The structural differences between the Izu-Bonin and Mariana forearcs have led to interesting hypotheses regarding the formation of the serpentinite seamounts in these regions. In seismic refraction profiles using ocean bottom seismometers across each system, investigators have identified bands of low-velocity material overlying the subducted slab and have interpreted them as zones of serpentinized mantle (Fryer et al., 1985; Taira et al., 1998; Takahashi et al., 1998; Kamimura et al., 2002). Detection of a potential serpentinized region overlying the subducting slab led Fryer et al. (1985) to suggest that fluids from the slab may have interacted with the overlying mantle. Alvin dives conducted in 1987 showed that active serpentinite mud volcanism was occurring at Conical Seamount (Fryer et al., 1990); Conical Seamount was subsequently drilled during Leg 125. As a result of drilling during Leg 125, Fryer (1992) suggested that the seamounts formed as a consequence of the mobilization of fault gouge on deep forearc faults during seismic events that were associated with the release of fluids.
Recent MCS surveys of the IBM system in the Izu arc region at ~31°N reveal the décollement interface and the Mohorovicic discontinuity (Moho) of the subducting oceanic plate. According to Kamimura et al. (2002), the mantle wedge above the slab has a velocity lower than that of average oceanic mantle, suggesting that it has been subject to interaction with slab-derived fluids. These authors describe a layer between the two plates, which they term the "plate boundary layer" (PBL), that has lower velocity than the forearc just beneath the serpentine seamounts and that thins and increases in velocity westward of the seamount, where it connects to the mantle wedge. The distribution of the low-velocity (serpentinite) zone suggests to these investigators that the seamounts are diapiric with a root that follows the low-velocity zone to depth beneath the forearc. A seismic line run along strike of the system showed a homogeneous velocity character (Kamimura et al., 2002). An interesting additional suggestion by Kamimura et al., looking at the MCS data, and Sato et al. (2004), looking at ocean bottom seismometer data, is that the PBL may contain significant amounts of chrysotile, the fibrous phase of serpentine, which may act as a lubricant and facilitate subduction of the plate, resulting in low seismic activity at the décollement.
Velocity and density of serpentinized peridotite clasts from Site 1200 were determined postcruise by Courtier et al. (this volume) and agree with values measured aboard ship (Shipboard Scientific Party, 2002). As noted by Courtier et al. (this volume), however, the 14 samples measured show a narrow range of velocity and density in comparison with the velocity and density range for the forearc mantle wedge and indicate that the degree of serpentinization of the clasts is greater than that of the mantle wedge. Specifically, in comparing the velocity of clasts from Torishima Forearc Seamount (Sites 783 and 784) measured by Ballotti et al. (1992) with velocities suggested by Kamimura et al. (2002) for a line run across Torishima Forearc Seamount, the percent serpentinization estimated from MCS data is 54%, which is lower than the value calculated by Courtier et al. (this volume) (66%) from laboratory measurements on serpentinized clasts. They conclude that the degree of serpentinization beneath the seamounts (Conical, Torishima Forearc, and South Chamorro) is greater than that of the surrounding mantle wedge, a conclusion consistent with the idea that the mud volcanoes tap a more highly serpentinized local source beneath the seamounts.