The present isotopic data are consistent with a previously published model that proposes: chemical variations in magmas of coexisting tholeiitic and calc-alkaline series are produced through crystal fractionation from mantle-derived magmas of basalt and magnesian andesite, respectively. Moreover, the tholeiitic series and the calc-alkaline series are isotopically identical. Thus, both magma series can be derived from a source mantle with the same isotopic composition, supporting the hypothesis of simultaneous generation of basalt and magnesian andesite magmas from a single diapir rising through the mantle wedge above the subduction zone. The differences of water content and temperature within the diapir are again thought to have been produced through dehydration and heating of an isotopically homogeneous hydrous diapir.
The isotopic data show that the high-SiO2 lavas have the same isotopic compositions as more mafic lavas. These data and liquid lines of descent of the Shirahama magmas suggest that even rhyolites can be produced by differentiation from mantle-derived magmas without crustal contamination.
Analyses from thirty-eight other arc volcanoes have been compiled to investigate the intravolcano variability of 87Sr/86Sr. Twelve of these display no intravolcano strontium isotopic variability, as is the case with the Shirahama Group, but others show greater variation of 87Sr/86Sr from individual volcanic centers, presumably reflecting crustal contamination. Most of the latter volcanoes are underlain by thick continental crust. It is noteworthy, however, that the greater variations of 87Sr/86Sr correlate with SiO2 content; andesites or dacites, not basalts, from the same volcano have the lowest 87Sr/86Sr, and these rocks are calc-alkaline in terms of FeO*/MgO and SiO2. Theoretically, assimilation of continental crust by the isotopically uniform Shirahama magmas could produce these relationships. Given that mantle-derived basalt and magnesian andesite both encounter continental crust on their ascent to the surface, the hotter basalt magma would assimilate more crustal wallrocks than the cooler andesite, resulting in the basalt being more radiogenic. Fractional crystallization, magma mixing, and/or assimilation-fractional crystallization of these magmas in crustal magma chambers could produce large compositional variations, but the derivatives of the hotter basaltic magmas (tholeiitic series in the broad sense) would display greater contamination than those derived from the cooler andesitic magmas (calc-alkaline series).
Current Contents
................
J. Petrology Editor's Page