WHAT CAN WE LEARN FROM ISOTOPES? - Chemical Fundamentals of Geology and Environmental Geoscience - page 191

Table 10.2 Rb, Sr, Nd and Sm element concentrations
(in ppm ) in the Earth's mantle and continental crust
(as cited in Henderson and Henderson, 2009)

• sufficient time must elapse between the mantle
melting event n and the later crustal melting event p
for significant crustal 87 Sr/ 86 Sr 'growth' to take place.
Had the interval n-p been short (e.g. ~100 Ma rather
than 1500 Ma), the initial ratio p obtained from r, s
and t would be barely distinguishable from mantle
values, regardless of the higher Rb/Sr ratio.

Element Z Average primitive mantle Average continental crust

Rb 37 0.605 49

Sr 38 20.3 320

Rb/Sr 0.0298 0.153

Nd 60 0.431 20

The solid residue from crustal melting, left behind at
depth, would evolve along a less steep path, leading to
a present-day 87 Sr/ 86 Sr ratio illustrated by q .
It follows that the initial Sr isotope ratio ( 87 Sr/ 86 Sr) 0
obtained from an isochron provides a sensitive tracer
for the involvement of old continental crust in magma
genesis. When compared to the mantle growth curve
(Figure 10.6a), the initial Sr isotope ratio tells us
whether an igneous magma has originated from the
Earth's mantle or whether, on the other hand, cont-
inental crust with higher 87 Sr/ 86 Sr has made a signif-
icant contribution to magma genesis. Using Figure 10.5b
as an example, the parent magma of the Grønnedal-Íka
intrusion when the complex formed 1299 Ma ago had
an initial ratio of 0.7032. This initial ratio lies very close
to the mantle growth curve at that time (see the co-
ordinates labelled 'G-I' in Figure 10.6a), leading
Blaxland et al (1978) to conclude that the parent magma
was essentially mantle-derived, with negligible crustal
involvement.
( 87 Sr/ 86 Sr) 0 can also shed light on different mantle source
regions, as illustrated by the depleted mantle represented
by dm in Figure 10.6a. As we shall see (Figure 10.8), such
depleted source signatures are characteristic of most
mid-ocean ridge basalts, suggesting that such basalts
tap a depleted 'reservoir' in the Earth's mantle.

Sm 62 1.327 3.9

Sm/Nd 0.325 0.195

event retain their primordial Rb/Sr ratio and continue
to evolve along the dashed extension of line m-n to pm
(representing the present-day 87 Sr/ 86 Sr ratio of 'prim-
itive mantle', i.e. that unaffected by the melting event).
What remains of the parcel that did undergo partial
melting event n , on the other hand, is a refractory solid
residuum with a Rb/Sr ratio even lower than the orig-
inal mantle (since Rb has been preferentially removed
into the partial melt and into the crust formed from it).
This 'depleted' parcel of mantle evolves hereafter along
a shallower trajectory in Figure 10.6a. Any further melt
extracted from this depleted mantle region today will
testify to the earlier melting event through its less radio-
genic 87 Sr/ 86 Sr ratio ( dm in Figure 10.6a).
Now suppose that the crustal rocks resulting from
the mantle melting event at n themselves undergo
partial melting in the crust at around 1.0 Ga BP ( p ),
and the magma formed fractionates into a range of
high-level intrusive rocks of varying compositions. As
before, partial melting in the crust generates melts
with the same 87 Sr/ 87 Sr as the source p but with Rb
concentrated relative to Sr, leaving the solid residue
with a lower Rb/Sr ratio. The high Rb/Sr ratios in the
intrusive rocks generate steeper 87 Sr/ 86 Sr evolution in
Figure 10.6a, and exhumed samples of these rocks, col-
lected and analysed today, would give 87 Sr/ 86 Sr values
similar to r, s and t . Plotted in 87 Sr/ 86 Sr versus 87 Rb/ 86 Sr
space (Figure 10.6b), these samples would define an
isochron with an initial ratio ≈ 0.7106, too high to be
consistent with direct derivation from the mantle. Two
requirements have to be satisfied to generate such a
high initial ratio:

Dating Cenozoic sediments using 87 Sr/ 86 Sr
Rb-Sr isochron dating as such does not lend itself to
dating the deposition of sedimentary rocks, because
the clastic component of sediments significantly pre-
dates deposition, and generally consists of minerals
too poor in Rb to generate measurable radiogenic 87 Sr
(a limitation that also applies to limestones). Rb-bearing
authigenic minerals such as glauconite may in some
cases provide Rb-Sr isochron dates, but such ages may
not accurately reflect the age of deposition that is being
sought.
The 87 Sr/ 86 Sr ratio does nonetheless provide a reliable
geochronological tool for the dating of Cenozoic marine
sediments. Those consisting of carbonate shells inherit

• the source region must have significantly higher Rb/Sr
than mantle rocks, providing steeper growth rate in
87 Sr/ 86 Sr;