Welcome to the new version of CaltechAUTHORS. Login is currently restricted to library staff. If you notice any issues, please email coda@library.caltech.edu
Published May 30, 2000 | Published
Journal Article Open

Primitive CaO-rich, silica-undersaturated melts in island arcs: Evidence for the involvement of clinopyroxene-rich lithologies in the petrogenesis of arc magmas


On the basis of the study of olivine-hosted melt inclusions in a calc-alkaline basalt from Batan Island (Philippines) we define a distinctive type of primitive, nepheline-normative island arc magma characterized by unusually high CaO contents (up to 19.0 wt %) that cannot be simply explained by melting of the metasomatized peridotitic mantle wedge above subducting oceanic lithosphere. CaO-rich melt inclusions with these characteristics are preserved in Fo85-90 olivine, and compositional variations among the inclusions are interpreted to reflect mixing between melts such as those found in the most CaO-rich inclusions (present in Fo90 olivine) and melts similar to primitive "normal" island arc magmas (trapped in Fo85 olivine). Compilation of primitive island arc magmas from the literature shows that whole rocks and olivine-hosted melt inclusions with CaO contents >13 wt % are found in many arc volcanoes from all over the world in addition to Batan. These inclusions occur in lavas ranging from CaO-rich ankaramites to basaltic andesites with low-CaO contents (i.e., <13 wt %). The globally occurring CaO-rich inclusions and whole rocks comprise a group that although defined on the basis of their CaO contents is compositionally distinctive when compared to island arc lavas that have lower CaO contents; for example, they have lower FeO at a given SiO2 content than most arc lavas, and they are all nepheline normative, with normative nepheline contents positively correlated with CaO contents. Variations in CaO content and normative compositions of experimental partial melts of lherzolite related to changes of pressure, temperature, and source composition suggest that there are no conditions under which partial melting of peridotite can generate melts having CaO contents and other properties comparable to those observed for the primitive, CaO-rich arc-derived melts identified here. Although melting of peridotite at high pressure in the presence of CO2 can produce CaO-rich, silica-poor liquids, we consider it unlikely that this is responsible for producing the CaO-rich, silica-undersaturated melts considered in this study because there are significant differences in nearly all other compositional characteristics between the CaO-rich arc magmas and melts known or thought to be produced by melting of carbonated peridotite. Model major element compositions of partial melts of clinopyroxene-rich lithologies (mantle pyroxenites, lower crustal pyroxenites, and eclogites) calculated using the MELTS algorithm suggest that the most CaO-rich, nepheline-normative melt inclusions and whole rocks identified here could represent intermediate to high degree (~10-40 wt %) partial melts of pyroxenites at lower crustal to upper mantle pressures. Such a hypothesis is supported by the comparison between the trace element compositions of model pyroxenite sources of the Batan CaO-rich melt inclusions and naturally occurring pyroxenites. The most likely source of the primitive CaO-rich, silica-undersaturated arc melts identified here is lower crustal and shallow upper mantle pyroxene-rich cumulates from arc environments because these cumulates have CaO concentrations at the upper end of the range observed for mantle pyroxenites. They are therefore more likely to yield partial melts with the restricted range of remarkably high CaO contents of the most CaO-rich inclusions and whole rocks identified here. Moreover, these cumulates often contain amphibole, which would lower their solidus temperatures relative to the anhydrous pyroxenite equivalents to values more consistent with those expected in deep crustal or shallow subarc environments.

Additional Information

Copyright 2000 by the American Geophysical Union Received November 16, 1999; Revised March 9, 2000; Accepted March 15, 2000; Published May 30, 2000. The authors benefited from numerous discussions with M. B. Baker, M. N. Ducea, M. M. Hirschmann, P. Kelemen, R. Maury, N. Metrich, and J. B. Saleeby. M. M. Hirschmann, R. Kessel, T. Plank, and R. L. Rudnick generously made their pyroxenite, eclogite, and arc lava compilations available to us. H. Sato, T. Plank, Y. Tatsumi, and W. M. White provided constructive reviews and comments. Electron microprobe analyses were done at Caltech with assistance from C. Ma. The ion microprobe analyses were done at the Lawrence Livermore National Laboratory with valuable assistance from D. L. Phinney and I. Hutcheon. Sc analyses were made by I. Hutcheon. This work was supported by NSF grants EAR-9706254 (EMS) and EAR-9805101 (JME) and is Caltech Division of Geological and Planetary Sciences contribution 5709.

Attached Files

Published - SCHIggg00.pdf


Files (1.6 MB)
Name Size Download all
1.6 MB Preview Download

Additional details

August 22, 2023
October 13, 2023