The boninite revisited

Abstract

Boninite occurring in Chichi-jima, Bonin Islands, is a rare rock, no exact equivalent known anywhere, and is a vesicular, feldspar-free glassy andesite according to SiO₂ content (56-59%), consisting of phenocrysts of olivine (Fo_90-85) and bronzite (En_89-80), microphenocrysts and microlites of augite, and hydrous glass, with small quantities of Cr-rich spinel (Cr₂O₃ 64-54%). Chemically it is characterized by high contents of MgO (12-6%) Cr (900-200 ppm), and Ni (300-100 ppm) similar to primitive basalts, apparently in ill accord with its intermediate SiO₂ content. The boninite was first described in some detail by KIKUCHI (1890) and named by _PETERSEN (1891) after Bonin Islands coming from the Japanese “Bu-nin-sima” for no-man's islands. Surprisingly enough, this peculiar rock has since been touched upon only cursorily. We suggest that the boninite was formed by separation at shallow depths from hydrous mantle and rapid quenching from relatively high temperatures.
The Bonin, or Ogasawara Islands, situated some 1000 km SSE of Tokyo, form the outer arc of the Idu-Marianas arc, and are composed of dominant volcanic rocks containing pillow lavas and subordinate sedimentary rocks with fossiliferous limestones, the age of which is Oligocene to early Miocene in Chichi-jima, the largest island, and Eocene in Haha-jima, the second largest one, lying 50 km south of Chichi-jima. The boninite, with subordinate bronzite-andesite, makes up the bulk of the pillow lavas in Chichi-jima, while in Haha-jima it has never been found. Volcanic breccias are composed mainly of hypersthene-andesite and-dacite,
The magma of the boninite composition could exist in equilibrium with mantle peridotite at 10-5 kb, 1100-1050°C under water-saturated conditions, based on the experimental results of GREEN (1976) and NICHOLLS (1974). During its rise to the surface, olivine (and chromite) crystallized first from it, joined immediately later by bronzite, and augite was the last phase to crystallize. Before plagioclase is precipitated, the magma was erupted rapidly on sea floor to form the plagioclase-free glassy boninite pillow lavas. Available experimental data suggest that the boninite was quenched rapidly from a temperature of more than 900°C. The bronzite-andesite may represent a slowly cooled, relatively unfractionated part of the boninite magma. Low pressure fractionation involving removal of olivine, bronzite, augite, plagioclase, and magnetite could give rise to the hypersthene-andesite and -dacite from a parental boninite.
The boninite bears some resemblance to the clinoenstatite-bearing volcanic rocks from Cape Vogel, Papua, which have crystallized first protoenstatite but not olivine, generated at higher temperatures than the boninite. The boninite may well be closely connected in genesis with magnesian andesites found in some orogenic calc-alkalic suites, suggesting a major role of water rising from the subduction zone.