Geology Reference
In-Depth Information
The OIBs in Figure 10.1, apart from Hawaii, have enrichment patterns compa-
rable to that of continental crust. A naive inference would be that OIBs also result
from about 1% partial melting of the mantle, but Hofmann and White [74] argued
that the degree of melting of OIB sources was larger (of the order of 5%), and
that the OIB sources must be enriched in incompatible trace elements relative to
MORB sources. Their interpretation is supported by the fact that Iceland shows
enrichments, even though it is located on a spreading centre and the degree of
melting is likely to be as great or greater than for MORBs.
The relatively low enrichment of Hawaii is interpreted as due in part to a relatively
high degree of melting, which tends to dilute the concentrations of incompatible
elements. This is consistent with Hawaii being much the strongest plume, as we
saw in Chapter 7, which would make it less prone to lose heat ascending through
the mantle, and thus more likely to melt more. However, the low enrichment may
also reflect a less enriched source than those of other OIBs.
Hofmann and White went on to propose that the enrichment of the OIB sources
is due to the incorporation of a higher proportion of old subducted oceanic crust
than in MORB sources. They proposed that subducted oceanic crust accumulated at
the base of the mantle, in the D
zone, and the plumes entrain some of this crust and
carry it to the top of the mantle, where it melts preferentially. This interpretation
is supported by the patterns of anomalous elements in Figure 10.1. Nb is low and
Pb is high relative to the general trend of continental crust. MORBs show the
reverse pattern, which is consistent with the MORB source being the complement
to continental crust. The OIB anomalies resemble the MORB anomalies, which
fits with them being due to an excess of old MORB. If OIBs were due just to
low-degree melting of normal mantle, the Nb and Pb anomalies might be expected
to resemble those of continental crust, or to have no anomaly at all.
This relationship of OIBs with MORBs was made more explicit by Hofmann
et al.
[191], who displayed the Nb/U ratio for MORBs and OIBs relative to primitive
mantle and continental crust. A later version of their plot is shown in Figure 10.7,
and a more sophisticated argument, reaching the same conclusion, is given by
Hofmann [60]. The plot shows that the OIBs resemble the MORBs in mostly having
a Nb/U ratio higher than primitive, rather than resembling continental crust, which
has a lower ratio. A few of the EM2 OIBs have ratios comparable to primitive, but
these can be interpreted as containing a few per cent of continent-derived sediment.
Figure 10.7 also argues against significant involvement of primitive mantle in
OIBs. Early views of the Nd isotopes of Figure 10.2(c) were that the spread of data
is due to mixing between a highly depleted component and a primitive component.
However, most OIBs have a distinctly higher Nb/U ratio than primitive, and the
few that are similar are better explained by mixing of continental components than
by primitive components.
