regions. At mid-ocean ridges, shallow mantle rises passively in response to plate spreading and melts partially
through adiabatic decompression. The latter process generates the oceanic crust at mid-ocean ridges and is responsible for about 80 % of Earth’s volcanism. At hot spots,
where hot buoyant mantle rises from the deeper mantle,
some melting occurs because temperatures exceed those
of the ambient shallow mantle. At island arcs, basaltic
and andesitic crust is generated by fluid-induced melting
of the mantle wedge overlying the subducted slab. The
newly generated basaltic crust is eventually subducted at
convergent plate margins, establishing a large-scale plate
tectonic cycle of ocean crust generation and its return into
the mantle. In this way, the mantle has evolved continuously since the onset of plate tectonics.
Formation and evolution of the “depleted” mantle
During partial mantle melting at mid-ocean ridges, hot
spots, and island arcs, elements that do not fit into the crystal lattices of mantle minerals, the so-called incompatible
elements (e.g., Cs, Rb, Ba, Th, U, Nb, Ta, K, Sr, Zr, Pb,
Hf, Ti, and rare earth elements), are extracted preferentially from the mantle and incorporated into the newly
formed basaltic crust. Left behind is a residual mantle
whose mineralogical and chemical composition has
changed by melting at shallow depths (usually <100 km
depth), that is, it has become depleted in incompatible elements. This residual mantle is therefore referred to as the
“depleted mantle.”
The depleted mantle (DM) has undergone a complex
geological evolution. Depleted mantle that has formed at
mid-ocean ridges, hot spots, or island arcs is transported
with the overlying oceanic crust and eventually subducted
at convergent plate margins. Although part of the oceanic
crust, consisting of igneous rocks and their sedimentary
cover, is removed during subduction, most of the oceanic
crust is recycled back into the mantle. Once introduced
into the deeper mantle, the subducted oceanic crust and
mantle become stretched and intermingled by mantle convection. In this way, crustal material that is enriched in
incompatible elements is reincorporated into the
DM. These processes counteract depletion by melt extraction and re-enrich the DM in incompatible elements.
Over geologic time, the DM is thus expected to have
evolved into an assemblage of materials that is mineralogically and chemically heterogeneous, consisting of variably depleted peridotitic mantle and some remnants of
recycled crustal rocks (e.g., van Keken et al., 2014).
Owing to the different composition, size, and distribution
of its individual components (e.g., residual peridotitic
mantle, oceanic crust, continental crust, and marine sediment), the DM is heterogeneous on a range of scales, from
the 1000-km scale of ocean basins (Dupré and Allègre,
1983; Hart, 1984) down to the cm scale of single hand
specimens. Each individual component of the DM can
also vary compositionally: the composition of the residual
peridotitic mantle varies because of differing degrees of
depletion
(melt
extraction)
or
refertilization
(metasomatism) (e.g., Johnson et al., 1990; Bodinier and
Godard, 2014). Crustal components (oceanic crust, continental crust, and marine sediment) are compositionally
variable as a result of different conditions of formation
(e.g., Plank, 2014; Rudnick and Gao, 2014; White and
Klein, 2014) and varying extents of hydrothermal alteration (e.g., Staudigel, 2014) and modification during subduction (e.g., Bebout, 2014; Ryan and Chauvel, 2014).
Composition of the depleted mantle
Direct investigation of mantle composition is a complicated task because mantle rocks are rarely exposed at the
surface. Mantle peridotites are sporadically found on the
ocean floor, in ophiolite complexes and peridotite massifs
on the continents, and as xenoliths in lavas (e.g., Bodinier
and Godard, 2014). Fragments of the DM recovered from
the ocean floor, abyssal peridotites, are often highly
altered through interaction with extracted melt and seawater. Peridotites from ophiolites, peridotite massifs, and
mantle xenoliths are also generally affected by melt-rock
interaction (metasomatism) and weathering and are sometimes metamorphosed (Bodinier and Godard, 2014).
Owing to their scarcity, small scale, and prevalent modification by secondary processes (metamorphism, metasomatism, alteration), it is difficult to derive a
representative composition of the DM by direct investigation of the available mantle samples. Current estimates of
average DM composition (Salters and Stracke, 2004;
Workman and Hart, 2005) are therefore mostly indirectly
inferred from mantle-derived melts: basalts generated at
mid-ocean ridges (MORB).
Mantle melting integrates over different components of
the DM, and one strategy of constraining DM composition
is to use compositional parameters that are invariant in
MORB (Salters and Stracke, 2004; Workman and Hart,
2005). MORBs, however, have highly heterogeneous
chemical compositions (Hofmann, 2014; Arevalo and
McDonough, 2010; Jenner and O’Neill, 2012; Gale
et al., 2013) that reflect the different conditions during
melting and the compositional diversity of their DM
source. Estimates of DM composition thus span a range
of compositions, mirroring the compositional variability
observed in MORB. In detail, there may be a systematic
bias between MORB and the DM because some components melt preferentially and may thus become overrepresented in the melt (MORB) relative to their mantle source,
the DM (Stracke, 2012).
Although estimating the composition of the DM is thus
complex, current estimates should adequately account for
the compositional range of the DM. Even more difficult
than estimating its composition is assessing the proportion
of the DM relative to the total mantle. Current estimates
are based mainly on geochemical mass balances or
“box” models and range anywhere from 30 % to 100 %
(e.g., Allègre et al., 1983; Boyet and Carlson, 2006;
DePaolo, 1980; Hofmann, 1988; Jacobsen and
DEPLETED MANTLE
183
through adiabatic decompression. The latter process generates the oceanic crust at mid-ocean ridges and is responsible for about 80 % of Earth’s volcanism. At hot spots,
where hot buoyant mantle rises from the deeper mantle,
some melting occurs because temperatures exceed those
of the ambient shallow mantle. At island arcs, basaltic
and andesitic crust is generated by fluid-induced melting
of the mantle wedge overlying the subducted slab. The
newly generated basaltic crust is eventually subducted at
convergent plate margins, establishing a large-scale plate
tectonic cycle of ocean crust generation and its return into
the mantle. In this way, the mantle has evolved continuously since the onset of plate tectonics.
Formation and evolution of the “depleted” mantle
During partial mantle melting at mid-ocean ridges, hot
spots, and island arcs, elements that do not fit into the crystal lattices of mantle minerals, the so-called incompatible
elements (e.g., Cs, Rb, Ba, Th, U, Nb, Ta, K, Sr, Zr, Pb,
Hf, Ti, and rare earth elements), are extracted preferentially from the mantle and incorporated into the newly
formed basaltic crust. Left behind is a residual mantle
whose mineralogical and chemical composition has
changed by melting at shallow depths (usually <100 km
depth), that is, it has become depleted in incompatible elements. This residual mantle is therefore referred to as the
“depleted mantle.”
The depleted mantle (DM) has undergone a complex
geological evolution. Depleted mantle that has formed at
mid-ocean ridges, hot spots, or island arcs is transported
with the overlying oceanic crust and eventually subducted
at convergent plate margins. Although part of the oceanic
crust, consisting of igneous rocks and their sedimentary
cover, is removed during subduction, most of the oceanic
crust is recycled back into the mantle. Once introduced
into the deeper mantle, the subducted oceanic crust and
mantle become stretched and intermingled by mantle convection. In this way, crustal material that is enriched in
incompatible elements is reincorporated into the
DM. These processes counteract depletion by melt extraction and re-enrich the DM in incompatible elements.
Over geologic time, the DM is thus expected to have
evolved into an assemblage of materials that is mineralogically and chemically heterogeneous, consisting of variably depleted peridotitic mantle and some remnants of
recycled crustal rocks (e.g., van Keken et al., 2014).
Owing to the different composition, size, and distribution
of its individual components (e.g., residual peridotitic
mantle, oceanic crust, continental crust, and marine sediment), the DM is heterogeneous on a range of scales, from
the 1000-km scale of ocean basins (Dupré and Allègre,
1983; Hart, 1984) down to the cm scale of single hand
specimens. Each individual component of the DM can
also vary compositionally: the composition of the residual
peridotitic mantle varies because of differing degrees of
depletion
(melt
extraction)
or
refertilization
(metasomatism) (e.g., Johnson et al., 1990; Bodinier and
Godard, 2014). Crustal components (oceanic crust, continental crust, and marine sediment) are compositionally
variable as a result of different conditions of formation
(e.g., Plank, 2014; Rudnick and Gao, 2014; White and
Klein, 2014) and varying extents of hydrothermal alteration (e.g., Staudigel, 2014) and modification during subduction (e.g., Bebout, 2014; Ryan and Chauvel, 2014).
Composition of the depleted mantle
Direct investigation of mantle composition is a complicated task because mantle rocks are rarely exposed at the
surface. Mantle peridotites are sporadically found on the
ocean floor, in ophiolite complexes and peridotite massifs
on the continents, and as xenoliths in lavas (e.g., Bodinier
and Godard, 2014). Fragments of the DM recovered from
the ocean floor, abyssal peridotites, are often highly
altered through interaction with extracted melt and seawater. Peridotites from ophiolites, peridotite massifs, and
mantle xenoliths are also generally affected by melt-rock
interaction (metasomatism) and weathering and are sometimes metamorphosed (Bodinier and Godard, 2014).
Owing to their scarcity, small scale, and prevalent modification by secondary processes (metamorphism, metasomatism, alteration), it is difficult to derive a
representative composition of the DM by direct investigation of the available mantle samples. Current estimates of
average DM composition (Salters and Stracke, 2004;
Workman and Hart, 2005) are therefore mostly indirectly
inferred from mantle-derived melts: basalts generated at
mid-ocean ridges (MORB).
Mantle melting integrates over different components of
the DM, and one strategy of constraining DM composition
is to use compositional parameters that are invariant in
MORB (Salters and Stracke, 2004; Workman and Hart,
2005). MORBs, however, have highly heterogeneous
chemical compositions (Hofmann, 2014; Arevalo and
McDonough, 2010; Jenner and O’Neill, 2012; Gale
et al., 2013) that reflect the different conditions during
melting and the compositional diversity of their DM
source. Estimates of DM composition thus span a range
of compositions, mirroring the compositional variability
observed in MORB. In detail, there may be a systematic
bias between MORB and the DM because some components melt preferentially and may thus become overrepresented in the melt (MORB) relative to their mantle source,
the DM (Stracke, 2012).
Although estimating the composition of the DM is thus
complex, current estimates should adequately account for
the compositional range of the DM. Even more difficult
than estimating its composition is assessing the proportion
of the DM relative to the total mantle. Current estimates
are based mainly on geochemical mass balances or
“box” models and range anywhere from 30 % to 100 %
(e.g., Allègre et al., 1983; Boyet and Carlson, 2006;
DePaolo, 1980; Hofmann, 1988; Jacobsen and
DEPLETED MANTLE
183