., 2006, Neoproterozoic arc–back-arc system ana- nal of Geophysical Research, v. 97, no. B13, p. 19,749– Pacifi c plate motion: Science, v. 313, p. 1281–1284, log to modern arc–back-arc systems: Evidence from 19,770, doi:10.1029/92JB01769. doi:10.1126/science.1128489. tholeiite-boninite association, serpentinite mudfl ows Pyle, D.G., Duncan, R., Wells, R.E., Graham, D.W., Harrison, Shervais, J.W., 2001, Birth, death, and resurrection: The life and across-arc geochemical trends in Eritrea, southern B., and Hanan, B., 2009, Siletzia: An oceanic large igne- cycle of suprasubduction zone ophiolites: Geochem- Arabian-Nubian Shield: Precambrian Research, v. 145, ous province in the Pacifi c Northwest: Geological Soci- istry Geophysics Geosystems v. 2, no. 1, doi:10.1029 p. 81–92, doi:10.1016/j.precamres.2005.11.015. ety of America Abstracts with Programs, v. 41, no. 7, /2000GC000080. Varne, R., Brown, A.V., and Falloon, T., 2000, Macquarie p. 369. Stein, C., 2003, Heat fl ow and fl exure at subduction zones: Island; its geology, structural history, and the timing Reagan, M.K., Ishizuka, O., Stern, R.J., Kelley, K.A., Ohara, Geophysical Research Letters, v. 30, doi:10.1029 and tectonic setting of its N-MORB to E-MORB mag- Y., Blichert-Toft, J., Bloomer, S.H., Cash, J., Fryer, P., /2003GL018478. matism, in Dilek, Y., et al., eds., Ophiolites and Oceanic Hanan, B.B., Hickey-Vargas, R., Ishii, T., Kimura, J.I., Stern, R.J., 2002, Subduction zones: Reviews of Geophys- Crust, New Insights from Field Studies and the Ocean Peate, D.W., Rowe, M.C., and Woods, M., 2010, Fore- ics, v. 40, doi:10.1029/2001RG000108. Drilling Program: Geological Society of America Spe- arc basalts and subduction initiation in the Izu-Bonin- Stern, R.J., 2004, Subduction initiation: Spontaneous and cial Paper 349, p. 301–320. Mariana system: Geochemistry Geophysics Geosys- induced: Earth and Planetary Science Letters, v. 226, Wakabayashi, J., and Dilek, Y., 2003, What constitutes tems, v. 11, Q03X12, doi:10.1029/2009GC002871. p. 275–292. ‘emplacement’ of an ophiolite?: Mechanisms and rela- Redfi eld, T.F., Scholl, D.W., Fitzgerald, P.G., and Beck, M.E., Stern, R.J., 2005, Evidence from ophiolites, blueschists, tionship to subduction initiation and formation of met- Jr., 2007, Escape tectonics and the extrusion of Alaska: and ultrahigh-pressure metamorphic terranes that amorphic soles, in Dilek, Y., and Robinson, P.T., eds., Past, present, and future: Geology, v. 35, p. 1039–1042, the modern episode of subduction tectonics began Ophiolites in Earth History: Geological Society of Lon- doi:10.1130/G23799A.1. in Neoproterozoic time: Geology, v. 33, p. 557–560, don Special Publications 218, p. 427–447, doi:10.1144 Saccani, E., and Photiades, A., 2004, Mid-ocean ridge doi:10.1130/G21365.1. /GSL.SP.2003.218.01.22. and supra-subduction affi nities in the Pindos Massif Stern, R.J., and Bloomer, S.H., 1992, Subduction zone Whattam, S.A., and Stern, R.J., 2011, The ‘subduction-initi- ophiolites (Greece): Implications for magma genesis infancy: Examples from the Eocene Izu-Bonin-Mariana ation rule’: A key for linking ophiolites, intra-oceanic in a proto-forearc setting: Lithos, v. 73, p. 229–253, and Jurassic California arcs: Geological Society of forearcs and subduction initiation: Contributions doi:10.1016/j.lithos.2003.12.002. America Bulletin, v. 104, p. 1621–1636, doi:10.1130/0016- to Mineralogy and Petrology, v. 162, p. 1031–1045, Scholl, D.W., 2007, Viewing the tectonic evolution of the 7606(1992)104<1621:SZIEFT>2.3.CO;2. doi:10.1007/s00410-011-0638-z. Kamchatka-Aleutian (KAT) connection with an Alaska Stern, R.J., Morris, J., Bloomer, S.H., and Hawkins, J.W., Whittaker, J.M., Müller, R.D., Leitchenkov, G., Stagg, H., crustal extrusion perspective, in Eichelberger, J., Gor- Jr., 1991, The source of metasomatic fl uids and the Sdrolias, M., Gaina, C., and Goncharov, A., 2007, Major deev, E., Izbekov, P., Kasahara, M., and Lees, J., eds., generation of arc melts: Trace element and radiogenic Australian-Antarctic plate reorganization at Hawaiian- Volcanism and Subduction: The Kamchatka Region: isotope evidence from Eocene boninites, Mariana Emperor Bend time: Science, v. 318, p. 83–86, doi:10.1126 American Geophysical Union Geophysical Mono- forearc: Geochimica et Cosmochimica Acta, v. 55, /science.1143769. graph 172, p. 3–35. p. 1467–1481, doi:10.1016/0016-7037(91)90321-U. Scholl, D.W., and von Huene, R., 2007, Crustal recycling at Sutherland, R., Barnes, P., and Uruski, C., 2006, Miocene– MANUSCRIPT RECEIVED 29 SEPTEMBER 2011 modern subduction zones applied to the past—Issues Recent deformation, surface elevation, and volcanic REVISED MANUSCRIPT RECEIVED 28 MARCH 2012 of growth and preservation of continental basement intrusion of the overriding plate during subduction MANUSCRIPT ACCEPTED 11 APRIL 2012 crust, mantle geochemistry, and supercontinent recon- initiation, offshore southern Fiordland, Puysegur mar- Printed in the USA

LITHOSPHERE | Volume 4 | Number 6 | www.gsapubs.org 483

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/4/6/469/3038655/469.pdf by guest on 01 October 2021