PORE FLUID STUDIES

Ever since the unusual composition of the pore fluids from these seamounts was first noted (Fryer et al., 1990) and attributed to a slab origin, the fluids have been a major focus for study. The muds recovered in drilling operations during Legs 125 and 195 from two active serpentinite seamounts (Conical and South Chamorro) contain pore fluids with compositions that indicate a slab derivation (Fryer et al., 1990, 1999; Sakai et al., 1990; Mottl, 1992; Mottl and Alt, 1992; Ryan et al., 1996; Benton, 1997; Benton et al., 2001). One of the objectives of drilling during both Legs 125 and 195 was to characterize the nature of fluid flux through the forearc region of the IBM convergent margin.

For decades it has been recognized that the output from subduction zone arc volcanoes is not in balance with the input flux of volatiles at subduction zones and that this imbalance can be significant (Peacock, 1990; Kerrick and Connolly, 2001; Sadofsky and Bebout, 2003). Estimates regarding the flux of volatiles vary regionally, but only ~16% of the H₂O and <20% of the CO₂ delivered to the IBM trench system can be accounted for in the arc magmatic output (smaller than the global average of 50% CO₂ flux through arcs [e.g., Marty and Tolstikhin, 1998; Shaw et al., 2003]). The rest of the H₂O and CO₂ must be released either through the forearc, incorporated into the mantle of the overriding plate, lost as exsolved magmatic gases, stored in subcrustal intrusions, or returned to the deep mantle. With respect to large-ion lithophile (LIL) elements, only 20%–30% of the to-the-trench inventory of these elements can be accounted for by the magmatic outputs of volcanic arcs (Morris and Ryan, 2003). It is important that the subduction-related processes by which slab constituents are recycled be adequately quantified because of their obvious importance for the geochemical cycles of volatile species.

Nonaccretionary convergent margins (Uyeda, 1982), where faulting of the forearc is prevalent and where the possibility of interaction with large sediment wedges is minimal, have been suggested as ideal localities in which to study the composition of slab-derived fluids (Fryer, 1996b). Upwelling water and seafloor seeps in a nonaccretionary forearc provide a unique window into processes of devolatilization of the subducting slab (Fryer et al., 1990, 1999). Fluids migrating upward through such a forearc mantle wedge pass through a matrix of already serpentinized depleted-mantle peridotite that is chemically simpler than typical crustal rocks and sediments. Therefore, a deep slab-derived component in the upwelling water is easier to recognize than it is in accretionary convergent margins where fluids must migrate through, and react with, a large sediment wedge.

Some of the processes that release water from the subducting plate include:

We can also anticipate that certain elements, such as K, Rb, and B, are removed from seawater and sequestered in secondary minerals during alteration at low temperature but are leached from rock into solution at high temperature (e.g., Seyfried and Bischoff, 1979; Seyfried and Mottl, 1982). These predictions are consistent with observations of unusual fluid composition (high pH, B, Rb, and alkalinity and low Cl, and Li) in active chimneys on Conical Seamount (Fryer et al., 1990) and in pore waters recovered during Leg 125 from Conical and Torishima Forearc Seamounts (Mottl, 1992).

Conical Seamount, located 90 km west of the trench near 19.5°N, was the first serpentinite mud volcano drilled in the Mariana forearc. In 1987 submersible dives with Alvin revealed chimneys at the summit composed of aragonite, calcite, and amorphous Mg silicate; when one chimney was disturbed it began to emit a weak flow of cold (1.5°C) water with a pH of 9.3 and elevated dissolved carbonate, methane, sulfate, and reduced sulfur relative to the surrounding seawater (Fryer et al., 1990). Drilling at the summit recovered unusual pore fluids that had less than half the chloride and bromide of seawater, a pH of 12.6, and were highly enriched in dissolved carbonate, light hydrocarbons, sulfate, bisulfide, Na/Cl, K, Rb, and B (Mottl, 1992). Near-surface gradients in chloride indicate that this water was upwelling at 1–10 cm/yr (Mottl, 1992). Pore water recovered from Torishima Forearc Seamount, an inactive serpentinite mud volcano in the Izu-Bonin forearc near 31°N, reflects reaction of cool (4°–11°C) seawater with harzburgite and contrasts greatly with that from Conical Seamount (Mottl, 1992). The distinctive composition of water upwelling at Conical Seamount implies that it originates by dehydration of the subducting Pacific plate. Based on earthquake depths, this occurs ~29 km below the seafloor (Hussong and Fryer, 1982; Seno and Maruyama, 1984) and the compositions of the ascending fluids probably originate at temperatures of 150°–250°C (Mottl et al., 2003, 2004). The unusual oxygen, carbon, and strontium isotopic signatures (Haggerty, 1991; Haggerty and Chadhuri, 1992) and the presence of organic acids (Haggerty and Fisher, 1992) in Conical Seamount pore fluids also suggest derivation of the fluids from the subducting slab and require that they have interacted with the ultramafics through which they migrated. The upwelling H₂O is in excess of that which serpentinizes the overlying mantle wedge during its ascent and represents one of the first returns of slab-derived volatiles to the oceans during the subduction process (Mottl et al., 2003, 2004).

Pore fluids recovered from the summit of South Chamorro Seamount at Site 1200 have a composition that is similar to that of pore fluids from Site 780 on the summit of Conical Seamount (Table T1) (Mottl et al., 2003). The two seamounts are 630 km apart but are about the same distance from the trench axis (85 km for South Chamorro Seamount and 90 km for Conical Seamount). Mottl et al. (2003) point out that differences between the seamounts include the fact that the waters upwelling through Conical Seamount have lower chlorinity and higher sulfate than those at South Chamorro. The composition of pore fluids sampled by gravity and piston coring from active seeps on several other serpentinite mud volcanoes varies across the Mariana forearc with distance from the trench, indicating that there is a progressive devolatilization of the subducting slab (Mottl et al., 2004). The nature of chemical cycling can be determined by species that partition strongly into the fluid phase, and this gives us an insight into the nature of fluid-mantle interaction in the suprasubduction zone region. Recent studies of some of these constituents show that the process may be quite complex.

Recent detailed studies of the compositions of drill samples comparing pore fluids, muds, and the included rock clasts indicate aspects of this complexity. Work by Peacock and Hervig (1999) suggests that a decrease in ¹¹B in slab sediments enriches ¹¹B in derivative fluids. Boron isotopic systematics in fluid samples from Conical Seamount support the observation that some degree of B isotopic mass fractionation occurs during release of structurally bound B from the slab and enriches the slab-derived fluids in ¹¹B (Benton et al., 2001). The temperature effect in the range predicted for the Mariana forearc beneath both Conical and South Chamorro Seamounts (~100°–250°) and pressures of ~0.8 GPa (Peacock, 1996; Mottl et al., 2003) is to increase ¹¹B (Wei et al., this volume). In order to produce the observed ¹¹B signatures of the Conical Seamount serpentinites, Benton et al. (2001) suggested that shallow devolatilization removes a slab-derived B component that is isotopically heavier than any specific slab component.

Wei et al. (this volume) focused studies on details of B concentrations and isotopic composition in 28 pore fluid samples from Site 1200 at South Chamorro Seamount. Isotopically, serpentine concentrates ¹⁰B because ¹⁰B-rich B(OH) is preferentially incorporated into the low-temperature serpentine structure (Benton et al., 2001). B is derived from several sources in the downgoing plate (e.g., compaction of sediment, clays, mineral dehydration, alteration of igneous rocks, and possible release of bound B from carbonates). Wei et al. (this volume) show that for very shallow depths below the seafloor B concentrations are significantly elevated. The samples also have lower than normal Br/Cl and ¹¹B. Wei et al. (this volume) note that this may be a consequence of pressure control because of the size difference between the halogens and OH^– (Vanko, 1986). Halogens chiefly substitute for OH^–, and Zhu (1993) predicted that they are a sensitive indicator of pH. Wei et al. (this volume) noted very high concentrations of iodine in the pore fluids with even higher concentrations in the muds. The iodine results indicate the potential for significant recycling of this element in the subduction process along the IBM system. Wei et al. (this volume) note that the iodine concentrations they measured were the highest ever measured in nonsedimentary rock types. Effluent from a nonserpentine mud volcano in the Black Sea also shows high I contents and Wei et al. (this volume) suggest the possibility that mud volcanism of all types (serpentinite and sedimentary) may be an important contributor of recycled I into the world's oceans. No difference in I content was noted between the Conical Seamount and South Chamorro Seamount samples. The data from Wei et al. (this volume) for ¹¹B concentrations compared with data from Site 780 at the summit of Conical Seamount (Benton et al., 2001) show that ¹¹B is a little higher, ~16% vs. 13%, at Site 1200 on South Chamorro Seamount. Wei et al. (this volume) suggest a mixing with seawater is required at shallow depths in a ratio of 1:2 to produce the observed concentrations and ¹¹B. The estimated total B recycled at the Mariana subduction zone is likely a significant proportion of the discrepancy between the large estimated output flux (2.5 x 10¹⁰ mol/yr) and the estimated input flux (0.6 x 10¹⁰ mol/yr). Wei et al. suggest an output of 0.03 x 10¹⁰ mol/yr for nine known seamounts for a total of 0.3 x 10¹⁰ mol/yr.

Benton et al. (2004) examined Li systematics in pore fluid and serpentinite mud samples from South Chamorro Seamount and Conical Seamount sites. They note that Conical Seamount muds have high Li contents compared to mantle values (3–7 ppm) and a mean ⁷Li value of +6%. Their serpentinized ultramafic clasts, however, generally have lower Li content, with a range of ⁷Li (6% to +10%). Higher ⁷Li clasts generally have higher overall Li contents. They conclude that there may be a depth control over Li exchanges between forearc mantle and slab-derived fluids. Furthermore, they point out that the ⁷Li variability does not occur in Mariana arc lavas or, in fact, in any other mature arc. They also note that there is no large difference between Li outputs in the forearc versus those in the volcanic front, despite the fact that other alkaline species (e.g., Ba, Sr, and K) do show changes across the Mariana subduction system. Their interpretation of these systematics is that Li is released in the shallow part of the subduction system and may be transported in down-dragged hydrated mantle to depths where arc magma is generated. A similar conclusion was offered for B isotopic variations in Izu arc volcanics (Straub and Layne, 2002). Savov et al. (2005) suggested the same sort of process might be involved in recycling of LIL elements in the Mariana system.

The work of Savov et al. (2005) on LIL elements concentrated on drill samples from Conical Seamount in which they noted that serpentinized peridotites from the drill sites at the summit and on the flanks of the edifice have U-shaped rare earth element (REE) patterns (as do boninites). They observe a high depletion in U, Th, and the high–field strength elements (HFSE), which vary in concentration by up to 2 orders of magnitude. Their data and subsequent pore fluid analyses (Mottl et al., 2004) show that the fluids from the subducting Pacific slab were reducing. Some fluid-mobile elements are substantially enriched in the serpentinized peridotite clasts that Savov et al. (2005) analyzed. They were able to calculate very large slab inventory depletions in B (79%), Cs (32%), Li (18%), As (17%), and Sb (12%). They conclude that if these highly enriched serpentinized peridotites dragged down to depths of arc magma generation they may represent an unexplored reservoir that could help balance the input-output deficit of these elements as observed by previous workers. Some species, thought to be mobile in fluids, such as U, Ba, Rb, and to a lesser extent, Sr and Pb, are, however, not enriched relative to the depleted mantle peridotites, and they estimate that <2% of these elements leave the subducting slab in the outer forearc (to a depth of 40 km). They suggest that enrichments of the latter elements in the volcanic arc and backarc basin indicate changes in slab fluid composition at greater depths.