Geoscience Reference
In-Depth Information
Figure 9.63.
Initial
87
Sr/
86
Sr ratio versus age
for volcanic and plutonic
rocks from the central
Andes, 26
◦
S-29
◦
S. (Data
from McNutt
et al
.(1975).)
0.800
W
E
0.700
100
0
Age (Ma)
decreases because much of the water has been lost from the subducting plate. This
means that the magmas produced are likely, and are observed, to be more alkaline
(more andesitic). It is also possible that they will be altered on their ascent through
the greater thickness of mantle and crust. Ultramafic rocks found above subduc-
tion zones are presumably tectonically emplaced pieces of the residue remaining
after partial melting of the overriding mantle. This subject is discussed more fully
in Section 10.2.1.
The chemical compositions of lavas in island arcs are spatially zoned with
respect to the subduction zone. The strontium isotope ratio exhibits some cor-
relation with the depth to the subduction zone. In the Indonesian Arc, this ratio
appears to increase slightly with depth to the subduction zone. In the central
Andes, the ratio increases with distance from the trench; but this is an age effect
as well as a depth effect since the youngest rocks are inland and the oldest on the
coast (Fig. 9.63). In this case, the isotopic data indicate that the magma source
moved progressively eastwards with time. The smallest, and oldest, initial ratio
of 0.7022 is in good agreement with the 0.702-0.704 which would be expected
for the oceanic crustal basalts and mantle that presumably melted to form the
sampled rocks. As time progressed and the magma source moved eastwards, the
rising magma would have to migrate through increasing thicknesses of crust; thus,
the likelihood of contamination of the strontium isotope ratio is high. Subducted
sea water, which would have a ratio of about 0.707, could also contaminate the
magma. Thus, the greatest measured value of 0.7077 is still entirely consistent
with a mantle or oceanic-crust origin for these rocks.
Back-arc spreading centres are necessarily influenced by the subduction zone.
Results from the Tonga arc and the spreading centres in the Lau Basin immediately
behind it show that lavas from the central Lau spreading centre were generated
by decompression melting in the garnet stability field and are indistinguishable
from MORB. In contrast, lavas from the Valu Far spreading centre to the south,
which is within 50 km of the arc volcanoes, are similar to the arc lavas. Results
of geochemical studies indicate that, like the arc lavas, they were derived from
the mantle wedge, but with lower levels of fluid from the subducted slab.
