Backbone of the Americas—Patagonia to Alaska, (3–7 April 2006)

Paper No. 7
Presentation Time: 10:50 AM

SUBDUCTION EROSION AND RECYCLED CRUST AT CONVERGENT MARGINS: THE ADAK ADAKITE EXAMPLE


KAY, Robert W., Earth and Atmospheric Sciences, Cornell University, Snee Hall, Ithaca, NY 14853-1504, rwk6@cornell.edu

At convergent margins, production of new andesitic continental crust involves magmatic addition and differentiation, but also three processes that transfer crustal mass to the mantle: 1) subduction, which transfers oceanic crust together with continent-derived components in sediment and alteration products, 2) delamination, which transfers dense basal crust in contractionally thickened regions of Andean-type arcs. 3) subduction erosion, which transfers upper plate crust to the lower plate and thence to the mantle in the forearc of non-accreting margins. Globally, the mass-transfer rate of each process is estimated to be a few Armstrongs (units of “A”: defined as 1 km3 /yr), which is the same order as the global arc magma (dominantly basalt) production rate. Note that process 2) also occurs following non-subduction-related contractional crustal thickening, as in continental collision events. The case for 3) as applied to the type locality of “adakite” (magnesian andesite) in the Aleutian arc illustrates a specific instance of the possible mis-identification of arc magma components where oceanic crust is the source, yet compositions of oceanic crust in 1) and 3) are similar. Although the 12 Ma adakites from Adak Is, are neatly modeled as melts of subducted Pacific oceanic plate (Kay, 1978), this source is not unique, and slab melting fails to explain the restriction of the Adak adakites to the time of northward arc relocation. An alternative origin to slab melting is suggested by evidence for episodic Aleutian subduction erosion, coupled with a Miocene northward relocation of the volcanic arc. Mafic crust, formerly oceanic crust, now trapped in the forearc was plucked from the hanging wall (subduction erosion) and transported together with hanging wall peridotite into the mantle wedge where it melted at high pressure in the mantle under the arc. More generally, the role of subduction erosion has been unwisely ignored in geochemical models. Furnishing a wide compositional range of components (even LIPS), all depending on the composition of the forearc crust, subduction-eroded crust may not be easily distinguished from that of subducting oceanic plate.