Why are S-type plutonic rocks more chemically variable than S-type volcanic rocks and S-types more variable than I-types?

Abstract

S-type granitic magmas form essentially through partial melting of crustal metasedimentary rocks, so this eliminates many possibilities for the origins of chemical variations in their parental magmas. Global, whole-rock, major-oxide data show that S-type volcanic series have far less chemical variability (analytical scatter) than S-type plutonic rocks. On the other hand, published data show far smaller contrasts between metaluminous ‘I-type’ silicic plutonic and volcanic rocks. There are several possible explanations for these features, but sampling bias and analytical errors can be discounted. Likewise, the differences cannot be explained by plutonic S-types being crystal cumulates or products of wall-rock assimilation, nor can they be due to volcanic suites being products of magma mixing or having compositions controlled by a different differentiation mechanism to that which might control S-type plutonic compositions. Variations in major-oxide concentrations suggest that the most probable explanation is that S-type plutonic magmas are derived from somewhat more aluminous metasedimentary source rocks, with higher ratios of metapelites to metagreywackes. If so, there are complementary implications that plutonic S-type magmas may be formed at lower temperatures, that they may be more hydrous than the S-type volcanic magmas, and that the differences in emplacement mechanism reflect this. The weaker contrasts between ‘I-type’ volcanic and plutonic rocks may be due to their crustal magma sources being more homogeneous and the melting reactions stoichiometrically similar, regardless of source-rock type.