Causes of along- and across-arc geochemical variations in the Southern Volcanic Zone (33°-43°S) in Chile and Argentina.

Jacques, Guillaume (2013) Causes of along- and across-arc geochemical variations in the Southern Volcanic Zone (33°-43°S) in Chile and Argentina. Open Access (PhD/ Doctoral thesis), Christian-Albrechts-Universität, Kiel, 190 pp.

[thumbnail of Thesis Jacques-2.0.pdf]
Preview
Text
Thesis Jacques-2.0.pdf - Published Version
Available under License German copyright act UrhG.

Download (12MB) | Preview

Abstract

The Southern Volcanic Zone (SVZ) in Chile is a 1,200 km long subduction zone. Along-arc variation in the magma geochemistry is a common characteristic of such long subduction zones. Many factors can control the magma composition. First, the slab input can vary in composition. The subducted sediments composition can vary, as they are a function of climate and therefore erosion and material supply in the trench, whereas the composition of the oceanic crust can vary in age (and consequently in slab surface temperature) and depends also at which midocean ridge it has formed. The local tectonic settings of the incoming plate (e.g. the presence or absence of major fracture zones or bend faults) will affect the degree of alteration and hydration of the lower oceanic crust and the lithospheric upper mantle, increasing the potential for serpentinization. The slab geometry can also vary. A deeper slab will be hotter and could possibly melt, which will increase the potential of carrying incompatible but fluid-immobile elements. On the upper plate, the tectonic will also play an important role. Increasing crustal thickness may increase the potential for crustal assimilation or fractional rystallization. In order to evaluate the effect of these factors on magmatic geochemistry, this PhD thesis presents a comprehensive geochemical data set (major and trace elements and O-Sr-Nd-Hf-Pb isotopes) from Holocene primarily olivine-bearing volcanic arc rocks, Chilean trench sediments and Andean crustal basement from 33°S to 43°S and extending to 300 km into the backarc in Argentina. All the volcanic arc samples show typical subduction zone trace element enrichment in highly incompatible and fluid-mobile elements and depletions in High Field Straight Elements and Heavy Rare Earth Elements. The backarc show the least enrichments and depletions, consistent with less of a slab-derived component, and consequently lower degrees of melting. The Transitional (T) SVZ (34.5-38°S) overlap the backarc samples in Sr and Nd isotopic composition, whereas the Northern (N) SVZ (33-34.5°S) extends to higher Sr and lower Nd and Hf isotope ratios. The Central (C) SVZ (38-43°S) stratovolcano samples are slightly shifted to higher Sr and/or Nd isotope ratios, whereas the monogenetic cinder cones between them are similar to the TSVZ samples. All samples form a tight correlation on the Pb isotopes diagrams. The volcanic arc samples plot at the radiogenic end of the array formed by the backarc samples. And largely overlap the trench sediments. This correlation indicates mixing between a South
Atlantic Mid-Ocean Ridge Basalt (MORB) source and a slab component derived from subducted trench sediments and altered oceanic crust. The difference in ratios of more- to lessincompatible fluid-immobile element ratios between the volcanic arc, backarc and trench sediments indicate that the slab component is a hydrous melt. The lack of correlations between MgO (or SiO2) and isotopes and the tight correlations of the Pb isotopes preclude significant assimilation of the old sialic crust in the TSVZ and CSVZ. Hf-Nd isotope ratios define separate linear arrays for the volcanic arc and backarc, neither of which trend toward subducting sediment, possibly reflecting a primarily asthenospheric mantle array for the volcanic arc and involvement of enriched Proterozoic lithospheric mantle in the backarc. Some backarc samples show an extra-enriched component with slightly higher Sr but lower Nd and Hf isotope ratios, and elevated delta 7/4 and delta 8/4, which are interpreted to be a different enriched mantle component, possibly subcontinental lithospheric mantle, than found in the other backarc samples. Although the CSVZ monogenetic cones are similar to the TSVZ stratovolcano samples, the CSVZ statovolcanoes have higher fluid-mobile to fluid-immobile element ratios and lower more- to less-incompatible fluid-immobile element ratios, consistent with an overall higher fluid flux and higher degrees of melting for the CSVZ. The higher Hf-Nd isotope ratios of the stratovolcanoes suggest a greater contribution from a more depleted source. Combined with geomorphologic data and geophysical data, this indicates derivation of a slab component carrying a combined signature of trench sediments and seawater altered oceanic rust and possibly serpentinized upper mantle, due to the larger hydration of the incoming plate caused by the more prominent and numerous fracture zones and bend faults in this segment. δ18O(melt) from groundmass or converted from olivine (by adding 0.6‰) yields values that extend below and above the MORB mantle range (see Bindeman, 2008 and references therein). The TSVZ, CSVZ monogenetic cone and backarc samples largely overlap the MORB mantle range and extend slightly above, but remain within the Island arc range. In contrast, the CSVZ stratovolcano samples fall below the MORB mantle range, indicating the possible influence from a depleted source such as serpentinites. The NSVZ samples, surprisingly, fall within the MORB range and show no correlation with parameters of differentiation (e.g. SiO2 or MgO) or isotopes precluding significant upper crustal assimilation. A quantitative mixing model between a mixed-source (slab-derived melt and a heterogeneous mantle beneath the volcanic arc) is consistent with local geodynamic parameters, assuming water-saturated conditions within the slab. This model predicts melting of the top several km of the slab (including sediments and oceanic crust) in both TSVZ and CSVZ. The SED:AOC ratio differs in the TSVZ (60:30) and in the CSVZ (30:70), whereas the amount of slab-derived melt added to the mantle wedge is higher in the CSVZ than in the TSVZ, consistent with the observed geochemical variations. The NSVZ samples are shifted to higher Sr isotope ratios and delta 8/4, and lower 206Pb/204Pb and Nd isotope ratios, but have similar delta 7/4 than the other volcanic arc samples. This is inconsistent with the pre-existing models of crustal assimilation or subduction erosion of the old sialic upper crust. Looking at a larger scale, there is evidence that both Southern and Northern American plates shared a common history since the Proterozoic (Ramos, 2010). The lower crust in Arizona for example has Sr, Nd and Pb isotope ratios appropriate to explain the NSVZ geochemical isotopic variations. Therefore the lower crust beneath the NSVZ may be similar to the one in Arizona. Another plausible scenario would be the flow of enriched asthenosphere derived from a plume component (Gough-type) that has been dragged into the Andes by convection of the South Atlantic mantle and then pushed southward via trench-parallel flow, consequently to the slab flattening and eastward arc migration during the Miocene. Finally, at least three mantle components have been identified in the SVZ: 1) depleted South Atlantic MORB (SAM-D) and 2) enriched South Atlantic MORB (SAM-E), which both were inferred from the normal backarc array in Sr-Nd isotopes, 3) an extra-enriched component in some CSVZ backarc, with high 87Sr/86Sr, delta 7/4 and delta 8/4, possibly reflecting enriched plums in the lithosphere. There may be an additional component that reached the NSVZ segment through flow of enriched (OIB-type) asthenospheric mantle, with higher 87Sr/86Sr and delta 8/4, lower 143Nd/144Nd and 206Pb/204Pb and similar delta7/4 than the TSVZ.

Document Type: Thesis (PhD/ Doctoral thesis)
Thesis Advisor: Hoernle, Kaj and Gill, James
Keywords: Subduction Zones, Chile, SVZ, Sr-Nd-Pb-Hf-O isotopes
Research affiliation: OceanRep > GEOMAR > FB4 Dynamics of the Ocean Floor > FB4-MUHS Magmatic and Hydrothermal Systems
Date Deposited: 10 Dec 2013 09:21
Last Modified: 26 Aug 2024 12:28
URI: https://oceanrep.geomar.de/id/eprint/22561

Actions (login required)

View Item View Item