UVM Theses and Dissertations
Format:
Print
Author:
Price, Robert P. W.
Dept./Program:
Geology
Year:
2005
Degree:
MS
Abstract:
Adakites are proposed to be generated by melting the down-going, oceanic lithosphere (slab) within subduction zones. Numerical modeling combined with experimental data shows it is not possible to melt the down-going slab unless the slab is very young (<5Ma) and thus warmer. An alternative hypothesis is the partial melting of a basaltic underplate at the base of the arc generating a felsic magma, which is buoyantly removed from the system forming igneous suites. To test the hypothesis that magma generated from the partial melting of an underplate can produce the same compositions as slab-melts, partial melts have been experimentally produced from two compositionally different, unaltered, basaltic dikes. One dike is a monomineralic hornblendite the second has a diverse assemblage of hornblende, plagioclase plus accessory phases. The dike materials were selected to represent basaltic underplate material. Melts, a product phase from these experiments, were characterized by their major and trace elements. The granitic classifications and trace element signatures of these melts have then been compared to those of Fiord land, New Zealand's granitoid suites.
Fiordland, New Zealand was selected for this project because it is well documented that the region was generated from a subduction regime. The granitoid suites of Fiordland are composed of two generations: late Jurassic and early Cretaceous, which are separated by a distinct 10Ma period of reduced magmatism. While both are characterized by enriched light REE and depleted heavy REE, only the younger suites have high Sr/Y and Dy/Yb. These characteristics indicate: 1) magma formation was dominated by partial melting rather than fractional crystallization, 2) garnet is a product phase from the partial melting of basalt and 3) only the younger suites are classified as adakitic. We contend that while the two generations differ chemically, their genesis is identical.
To simulate lower-crustal conditions, piston cylinder experiments were conducted at 1.4 GPa based upon barometric estimates from the region. To determine the solidus, the liquidus and garnet-bearing zones two basalts, the experiments were conducted at 800-1200C̊. Results show that partial melts produced from a basaltic source, can be either HREE depleted with an adakitic signature or HREE depleted without the adakite signature and are dependent upon the composition of the source. While both melt signatures have been produced by partially melting basalts with garnet residue, the "nonadakitic" melts were produced from a plagioclase-free, and thus strontium-free, source.
In relation to Fiordland, these findings show granitoids there may have been produced by partial melting of basalt leaving behind a garnet residue, as opposed to a few select unique suites whose interpretations were based upon trace element ratios, such as Sr/Y. Applied globally, these results indicate that both the implication of adakite genesis by melting of the down-going slab, may be incorrect, and the compositional definition of adakites is too narrow. Future research should consider the isotopic and REE signature of the rocks in addition to other trace element distribution when attempting to determine genetic signatures. Re/Os may be a new method to help distinguish between basaltic underplate melts and those from the down going slab.
Fiordland, New Zealand was selected for this project because it is well documented that the region was generated from a subduction regime. The granitoid suites of Fiordland are composed of two generations: late Jurassic and early Cretaceous, which are separated by a distinct 10Ma period of reduced magmatism. While both are characterized by enriched light REE and depleted heavy REE, only the younger suites have high Sr/Y and Dy/Yb. These characteristics indicate: 1) magma formation was dominated by partial melting rather than fractional crystallization, 2) garnet is a product phase from the partial melting of basalt and 3) only the younger suites are classified as adakitic. We contend that while the two generations differ chemically, their genesis is identical.
To simulate lower-crustal conditions, piston cylinder experiments were conducted at 1.4 GPa based upon barometric estimates from the region. To determine the solidus, the liquidus and garnet-bearing zones two basalts, the experiments were conducted at 800-1200C̊. Results show that partial melts produced from a basaltic source, can be either HREE depleted with an adakitic signature or HREE depleted without the adakite signature and are dependent upon the composition of the source. While both melt signatures have been produced by partially melting basalts with garnet residue, the "nonadakitic" melts were produced from a plagioclase-free, and thus strontium-free, source.
In relation to Fiordland, these findings show granitoids there may have been produced by partial melting of basalt leaving behind a garnet residue, as opposed to a few select unique suites whose interpretations were based upon trace element ratios, such as Sr/Y. Applied globally, these results indicate that both the implication of adakite genesis by melting of the down-going slab, may be incorrect, and the compositional definition of adakites is too narrow. Future research should consider the isotopic and REE signature of the rocks in addition to other trace element distribution when attempting to determine genetic signatures. Re/Os may be a new method to help distinguish between basaltic underplate melts and those from the down going slab.