= Solution
\b[Today the usual first melt is <basaltic> melt from the <mantle wedge>; in a hotter <Archean> <subduction zone>, melting of hydrated <basaltic crust> itself could generate sodic <felsic> melt of <tonalite-trondhjemite-granodiorite> composition.] Distinguish a primary melt from the evolved melt ultimately added to the upper <continental crust>.
On the modern <Phanerozoic> <Earth>:
* <Hydrothermal alteration> introduces <water> into the incoming <oceanic crust> and locally into <serpentinized mantle>. <Sediments> also carry <water> in <pore space> and hydrous <minerals>.
* As the <slab> descends, heating and <metamorphism> break down hydrous <minerals>. Released fluids enter the hotter <peridotite> of the <mantle wedge> and lower its <solidus>, causing <flux melting>. Most ordinary modern <slabs> release fluids before extensive melting of their <basaltic crust>.
* The primary melt is predominantly hydrous <basaltic> <magma>. During ascent, storage and <fractional crystallization>, <mafic minerals> are removed and the remaining liquid evolves toward <andesite>, <dacite> and <rhyolite> compositions. <Partial melting> of recently emplaced <mafic> lower <continental crust>, <magma mixing> and some <crustal contamination> can contribute to <felsic> <magma>. Intrusive equivalents include <diorite>, <tonalite> and <granodiorite>.
This adds juvenile material derived from the <mantle>; simply remelting old <continental crust> is reworking, not net juvenile addition. The amount and depth of <slab> <water> release depend on its thermal history, as quantified in <Subduction factory: 4. Depth-dependent flux of H2O from subducting slabs worldwide by Peter E. van Keken and others>.
For the hotter <Archean> endmember posited in the question, a hotter <mantle> and a warm descending plate make it easier for hydrated <basaltic crust> to reach its <solidus>. Its <partial melting> can directly make <silica>-rich, <sodium>-rich <tonalite-trondhjemite-granodiorite> liquids, leaving a more <mafic> residue. Residual <garnet> and <amphibole>, and the presence or absence of residual <plagioclase>, help determine the <rare-earth element> and strontium signatures. These are not ultramafic liquids merely because the surrounding <mantle> was hotter.
\b[The contrast is therefore commonly fluid-fluxed <mantle> melting followed by differentiation today, versus a greater potential contribution from direct melting of hydrated <mafic crust> in the hotter case.] <Tonalite-trondhjemite-granodiorite> is not unique evidence for <subduction>: melting of thickened <mafic crust> can produce similar liquids. Nor are all modern arcs devoid of <slab melting>. These qualifications are central to <The composite Archaean grey gneisses: Petrological significance, and evidence for a non-unique tectonic setting for Archaean crustal growth by Jean-François Moyen>.
Back to article page
