Anomalous Subsidence of the Ontong Java Plateau (ODP Leg 192)

The Ontong Java Plateau (OJP) in the western Pacific is the largest volcanic oceanic plateau and is anomalous compared to other oceanic large igneous provinces like the Kerguelen plateau in that it never formed a subaerial landmass. To estimate paleo-eruption depths and thus better constrain the subsidence history of the OJP, we analyzed dissolved H₂O and CO₂ concentrations in glasses from basaltic pillow rims recovered during ODP Leg 192. The glasses are from four sites (1183, 1185, 1186, 1187) in the central and eastern parts of the plateau. In addition, we analyzed non-vesicular glass shards in volcaniclastic rocks from Site 1184 located on the southeastern part of the OJP. Using the pressure-dependent solubilities of H₂O and CO₂, our data suggest original eruption depths (at 122 Ma) varying from ∼1100 m below sea level (mbsl) on the central part of the plateau to 2200-2600 mbsl on the eastern edge. The glass shards from Site 1184 suggest a quenching depth of ~700 mbsl. These estimated depths must be viewed with caution, however, because submarine basaltic pillows (MORB) are typically supersaturated with CO₂, and in the case of Site 1184, because equilibrium degassing may not occur during ascent and quenching in shallow-water hydrovolcanic eruptions. By correcting the present-day depth of the top of the igneous basement for sediment loading, we use our eruption depths to estimate how much subsidence has occurred on the OJP since the time it formed. The results suggest subsidence of 1500 +/- 300 m over much of the OJP since 122 Ma. Given the consistency of the subsidence estimates between different sites, we think it unlikely that the CO₂ values reflect strong supersaturation as is found in MORB. The much larger volumes and likely long flow distances on the OJP compared with MORB flows probably allow dissolved CO₂ to reach equilibrium values for the appropriate seafloor depth before final quenching. Our estimates of subsidence for the OJP are significantly less than predictions from thermal models of hotspot-affected lithosphere. These results are paradoxical because partial melting models based on data for high-Mg basalts from Sites 1185 and 1187 indicate melting of peridotitic mantle with a high ( ∼1500° C) potential temperature.