Implications for magmatic processes at Ontong Java Plateau from volatile and major element contents of Cretaceous basalt glasses

Abstract

Major elements, Cl, CO2, and H2O were determined in unaltered tholeiitic basalt glasses from Drilling Program Sites 803D and 807C on Ontong Java Plateau (OJP), and Sites 462A in Nauru Basin (NB) and 802A in East Mariana Basin (EMB). Glasses are moderately evolved (Mg# = 53–60) with low Na2O, indicating they formed by large extents of melting. Extent of melting is less for glasses from OJP than from EMB, requiring faster or more focused mantle upwelling or a longer period of activity for OJP.Glasses within each hole define 1–3 different compositional types. Each compositional type is made up of many glass‒bounded flow units whose chemistry is identical within analytical uncertainty, suggesting that each compositional type formed during a single eruptive episode with multiple magma surges. The lack of diversity within each type also suggests that the magmatic system was large and well buffered. Major element chemistry also requires that significant shallow crystallization occurred during plateau formation and limits the amount of deep crystallization that might be responsible for the thick crust of the plateau. Chlorine contents are constant within each compositional type and are 580 ppm in type A and 850 ppm in types C–G from Hole 807C, and 360 ppm in glasses from Hole 803D. Glasses from 802A and 462A have ∼240 ppm and 1975 ppm Cl, respectively. Cl/K ratios range from 0.2 to 2.33, well above mid‐ocean ridge basalt (MORB) mantle values (0.07), suggesting that most liquids assimilated hydrothermally altered material as they ascended and crystallized. High Cl/K ratios also suggest that abundant hydrothermal activity may have accompanied plateau formation. Low dissolved CO2 (i.e., CO32−) contents in one OJP compositional type (54 ppm in 807C type A) are similar to those on shallow sections of the mid‐ocean ridge and are consistent with a maximum depth of eruption of 1320 ± 250 m, in contrast to the greater eruption depths inferred from fossils in the overlying sediments. Our proposed eruption depth falls in the MORB field on a plot of depth versus Na8.0 supporting a ridge or near‐ridge origin. The shallow depth estimate allows for more postformation subsidence and diminishes the importance but does noteliminate the need for additional processes such as prolonged underplating to account for OJP's shallow depth. Higher CO2 in other glasses, including types C–G from Hole 807C, suggests that some liquids were oversaturated with CO2 when emplaced and therefore did not travel far from their point of eruption.H2O contents are similar to MORB and range from 0.13 wt% for NB, to 0.25 wt% (807C, types C–G) to 0.49 wt% (807C type A). Ratios of H2O to rare earth elements (REE) are slightly higher than most MORB and are similar to those from northern Mid‒Atlantic Ridge (MAR), which has distinctively high H2O. H2O/Ce (constant in depleted through enriched MORB) ranges from 200 to 400, compared to 220–380 for northern MAR glasses and 150–250 for MORB elsewhere and >500 for arc‒related glasses. The slightly higher H2O contents in the OJP mantle source could have contributed very little to the extensive melting that built the plateau, so a substantial temperature anomaly is still required.