Geology Reference
In-Depth Information
each from the eclogite melt and the peridotite. For the sake of conciseness, the term
hybrid pyroxenite
will be used here.
There is by now significant observational evidence that some melts from eclogitic
or pyroxenitic sources reach the surface without fully equilibrating to the homo-
geneous composition. Takahashi
et al.
[210] concluded that the Columbia River
Basalts are derived from shallow melting of a pyroxenite source in a mantle plume
head. Sobolev
et al.
[209] argue that unusually high Ni and Si contents of Hawaiian
shield basalts are consistent with derivation from a secondary, olivine-free pyrox-
enitic source produced when melt from recycled oceanic crust hybridises with
peridotite in the Hawaiian plume [74]. Others have argued for some time that small
near-ridge seamounts are produced by melting from heterogeneities, plausibly of
recycled oceanic crust, that only pass through the edge of the sub-ridge melting
zone [193, 194]. Salters and Dick [211] show that abyssal peridotites from the
southwest Indian Ridge cannot explain the neodymium isotopes of nearby basalts
without invoking a more enriched component, plausibly pyroxenite or eclogite,
that has been completely melted out of the residual peridotites.
Osmium isotopes provide some of the strongest evidence for the survival in
erupted basalts of unequilibrated signatures from eclogites [204, 209]. Osmium
isotopes correlate nearly linearly with Sr, Nd and Pb radiogenic isotopes, and the
sublinear correlations have been interpreted as indicating mixing between liquids,
rather than reaction between a liquid and a solid. This would imply that the eclogite-
derived melts survive their passage through the peridotite matrix until they reach
the peridotite melting zone, or even near-surface magma chambers.
Kogiso
et al.
[204] have considered in some detail the physical circumstances
in which eclogite-derived melts might survive both the initial melting process
and then the passage through the peridotite matrix, taking account of the size
of the eclogite body, diffusion rates in solids and liquids, and whether the melt is
saturated or undersaturated in silica. Their general conclusion is that some eclogite-
derived melt might reach the shallower peridotite melting zone if its source pods of
eclogite are thicker than 1-10 m, whereas other melt will refreeze and be trapped
in the mantle. Melts from smaller eclogite bodies or silica-undersaturated melts are
the most likely to be trapped. The melt that does survive migration may do so by
passing through relatively narrow channels that become insulated from surrounding
peridotite by a reaction zone formed by preceding eclogite melt. A sketch of the
resulting possibilities is shown in Figure 10.13.
If eclogite-derived melt passes through a channel without reacting with surround-
ing peridotite, then some of the chemical disequilibrium between the eclogite and
the peridotite will remain. This means, on the one hand, that the eclogite-derived
melt will retain some signature of its eclogitic source, which the observations just
cited support. On the other hand, it means that some of the peridotite will not
