Carbon subduction slabs

In the case of arc volcanism, the critical observation is that the C02/ He ratio of arc-related volatiles is significantly greater than that of MORBs (Marty et ai, 1989 Sano and Marty, 1995 Sano and Williams, 1996 Varekamp et ai, 1992). In a compilation of subduction zone type gases, Marty and Tolstikhin (1998) report a median value of 11.0 3.3 X 10, or 5 times that of MORBs. Consequently, an estimate of the proportion of carbon from nonmantle sources (—80%), presumably the subducted slab, can be inferred by scaling to the (upper mantle) C02/ He value. 

We have a reasonable picture of the geochemical cycle of water and to a lesser extent that of carbon. To return to the question of the fate of volatiles in subducting slabs, it is clear that there are a number of viable hosts for volatiles in the mantle. The crux is whether they actually contain H2O or CO2, or whether essentially all the volatiles are stripped out of the slab during subduction, leaving a desiccated slab to continue down into the mantle. There is no consensus on this issue. One extreme is represented by Dixon et al. (2002), who argue that water is extracted from the subducting slab with 92-97% efficiency. 

The output of carbonate-derived CO2 (L) cannot be balanced with the amount of sedimentary carbonate CO2 being subducted except for the Central American and Sunda (Indonesia) volcanic arcs. This would suggest that, in both these localities, sediment-derived carbonate may be efficiently transported to the zones of magma generation. When carbonate CO2 from the subducting altered oceanic crust is taken into account, the output can be supplied solely from the slab indeed, at all localities, except Japan, only a fraction of the input CO2 is necessary to supply the output. These figures reinforce the notion that subduction zones act as conduits for the transfer of carbon into the mantle (Kerrick and Connolly 2001). 

---