Chemical discontinuities in Archean metavolcanic terrains and the development of Archean crust

Abstract

Most large Archean greenstone belts ( ⋍ 2.7 Ga ), comprise thick (12–15 km) mafic to felsic metavolcanics sequences which exhibit consistent but discontinuous geochemical patterns resulting from mantle-crust processes. In a typical Archean metavolcanic sequence, thick (5–8 km) uniform tholeiitic basalt is followed by geochemically evolved rock units (4–7 km thick) containing intermediate and felsic calc-alkaline rocks. This major geochemical discontinuity is marked by a change from LIL-element depleted basalts which show unfractionated REE abundance patterns, to overlying andesites with higher LIL-element contents, fractionated REE patterns and relatively depleted HREE. A less well marked discontinuity separates andesitic rocks from still later more felsic dacite-rhyolite extrusive assemblages and their intrusive equivalents, and is identified by a further increase in LIL element content and REE fractionation. The major geochemical discontinuity apparently separates rocks derived by partial melting of mantle (either directly or through shallow fractionation processes) from those which originated either by partial melting of mantle material modified by crustal interactions or by partial melting of crustal material. We suggest that accumulation of a great thickness of mantle derived volcanic rocks can lead to sagging and interaction of the lower parts of the volcanic piles with upper mantle material. The resulting modified mantle acts as a source for some of the geochemically evolved rocks observed in volcanic successions. Subsequent direct melting of the volcanic pile produces the felsic magmas observed in the upper parts of Archean volcanic successions. This process, termed sag-subduction, is the inferred tectonic process operating in the comparatively thin, hot Archean crustal regime. By this process, large masses of ultimately mantle-derived material were added to the crust.