 |
| Titel |
Alkaline magma- oceanic lithosphere interaction: a key to understand the nephelinite-alkali basalt transition observed in oceanic islands |
| VerfasserIn |
Sébastien Pilet, Michael B. Baker, Edward M. Stolper, Othmar Muntener |
| Konferenz |
EGU General Assembly 2010
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 12 (2010) |
| Datensatznummer |
250034903
|
|
|
|
|
|
| Zusammenfassung |
| An important question in the petrogenesis of oceanic island basalts is related to the location
of the different mantle components which interact during their formation. Most
models suggest that all components are located within the convecting mantle and
therefore neglect the potential role of the oceanic lithosphere [e.g. 1]. Here we show
that lithospheric mantle plays a fundamental role in the process responsible for the
range of parental melt (i.e. from nephelinite to tholeiite) observed in intraplate
volcanoes.
Alkaline lavas from continental volcanoes or oceanic islands characterized by thick
lithosphere (>50 km) define a compositional continuum from nephelinites to alkali olivine
basalts and often to tholeiites. The decrease in incompatible trace-element concentrations
from nephelinitic to tholeiitic magmas in single volcanoes is consistent with this continuum
reflecting an increase in the degree of partial melting of a common source [2]; however, no
experiments on mantle lithologies (peridotite, pyroxenite) have reproduced the observed
compositional continuum (nor even the observed range of silica contents: ~40 to 48 wt.
% SiO2). Alternatively, this continuum could be explained by reaction between
nephelinitic/basanitic liquid and surrounding peridotite [3, 4, 5]. To test this latter hypothesis,
“sandwich” experiments were performed in which a layer of hornblendite (producing
nephelinitic magmas [5]) was packed between layers of moderately depleted peridotite.
Experiments were done at 1.5 and 2.5 GPa, with temperature ranging from 1225 to 1425Ë
C. At the same temperature (1250-1300Ë C), the SiO2 contents of partial melts
produced in the sandwich runs are up to 4-5 wt. % higher than liquids from the
hornblendite-only experiments. This difference reflects the dissolution of orthopyroxene
in the peridotite layers in the sandwich runs. For both major and trace elements,
the compositional trends defined by glasses from the hornblendite-only melting
experiments and from the sandwich experiments are similar to trends observed in natural
basanite - alkali basalt suites. These results suggest that compositional trends from
nephelinite/basanite to alkali basalt observed in intraplate setting are related to reaction
between nephelinitic/basanitic liquids and peridotite rather than, for example, a
pressure effect and/or an increase in the degree of partial melting of peridotitic
sources.
Although we do not exclude that the alkaline magma-peridotite interaction is an
important process in the convecting mantle (i.e. at pressure higher that 2.5-3 GPa), we
suggest that the main interaction which produces the nephelinite/basanite to alkali
basalt/tholeiite composition ranges observed in oceanic islands appends in the lithospheric
mantle. These experiments indicate also that the temperature at which alkali melts
interact with peridotites could be significantly lower that the solidus temperature of
theses peridotites. This provides an explanation for the implication of lithospheric
components during the generation of alkaline lavas without requiring that these
components reach their melting temperature. We conclude that lithospheric mantle
needs to be considerate as an important component in the petrogenesis of alkaline
lavas.
[1] Ito and Mahoney (2005) EPSL 230, 29– 46;
[2] Frey et al. (1978) J. Petrol. 19, 463-513;
[3] Shaw et al. (1998) Contrib. Mineral. Petrol. 132, 354-370;
[4] Lundstrom (2000) Nature 403, 527-530;
[5] Pilet et al. (2008) Science 320, 916-919. |
|
|
|
|
|
|