HR: 14:40h
AN: V33D-05 [Abstracts]
TI: Trace element variations along the Central Indian Ridge: deciphering Indian mantle
heterogeneities
AU: * Escrig, S
EM: escrig@eps.harvard.edu
AF: Harvard University, 20 Oxford Street, Cambridge, MA 02138
United States
AU: Bezos, A
EM: bezos@eps.harvard.edu
AF: Harvard University, 20 Oxford Street, Cambridge, MA 02138
United States
AU: Bezos, A
EM: bezos@eps.harvard.edu
AF: Laboratoire de G‚osciences Marines, Institut de Physique du Globe
Place Jussieu, Paris, 75005
France
AU: Langmuir, C H
EM: langmuir@eps.harvard.edu
AF: Harvard University, 20 Oxford Street, Cambridge, MA 02138
United States
AU: Humler, E
EM: humler@ipgq.jussieu.fr
AF: Laboratoire de G‚osciences Marines, Institut de Physique du Globe
Place Jussieu, Paris, 75005
France
AU: Su, Y J
EM:
AF: Lamont-Doherty Earth Observatory of Columbia University, 61 Rt. 9W, Palisades, NY 10964
United States
AB:
We report more than 50 ICP-MS and 70 LA-ICP-MS new analyses of basalt glasses along the Central Indian (CIR) and Carlsberg
Ridges. The new high precision data further document the chemical characteristics, geographical distribution and origin of
mantle heterogeneity beneath the Indian Ocean. Comparison with other data shows the importance of interlab differences in
interpreting ICP-MS and particularly LA-ICP-MS data. Samples from the Carlsberg Ridge and north of 10°S along the CIR
display both incompatible element enriched and depleted compositions. Between 10°S and 20°S, all compositions are
enriched. South of 20°S, bimodal compositions abruptly reappear. The most-enriched trace element compositions are found
within and south of the Marie Celeste FZ, confirming the effect of the fracture zone on the mantle flow originating from the
neighboring Reunion plume. The progressive diminishment of the incompatible enrichment towards the north of the fracture zone
suggests a plume influence over more than 1000km of ridge length. Samples from the Carlsberg Ridge are virtually
indistinguishable from the large set of EPR samples analyzed in our laboratory, and reflect two component mixing of an
enriched end member generated by low degree melt metasomatism with a depleted end member. CIR basalts also demonstrate linear
mixing arrays, but both enriched and depleted end members are distinct from the EPR. The enriched component is marked by
high Ba, Rb and Th (but not U) relative to the REE. The depleted component is less depleted in light REE and relatively more
depleted in U and high field strength elements, including Nb, Ta, Zr. The combination of these two effects creates data
arrays distinct from the EPR for trace element ratios such as Ba/Nb, Th/U, Th/Ta and Th/Nb. Th/Nb varies substantially in
this region, in contrast to the almost constant ratio on other ocean ridges. The enriched member is not accounted for by
recycled sediment or ocean crust, and does not have significant Nb depletion. Therefore recycled sediment or other
contamination by continental materials does not account for the distinct flavor of Indian heterogeneity. The linear data
arrays and strong fractionation of highly incompatible elements also rule out recycled ocean crust. While there are clearly
multiple enriched components along the CIR, the fact that the data have a similar topology to the other ocean ridges suggests
the addition and substraction of low degree melts to the mantle sources. The details must differ, however, perhaps owing to
differing in residual phases and oxidation state over Earth history. The generally high Th/U ratio and relative Rb enrichment
and U depletion in this region appear to reflect the overall isotopic characteristics of the Indian Ocean. The abrupt
changes in signature associated with fracture zones reaffirm the significance of these surface features to mantle composition
and flow. It is not entirely clear how the trace element depletion is related to the event responsible for the enriched
isotopic signature observed in the Indian Oceans. Higher density regional coverage of other Indian ridges should lead to much
higher definition of mantle domain boundaries and place further constraints on the cause of their formation.
DE: 1025 Composition of the mantle
DE: 3610 Geochemical modeling (1009, 8410)
DE: 3614 Mid-oceanic ridge processes (1032, 8416)
DE: 3619 Magma genesis and partial melting (1037)
DE: 3621 Mantle processes (1038)
SC: Volcanology, Geochemistry, Petrology [V]
MN: Fall Meeting 2005