WHERE IS THE SUBDUCTION CREW GOING -AND WHY DOES IT MATTER?
Dr. R.J. Stern

The Subduction Crew will be working in the southernmost part of the Marianas. This is the southernmost part of a convergent plate margin which includes the Izu and Bonin islands as well as the Marianas. The Izu-Bonin-Mariana (IBM) arc system extends over 3500 km south from near Tokyo, Japan, to beyond Guam, U.S.A. (Fig. A-3). The Marianas are part of the United States of America. Guam is the largest of the Mariana Islands (http://www.gov.gu/index.html ) and it is where our cruise will begin and end. Guam is approximately 6,000 miles west of San Francisco; 3,700 miles west-southwest of Honolulu; 1,500 miles southeast of Tokyo; 2,100 miles southeast of Hong Kong, China; 1,500 miles east of Manila, Philippines; and 3,100 miles northwest of Sydney, Australia. The other islands of the Marianas are smaller than Guam and are part of the Commonwealth of the Northern Mariana Islands (http://lumahai.soest.hawaii.edu/Enso/map/usapi/cnmi.html). The IBM arc system is famous because this is where the greatest depth in the world is found, where the Challenger Deep in the Mariana Trench SW of Guam is about 11km deep (Fig. III-3).

Much of the cruise will be spent with the Mariana islands visible on the horizon. You can see a picture of the Mariana islands from space (Fig. III-A); more satellite images of Earth can be found on NASA’s Visible Earth website (http://visibleearth.nasa.gov/ ). Most of our cruise will be around the islands of Guam and Saipan. Figure III-B shows a simplified map of the Mariana arc system with the important elements that will be discussed in the following brief overview. Compare the map view of Fig. III-B with the simplified cross-section of Fig. II-6 to get a better 3-dimensional perspective of the Mariana convergent margin. Especially compare the locations of the forearc, arc volcano (Mariana arc), and back-arc basin (Mariana Trough) in the two figures.

Before we describe the Mariana arc system, which is the roof of the Subduction Factory, let’s discuss what the factory is utilizing. We know quite a bit about what is being fed into the Subduction Factory and how fast. In order to better understand what is being subducted, the Ocean Drilling Program drilled the seafloor east of the Mariana Trench during Leg 185 in Spring, 1999. You can learn more about what was found in this drill hole at http://www-odp.tamu.edu:80/publications/prelim/185_prel/185toc.html . This hole penetrated almost 500 m of sediments, not really very much considering that the oceanic crust here is almost 170 million years old (Fig. III-C). The drill hole could not go any deeper than about another 500 m into the oceanic crust, only a small fraction of the total thickness of oceanic crust, which is typically about 6000m.

Sediments, oceanic crust, and underlying mantle lithosphere is being fed into the Mariana Trench at a rate of about 4 cm/year, a little under two inches every year. That is somewhat faster than your fingernails grow but not as fast as your hair grows. We can trace the subduction zone down to about 700 km using the location of earthquakes (Fig. III-D). A rate of 4cm/y corresponds to 40km/million years, so the material now at 700 km depth was last at the bottom of the ocean about 20 million years ago.

Recall what was said about forearcs in Section I.C "Forearc Processes". The forearc is that part of the arc system between the trench and the volcanic arc. The forearc includes the chain of non-volcanic islands where most of the people in the Marianas live, called the ‘frontal arc’. The Mariana frontal arc consists of Eocene (~45 million year old) igneous basement surmounted by younger uplifted reefs. Because it is far away from most sources of sediment such as glaciers or rivers, very little sediment has been deposited on the Pacific Plate in spite of the fact that this is the oldest seafloor anywhere on Earth. As a result of the small sediment influx, there is no accretionary prism associated with the IBM forearc or trench. The Mariana forearc is a ‘non-accretionary’ forearc as shown in Fig. I-3. The Mariana forearc is quite fractured because it has been ‘bowed-out’ by seafloor spreading and widening of the Mariana Trough back-arc basin. Fluids released from subducted sediments and oceanic crust percolate upwards beneath the forearc and alter the rocks of the mantle, known as peridotite. Peridotite reacts with water at low temperatures to form serpentine, so named because its scaly green appearance reminded early geologists of snakeskins. The addition of water to form serpentine results in an expansion in volume and decrease in density, so that places in the mantle that are serpentinized will tend to rise towards the seafloor if given an opportunity. When these masses of serpentine break through the crust, they ‘erupt’ and flows of serpentine mud volcanoes spill down the sides of these mud volcanoes. Fluids vent from the summits of these serpentine mud volcanoes, and biological communities feast on the thriving bacteria. That is what has happened in the Mariana forearc (Fig. III-1). In fact, the Mariana forearc is the only place in the world where serpentine mud volcanoes are active today. While the subduction crew research cruise will mostly concentrate on the arc system that lies to the west of Guam, we will spend about 5 days studying the forearc. We will pick up Dr. Patricia Fryer who is leading an effort by the drillship JOIDES Resolution ( http://www-odp.tamu.edu:80/resolutn.html ) to drill S. Chamorro seamount, which is an excellent example of a serpentine mud volcano (Fig. III-2). Drilling at the South Chamorro Seamount will (1) examine the processes of mass transport and geochemical cycling in the subduction zones and forearcs of nonaccretionary convergent margins; (2) ascertain the spatial variability of slab-related fluids within the forearc environment as a means of tracing dehydration, decarbonation, and water/rock reactions in subduction and supra- subduction zone environments; (3) study the metamorphic and tectonic history of nonaccretionary forearc regions; (4) investigate the physical properties of the subduction zone as controls over dehydration reactions and seismicity; and (5) investigate biological activity associated with subduction zone material from great depth. After Dr. Fryer is aboard the Melville, we will survey the forearc region between Guam and the trench.

The magmatic axis of the arc is well defined from Japan to Guam. The volcanic arc is often submarine, reflecting the fact that the platform on which the arc volcanoes are built fluctuates in depth between about 1 and 4 km water depth. The magmatic arc in the Marianas is mostly defined by volcanic islands in the Central Island Province, and again becomes submarine south of Anatahan. The part of the magmatic arc that we will be studying is mostly underwater, although we expect to work north near the volcanic islands of Anatahan, Sarigan, and Guguan. You can learn more about the volcanic islands of the Marianas at http://volcano.und.nodak.edu/vwdocs/volc_images/southeast_asia/mariana/basic_geology.html

The back-arc region is characterized by an actively spreading back arc basin known as the Mariana Trough. This region is characterized by slow seafloor spreading, at a full rate of about 4 cm/year. This is a region characterized by young pillow lavas and several vigorous hydrothermal vent fields.

The area where our work will be focussed is shown in Fig. III-3. The first part of our cruise will be devoted to surveying the seafloor west of Guam and Saipan, including the underwater arc volcanoes and the back-arc basin spreading axis. We will be using an instrument called HAWAII MR1. The HAWAII MR1 (H-MR1 for short) is a shallow towed sidescan sonar system that collects digital bathymetry and acoustic imagery data in all ocean depths. The collection of sidescan sonar data is one of the most efficient methods for imaging large areas of the ocean floor. H-MR1 is operated by the University of Hawai Mapping Research Group and will be towed behind the Melville. The H-MR1 ‘towfish’ transmits an acoustic beam perpendicular to the research vessel's path out to each side of the craft. Some of the sound energy is reflected or backscattered in the direction of the sonar system off the surface of the ocean floor. The sidescan sonar system receives the backscattered energy, amplifies it, and sends it to a data acquisition system located on the ship. With acoustic imagery swaths up to 25 km wide and a survey speed of 9 knots, H-MR1 can image up to 415 square kilometers per hour. At this rate, it would take about 5 hours to survey an area the size of Dallas County. This high survey rate, along with H-MR1's high resolution, makes it an ideal wide-area seafloor survey tool.

HAWAII MR-1 simultaneously acquires digital bathymetry — that is, it maps the water depth to produce a sort of underwater topographic map - and sidescan sonar imagery. Sidescan sonar imagery reveals the distribution of areas that are rough and smooth to the H-MR1 sonar. Places that are ‘rough’ around the summits of volcanoes are most likely young lava flows, where as ‘smooth’ areas are usually places covered with sediment. You can learn more about H-MR-1 at http://www.soest.hawaii.edu/HMRG/MR1/mr1_online.htm . Some images acquired by H-MR1on other research cruises can be viewed at http://www.soest.hawaii.edu/HMRG/MR1/mr1images/mr1images.html . Side-scan sonar works a lot like side-scan radar which is carried aboard airplanes, space shuttles, or satellites. You can learn how orbital imaging radar works at http://southport.jpl.nasa.gov/html/introduction

For purposes of sampling young lava flows during the second part of the cruise, we want to know where the sonar-rough areas are. We also want to know where the conical features are, and where we find conical features that are sonar-rough, we will probably have found a young volcano. Although we are most interested in using the snar-generated bathymetry and imagery to map the distribution of young volcanoes, we expect to see many new things in the imagery, including underwater faults and sedimentary deposits of various kinds. This part of the ocean floor has never been imaged with this technology before, so plenty of surprises are bound to appear. We will image as much of this region as possible before we leave to pick up Dr.Fryer from the Resolution on March 17. After completing surveying of the forearc, we will return to Guam on March 22 to offload the MR-1 and Dr. Fryer. We will pick up some more scientists and students, mostly from Oregon State University, and return to sea to begin the sampling portion of the cruise. We will sample the seafloor from March 23 until the cruise ends April 12. We will use the images that are produced from the HAWAII MR-1 survey of the area to choose our sampling sites. We will look for features that look like recently active volcanoes and young lava flows. We will sample the features that we identify on the H-MR1 images by dredging. Dredging is a ‘low-technology’ method of sampling that entails lowering a heavy, chain-bag to the ocean floor, dragging this over the seafloor so that it collects some rocks, and then bringing the dredge and its load of rocks back to the surface for examination. We plan to carry out somewhere between 60 and 80 samplings of the seafloor in the study area using this technique.

The samples that we collect will be examined aboard ship and a few representative samples will be sawed using a rock saw that will be onboard. This will yield rock slabs and these slabs will be carried back to geochemistry laboratories at Oregon State University and The University of Texas at Dallas. There, the samples will be pulverized and the powders analyzed for chemical and isotopic compositions. We are especially interested to use this data to understand the nature of the mantle source that was melted to generate the back-arc basin and volcanic arc magmas. One of the most important things that we hope to realize from this study is why do arc magmas have chemical features indicating that they are derived from depleted mantle? One possibility is that arc magmas are generated from mantle which has already been melted beneath the back-arc basin axis. Figure III-4 presents one way in which this might happen. Subduction of the Pacific plate induces convection in the overlying mantle asthenosphere, dragging mantle down beneath the volcanic arc. Mantle rises beneath the back-arc basin spreading ridge where it melts and erupts at the surface as lava. Some of the depleted mantle is carried by the induced convection into the region beneath the volcanic arc where the addition of large amounts of water from the subducted Pacific plate causes this mantle to melt again.

By mapping volcanoes and lava fields on the seafloor between the backarc basin spreding ridge and the volcanic arc, sampling these lavas, and analyzing them for chemical and isotopic compositions, we plan to test this hypothesis. This will tell us more about what is being processed in the Subduction Factory in addition to the subducted sediment, and so lead to a better undestanding of how the Subduction Factory operates.