sign in sign up
<<
Home , Volcanic arc

or collective redistirbution other or collective or means this reposting, of machine, is by photocopy article only anypermitted of with portion the approval The O of This article has been published in published been This has article Mountains in the

By Hubert Staudigel, Anthony A.P. Koppers, Terry A. Plank, and Barry B. Hanan , Volume 1, a quarterly 23, Number The O journal of in the Factory ceanography S ociety. ©

Figure 1. Simplified model of a subduction zone. The oceanic plate, including (altered) 2010 by The O , sediments, and seamounts, is subducted beneath a volcanic arc where it is heated from above by the overlying and subarc . Volatiles and melts expelled from the downgoing slab interact with and melt the subarc mantle to produce a volcanic ceanography S arc. Although not depicted, some seamounts may also be decapitated in the subduction (see Watts et al., 2010) and be added wholesale to the . Also, seamounts typi- cally have deep flexural moats filled up with more volcanic materials, which will significantly ceanography S add to the total volume of oceanic intraplate volcanic materials being subducted. ociety. A

Back-arc Volcanic Arc Permission is reserved. ll rights granted to in teaching copy this and research. for use article R ociety. S Spreading Accetionary Wedge Center Fore-arc end all correspondence to: [email protected] Th e O or

Magma Feeding Arc Volcanoes ceanography S ociety, P Sub-arc Sediment Mantle 1931, R O Box epublication, systemmatic reproduction,

Oceanic Crust M D 20849-1931, USA ockville, Lithospheric Mantle Asthenospheric Mantle .

176176 OceanographyOceanography Vol Vol.23,.23, No No.1.1 Abstract. The “Subduction Factory” is a metaphor for the geochemical since the onset of on processing of subducted oceanic crust and sediment into components that are either Earth. In its “standard model,” the incorporated into the volcanic arc or recycled into Earth’s mantle. Seamounts may be Subduction Factory consumes three a significant source of material to the Subduction Factory, in particular, by providing input components to build an arc: trace elements such as K, Ba, La, Ce, U, Th, Pb, Rb, and Cs. Seamount subduction (1) subducted sediments, (2) altered might also play a role in the global distribution of chemical mantle heterogeneities. oceanic crust, and (3) the subarc Neither one of these effects of seamount subduction is well understood. The mantle (Figure 1). Although seamount Izu-Bonin-Marianas (IBM) volcanic arc is a region where the potential impact of subduction has been considered previ- seamount subduction may be explored most effectively. There, sections of the IBM arc ously, there is no systematic study and many of the incoming seamounts display unusually high 206Pb/204Pb ratios, which that explores this effect quantitatively, offer a particularly promising geochemical tracer that may help quantify seamount and, hence, it is typically not part of input into the Subduction Factory. Although this process remains to be explored Subduction Factory geochemical mass in a quantitative manner, it is apparent that the demise of seamounts in subduction balances. Although the seamounts’ total zones offers an exciting research target with important consequences for globally volume is much smaller than that of relevant geochemical processes. the oceanic crust, seamounts neverthe- less could play an important role in Introduction in models explaining the chemical the Subduction Factory: Every year, almost 3 km2 of oceanic processes occurring in subduction zones • Rocks from oceanic intraplate crust are produced at mid- ridges (e.g., Kelley et al., 2005). volcanics (OIVs) introduce into the (Williams and von Herzen, 1974; Cogné Subduction Factory larger inventories and Humler, 2004) and subducted The Subduction Factory of the trace elements needed to make at convergent plate boundaries. This Geochemical interactions between the (e.g., K, Ba, La, Ce, subduction process involves the whole subducting plate and the arc system have U, Th, Pb, Rb, Cs) when compared to plate: the oceanic crust, including its been popularized through the concept normal oceanic crust, which is made volcanic and plutonic foundation, over- of a “Subduction Factory” that processes up of mid-ocean ridge (MORB) lying sediments, and any seamounts the subducting oceanic crust and sedi- that is depleted in these elements. The that are riding on top (Figure 1). As the ment into two fractions—a portion that concentration of these trace elements incoming plate is subducted, it interacts interacts with the convecting subarc in OIVs can be larger by an order of physically and chemically with the mantle and that rises as fluids or melts magnitude or more, and therefore an overriding plate. Chemical interactions to make the volcanic arc, and a residual equivalent mass of seamount mate- between these two plates are mostly portion that sinks deeply into the mantle rials contributes substantially more to based on dehydration reactions and, with the downgoing slab. The mass this budget than does MORB. possibly, melting of the descending crust balance of these processes is key to • Large seamounts (and volcanic as it is squeezed and heated top-down global chemical dynamics because the ) are surrounded by aprons through contact with the much hotter, volcanic arc ultimately contributes to the of volcaniclastic sediments that are convecting subarc mantle. Escaping growth of continents. Everything that shed mainly during their explosive slab fluids or melts permeate into the is not added to the volcanic arc system emerging growth periods or from mantle wedge and may cause additional determines much of the chemical evolu- their collapses. These sediments carry melting, yielding that feed the tion of the mantle and the development OIV chemical signatures that are volcanoes of an arc system (Figure 1). of its geochemical heterogeneities. modified by intense interactions with The role of subducted seamounts, This cycling creates chemical fluxes . Such aprons are particularly and the volcaniclastic sediments they that have been a major control on thick in a moat around the seamount deposit around them, often is neglected planetary-scale chemical fractionation formed by gravitational loading,

Oceanography March 2010 177 and they may extend out for several the subducting seamount, possibly where the related bulk geochemical

hundred kilometers (Staudigel and past the ~ 120-km depth of arc fluxes (K, Rb, Cs, CO2) are much Clague, 2010). At Ocean Drilling generation. This would diminish the larger than in the nearby abyssal Program (ODP) Site 801 in the role of subducted oceanic crust in the seafloor at Sites 417D and 418A western Pacific, without any nearby Subduction Factory and increase the (Staudigel et al., 1996). seamounts, volcaniclastic material volume of altered crust subducted into makes up 40% of the mass of the sedi- the deeper mantle. Altered oceanic All of these geophysical and geochem- ments, and provides more than 50% of crust is enriched in alkali metals (K, ical consequences, in combination with the elemental budget for important Rb, Cs), carbonate, water, Fe+3, and the large numbers of seamounts, suggest subduction tracers, such as Sr, K, radiogenic Sr (Staudigel et al., 1996), that seamount subduction may be a and Ce (Plank and Langmuir, 1998). and, hence, these elements are likely very significant player in the dynamic This contribution is likely to be more to be recycled back into the deep geochemical system of the solid Earth. pronounced closer to large seamounts. mantle or contribute to back-arc basin There are several instances where • The actual volume of OIVs subducted magmatic processes and volatilization seamount subduction has been invoked

is larger than revealed by the small of the mantle by CO2 and H2O-rich as part of the Subduction Factory. For tops of seamounts sticking out above fluids (i.e., metasomatism). example, it has been argued that the the seafloor sediment surface because • Seamounts, and possibly the (directly) subduction of the Louisville Ridge of the presence of OIV material in the underlying oceanic crust, are likely seamounts and their associated volca- volcaniclastic aprons and because a to be extensively altered, partly due niclastics, which have the radiogenic substantial amount of OIV material to the more extensive exposure to isotope and trace element signatures is covered by (pelagic) sediments. water circulation and partly due to associated with the Louisville mantle Furthermore, seamount edifices are their larger abundance of volcanic plume, can be traced through the partially sunk by the deflection of the glass and clastic (fragmental) rocks subduction zone (Hawkesworth et al., from seamount loading (see Staudigel and Clague, 2010). 1997; Turner and Hawkesworth, 1998; (Koppers and Watts, 2010), and Prolonged access to seawater is a Hanan et al., 2008). More recently, additional quantities of OIV magmas consequence of seamounts not being Hoernle et al. (2008) suggested also may have been intruded into the sealed by pelagic sediments (which that the subduction of the Cocos oceanic lithosphere. serve to seal normal oceanic crust) seamounts impacts mantle flow beneath • Seamount loading pushes the upper and of their role as hydrothermal the Costa Rica arc. 500 m of intensely low-temperature- siphons (Fisher and Wheat, 2010). altered oceanic crust down by several Some indication of the larger degree Seamount Subduction in kilometers. Hence, a seamount shields of alteration in seamounts is seen in the Izu-Bonin-Mariana Arc the most-altered upper oceanic the extreme alteration of a seamount/ The Izu-Bonin-Mariana (IBM; Figure 2) crust from top-down heating during abyssal hill at Drilling arc offers a particularly interesting subduction, resulting in delayed Project (DSDP) Site 417A in the scenario for studying seamount subduc- heating of the oceanic crust beneath North Atlantic on the Bermuda Rise, tion and the relevant arc mass balances. It is a relatively simple subduction zone Hubert Staudigel ([email protected]) is Research Geologist and Lecturer, Institute that does not have an accretionary wedge of Geophysics and Planetary Physics, Scripps Institution of Oceanography, University of that accumulates sediments scraped off California, San Diego, La Jolla, CA, USA. Anthony A.P. Koppers is Associate Professor, the downgoing plate. Hence, at the IBM College of Oceanic and Atmospheric Sciences, Oregon State University, Corvallis, OR, USA. arc, all of the oceanic crust, sediments, Terry A. Plank is Professor, Lamont-Doherty Earth Observatory of Columbia University, volcanic basement, and seamounts are Palisades, NY, USA. Barry B. Hanan is Staff Scientist, Department of Geological Sciences, subducted and thus enter the Subduction San Diego State University, San Diego, California, USA. Factory. South of 25°N, approximately

178 Oceanography Vol.23, No.1 Izu-Bonin-Mariana Arc West Pacific Seamount Province 35

Seamount Trails Izu Studied SeamountTrails Geochemical Signatures Shatsky Rise 30 1149 206Pb/204Pb > 19.5

206Pb/204Pb 19.5 - 18.8

n Boni 206Pb/204Pb < 18.8

ODP/DSDP Basement Site 25 NMSP Ogasawara FZ Latitude (°N) 800 20 Mid-Pacific

801 Mountains Marianas 15

24 18 19 Nd/Sm 206Pb/ 204Pb Magnetic Lineations 118 Ma M0 Figure 2. Nd/Sm and 206Pb/204Pb along the Izu-Bonin-Mariana arc system (Stern et al., 2003) 135 Ma M12 206 204 149 Ma M21 versus latitude. Pb/ Pb of seamounts on the 5 Ontong Java Pacific Plate are color coded (Koppers et al., 2003; 160 Ma M29 Plateau Konter et al., 2008). Arrows (yellow) indicate plate motion directions. Scientific ocean drilling sites (stars) have provided data for geochemical 140 145 150 155 160 165 170 175 180 mass balance calculations (Kelley et al., 2005). Longitude (°E) NMSP = Northern Marianas Seamount Province.

at the location of the Northern geochemical data summarized in those responsible in the formation of Marianas Seamount Province (NMSP; Figure 2. Here, two geochemical the South Pacific Isotope and Thermal Figure 2), the incoming plate is littered parameters may be used to trace the Anomaly (SOPITA). SOPITA is an with seamounts, but there are only a impact of seamount subduction on arc east-west elongated region in the mantle few north of this latitude. Hence, the volcanoes: (1) the rare earth element characterized by an unusual abundance Mariana subduction zone (including that abundance ratio of Nd/Sm, and of subducted components (Staudigel beneath the NMSP) receives a substan- (2) the lead isotope ratio 206Pb/204Pb. et al., 1991). These subducted compo- tial influx of seamounts, while the Izu High Nd/Sm ratios are geochemical nents include both recycled modified and Bonin arcs receive almost none. fingerprints that indicate enrichments oceanic crust and sediments. As a result, Some physical evidence for seamount in light rare earth element abundances 206Pb/204Pb ratios serve as a character- subduction has been suggested from the that are typical in the generation of istic fingerprint for the subduction of a formation of serpentinite mud volcanoes OIV magmas through small degrees of particular subset of OIVs; thus, they may in the Mariana forearc (see Spotlight 9 melting of mantle materials. MORB, also be used to quantify seamount input on page 174 of this issue [Wheat et al., on the other hand, has lower Nd/Sm into an arc system. 2010]; Stern et al., 2006; Oakley et al., ratios and is formed by higher degrees The IBM arc shows some rather 2008; Watts et al., 2010). of mantle melting. High 206Pb/204Pb distinct along-arc latitudinal varia- This physical evidence for seamount ratios (> 19.5) are rare isotopic signa- tions in Nd/Sm and 206Pb/204Pb ratios subduction is also supported by tures of OIV mantle sources except for (Figure 2). Both tracers are elevated

Oceanography March 2010 179 in the Marianas, they spike up at the high 206Pb/204Pb ratios outboard of the opportunity to explore the potential NMSP, and they decline to lower values Mariana arc may also extend into the involvement of OIVs in the Subduction in the Izu-Bonin arc. NMSP’s origin subduction zone beneath this arc system. Factory, because 206Pb/204Pb ratios are has been explained by ideas as varied Although this scenario is very likely, we a particular distinctive geochemical as “unzipping” of the arc by coincident also caution that we know little about the fingerprint. Substantiation of the back-arc rifting (Stern et al., 1984), OIV seamounts next to the trench and connection and its quantification would tapping enriched (Stern et al., 1993) hence our extrapolations are tenuous. provide profound insights into our or lithospheric mantle (Pearce et al., Furthermore, there can never be abso- understanding of chemical geodynamics (see also Koppers and Watts, 2010). Most of the seamounts in Figure 1 originate in SOPITA (Staudigel et al., 1991; Koppers …seamounts offer exciting research et al. 2003; Konter et al., 2008), and opportunities with potential implications plate motion models suggest that these for our understanding of the chemical seamounts are heading for their ultimate “graveyard” beneath the IBM subduction “e volution of Planet Earth. zone. Subduction of these seamounts may ultimately lead to the formation of another geochemical anomaly that includes high 206Pb/204Pb ratios in 2005), or an increased subduction flux lute certainty as to what geochemical Earth’s mantle, just like those found of OIVs (Peate and Pearce, 1998; Sun component” is currently located in the today in SOPITA. Could that mean that and Stern, 2001; Sumino et al., 2004; root zones of these arc volcanoes, and SOPITA itself formed by recycling of Tollstrup and Gill, 2005; Ishizuka inferences have to be extrapolated from OIVs? This possibility poses the question et al., 2007). We suggest here that the what can be sampled outboard of the of what role subducted seamount grave- latter is more likely, largely due to the subduction zone. However, successful yards play in the origin of major mantle characteristic 206Pb/204Pb ratios of the tracing of seamounts into the arc, in chemical anomalies such as SOPITA seamounts on the subducting plate combination with an understanding in the Pacific. We believe that the cycle (Figure 2). There, the highest206 Pb/204Pb of the temporal and spatial variation of birth, destruction, and rebirth of ratios (in red, Figure 2) can be found of 206Pb/204Pb ratios within the arc, OIVs might play a fundamental role in closest to the northern Marianas will help us understand the extent to chemical geodynamics. Even as they (i.e., the NMSP region) with near- which seamounts can play a role in arc meet their final demise in subduction identical values in the seamounts and magmatic processes. These geochemical zones, seamounts offer exciting research the arc. Plate motion vectors (in yellow, constraints will, in turn, help us ulti- opportunities with potential implications Figure 2) point these high-206Pb/204Pb mately better understand key geophysical for our understanding of the chemical seamounts directly toward similarly high data sets from this arc, such as seismic evolution of Planet Earth. values in the IBM arc. anisotropy observations (Pozgay et al., Recently, Konter et al. (2008) argued 2007), and inform geodynamic models Acknowledgements that high 206Pb/204Pb ratios appear to (Kneller and van Keken, 2007). We thank R.J. Stern and J.B. Gill for be a long-lived characteristic of some insightful and constructive reviews of Pacific OIV seamounts that can be Outlook this manuscript. traced all the way from their original The association of high-206Pb/204Pb arc locations in the South Pacific to volcanoes with subduction of seamounts the seamounts shown in Figure 2. Given that have high 206Pb/204Pb ratios at this longevity, the currently known the IBM arc offers an exceptional

180 Oceanography Vol.23, No.1 References Pearce, J.A., R.J. Stern, S.H. Bloomer, and P. Fryer. Sun, C-H., and R.J. Stern. 2001. Genesis of Cogné J.P., and E. Humler. 2004. Temporal varia- 2005. Geochemical mapping of the Mariana Mariana shoshonites: Contribution of the tion of oceanic spreading and crustal produc- arc-basin system: Implications for the nature subduction component. Journal of Geophysical tion rates during the last 180 My. Earth and and distribution of subduction components. Research 106:589–608. Planetary Science Letters 227:427–439. Geochemistry, Geophysics, Geosystems 6, Turner, S., and C. Hawkesworth. 1998. Using Fisher, A.T., and C.G. Wheat. 2010. Q07006, doi:10.1029/2004GC000895. geochemistry to map mantle flow beneath the Seamounts as conduits for massive fluid, Peate, D.W., and J.A. Pearce. 1998. Causes of Lau Basin. Geology 26:1,019–1,022. heat, and solute fluxes on ridge flanks. spatial compositional variations in Mariana Tollstrup, D.L., and J.B. Gill. 2005. Hafnium Oceanography 23(1):74–87. arc : Trace element evidence. The systematics of the Mariana arc: Evidence Hanan, B., T. Rooney, A. Pietruszka, L. Tisn, Arc 7:479–495. for sediment melt and residual phases. D. Hahm, P. Castillo, D. Hilton, and J. Hawkins. Plank, T., and C.H. Langmuir. 1998. The Geology 33:737–740. 2008. Hf and Pb isotope constraints on the chemical composition of subducting sediment: Watts, A.B., A.A.P. Koppers, and D.P. Robinson. source origin of Northern Lau Basin back-arc Implications for the crust and mantle. Chemical 2010. Seamount subduction and . basin . Geochimica et Cosmochimica Geology 145:325–394 Oceanography 23(1):166–173. Acta 72(12):A347. Pozgay, S.H., D.A. Wiens, J.A. Conder, H. Shiobara, Wheat, C.G., P. Fryer, K. Takai, and S. Hulme. Hawkesworth, C.J., S.P. Turner, F. McDermott, and H. Sugiok. 2007. Complex mantle flow 2010. Spotlight 9: South Chamarro Seamount. D.W. Peate, and P. van Calsteren. 1997. in the Mariana subduction system: Evidence Oceanography 23(1):174–175. U-Th isotopes in arc magmas: Implications from shear wave splitting. Geophysical Journal Williams, D.L., and R.P. von Herzen. 1974. for element transfer from the subducted crust. International 170(1):371–386. Heat loss from the Earth: new estimate. Science 276:551–555. Staudigel, H., K.-H. Park, M.S. Pringle, Geology 2:327–328. Hoernle, K, D.L. Abt, K.M. Fischer, H. Nichols, J.L. Rubenstone, W.H.F. Smith, and A. Zindler. Zindler, A., and S. Hart. 1986. Chemical geody- F. Hauff, G.A. Abers, P. van den Bogaard, 1991. The longevity of the south Pacific isotope namics. Annual Review of Earth and Planetary K. Heydolph, G. Alvarado, M. Protti, and thermal anomaly. Earth and Planetary Sciences 14:493–571. W. Strauch. 2008. Arc-parallel flow in Science Letters 102:24–44. the mantle wedge beneath Costa Rica Staudigel, H., T. Plank, W. White, and and Nicaragua. Nature 451:1.094–1,097, H.-U. Schmincke. 1996. Geochemical fluxes doi:10.1038/nature06550. during seafloor alteration of the basaltic Ishizuka, O., R.N. Taylor, M. Yuasa, J.A. Milton, upper oceanic crust: DSDP Sites 417–418 R.W. Nesbitt, K. Uto, and I. Sakamoto. 2007. (Overview). Pp. 19–38 in Subduction Top to Processes controlling along-arc isotopic Bottom. G.E. Bebout, D.W. Scholl, S.H. Kirby, variation of the southern Izu-Bonin arc. and J.P. Platt, eds, American Geophysical Union Geochemistry, Geophysics, Geosystems 8, Monograph Series 96, Washington, DC. Q06008, doi:10.1029/2006GC001475. Staudigel, H., and D.A. Clague. 2010. The geolog- Kelley, K.A., T. Plank L. Farr, J. Ludden, and ical history of deep-sea volcanoes: Biosphere, H. Staudigel. 2005. Subduction cycling of hydrosphere, and lithosphere interactions. U, Th and Pb. Earth and Planetary Science Oceanography 23(1):58–71. Letters 234:369–383. Stern, R.J., N.C. Smoot, and M. Rubin. 1984. Kneller, E.A., and P.E. van Keken. 2007. Trench- Unzipping of the arc, Japan. parallel flow and seismic anisotropy in the Tectonophysics 102:153–174. Mariana and Andean subduction systems. Stern, R.J., M.C. Jackson, P. Fryer, and E. Ito. 1993. Nature 450(7173):1,222–1,225. O, Sr, Nd and Pb isotopic composition of the Konter, J.G., B.B. Hanan, J. Blichert-Toft, Kasuga cross-chain in the Mariana arc: A new A.A.P. Koppers, T. Plank, and H. Staudigel. perspective on the K-h relationship. Earth and 2008. One hundred million years of mantle Planetary Science Letters 119:459–475. geochemical history suggest the retiring Stern, R.J., E. Kohut, S.H. Bloomer, M. Leybourne, of mantle plumes is premature. Earth and M. Fouch, and J. Vervoot. 2006. Subduction Planetary Science Letters 275:285–295. factory processes beneath the Guguan cross- Koppers, A.A.P., H. Staudigel, M.S. Pringle, and chain, Mariana Arc: No role for sediments, J.R. Wijbrans. 2003. Short-lived and discon- are serpentinites important? Contributions to tinuous intraplate in the South Mineralogy and Petrology 151(2):202–221. Pacific: Hot spots or extensional volcanism? Stern, R.J., M.J. Fouch, and S.L. Klemperer. Geochemistry, Geophysics, Geosystems 4(10), 2003. An overview of the Izu-Bonin-Mariana 1089, doi:10.1029/2003GC000533. subduction factory. Pp. 175–222 in Inside the Koppers, A.A.P., and A.B. Watts. 2010. Intraplate Subduction Factory. J. Eiler, ed., Geophysical seamounts as a window into deep Earth Monograph Series Volume 138, American processes. Oceanography 23(1):42–57. Geophysical Union, Washington, DC. Oakley, A.J., B. Taylor, and G.F. Moore. 2008. Sumino, H., K. Notsua, S. Nakaib, M. Satoa, Pacific Plate subduction beneath the central K. Nagaoa, M. Hosoec, and H. Wakitaa. 2004. Mariana and Izu-Bonin fore arcs: New Noble gas and carbon isotopes of fumarolic insights from an old margin. Geochemistry, gas from Iwojima Volcano, Izu–Ogasawara arc, Geophysics, Geosystems 9, Q06003, Japan: Implications for the origin of unusual arc doi:10.1029/2007GC001820. . Chemical Geology 209:153–173.

Oceanography March 2010 181