Geology Reference
In-Depth Information
and Table 10.1 shows that it would have a concentration of about 10
3 ng/g. This is
significantly higher than traditional estimates of 3-4.7 ng/g [184-186]. The reasons
for the previous answers will be discussed later, though we can note from Figure 9.8
that abundances estimated from MORB concentrations may be too low by a factor
of up to 2 or so, because of the gradient of basalt tracers through the upper mantle.
Similar mass balances apply to other refractory incompatible elements.
±
10.3 Incompatible trace elements in the mantle
We are now in a position to interpret some features of Figure 10.1, starting with
the enrichment of MORBs in incompatible trace elements, again using uranium as
an example. MORB, at about 75 ng/g in the figure, is enriched relative to both the
primitive mantle (20 ng/g) and our inferred MORB-source mantle (10 ng/g). The
enrichment of MORB relative to its source is interpreted as resulting from the con-
centration of U into the melt phase during melting, because of the incompatibility
of U in the crystal structures of mantle minerals. The concentration in the present
interpretation is by a factor of 7.5 (or a range of 5-8).
At this point in conventional interpretations, this concentration factor would be
used to infer that the MORB source underwent about 13% partial melting, but we
cannot make this inference if the source is heterogeneous. This will be discussed in
Section 10.4. Actually, in the conventional interpretation, the concentration factor
is 16-25 (75 ng/g relative to 3-4.7 ng/g) and the inferred degree of melting would
be 4-6%, or 6-10% if MORB is taken to average 50 ng/g.
The continental crust is enriched in the most incompatible elements by a factor
of 80-100 relative to primitive (Figure 10.1). It was recognised early that the
depletion of the MORB source is likely to be complementary to the enrichment of
the continental crust [176]. This interpretation is implicit in the mass balance of
Table 10.1. However, previous mass balances, based on what are argued here to
be underestimates of the MORB-source abundances, indicated that only 30-60%
of the mantle had been so depleted, and the rest was assumed to be primitive. It
will be shown in the next section that we cannot expect a significant amount of
primitive mantle to have survived. Nevertheless the basic complementarity of the
mantle and the continental crust is a straightforward inference from the identified
reservoirs in the Earth, as was illustrated in compiling Table 10.1.
Hofmann [190] demonstrated that the general pattern of the continental crust
in Figure 10.1 could be accounted for by about 1% partial melting of a primitive
mantle (see also [60]). He stressed that this is clearly not the process by which the
continental crust was formed, but the effect of the relative compatibilities persists
through the real process, which involves multiple stages. This conclusion will
remain true in the present interpretation.
