Mantle Mixing: Processes and Modeling - Physics and Chemistry of the Deep Earth - page 352

Herzberg, 2011, Wang et al ., 2010) which is most
logically derived from recycled oceanic crust. The
presence of trace elements that suggest hydrother-
mal alteration of metasomatism (Dixon et al .,
2002; Stroncik & Haase, 2004; Nishio et al .,
2004) further indicate the presence of oceanic
crust in the source. Hafnium and neodymium
isotopes suggest that the mixing of sediments
and oceanic crust is important in the formation
of OIBs (Chauvel et al ., 2008). While modeling
suggests that the recycling of oceanic crust alone
is not sufficient to explain Pb isotope systematics
(Christensen & Hofmann, 1994), it is quite possi-
ble that alteration of the crust before subduction
is sufficient to explain the Pb discrepancy (Kelley
et al ., 2005). Variations in Nb/U suggests that the
MORB source is not yet homogenized which con-
firms the long term role of the recycling of oceanic
crust. The lack of constancy suggests that oceanic
crust is inefficiently mixed back into the mantle
on timescales on the order of billions of years
(Sun et al ., 2008).
The OIB isotope systematics indicates secular
changes in the state of the recycled material. For
example, Gurenko et al . (2009) suggested that
the presence of eclogite in the source is lim-
ited to recent ( < 1 Ga) subduction but that the
older HIMU signature is principally hosted by
the peridotite. There are strong indications that
EM-I can be generated by the recycling of oceanic
crust in a warmer Earth, which renders the ef-
fect of water in subduction zones less important
(Hart et al ., 1986; Elliott et al ., 2007; Brandenburg
et al ., 2008).
The effects of water may be profound on the
chemical evolution of the Earth. The isotopic het-
erogeneity that is added by alteration of oceanic
crust and mantle, and subsequent metasomatism
in the mantle wedge (e.g., Turner et al ., 2012)
has profound implications on the elemental and
isotopic budget of the downgoing lithosphere.
Metasomatism alone has been suggested as an
alternative to continental delamination in the
formation of EM-II (Workman et al ., 2004).
Elemental differentiation caused by water circu-
lation and hydration-dehydration reactions have
also been suggested to be important in creating

isotopic heterogeneity (Iwamori & Albar ede,

2008; Iwamori et al ., 2010).
Detailed modeling of the water budget in
present-day subduction zones suggest that while
large amounts of water are released from the slab
(Figure 12.3; van Keken et al ., 2011), a significant
amount of the water can be subducted to the deep
Earth. In van Keken et al . (2011) we estimated
that 30% of the mineralogically bound water

(or 15% of total water) entering the trench is
subducted past the volcanic front. This translates
toarateof3

∼

10 8 Tg/Myr. This estimate is at
the low end of other estimates (e.g., Ito et al .,
1983; Iwamori, 2007; see also the references in
van Keken et al ., 2011, table 3). Interestingly, the
relatively low water return flux to the mantle
is fairly consistent with independent estimates
that take into account Phanerozoic sea level
variations and the concentration of water in the
sources that form the mid-oceanic ridge, arc and
back-arc basalts (Parai & Mukhopadhyay, 2012).
The deep water subduction provides a pathway
for the addition of water to the transition zone
(Richard et al ., 2007; Chapter 13, this volume,
below). Since the transition zone minerals can
retain significant amounts of water there is the
suggestion that the transition zone can form an
effective chemical filter to whole mantle con-
vection. Melting under deep hydrated conditions
(e.g., Jing and Karato, 2009; Karato, 2011) can
filter out incompatible elements. This has been
suggested to explain how the MORB and OIB
variations are formed by whole mantle convec-
tion without the need for invoking geographically
separated reservoirs (Bercovici & Karato, 2003).
The processes discussed above provide near-
continuous and long-term generation of hetero-
geneity. In addition, several more punctuated
processes are important. For example, the
episodic nature of the continental crust formation
(Parman, 2007; Rollinson, 2008; Hawkesworth
et al ., 2010) will cause a complementary episod-
icity in the nature of the flux of material to
the Earth's deep interior. A similar episodicity
may occur in the foundering of deep continental
crust, which is generally attributed to explain
the bulk characteristics of the continental crust

×