Sie sind auf Seite 1von 18

©2013 Society of Economic Geologists, Inc.

Economic Geology, v. 108, pp. 1953–1970

The Lithospheric Mantle Plays No Active Role in the Formation of


Orthomagmatic Ore Deposits
Nicholas Arndt†
ISTerre UMR 5275 CNRS, Université Grenoble Alpes, BP 53, 38041 Grenoble cedex 9, France

Abstract
The hypothesis that the metals in certain orthomagmatic ore deposits come from a source in the subconti-
nental lithospheric mantle is evaluated in this paper. According to this hypothesis, parts of the mantle beneath
the continents are metasomatically enriched in metals like Ni, Cu, and the platinum group elements (PGE). It
is proposed that under some circumstances, these metals are transported into the crust where they become con-
centrated in orebodies. An examination of the compositions of xenoliths from the lithospheric mantle reveals
little evidence, however, of components that could represent the source of metal-enriched magmas. In addition,
the mechanism whereby metals are brought from the source to the surface is very unclear. The lithosphere
is the coldest part of the mantle and it only melts under special circumstances. The normal product is a low-
degree melt, an alkaline, Si-undersaturated magma of the type that only rarely contains ore deposits. Major
magmatic orebodies normally form from high-volume, high-flux magmas that are produced by high-degree
melting in deeper, hotter parts of the mantle—in the asthenosphere or a mantle plume. For melting to occur
in the lithosphere, rather than in the hotter parts of the mantle, the melting point of the source must be drasti-
cally reduced by the presence of volatiles. On the other hand, there is ample evidence that the host magmas
of ore deposits were abnormally rich in water or CO2 as would have been the case if they came from a volatile-
rich metasomatized source. Magmas from sublithospheric sources could have interacted with the lithospheric
mantle as they ascended toward the surface and they could have picked up some metals through this interac-
tion. This process could have contributed to the formation of some ores, a notable example being the PGE
deposits in Bushveld Complex. There is ample geological and geochemical evidence, however, that the majority
of magmatic deposits form when magmas from sublithosphere sources assimilate material from the continental
crust and that the latter process is instrumental in the formation of the deposits.

Introduction deposits are extracted from a source in the subcontinental


In what might be called the “standard model” for the for- lithospheric mantle (SCLM) and not in the asthenosphere or
mation of magmatic sulfide deposits (Naldrett, 1992, 1999, in a plume. The model builds on a concept—well entrenched
2004; Lightfoot and Naldrett, 1994; Arndt et al., 2004; Li et in the geologic literature—that certain types of continental
al., 2009; Arndt, 2011), magma, usually tholeiitic and resulting basalts result from partial melting of metasomatized portions
from relatively high degrees of melting in the asthenosphere of the SCLM (Hawkesworth et al., 1984b, 1988; Sweeney and
or in a mantle plume, assimilates material from the conti- Watkeys, 1990; Hergt et al., 1991; Menzies, 1992b; Lightfoot
nental crust, a process that leads to the segregation of sulfide et al., 1993, 1994; Lassiter and DePaolo, 1997; Shirey, 1997;
liquid. The chalcophile ore metals partition from the magma Jourdan et al., 2007; Zhang et al., 2009). The foundation of
into the sulfide liquid, and an ore deposit forms if the sulfide this model lies in the trace element and isotopic composi-
accumulates in sufficient quantity and with a sufficiently high tion of these magmas, which differ from that of the sublitho-
grade. We know within reasonable limits the likely compo- spheric mantle and which, in the opinion of many authors,
sitions of the magmas that are produced by melting in the cannot have resulted from crustal contamination.
upper mantle: these magmas contain moderate to high levels In the papers advocating an SCLM source of ore metals
of Ni, Cu, and platinum group elements (PGE; Keays, 1995; (e.g., Garuti et al., 1997; Hong et al., 2002; Maier and Barnes,
Rehkämper et al., 1999) and, given an efficient concentra- 2004; Fiorentini and Beresford, 2008; Fiorentini et al., 2008;
tion process, they can produce high-tenor ore. There is some Richardson and Shirey, 2008; Zhang et al., 2008) it is proposed
discussion about the exact nature of the ore-forming process, that the metasomatized portions of the SCLM are enriched in
particularly the question of whether sulfur must be added chalcophile and highly siderophile elements and that partial
from an external source (Lehmann et al., 2007; Godel et al., melting of this part of the mantle, or contamination within the
2011), but the general lines of the model, as applied to depos- lithosphere of magma from sublithospheric sources, produces
its such as Norilsk-Talnakh in Russia (Fig. 1; Naldrett, 1992; magmas of unusual compositions that are particularly capable
Lightfoot and Hawkesworth, 1997; Li et al., 2009), Voisey’s of forming magmatic ore deposits (Fig. 2). The evidence cited
Bay in Canada (Li and Naldrett, 1999), Jinchuan in China in support of a metal-enriched source in the SCLM is diverse,
(Chai and Naldrett, 1992; Tang, 1993; De Waal et al., 2004), ranging from concrete observations such as the mineralogy
and Kambalda in Australia (Lesher and Keays, 2002), are well and chemical compositions of xenoliths from the lithospheric
established. mantle, through to estimation of unusually high metal con-
Many recent papers have promoted an alternative hypoth- tents or abnormal metal ratios in the parental magmas of ore
esis according to which the ore metals in magmatic ore deposits. A summary of these arguments is given in Table 1.
In this paper the arguments that have been advanced to
†E-mail, Nicholas.Arndt@ujf-grenoble.fr support of an SCLM source of ore metals are evaluated first,

0361-0128/13/4173/1953-18 1953
1954 NICHOLAS ARNDT

Pechenga
Slave
Craton Norilsk
Voisey’s Udachnaya
Bay Kamchatka
Ivrea zone
Lherz
Rhonda

Gorgona

Lihir

Reunion
Bushveld Kaapvaal Kambalda
Craton
Complex

Fig. 1. Map showing the localities of ore deposits (in bold) or other geologic units (smaller font) mentioned in the text.

a flood basalts distinguishing those that provide direct evidence of derivation


of metals from the SCLM from those in which a lithospheric
continental shallow intrusions with ore deposits
source is adopted by default, after evaluation and rejection of
crust other explanations. The SCLM model itself is then critically
evaluated, and the range of compositions that have been pro-
mid- and lower crustal
posed for the mantle beneath the continents are discussed. In
magma chambers this discussion, the emphasis is on models in which the SCLM
plays an active role; i.e., as a source of metals and/or volatiles,
depleted lithospheric and those models in which the lithospheric mantle serves only
mantle (DL) as a passive guide or impediment to the ascent of the mantle
source or of magmas derived from this source are excluded
metasomatized layer (MM & CM) (e.g. Begg et al., 2011). The questions of how these reservoirs
could have been produced and, crucially, how the metals and
other components could have been extracted and transported
into the crust are discussed. Finally, the “standard model” is
eclogite ancient shown to satisfactorily explain most of the features of mag-
veins (EC) mantle wedge (SC) matic Ni-Cu sulfide deposits. The conclusions that emerge
are that most magmatic ore deposits probably develop as
magma of normal composition and sublithospheric origin
b interacts with continental crust, and that the latter process is
instrumental in the formation of the ore deposits.
sub-lithosphere source of
magma and/or heat Arguments Advanced in Support of an
SCLM Source of Ore Metals
Portions of the SCLM are enriched in chalcophile elements
and provide a likely source of metals in ore deposits
Fig. 2. Sketch showing the possible architecture of the continental litho-
sphere. Two alternative models for the formation of magmatic ore deposits Several authors (Sharpe, 1982; Sun et al., 1991; Zhang et al.,
are (a) Extraction of metals from a portion of the subcontinental mantle with 2008) have proposed that the parental magmas of ore deposits
appropriate composition. Abbreviations: CM = carbonate-metasomatized, contained high concentrations of ore metals. In many such
MM = melt-metasomatized, SC = subduction component. (b) Derivation of
magma from a sublithosphere source (plume or asthenosphere) followed by cases, these magmas are also said to have unusual trace ele-
concentration of metals by processes within magma chambers in the conti- ment or isotopic compositions that have been interpreted to
nental crust. indicate derivation from the SCLM. Perhaps the best-known
LITHOSPHERIC MANTLE: FORMATION OF ORTHOMAGMATIC ORE DEPOSITS 1955

Table 1. Arguments for a Source of Ore Metals in the Subcontinental Lithospheric Mantle

Argument Justification References

Portions of the SCLM are enriched in Some parental magmas of ore deposits are enriched in metals; Sharpe (1982), Sun et al. (1991),
chalcophile elements and provide a trace-element or isotopic compositions resemble those of an Garuti et al. (1997), Kerrich et al. (2000),
likely source of metals in ore deposits SCLM source; metasomatized SCLM is enriched in ore metals Hong et al. (2002), Maier and Barnes (2004),
Zhang et al. (2008), Richardson and
Shirey (2009), Hronsky et al. (2012)
Magma or metals are derived from Magmas that form some ore deposits have near-chondritic Mitchell and Keays (1981), Keays (2012)
an S-undersaturated source PGE ratios which requires that they were undersaturated in
sulfur; some parts of the SCLM have this characteristic
Some Ni-Cu-PGE ores contain The association of minerals such as amphibole and biotite with Fiorentini and Beresford (2008),
hydrous minerals from a meta- ore sulfides indicates that the magma was hydrous; trace element Fiorentini et al. (2008)
somatized mantle source and isotope data point to a source in metasomatized SCLM
Default arguments The parental magma is not contaminated by continental crust Hawkesworth et al.(1984b), Hergt et al.
and its composition differs from that of the asthenosphere or (1991), Lightfoot et al. (1994), Lassiter and
plume; therefore the source of magma and/or metals is in DePaolo (1997), Hong et al. (2002), Maier
the SCLM and Barnes (2004), Zhang et al. (2008)

and most convincing example is the Bushveld Complex in In most arguments for an SCLM source of ore metals, ref-
South Africa whose parental magmas, as estimated from the erence is made to metasomatism of parts of the subcontinen-
compositions of peripheral sills or a basal chill zone (Fig. 3), tal lithosphere: it is proposed that, in addition to lithophile
contained unusually high platinum concentrations (Sharpe,
1982; Maier and Barnes, 2004; Wilson, 2012). It is suggested 20
that the high metal contents were crucial to the formation of Continental tholeiites
Bushveld sills
the PGE ores of, for example, the Merensky Reef and UG2 Bushveld sills Mantle peridotite
16
deposits. Another example is provided by the continental Komatiites

flood basalts that host Ni-Cu-PGE sulfide deposits, like those Meimechites
Oceanic basalts
of the Norilsk-Talnakh region in Russia. These basalts have 12
Pt (ppb)

Boninites
high PGE contents and near-chondritic ratios of Pt/Pd, and
in this respect they differ from basalts of mid-ocean ridges 8
and ocean islands, which contain lower contents of the PGE
(Fig. 3). To consider another tectonic setting, many authors 4
argue that the source of Cu or Au in porphyry deposits are MARID
extracted from a part of the suprasubduction mantle wedge
0
that was abnormally enriched in these metals (McInnes et al., 0 10 20 30 40 50
10
1999; Mungall, 2002; Hronsky et al., 2012). And on a broader Mgo (wt%)
scale, the notion of a “metallogenic province” is often associ-
ated with the hypothesis that these clusters of ore deposits are
linked to metal-rich parts of the mantle (Kerrich et al., 2000; 1
Hong et al., 2002; Hronsky et al., 2012). MARID
Ir (ppb)

Isotopic data provide intriguing evidence that at least a por-


tion of the PGE in the Merensky Reef in the Bushveld were
derived from the SCLM. Richardson and Shirey (2008) meas- 0.1

ured the Re-Os isotope compositions of inclusions in diamonds


transported in kimberlites from the Kaapvaal subcontinental
lithospheric mantle. They obtained relatively radiogenic iso- 0.01
tope ratios similar to those previously measured in platinum 0 10 20 30 40 50
group minerals (PGM) from the Merensky Reef by Hart and
Kinloch (1989) and on this basis argued that the PGE in Bush- MgO (wt%)
veld ores came from an enriched component of the SCLM. It Fig. 3. Platinum and iridium contents, plotted against MgO of various
must be noted, however, that (1) in Richardson and Shirey’s types of volcanic and intrusive rocks and mantle xenoliths. The Pt contents
model, the bulk of the magma comes from a sublithospheric of continental tholeiites, and particularly those of marginal sills of the Bush-
veld Complex, are far higher than those in basaltic rocks from oceanic set-
source that entrains the PGE as it passes through the SCLM, tings with similar MgO contents. They are comparable to levels in post 3 Ga
and (2) these data do not necessarily require that the primary komatiites and higher than in mantle peridotites, reflecting the incompatible
magma was abnormally enriched in PGE. As explained in a behavior of Pt and an absence of sulfide in the residue of melting. Ir con-
later section, extensive crustal contamination could account tents of continental tholeiites are also higher than in most oceanic basalts, but
lower than in mantle peridotite, reflecting the more compatible behavior of
for the radiogenic Os in the Merensky samples, and within- this element. Data from Sharpe (1982), Arndt et al. (2004), Maier and Barnes
crust processes may have boosted the concentrations of PGE (2004), Barnes et al. (2010), Maier et al. (2012). MARID = rock composed
in magmas that initially had only a modest PGE content. mainly of mica, amphibole, rutile, ilmenite, and diopside (discussed in text).
1956 NICHOLAS ARNDT

elements and volatiles, the metasomatized regions also volatiles in ore-bearing ferropicrites in the Pechenga green-
became enriched in chalcophile and/or highly siderophile ele- stone belt of Russia came from a source located either in
ments. Although the idea is not unreasonable, closer evalu- metasomatized subcontinental mantle or in a volatile-bearing
ation provides little direct evidence that the metasomatized plume. To explain how the metals were transported into the
parts of the SCLM are enriched in metals such as Ni, Cu and continental crust, they proposed (p. 339) that “Metasomatism
the PGE. This aspect of the problem is discussed in a later introduced alkalis, Cu, PGEs and S into the depleted man-
section when the compositions of the diverse components tle…. Increased water activity caused the harzburgite to melt,
that are known or inferred to exist in the subcontinental litho- producing volatile-rich sulphide-bearing ultramafic magma
spheric mantle are explored. that evolved to intrusions that host Ni-Cu-PGE mineraliza-
tion.” In a third paper, Fiorentini et al. (2012) proposed that
Magma or metals are derived from an komatiites from the Ni sulfide belt of the Yilgarn craton in
S-undersaturated source Western Australia were hydrous and that their degassing may
Mitchell and Keays (1981) and Keays (1995, 2012) argued be implicated in the formation of the ore deposits.
that the magmas that produce ore deposits are undersatu- These ideas are interesting and, particularly in the case of
rated in sulfide and that these magmas came from a source the Pechenga deposits, have considerable merit. However,
with very low sulfide contents such as the depleted peridotite they are open to the following criticisms:
of the SCLM. While it is probably true that most of the mag-
mas that yield ore minerals were indeed S undersaturated, 1.  In many magmatic deposits hydrous phases are indeed
this may merely be the consequence of the manner in which associated with ore sulfides, but the total concentration of
the magmas have formed. Wendlandt (1982), Keays (1995), these phases is very low. As pointed out by Barnes and Camp-
and Mavrogenes and O’Neill (1999) have demonstrated bell (1988), water behaves as an incompatible element and
that the solubility of sulfide in mafic or ultramafic magmas becomes concentrated in the last portion of the silicate liquid
decreases with increasing pressure. The magmas parental to to solidify; the hydrous minerals then crystallize from this liq-
ore deposits are produced by melting at great depth, >90 km uid. Sulfide liquid, which remains in the liquid state to tem-
for basalts and picrites in continental settings. At these depths peratures below the solidus of the silicate liquid, accumulates
only a small fraction of sulfur dissolves in the melt, but as the together with the last fractions of silicate liquid: the associa-
magma rises through the lithosphere, its capacity to dissolve tion between sulfide and hydrous minerals may therefore be
sulfur increases. The consequence is that almost every deep- largely coincidental and does not imply that water was directly
sourced magma should be S undersaturated when it reaches associated with the formation of the ores.
the crust. An additional process, such as crustal contamina- 2. The mechanism proposed for the transport of metals
tion or the addition of S from an external source, is required from the metasomatized zone toward the surface is poorly
to induce the segregation of an immiscible sulfide liquid. It known, as discussed in a later section.
therefore seems unnecessary, on this basis alone, to invoke
derivation from the SCLM. Default arguments for an SCLM source
Furthermore, it is by no means certain that those parts of In a great number of papers advocating a lithospheric source
the SCLM that are invoked as the source of ore-forming mag- of magmas or metals (e.g. (Hawkesworth et al., 1984a; Hergt
mas have very low sulfide contents. It is true that the har- et al., 1991; Gallagher and Hawkesworth, 1992; Lightfoot et
zburgite or dunite that forms the building block of ancient al., 1994; Fedorenko et al., 1996; Molzahn et al., 1996; Las-
SCLM is highly depleted in S (Lorand et al., 2003; Pearson siter and DePaolo, 1997), the authors first test whether the
et al., 2004), but this material is highly refractory and is an host magmas have been contaminated by continental crust
unlikely source of partial melt. The more fertile parts affected and conclude that the amount of contamination was negli-
by metasomatism may or may not be S undersaturated. This gible. They then observe that the compositions of the rocks
issue is discussed later in the paper, following evaluation of in question are unlike those of magmas from depleted upper
the compositions of various components of the SCLM. mantle and conclude, with no further ado, that the source was
in the SCLM. There are three major problems with this type
Some Ni-Cu-PGE ores contain hydrous minerals that could of argument: First, the criteria used to dismiss crustal contam-
have come from a metasomatized mantle source ination are based on oversimplified assumptions or are totally
In several papers the presence of hydrous phases in Ni-Cu- invalid; second, even in cases where crustal contamination is
PGE ores has been linked to derivation from a source in not important, the source is not necessarily in the lithosphere
metasomatized SCLM. The argument is that metasomatism is but could be a component in a mantle plume or another part
caused by the introduction into the mantle of hydrous or other of the sublithospheric mantle; third, and most significantly,
volatile-bearing fluids and that the presence of volatiles would only rarely is the SCLM model evaluated objectively—instead
promote melting and facilitate the extraction of metals from it is commonly adopted by default after dismissal of other
the metasomatized source. Fiorentini and Beresford (2008) alternatives. Here, the tests that are made of crustal contami-
noted the presence of hydrous phases and Fe-Ni sulfides in nation models are considered first, and then the SCLM model
the Valmaggia ultramafic pipe in the Ivrea-Verbano zone of itself is scrutinized.
Italy. They invoked a link between hydrous phases (amphi- Crustal contamination should be accompanied by frac-
bole and biotite) in Ni-Cu-PGE ores in the Ivrea region and tional crystallization: If this argument is correct, indices
a hydrated metasomatized source in the SCLM. In another like SiO2, Th/Nb or 87Sr/86Sr, which monitor the amount of
study (Fiorentini et al., 2008) the same authors proposed that contamination, should correlate with MgO or Mg numbers,
LITHOSPHERIC MANTLE: FORMATION OF ORTHOMAGMATIC ORE DEPOSITS 1957

which monitor the amount of crystallization. The rationale (Fig. 4). Similar patterns are seen in many komatiites, which
behind this argument is the (correct) observation that, unless fractionate abundant olivine following an earlier period of
the magma is superheated, crystallization is required to pro- contamination (Arndt et al., 2008). In the Lower and Main
vide the heat needed to heat and assimilate the contaminant zones of the Bushveld Complex, within-chamber accumula-
(DePaolo, 1981; Bohrson and Spera, 2001). This combination tion of olivine, pyroxene, or plagioclase produces wide varia-
of processes leads to the well-known assimilation-fractional tions in composition that are superimposed on the effects of
crystallization (AFC) models of crustal contamination. The contamination which took place deeper in the crust (Maier et
pitfall in the argument is the fact that, in many cases, multiple al., 2000; Wilson, 2012).
types and compositions of crustal rocks are assimilated, or a Crustal contamination should produce heterogeneous
period of fractional crystallization unaccompanied by contam- hybrid magmas: Here it is argued that the continental crust
ination follows an earlier period of AFC. In both cases the is heterogeneous and that any mixing between crust and
correlation between fractionation and contamination indices magma, itself an erratic process, should produce hybrid mag-
will be obscured or lost. mas with a wide range of compositions. In cases where the
A well-known example is illustrated in Figure 4, which magma has a uniform basaltic composition, as in the upper
shows the compositions of rocks in ore-bearing and weakly two-thirds of the Siberian traps (Wooden et al, 1993), it has
mineralized intrusions of the Norilsk-Talnakh region in Russia been argued that the source must be in the mantle (because
(Naldrett, 1992). There is sound geochemical evidence that no other source can yield a large volume of homogeneous
the magma that fed into the weakly mineralized intrusions had mafic magma) and since this composition differs from that
become heavily contaminated with granitic material in a mid- inferred for the asthenosphere, the mantle in question must
crustal magma chamber to impart the distinctive geochemi- be the SCLM (Hawkesworth et al., 1984a; Lightfoot et al.,
cal signature observed—high La/Sm, Th/Nb, and 87Sr/86Sr 1993).
(Fig. 4; Lightfoot et al., 1990; Wooden et al., 1993). The ore- A flaw in this argument is revealed by closer examination of
bearing intrusions were less contaminated in the mid-crustal the uniform compositions of the Siberian flood basalts (Fig. 5).
chamber and they have lower values of these contamination These magmas have evolved compositions characterized by
indices (Arndt et al., 2003). Both magmas subsequently crys- high SiO2 and low MgO or Mg numbers (Wooden et al., 1993;
tallized and accumulated olivine and other minerals at the Arndt et al., 1998). Their compositions are very different from
shallow level of intrusion, producing large variations of MgO the primary picritic magmas that are produced by melting at
and SiO2 that do not correlate with the contamination indices depths at or near the base of the subcontinental lithosphere.
Furthermore, the covariation of Al2O3 with MgO and types of
5.0
phenocrysts observed in these lavas indicate that plagioclase
L Talnakh
was part of the crystallizing assemblage (Wooden et al., 1993).
4.5
Norilsk The involvement of this low-pressure phase places the frac-
4.0
tionation at shallow, crustal levels. The conclusion is that the
uniform composition of the basalts was also acquired at these
3.5
depths—the homogenization evidently was a shallow-level,
La/Sm

3.0 late-stage process and the compositions of erupted lavas are


far removed from those of their mantle sources.
2.5
Another example of within-crust homogenization is pro-
2.0 vided by the Mafic Complex in the Ivrea zone in Italy (Sini-
goi et al., 1994; Quick et al., 2009). Figure 6 shows that the
1,5
0.710 5 10 15 20 25 magmas that first fed into the lower part of this deep-crustal
MgO (wt%) magma chamber had relatively high eNd and low 87Sr/86Sr,
0.709 combined with trace element characteristics that indicate
derivation from a depleted mantle source (Voshage et al.,
1990; Sinigoi et al., 1994). Gabbros higher in the Complex, in
Sr/ 86 Sr

0.708
contrast, have mineral assemblages (abundant spinel, garnet,
0.707 and orthopyroxene) and the combination of low eNd and high
87

Th/Nb that signal massive contamination with crustal rocks.


0.706 Partial melting of the metasediments that surround the intru-
sion and the presence of rafts of melt-depleted metasediment
0.705 within the gabbros point to local wall rocks as the contami-
5 10 15 20 25 nant. Throughout the upper 90% of the intrusion, the isotopic
MgO (wt%) and trace element composition is uniform, illustrating once
Fig. 4. Diagrams showing variations of La/Sm and 87Sr/86Sr with MgO in again the efficacy of mixing and homogenization in a crustal
intrusions of the Norilsk-Talnakh region (Arndt et al., 2003). The isotopic and magma chamber.
trace element ratios of the ore-bearing “Norilsk-type” and weakly mineral- In a later section it is argued that the SCLM, being het-
ized “L. Talnakh-type” intrusions differ because the latter were more highly erogeneous at all scales, is most unlikely to produce homoge-
contaminated by granitic rock in mid-crustal magma chambers; subsequently,
the magmas in both intrusions differentiated with no further contamination.
neous magma.
As a consequence, there is no correlation between indices of contamination The calculated amount of assimilated crust is unreasonably
and fractional crystallization. high: In many papers the amount of crustal contamination is
1958 NICHOLAS ARNDT

Fig. 5. Variations of selected major and trace elements and Sr isotope compositions in the Siberian Traps (data from
Wooden et al. 1993). The different symbols distinguish the lithologic units and are not relevant to the issues discussed
here. The compositions of lavas in the upper two-thirds of the volcanic pile have remarkably uniform compositions, despite
elevated values of indices such as Th/Nb, La/Sm, and 87Sr/86Sr that indicate the presence of a significant component of conti-
nental crust (Wooden et al., 1993; Arndt et al., 2003). The diagram shows that magmas that are heavily contaminated through
assimilation of crustal rocks can maintain uniform compositions.

Ivrea Zone – mafic complex calculated using geochemical data and values exceeding 30
to 50% are obtained. On thermal grounds, it is argued that
contamination must be accompanied by crystallization in the
DIO proportion of at least 1:3 (DePaolo, 1981; Bohrson and Spera,
2001), which requires that the magma would totally crystallize
before 30 to 50% of crust was added. It is also claimed that if
MG so much crust were added, the contaminated magma should
have a silica-rich, highly evolved composition, not the basaltic
7.5 km

composition of common flood volcanics.


What is wrong with these arguments? First, in many studies
the choice of parental magma is open to question. A com-
UZ mon error is the adoption of the most magnesian basalt in the
sample suite (e.g., Molzahn et al., 1996), even though it can
often be demonstrated that such basalts are too evolved to
be primary magmas. Melting beneath thick SCLM produces
picritic to komatiitic magmas (Wooden et al., 1993; Herz-
berg, 1999), to which large amounts of crust can be added
IZ without exceeding a basaltic composition. Second, because
800 m

BZ Fig. 6. Variation of Nd isotope composition in the mafic complex of the


Ivrea zone. In the upper 7.5 km of the complex, compositions are uniform
despite abundant mineralogical and geochemical evidence of crustal contam-
ination (Sinigoi et al., 1994; Quick et al., 2009). The strongly negative eNd(t)
-10 -5 0 5 10 indicate that the magmas are heavily contaminated but their compositions
are homogeneous. This is another illustration that crustal contamination of
εNd (270 Ma) mantle-derived magmas does not produce heterogeneous hybrid magmas.
LITHOSPHERIC MANTLE: FORMATION OF ORTHOMAGMATIC ORE DEPOSITS 1959

of the large difference between the liquidus temperature are relatively depleted (in the case of relict ocean crust) or
of picrite and the temperature of appearance of the second enriched (in solidified melt from the asthenosphere).
major phase (usually pyroxene), these highly magnesian mag- Component 3: Metasomatized mantle (MM and CM): Meta-
mas can accommodate a large proportion of assimilated crust somatism has changed the composition of large parts of the
(Mungall, 2007). Third, the contaminant is probably not bulk SCLM. Clear evidence of metasomatism is revealed by Sr-Nd-
continental crust but a partial melt of this crust that contains Hf-Pb isotope ratios that depart considerably from those of
far higher concentrations of incompatible trace elements basalts in an oceanic domain. It should be noted, however,
(Thompson et al., 1984; Williams et al., 2011). In the Ivrea that even where extremely enriched Nd and Sr isotope com-
zone, there is strong geological and geochemical support positions exist in the SCLM; they are usually accompanied by
for the argument that partial melt from metasediment was unusually low Nd and Sr concentrations (Pearson and Nowell,
added to the mafic magma to produce the distinctive min- 2002).
eral assemblages and geochemical compositions observed in Volatile-bearing rocks such as phlogopite-K richterite peri-
the Mafic Complex (Voshage et al., 1990; Sinigoi et al., 1991; dotite and pyroxenite record interaction with two types of
Quick et al., 2009). In other cases, it can be demonstrated that fluid: high-temperature silicate melt, and lower temperature
the assimilated material does not melt completely but breaks fluid or melt composed dominantly of CO2 and H2O (Hawkes-
down to a molten fraction, which mixes into the host magma, worth et al., 1984b; Menzies and Hawkesworth, 1987; Griffin
and a solid residue that reacts with the magma or remains et al., 1999; Ionov et al., 2002; Simon et al., 2007; Konc et al.,
as xenocrysts. The heat required to absorb a contaminant in 2012). Interaction with silicate melt changes the mineralogy
this way is less than the heat required for total melting. When of depleted lithospheric mantle by introducing components
such factors are taken into account, the calculated amount of such as clinopyroxene and potassic amphibole and mica to
assimilated crust in most suites of continental basalts remains transform the original harzburgite to phlogopite-K-richterite
within reasonable physical and chemical limits. lherzolites (Erlank et al., 1987; Gregoire et al., 2003). The
extreme example of this metasomatism is MARID (Figs. 7,
Composition of the SCLM 8), a rock composed mainly of mica, amphibole, rutile, ilmen-
The compositions of various components of the SCLM, ite, and diopside. Strongly metasomatized mantle is enriched
using information from mantle xenoliths, peridotite massifs, in incompatible trace elements, but the signature from the
and mantle-derived magmas, are evaluated in this section. two types is different. Melt-metasomatized mantle (MM) is
Four components can be identified, as illustrated in Figure 7. broadly enriched in all incompatible elements and it has a
Component 1: Depleted lithospheric mantle: The mantle chemical signature like that of the end members of an array
beneath Archean cratons (DL, Figs. 2, 7) is composed mainly defined by the compositions of ocean island basalts. It is rela-
of harzburgite or dunite containing olivine and orthopyrox- tively enriched in the majority of incompatible trace elements
ene with high Mg numbers and only small amounts of easily but not in those with the lowest partition coefficients. Man-
fusible minerals (Boyd and Mertzman, 1987; Menzies, 1990a; tle-normalized patterns commonly show a peak around La
Ionov et al., 1992; Griffin et al., 1999, 2003, 2009; Pearson with, in many cases, a positive Nb-Ta anomaly (McDonough,
and Nowell, 2002; Pearson et al., 2004; Simon et al., 2007). 1990; Pearson and Nowell, 2002). Lower temperature meta-
The trace element compositions are difficult to establish somatism by CO2-H2O fluids or melts (CM) produces more
because of common contamination of mantle xenoliths by the subtle changes in mineralogy but imparts a distinctive trace
host kimberlite or alkali basalt, but whole-rock compositions element signature distinguished by enrichment in elements
reconstructed from mineral compositions reveal low concen- such as Rb, Ba, Sr, and the REE and pronounced depletion of
trations of incompatible trace elements (Fig. 8a, b; Simon et all the high field strength elements, including Zr-Hf as well as
al., 2007). Isotopic compositions reflect the trace elements, Nb-Ta, and Ti (Yaxley et al., 1991, 1998; Ionov and Harmer,
with high eNd, and low 87Sr/86Sr and gOs. Depleted harzbur- 2002; Pearson et al., 2003; Fig. 8d).
gite or dunite was the initial building block of the lithospheric Because the sources of the fluids responsible for both types
mantle, but through geologic time it has been modified by the of metasomatism come from deep in the mantle, this type of
addition of metasomatic material (Menzies, 1990a, b; Griffin metasomatism might be expected to affect preferentially the
et al., 1998, 2003; Simon et al., 2007; Foley, 2008). Depleted lower part of the lithospheric mantle. However, K-richterite
harzburgite and dunite is also a common component of peri- is not stable at the base of the SCLM and while it is difficult
dotite massifs like Rhonda or Lherz (Maaloe and Aoki, 1977; to estimate directly the depth of equilibration from MARID
Bodinier et al., 2004). assemblages, they are widely believed to form at depths of
Component 2: Eclogite and pyroxenite (EC, Figs. 2, 7): Sev- no more than 100 to 120 km. Because of the low solidi of
eral origins are attributed to this component: fragments of most hydrous metasomatic components, they do not freeze
ocean crust that were trapped in the lithosphere during its for- at the high temperatures at the base of the lithosphere but
mation, solidified intrusions of melt from deeper in the man- would instead penetrate to mid-lithospheric depths (McKen-
tle, or the products of reaction between melt and older zie, 1989).
lithosphere (MacGregor and Manton, 1981; Jacob, 2004; Pear- Component 4: The “subduction” component (SC). This com-
son et al., 2004; Bodinier et al., 2008; Gréau et al., 2011). The ponent is produced in the mantle wedge when hydrous fluid
eclogitic component has a basaltic composition and a high-pres- or melt introduces a distinctive suite of major and trace ele-
sure assemblage consisting mainly of garnet and sodic pyrox- ments from dehydrating subducting ocean crust. The mantle
ene. The levels of lithophile trace elements are higher than in wedge has high Si and low Fe, giving it a distinctive harzbur-
depleted lithosphere, and the more incompatible elements gitic mineralogy, and it is strongly enriched in fluid-mobile
1960 NICHOLAS ARNDT

2.5
2,5
Low-Ti basalt High-Ti basalt
MARID Perid xenolith
2.0 Carbon xenolith Continental crust
2,0
Mantle Bushveld

(Nb/La)N 1.5
1,5
OIB
CMM MARID
(Nb/La)N

1.0
1,0 DM

0.5
0,5
DL LCC UCC
Bushveld
0.0
0,0
0.0
0,0 1.0 0,5 2.0 1,0 3.0 1,5 4.0 2,05.0 6.0
2,5
(La/Sm)N
(Zr/Sm)N

2.5

2.0

OIB MARID
1.5
(Nb/La)N

1.0 DM

0.5 CMM
LCC UCC

DL Bushveld
0.0
0.0 0.5 1.0 1.5 2.0 2.5

(Zr/Sm)N
Fig. 7. Comparison between the compositions of basalts and possible components of the subcontinental lithospheric man-
tle. The parental magmas of magmatic ore deposits have compositions similar to low Ti basalt, as illustrated by flood basalts
from the Siberian and Karoo provinces (Wooden et al., 1993; Jourdan et al., 2007). High Ti basalt represented by examples
from the Siberian and Ethiopian provinces (Wooden et al., 1993; Kieffer et al., 2004) do not normally contain ore deposits.
Mantle components are represented by: MARID = strongly metasomatized peridotite from Grégoire et al. (2002); Perid xeno-
lith (DL) = reconstructed compositions of peridotite xenoliths from Simon et al. (2007); carbon xenolith (CMM) = carbonated
xenoliths. UCC = upper continental crust, LCC = lower continental crust (Rudnick and Fountain, 1995); DM = depleted
upper mantle, OIB = source of Hawaiian basalts (Hofmann, 1988); Bushveld = marginal sills (Sharpe, 1982; Wilson, 2012).
Notes: (a) The compositions of flood basalts are aligned between mantle sources, either DM or OIB and the compositions
of upper or lower continental crust. They do not coincide with compositions of components in the SCLM which show a much
larger spread of compositions. (b) The wide diversity of compositions of components in the SCLM is unlikely to give rise to
magmas of uniform compositions such as those illustrated in Figures 4 and 5.

elements such as Cs, Rb, and Sr, moderately enriched in Th moderate to high 87Sr/86Sr and gOs values while eNd varies
and the REE, and deficient in Nb-Ta and Ti (Hawkesworth from distinctly positive to slightly negative. The fact that very
et al., 1994; Eiler, 2003; Pearce et al., 2005). Similar charac- few lavas in subduction settings record eNd values less than
teristics are ascribed to the portions of the SCLM that are about −3 is important for the following discussion.
said to give rise to continental flood basalts (Hawkesworth et Although it is commonly proposed that parts of the SCLM
al., 1984b, 1988; Sweeney and Watkeys, 1990; Hergt et al., formed in a subduction setting, or that ancient mantle wedge
1991; Menzies, 1992b; Lightfoot et al., 1993, 1994; Lassiter is incorporated at a later stage into the SCLM, the subduc-
and DePaolo, 1997; Shirey, 1997; Zhang et al., 2009; Ionov, tion component is only poorly represented in xenolith suites
2010). Unlike carbonate-metasomatized mantle, this compo- (McDonough, 1990; Arndt and Christensen, 1992; Pearson
nent is not deficient in Zr-Hf relative to the REE. It contains and Nowell, 2002; Pearson et al., 2004). Analyses of an ancient
LITHOSPHERIC MANTLE: FORMATION OF ORTHOMAGMATIC ORE DEPOSITS 1961

(c) MARID
(a) Depleted peridotite

(d) Carbonate

(b)

(f) Amphibole harzburgites

(e) Lherzolites

W"
Websterites

Fig. 8. Mantle normalized trace elements in probable components of the SCLM. (a) and (b) Compositions of peridotite
xenoliths from the Kaapvaal craton, reconstructed from the compositions of primary minerals by Simon et al. (2007); (c)
MARID-type xenoliths (Grégoire et al., 2002); (d) Carbonate minerals in xenoliths (diagram from Pearson et al. (2003) using
data from Ionov and Harmer (2002); (e) and (f) Peridotites from the Rhonda massif (from Bodinier et al., 2008).
1962 NICHOLAS ARNDT

mantle wedge, as sampled in xenoliths in volcanoes of the metals in mantle xenoliths or peridotite massifs (Pearson et
Kamchatka arc (Ionov, 2010), reveal a large range of composi- al., 2003), and (3) through measurements of the PGE con-
tions but lack the “subduction signature” (Fig. 9). Very few tents of high-degree mantle melts like komatiites (Fiorentini
samples show the over enrichment of fluid-soluble elements et al., 2011). Information obtained through measurements
like Ba, Rb, and Sr that distinguish magmas from subduction of PGE contents of mafic and more evolved volcanic rocks
zones. In addition, rather than being deficient in Nb and Ta is subject to major uncertainties because of the sensitivity of
(i.e., having negative Nb-Ta anomalies), these xenoliths dis- the PGE to the segregation or accumulation of sulfide during
play a wide range of Nb/La ratios, the majority higher than passage through the crust (Rehkämper et al., 1999). In addi-
the chondritic ratio. tion, Lorand et al. (2003), Pearson et al. (2003), and Wittig
et al. (2010) have pointed out that seemingly fertile mantle
Chalcophile and Highly Siderophile Elements peridotite may show the effects of multiple episodes of melt
in SCLM Components extraction and metasomatism, which would have significantly
The PGE contents of the mantle components can be esti- altered their trace element contents and obscured the signal
mated in several ways: (1) inference based on the geochemi- from the mantle source.
cal behavior of these elements during partial melting and Maier et al. (2012) recently summarized the PGE contents
crystallization (Keays, 1982, 1995; Rehkämper et al., 1999), of peridotite xenoliths from the Kaapvaal and Karelian cratons
(2) through measurements of the concentrations of these and reported the data summarized in Figure 10. To these can

4
(a) Udachnaya xenoliths
Mantle wedge
3 Peridotite massif
(Nb/La) N

convecting
1
mantle
low-Ti basalts
Upper continental crust
0
0 2 4 6 8 10

(La/Sm) N

4
(b)
3
(Nb/La) N

high-Ti basalts
1

Upper continental crust


low-Ti basalts
0
0.0 0.5 1.0 1.5 2.0 2.5

(Zr/Sm) N
Fig. 9. Comparison between the compositions of flood basalts (fields labelled “high-Ti” and “low-Ti” taken from Fig. 7)
and the compositions of little-altered xenoliths from the Udachnaya kimberlite (Ionov et al., 2010) and xenoliths from an
ancient mantle wedge (Ionov, 2010). Note the vast range of compositions of the samples from the SCLM, and the absence of
the “subduction signature,” i.e. (Nb/La)N <1 and high La/Sm, in samples from the mantle wedge.
LITHOSPHERIC MANTLE: FORMATION OF ORTHOMAGMATIC ORE DEPOSITS 1963

0.1 mineralogical and geochemical evidence of metasomatism, the


extreme example being the MARID rocks (Fig. 10c), Pt and to
Metals / C1 chondrite

a lesser extent Pd are slightly enriched compared to Ir, but the


overall levels of PGE remain low. Indeed, Lorand et al. (2003)
0.01 and Pearson et al. (2004) concluded that interaction with per-
colating melts commonly leads to a reduction of PGE abun-
A\-#J" dances, because intergranular sulfides are dissolved and
removed. In the data compiled by Maier et al. (2012) and other
0.001 authors, there is very little evidence of the enriched source
that is said to provide metals to magmatic ore deposits.
To further explore the models that invoke an SCLM source
Jagersfontein of ore metals, in the following section the probable chalco-
0.0001 phile element characteristics of the mantle components iden-
0.1 tified in the previous section are summarized.
Depleted lithospheric mantle
Metals / C1 chondrite

This refractory material contains only moderate concen-


0.01 trations of IPGE, usually within error of those of the primi-
tive mantle (Irvine et al., 2003; Pearson et al., 2004; Wittig
et al., 2010). These elements are retained in alloys and accu-
mulate with olivine. Concentrations of the PPGE are far
A\-#J" lower because these elements are removed with the melt
0.001
that extracted other incompatible elements. Most Ni will be
bound up in olivine but a very minor amount will be pres-
Frank Smith ent together with Cu in sulfides. The Pt and Pd contents of
MORB typically are lower than those of the primitive mantle
0.0001
(Keays, 1995; Rehkämpfer et al., 1999).
0.1
MARID Eclogite
Metals / C1 chondrite

The PGE contents of this component are difficult to estab-


lish because the behavior of PGE during subduction is not
0.01 well understood. Gréau et al. (2011) reported only small
amounts (0 to a maximum of 5%) of base metal sulfides in
A\-#J" eclogite xenoliths in South African kimberlites. PGE mea-
sured in the sulfides in eclogites and associated gabbros
0.001 interpreted to have cycled through a subduction zone have
unusual patterns, being enriched in Ru and Rh, and very low
Pt. Overall levels are higher than those of primitive mantle
and comparable to levels in mid-ocean ridge basalts, which
0.0001 are the probable source of these rocks (MacGregor and Man-
Os Ir Ru Rh Pt Pd Au ton, 1981; Jacob, 2004).
Fig. 10. Concentrations of chalcophile and highly siderophile elements, Metasomatized mantle (MM and CM)
normalized to chondrites, in selected examples of xenoliths from the SCLM
of the Kaapvaal craton (from Maier et al. 2012). The agents of metasomatism of the lithospheric mantle
most probably are volatile-rich, low-volume silicate melts or
carbonate-rich fluids derived from moderate to great depth in
be added analyses of xenoliths from the lithosphere beneath the mantle below the lithosphere (Hawkesworth et al., 1984a;
the Slave (Irvine et al., 2003; Pearson et al., 2003) and Siberian Griffin et al., 2003; Simon et al., 2007). The compositions of
(Ionov et al., 2010) cratons, from the mantle beneath younger the silicate melts likely resemble those of low-degree partial
continental crust (e.g. (Handler and Bennett, 1999), and from melts like meimechites and alkali picrites (Arndt et al., 2004):
subduction settings, the latter representing the composition if so they contain high concentrations of PGE and moderately
of the mantle wedge (McInnes et al. 1999; Ionov, 2010). fractionated patterns. If Mungall et al’s (2002) explanation
Inspection of these data shows that the measured PGE con- for high PGE contents in meimechites and other low-degree
tents are in most cases very low and with few exceptions do melts is correct, these fluids were relatively oxidizing. Data on
not have the patterns predicted for the sources of metals in the PGE contents of carbonatites are sparse but there is little
ore deposits. Most samples in Maier et al.’s (2012) compila- evidence that they contain abnormally high concentrations of
tion (Fig.10) have PGE levels that vary from a tenth to a hun- PGE. Analyses of carbonates in mantle xenoliths have rela-
dredth of those in primitive mantle and the palladium-platinum tively high concentrations and moderately fractionated pat-
group elements (PPGE) are depleted relative to the iridium- terns, but the PGE contents of carbonatites appear to be low
platinum group elements (IPGE). In xenoliths displaying (1 ppt Os, G. Pearson, pers commun.).
1964 NICHOLAS ARNDT

The “subduction” component (SC) The problem of extracting magma from cold
The PGE content of the fluid and/or melt transferred from subcontinental lithospheric mantle
subducting crust into the lithosphere is difficult to establish. As pointed out by Arndt and Christensen (1992) and
Woodland et al. (2002) concluded that the PGE are little emphasized by other authors (Menzies, 1992a; Sharma et al.,
affected by the extraction of fluid from the subducting slab. 1992; Wooden et al., 1993; Saunders et al., 2005), the sub-
If so, this portion of the mantle probably would have PGE continental lithospheric mantle is a highly improbable source
contents similar to that of refractory Archean lithosphere with of voluminous mafic magma. For the lithospheric mantle to
moderate concentrations of IPGE and very low concentra- have survived beneath the continents for billions of years,
tions of the PPGE. Lee (2002) reported these characteristics it must have had low density and high viscosity, characteris-
in xenoliths from western United States. Overall PGE con- tics imparted by its refractory composition, its relatively low
centrations are near chondritic levels, as inferred from stud- temperature, and its low water content. All these characteris-
ies of arc lavas (Mungall, 2002; Jenner et al., 2010). Dale et tics also make it particularly resistant to partial melting. Gal-
al. (2009, 2012) in their study of PGE contents in arc lavas lagher and Hawkesworth (1992) suggested that part of the
found that the PPGE were strongly enriched compared to lithospheric mantle, probably a basal layer, would melt if its
the IPGE; extraction of this melt would leave the residue composition had been modified through the introduction of
extremely depleted in PPGE. McInnes et al. (1999) showed volatiles and other components that reduced its melting tem-
that xenoliths from beneath the Lihir volcano, presumed to perature. They proposed that the Siberian traps and other
represent samples from the subarc lithosphere, are enriched continental flood basalts might have formed when heat from
in Au and PGE. a mantle plume induced melting in a metasomatized layer at
Importantly, if the subduction component does indeed the base of the lithosphere. They further inferred, on the basis
come from the mantle wedge, it will be relatively oxidized and of the trace element contents of continental flood basalts, that
much of the sulfur will be present as sulfate, not sulfide. The the metasomatism had taken place in a subduction setting
absence of sulfide saturation leads to high concentrations of to introduce the so-called subduction component. Likewise,
chalcophile elements in residual liquids (Jenner et al. 2010) other authors (Hatton and Sharpe, 1988) have proposed that
but such magmas will need to be reduced by assimilation of a the sources of the Bushveld Complex and other layered intru-
reductant such as organic matter if sulfide is to segregate at an sions were portions of the SCLM that had been enriched in
early stage of crystallization. subducted material such as detrital sediment. Does this model
stand close scrutiny?
Summary An immediate problem stems from the compositions of
Although there is no doubt that parts of the SCLM are mantle xenoliths, in which the subduction component is very
metasomatized and contain elevated concentrations of incom- poorly represented (Figs. 7−9). In the compilations made
patible trace elements and in some cases volatiles, there is by Ionov et al. (1992, 1997) or Maier et al. (2012) very few
little evidence that these regions are the source of metals in samples show the particular combination of trace element and
magmatic ore deposits. The component of the lithospheric isotopic features thought to characterize the mantle wedge
mantle that is most commonly cited as a source of continen- above a subduction zone; i.e., very high contents of fluid-
tal basalts—ancient subduction-affected lithosphere—is in soluble large-ion lithophile elements (LILE) like Cs, Rb, Pb,
fact very poorly represented in most xenolith suites and in and Sr, intermediate contents of Th and the REE, low con-
peridotite massifs, and it probably forms only a minor part centrations of Nb-Ta and Ti (but not Zr-Hf), high 87Sr/86Sr
of the cratonic lithospheric mantle. Even if ancient mantle and γOs, and moderate εNd. In the xenolith suites, extremely
wedge formed part of the original building blocks of the con- few samples are overenriched in LILE and most do not show
tinental lithosphere, its distinctive geochemical signature has negative Nb-Ta anomalies. Of those that do, the majority are
subsequently been overprinted by the influx of fluids from depleted as well in Zr-Hf and Ti, the tell-tale sign of carbon-
sources deeper in the mantle. The most abundant types of ate metasomatism (Figs. 7−9). These features are absent from
metasomatized SCLM have trace element and isotopic com- almost all continental flood basalts, including those that host
positions that are distinct from those of the parental magmas magmatic ore deposits.
of most magmatic ore deposits (Figs. 7, 9), excluding them There are, however, abundant examples of xenoliths that
as a potential source of these magmas. And finally, there is have been metasomatized by fluids or silicate melts from
little evidence that metasomatized portions of the mantle are deeper in the mantle. These peridotites do contain volatiles
abnormally enriched in chalcophile and highly siderophile in minerals such as amphibole, phlogopite, or carbonate that
elements. would reduce their melting temperatures; material with this
composition would melt preferentially if heat were conducted
The Plausibility of Models Invoking an into the lithosphere from a mantle plume. But would such a
SCLM Source source yield metal-enriched magmas of the type that are said
As mentioned in the Introduction, the validity of models to produce magmatic ore deposits? Here again the answer is
invoking extraction of ore metals from the lithospheric mantle negative, for several reasons. First, even in MARID rocks the
is very rarely discussed; in many if not all cases, an SCLM content of PGE (not to mention Ni, Cu, and Cr) is low. Second,
source is adopted by default. When such models are scruti- magmas issuing from such rocks would have high contents
nized, they are seen to be far less plausible than those invok- of highly to moderately incompatible elements and in terms
ing within-crust processes. of their trace elements (if not their isotopic compositions),
LITHOSPHERIC MANTLE: FORMATION OF ORTHOMAGMATIC ORE DEPOSITS 1965

they would resemble the high Ti suite of continental flood eclogite comprising the Archaean mantle keel of the Kaapvaal
basalts. In regions containing magmatic ore depositions such craton.” Given the unusual Os isotope and trace element char-
as Norilsk-Talnakh in Russia, the low Ti suite, not the high Ti acteristics of Bushveld magmas (Richardson and Shirey, 2008;
suite, contains the ores (Wooden et al., 1993; Lightfoot et al., Fig.6), and the unusual seismic characteristics of lithospheric
1994; Arndt et al., 1998). mantle beneath the complex (Webb et al., 2004), some inter-
Another source of information about compositions in parts action of this type cannot be ruled out. Whether or not the
of the continental lithosphere also provides little support process is more widespread remains uncertain.
for the SCLM model. Analyses of peridotite massifs such as Zhang et al. (2008, p. 167) suggested that “large-volume
Rhonda and Lherz (Reisberg and Zindler, 1987; Garuti et al., plume-derived hot magmas may scavenge osmium and other
1997; Bodinier et al., 2004, 2008) also reveals a large range of PGEs from the cratonic sulfides in old SCLM keels on their
compositions but nothing appropriate as a source of metals way to the surface.” Based on their study of the Ivrea zone
in ore deposits. In the Rhonda massif, for example, incom- of Italy, Fiorentini and Beresford (2008, p. 339) suggested
patible element contents are low, Pb anomalies are negative, that hydrous fluids from a sublithosphere source (a slightly
and most samples lack negative Nb-Ta anomalies (Bodinier hydrated plume?) both created a zone of Ni-Cu-PGE sulfide
et al., 2004,2008). Those samples that do have deficits in enrichment in the SCLM and caused it to melt. In their model,
Nb-Ta are also depleted in Zr-Hf, the signature of carbonatite “The lowering of the peridotite solidus because of increased
metasomatism. water activity caused the harzburgite in the zone percolated
Yet another problem arises from the uniform compositions by fluids to undergo partial melting, thus producing pock-
of flood basalts and other magmas parental to ore depos- ets of volatile-rich sulphide-bearing ultramafic magma that
its (Figs. 5, 6). As pointed out in the section in which these evolved to form independent intrusions.” In other words, they
compositions were discussed, the process that homogenizes assign two separate roles to the fluid: first, it introduces the
magmas in the continental crust operates in a shallow crustal metasomatic components, including the ore metals; second,
setting. The alternative—that the uniform major and trace it causes the peridotite to melt. Whether the two processes
element composition reflects that of the mantle source— can be separated in this way is contentious—if the metals are
is difficult to reconcile with the vast range of compositions introduced in the incoming hydrous fluids, which themselves
documented or inferred in the SCLM. Figure 9 contrasts the are the cause of partial melting, why is it necessary to invoke
restricted range of compositions of flood basalts, which are the presence of metal-rich zones in the SCLM? Could not the
confined to an array between the sublithosphere mantle and metals be transferred directly from source to surface, without
continental crust, with the enormous spread defined by analy- being stored in an SCLM reservoir? Some type of scavenging
ses of various types of xenoliths. of chalcophile and siderophile elements for the SCLM may be
An alternative way to address the problem is to consider possible, but at present the mechanism is very poorly under-
the likely distribution of metasomatized components within stood; the process as a whole requires further investigation.
the SCLM and manner in which they could have been tapped
by invading magma. As mentioned in the section on compo- Alternatives to the SCLM Model
nents in the SCLM, although the source of metasomatizing
fluids is the underlying mantle, these fluids penetrate to mid- Melting of recycled components in mantle plumes
lithospheric depths before solidifying. No matter whether the It is now well established that mantle plumes contain, in
cause of melting was advection of heat from a plume beneath addition to peridotite, a component of recycled material such
the lithosphere (as advocated by Gallagher and Hawkesworth, as oceanic crust or sediment, or fragments from the continen-
1992), invasion of hydrous fluid (Fiorentini and Beresford, tal lithosphere (Hofmann and White, 1982; Sobolev et al.,
2008), or ascent of the lithosphere during rifting, any melting 2007). This material has compositions different from that of
should take place preferentially at the base of the lithosphere. normal convecting mantle and in many respects it is similar to
The degree of melting may vary considerably but on average some of the measured or inferred components of the SCLM
it will be low and, if the composition were constant, it will (White, 1985; Zindler and Hart, 1986; White, 2010). In par-
decrease upward from the base of the lithosphere. The prod- ticular, the presence of recycled sediment, as in the Samoa
uct will be a mixture of melts derived by variable degrees of plume, imparts a “continental” trace element and isotopic sig-
melting of components spanning the entire range illustrated nature, i.e., enriched incompatible elements, a negative Nb
in Figures 7 to 9. The overwhelming characteristic of magma and a positive Pb anomaly, high 87Sr/86Sr, and low 143Nd/144Nd.
from the SCLM will therefore be extreme heterogeneity (as Partial melting of eclogite derived from recycled crust pro-
displayed by lamprophyres and alkalic magmas that demon- duces relatively siliceous magma. Rather than automatically
strably come from the SCLM; Rock, 1991), not the remark- assuming that every enriched mantle source must be in the
able homogeneity that typifies the parental magmas of ore SCLM, some consideration should be given to the possibility
deposits. that such components reside in mantle plumes.
Transport of metals by magma passing through the SCLM Interaction between magmas derived from beneath
Recognizing that the SCLM is an unlikely source of volu- the lithosphere and the continental crust
minous melt, Richardson and Shirey (2008, p. 912) proposed Rehkämper et al. (1999, p. 3925) noted that “the PGE are
that magma from the underlying convecting mantle “heats probably not well suited as geochemical tracers of basalt source
and preferentially removes the low-temperature-melting sul- compositions, because the PGE systematics of melts are so
phide and other metasomatic phases from harzburgite and readily modified during petrogenesis.” They were referring
1966 NICHOLAS ARNDT

to processes such as crystallization, segregation of sulfides, or alkali picrite. An eclogitic component would first convert to
interaction with wall rocks that can radically change the PGE pyroxenite, then undergo more extensive melting to produce
content of magma as it passes through the continental crust. tholeiitic picrite. These magmas, which are well represented
The key mechanism was outlined in an insightful paper by in continental flood volcanic provinces, contain high contents
Kerr and Leitch (2005) who described how the chalcophile of Ni, Cu, and the PGE.
element contents change dramatically as sulfides first accu- 2. There is ample evidence that such magmas interact
mulate then redissolve in an incoming flux of S-undersatu- with and assimilate continental crust as they ascend toward
rated magma. Through this process, the PGE contents in the surface. Indeed, it is difficult to imagine how these dense,
the incoming magma may reach high levels. Li et al. (2009) hot magmas could escape contamination. Several examples of
adopted these ideas to explain the high PGE tenors in ore of this interaction are cited in the earlier sections where crustal
the Norilsk-Talnakh deposits. Other authors (Campbell et al., contamination is discussed (see Figs. 5, 6). The only types of
1983; Czamanske et al., 1995; Beresford et al., 2002; Lesher magma that may escape contamination are volatile-charged
and Keays, 2002; Naldrett, 2004) have emphasized that ore magmas such as those that form by low degrees of partial
formation is essentially a matter of segregation of sulfide liq- melting of enriched sources, as is the case for magmas from
uid or oxides followed by remobilization of these materials metasomatized SCLM. The dissolved volatiles reduce signifi-
and upgrading of the metal contents of the sulfide liquid (i.e., cantly the density of the magma, which renders it less likely to
the tenor) through reaction with incoming magmas. In most be trapped at density discontinuities at the base of, or within,
cases the variability in tenor can be explained in terms of mass the crust (Arndt et al., 1998). In addition, the strong pressure
balance between silicate and sulfide melt in different physical dependence of CO2 solubility causes this phase to exsolve at
or dynamic settings, (R factors, N factors; Campbell and Nal- lower crustal depths, creating bubbles that propel the magma
drett, 1979; Lesher and Burnham, 1999; Naldrett, 2008) and to the surface. Such volatile-rich magmas, exemplified by the
that this source of variance swamps the variability in magma barren Gudchikhin picrites and meimechites from the Sibe-
compositions. rian volcanic province, only rarely contain ore deposits (Arndt
In most of these models, the source of the magma is con- et al., 2004).
sidered to be a mantle plume. In the view of many authors, 3.  Segregation of immiscible sulfide liquid, a crucial step in
there is no other credible way of producing a large volume of the formation of most orthomagmatic ore deposits, is trig-
mafic magma, including highly magnesian picrite or komati- gered by interaction with crustal rocks (Naldrett, 2008). Two
ite, in a short period in time, as was the case for the Siberian processes—the incorporation of sulfur from sedimentary
traps (Campbell et al., 1992) or the Bushveld Complex (Caw- rock and assimilation of siliceous material, which decreases
thorn, 1998; Wilson, 2012). Objections to plume models, for the sulfide solubility—play a role. In many examples, the
example, a lack of uplift preceding magmatism (Czamanske metal contents of the ore may be upgraded as magma moves
et al., 1998) or questions about the quantity of melt that can through the conduit system. Both crustal assimilation and
be extracted from a source at sublithospheric depths (e.g., upgrading operate most efficiently when the magma is hot
Anderson et al., 1992) are countered by recent work in which and fluid. With the exception of some extreme examples of
plumes are shown to contain a high proportion of recycled PGE reef deposits, almost the entire range of PGE, Ni, and
oceanic crust (Cordery et al., 1997; Sobolev et al., 2011). For Cu contents in magmatic sulfide are compatible with varia-
other Ni-Cu sulfide deposits, such as Jinchuan in China, the tions in the R factor coupled with normal magmatic pro-
tectonic context is poorly known. cesses. Such processes operate efficiently in high-temperature,
low-viscosity, volatile-poor magma, but not for volatile-rich
The Standard Model: Revisited magmas, which are transformed by the loss of their volatiles
In previous sections, the case was made that ore metals during approach to the surface to phenocryst-charged viscous
in orthomagmatic deposits are rarely if ever derived from mush.
sources in the SCLM model. It now remains to be shown that
the standard model provides a better explanation. Consider Conclusions
the following arguments: Very little evidence supports models that advocate an
SCLM source of the metals in magmatic ore deposits. In
1.  There is abundant evidence that large volumes of mag- many papers, an SCLM source is assumed by default, with
mas come from sources in the asthenosphere or in mantle no consideration of the plausibility of such a model. Examina-
plumes. The compositions of these magmas are known within tion of the probable compositions in the subcontinental litho-
reasonable limits—mid-ocean ridge basalt provides an esti- spheric mantle reveals a paucity of compositions resembling
mate of the composition of melts from ambient upper mantle those of ore-forming magmas. In addition, the processes that
and basalts, and picrites from localities such as Hawaii, Reun- extract metals from a putative SCLM source are very poorly
ion, and Gorgona give the compositions of magmas produced constrained.
by moderate to high degrees of melting in mantle plumes. Far more plausible are models in which the high concen-
Using constraints from experimental petrology we can infer trations of chalcophile and siderophile elements in orthom-
the compositions of melts produced in similar sources but agmatic ores result from interaction between magma from
at greater depths beneath the continental lithosphere. At sublithospheric sources and the continental crust. This inter-
high pressures, peridotite from a hot source like that which action results in the segregation and mobilization of sulfide, a
formed the Gorgona komatiites would undergo low to mod- process that under favorable conditions is capable of generat-
erate degrees of melting and would yield meimechite or ing large quantities of high-grade ore.
LITHOSPHERIC MANTLE: FORMATION OF ORTHOMAGMATIC ORE DEPOSITS 1967

Acknowledgments Cawthorn, R.G., 1998, Emplacement and crystallization time for the Bush-
veld Complex: Journal of Petrology, v. 39, p. 1669−1687.
The research reported here was supported indirectly by Chai, G., and Naldrett, A.J., 1992, Characteristics of Ni-Cu-PGE minerali-
grants from the French Agence Nationale de Recherche, zation and genesis of the Jinchuan deposit, northwest China: Economic
project M&Ms; the American National Research Foundation, Geology, v. 87, p. 1475−1495.
Cordery, M.J., Davies, G.F., and Campbell, I.H., 1997, Genesis of flood
Continental Geodynamics program; the French Centre basalts from eclogite-bearing mantle plumes: Journal of Geophysical
national de la recherche scientifique, and the Université Research, v. 102, p. 20,179−20,197.
Joseph Fourier in Grenoble. Special thanks go to the two Czamanske, G.K., Zen’ko, T.E., Fedorenko, V.A., Calk, L.C., Budahn, J.R.,
journal reviewers, Marco Fiorentini and Graham Pearson, for Bullock, J.H., Jr., Fries, T.L., King, B.S., and Siems, D.F., 1995, Petrog-
their very useful and constructive comments and to the journal raphy and geochemical characterization of ore-bearing intrusions of the
Noril’sk type, Siberia; with discussion of their origin: Resource Geology
editors, Larry Meinert and Steve Barnes for their assistance. Special Issue 18, p. 1−48.
Czamanske, G.K., Gurevitch, Fedorenko, V.A., and Simonov, V., 1998,
REFERENCES Demise of the Siberian plume: Paleogeographic and paleotectonic recon-
Anderson, D.L., Zhang, Y.-S., and Tanimoto, T., 1992, Plume heads, conti- struction from the prevolcanic and volcanic record, north-central Siberia:
nental lithosphere, flood basalts and tomography: London, Geological Soci- International Geology Review, v. 40, p. 95−115.
ety Special Publication 68, p. 99−124. Dale, C., Macpherson, C.G., Pearson, D.G., Hammond, S.J., and Arculus,
Arndt, N.T., 2011, Insights into the geological setting and origin of Ni-Cu R.J., 2012, Inter-element fractionation of highly siderophile elements in the
(PGE) sulfide deposits of the Norilsk-Talnakh region, Siberia: Reviews in Tonga arc due to flux melting of a depleted source: Geochimica et Cosmo-
Economic Geology, v. 17, p. 199–215. chimica Acta, v. 89, p. 202−225.
Arndt, N.T., and Christensen, U., 1992, Role of lithospheric mantle in con- Dale, C.W., Burton, K.W., Pearson, D.G., Gannoun, A., Alard, O., Argles,
tinental volcanism: Thermal and geochemical constraints: Journal of Geo- T.W., and Parkinson, I.J., 2009, Highly siderophile element behavior
physical Research, v. 97, p. 10967−10981. accompanying subduction of oceanic crust: Whole rock and mineral-scale
Arndt, N.T., Chauvel, C., Fedorenko, V., and Czamanske, G., 1998, Two man- insights from a high-pressure terrain: Geochimica et Cosmochimica Acta,
tle sources, two plumbing systems: Tholeiitic and alkaline magmatism of v. 73, p. 1394–1416.
the Maymecha River basin, Siberian flood volcanic province: Contributions De Waal, S.A., Xu, Z., Li, C., and Mour, H., 2004, Emplacement of viscous
to Mineralogy and Petrology, v. 133, p. 297−313. mushes in the Jinchuan ultramafic intrusion, western China: Canadian Min-
Arndt, N.T., Czamanske, G.K., Walker, R.J., Chauvel, C., and Fedorenko, eralogist, v. 42, p. 371−392.
V.A., 2003, Geochemistry and origin of the intrusive hosts of the Noril’sk- DePaolo, D.J., 1981, Trace element and isotopic effects of combined wall
Talnakh Cu-Ni-PGE deposits: Economic Geology, v. 98, p. 495−515. rock assimilation and fractional crystallisation: Earth and Planetary Science
Arndt, N.T., Lesher, C.M., and Czamanske, G.K., 2004, Mantle-derived Letters, v. 53, p. 189−202.
magmas and magmatic Ni-Cu-(PGE) deposits: Economic Geology 100th Eiler, J., 2003, Inside the subduction factory: Washington, D.C., AGU Geo-
Aniversary volume, p. 5−24. physical Monogram Series, v. 138, 311 p., doi:10.1029/GM138.
Arndt, N.T., Barnes, S.J., and Lesher, C.M., 2008, Komatiite: Cambridge, Erlank, A.J., Waters, F.G., Hawkesworth, C.J., Haggerty, S.E., Allsopp, H.L.,
Cambridge University Press, 487 p. Rickard, R.S., and Menzies, M.A., 1987, Evidence of mantle metasomatism
Barnes, S.J., and Campbell, I.H., 1988, Role of late magmatic fluids in Mer- in peridotite nodules from the Kimberley Pipes, South Africa, in Menzies,
ensky-type platinum deposits: A discussion: Geology, v. 16, p. 488−490. M.A., and Hawkesworth, C.J., eds., Mantle metasomatism: London, Aca-
Barnes, S.-J., Maier, W.D., and Curl, E.A., 2010, Composition of the mar- demic Press, p. 221−312.
ginal rocks and sills of the Rustenburg Layered Suite, Bushveld Complex, Fedorenko, V., Lightfoot, P.C., Naldrett, A.J., Czamanske, G.K., Hawkes-
South Africa: Implications for the formation of the platinum-group element worth, C.J., Wooden, J.L., and Ebel, D.S., 1996, Petrogenesis of the Sibe-
deposits: Economic Geology, v. 105, p. 1491−1511. rian Flood Basalt sequence at Noril’sk: International Geology Review, v. 38,
Begg, G.C., Hronsky, J.A.M., Arndt, N.T., Griffin, W.L., O’Reilly, S.Y., and p. 99−135.
Hayward, N., 2011, Lithospheric, cratonic, and geodynamic setting of Ni- Fiorentini, M.L., and Beresford, S.W., 2008, Role of volatiles and meta-
Cu-PGE sulfide deposits: Economic Geology, v. 105, p. 1057–1070. somatized subcontinental lithospheric mantle in the genesis of magmatic
Beresford, S., Cas, R., Lahaye, Y., and Jane, M., 2002, Facies architecture of Ni-Cu-PGE mineralization: Insights from in situ H, Li, B analyses of hydro-
an Archean komatiite-hosted Ni-sulphide ore deposit, Victor, Kambalda, magmatic phases from the Valmaggia ultramafic pipe, Ivrea-Verbano zone
Western Australia: Implications for komatiite lava emplacement: Journal of (NW Italy): Terra Nova, v. 20, p. 333−340.
Volcanology and Geothermal Research, v. 118, p. 57−75. Fiorentini, M.L., Beresford, S.W., Deloule, E., Hanski, E., Stone, W.E., and
Bodinier, J.L., Menzies, M., Shimizu, N., Frey, F.A., and McPherson, E., Pearson, N.J., 2008, The role of mantle-derived volatiles in the petrogenesis
2004, Silicate, hydrous and carbonate metasomatism at Lherz, France: of Palaeoproterozoic ferropicrites in the Pechenga greenstone belt, north-
Contemporaneous derivatives of silicate melt—harzburgite reaction: Jour- western Russia: Insights from in-situ microbeam and nanobeam analysis of
nal of Petrology, v. 45, p. 299−320. hydromagmatic amphibole: Earth and Planetary Science Letters, v. 268,
Bodinier, J.L., Garrido, C.J., Chanefo, I., Brigiier, O., and Gervilla, F., 2008, p. 2−14.
Origin of pyroxenite/peridotite veined mantle by refertilization reactions: Fiorentini, M.L., Barnes, S.J., Maier, W.D., Burnham, O.M., and Heggie, G.,
Evidence from the Ronda Peridotite (southern Spain): Journal of Petrology, 2011, Global variability in the platinum-group element contents of komati-
v. 49, p. 999−1025. ites: Journal of Petrology, v. 52, p. 83−112.
Bohrson, W.A., and Spera, F., 2001, Energy-constrained open system mag- Fiorentini, M.L., Beresford, S.W., Stone, W.E., and Deloule, E., 2012, Evi-
matic processes. II: Application of energy constrained assimilation/frac- dence of water degassing during emplacement and crystallization of 2.7
tional crystallization (EC-AFC) model to magmatic systems: Journal of Ga komatiites from the Agnew-Wiluna greenstone belt, Western Australia:
Petrology, v. 42, p. 1019−1041. Contributions to Mineralogy and Petrology, v. 164, p. 143−155.
Boyd, F.R., and Mertzman, S.A., 1987, Composition and structure of the Foley, S.F., 2008, Rejuvenation and erosion of the cratonic lithosphere:
Kaapvaal lithosphere, southern Africa: Geochemical Society Special Pub- Nature Geoscience, v. 1, p. 503−510.
lication 1, p. 13−24. Gallagher, K., and Hawkesworth, C., 1992, Dehydration melting and the gen-
Campbell, I.H., and Naldrett, A.J., 1979, The influence of silicate:sulfide eration of continental flood basalts: Nature, v. 258, p. 57−59.
ratios on the geochemistry of magmatic sulfides: Economic Geology, v. Garuti, G., Oddone, M., and Torres-Ruiz, J., 1997, Platinum-group element
74, p. 1503−1505. distribution in subcontinental mantle: Evidence from the Ivrea-zone (Italy)
Campbell, I.H., Naldrett, A.J., and Barnes, S.J., 1983, A model for the origin and the Betic-Rifean cordillera (Spain and Morocco): Canadian Journal of
of the platinum-rich sulfide horizons in the Bushveld and Stillwater Com- Earth Sciences, v. 34, p. 444−463.
plexes: Journal of Petrology, v. 24, p. 133−165. Godel, B., Seat, Z., Maier, W.D., and Barnes, S.-J., 2011, The Nebo-Babel Ni-
Campbell, I.H., Czamanske, G.K., Fedorenko, V.A., Hill, R.I., and Stepanov, Cu-PGE sulfide deposit (West Musgrave block, Australia): Pt. 2. Cnstraints
V., 1992, Synchronism of the Siberian Trap s and the Permian-Triassic on parental magma and processes, with implications for mineral explora-
boundary: Science, v. 258, p. 1760−1763. tion: Economic Geology, v. 106, p. 557−584.
1968 NICHOLAS ARNDT

Gréau, Y., Huang, J., Griffin, W.L., Renac, C., Alard, O., and O’Reilly, S.Y., Ionov, D.A., Kramm, U., and Stosch, H.-G., 1992, Evolution of the upper
2011, Type I eclogites from Roberts Victor kimberlites: Products of extensive mantle beneath the southern Baikal rift zone: A Sr-Nd isotope study of
mantle metasomatism: Geochimica et Cosmochica Acta, v. 75, p. 6927−6954. xenoliths from the Bartoy volcanoes: Contributions to Mineralogy and
Grégoire, M., Bell, D.R., and Le Roex, A.P., 2002, Trace element geochem- Petrology, v. 111, p. 235−247.
istry of glimmerite and MARID mantle xenoliths: Their classification and Ionov, D.A., Griffin, W.L., and O’Reilly, S.Y., 1997, Volatile-bearing minerals
relationship to phlogopite-bearing peridotites and to kimberlites revisited: and lithophile trace elements in the upper mantle: Chemical Geology, v.
Contributions to Mineralogy and Petrology, v. 142, p. 603−625. 141, p. 153−184.
——2003, Garnet lherzolites from the Kaapvaal craton (South Africa): Trace Ionov, D.A., Bodinier, J.L., Mukasa, S.B., and Zanetti, A., 2002, Mechanisms
element evidence for a metasomatic history: Journal of Petrology, v. 44, p. and sources of mantle metasomatism: Major and trace element composi-
629−657. tions of peridotite xenoliths from Spitsbergen in the context of numerical
Griffin, W.L., O’Reilly, S.Y., Ryan, C.G., Gaul, I., and Ionov, D., 1998, Secular modelling: Jounal of Petrology, v. 43, p. 2219−2259.
variation in the composition of the subcontinental lithospheric mantle, in Ionov, D., Doucet, L.S., and Ashchepkov, I.V., 2010, Composition of the lith-
Brann, J., Dooley, J.C., Goleby, B.R., van der Hilst, R.D., and Klootwijk, ospheric mantle in the Siberian craton: New constraints from fresh peri-
C.T., eds., Structure and evolution of the Australian continent: Washington, dotites in the Udachnaya-East Kimberlite: Journal of Petrology, v. 51, p.
Americal Geophysical Union, p. 1−26. 2177−2210.
Griffin, W.L., O’Reilly, S.Y., and Ryan, C.G., 1999, The composition and Irvine, G.J., Pearson, D.G., Carlson, R.W., Kjarsgaard, B.A., and Dreibus,
origin of subcontinental lithospheric mantle, in Fei, Y., Bertka, C.M., and G., 2003, Evolution of the lithospheric mantle beneath northern Canada:
Mysen, B.O., eds., Mantle petrology: Field observations and high-pressure A Re-Os isotope and PGE study of kimberlite-derived peridotite xenoliths
experimentation: Houston, TX, Geochemical Society, p. 13−46. from Somerset Island and a comparison to the Slave and Kaapvaal cratons:
Griffin, W.L., O’Reilly, S.Y., Abe, N., Aulbach, S., Davies, R.M., Pearson, Lithos, v. 71, p. 461−488.
N.J., Doyle, B.J., and Kivi, K., 2003, The origin and evolution of Archean Jacob, D.E., 2004, Nature and origin of eclogite xenoliths from kimberlites:
lithospheric mantle: Precambrian Research, v. 127, p. 19−41. Nature, v. 77, p. 295−316.
Griffin, W.L., O’Reilly, S.Y., Afonso, J.C., and Begg, G.C., 2009, The composi- Jenner, F., O’Neill, H.S.C., Arculus, R.J., and Mavrogenes, J.A., 2010, The
tion and evolution of lithospheric mantle: A re-evaluation and its tectonic magnetite crisis in the evolution of arc-related magmas and the initial con-
implications.: Journal of Petrology, v. 50, p. 1185–1204. centration of Au, Ag and Cu: Journal of Petrology, v. 51, p. 2445−2464.
Handler, M.R., and Bennett, V.C., 1999, Behavior of platinum-group ele- Jourdan, F., Bertrand, H., Schärer, U., Blichert-Toft, J., Féraud, G., and Kam-
ments in the subcontinental mantle of eastern Australia during variable punzu, A.B., 2007, Major and trace element and Sr, Nd, Hf, and Pb isotope
metasomatism and melt depletion: Geochimica et Cosmochimica Acta, v. compositions of the Karoo large igneous province, Botswana-Zimbabwe:
63, p. 3597–3618. Lithosphere vs. mantle plume contribution: Jounal of Petrology, v. 48, p.
Hart, S.R., and Kinloch, E.D., 1989, Osmium isotope systematics in Wit- 1043−1077.
watersrand and Bushveld ore deposits: Economic Geology, v. 84, p. Keays, R.R., 1982, Palladium and iridium in komatiites and associated rocks:
1651–1655. Application to petrogenetic problems, in Arndt, N.T., and Nisbet, E.G.,
Hatton, C.J., and Sharpe, M.R., 1988, Significance and origin of boninite-like eds., Komatiites: London, George Allen and Unwin, p. 435−458.
rocks associated with the Bushveld Complex, in Crawford, A.J., ed., Bonin- ——1995, The role of komatiite and picritic magmatism and S-saturation in
ites: London, Unwyn Hyman, p. 174−207. the formation of ore deposits: Lithos, v. 34, p. 1−18.
Hawkesworth, C.J., Marsh, J.S., Duncan, A.R., Erlank, A.J., and Norry, M.J., ——2012, The role of the interaction between asthenospheric and litho-
1984a, The role of continental lithosphere in the generation of the Karoo spheric mantles in the generation of giant Ni-Cu-PGE sulfide deposits
volcanic rocks: Evidence from combined Nd- and Sr- isotope studies: Geo- [abs.]: International Ni-Cu-(PGE) Symposium, 12th, Guiyang, China,
logical Society of South Africa Special Publication 13, p. 341−354. Abstracts, p. 3−6.
Hawkesworth, C.J., Rogers, N.W., Van Calsteren, P.W.C., and Menzies, M.A., Kerr, A., and Leitch, A.M., 2005, Self-destructive sulfide segregation systems
1984b, Mantle enrichment processes: Nature, v. 311, p. 331−336. and the formation of high-grade magmatic ore deposits: Economic Geol-
Hawkesworth, C., Mantovani, M., and Peate, D., 1988, Lithosphere remo- ogy, v. 100, p. 311−332.
bilization during Parana CFB magmatism.: Journal of Petrology, Special Kerrich, R., Goldfarb, R., Groves, D., Garwin, S., and Jia, Y., 2000, The char-
Lithosphere Issue, p. 205−223. acteristics, origins, and geodynamic settings of supergiant gold metallogenic
Hawkesworth, C.J., Gallagher, K., Hergt, J.M., and McDermott, F., 1994, provinces: Science in China, Series D, Earth Sciences, v. 43, p. 1−68.
Destructive plate margin magmatism: Geochemistry and melt generation: Kieffer, B., Arndt, N.T., and others, 2004, The transition from plateau to
Lithos, v. 33, p. 169−188. shield volcanism in Ethiopia: A petrological and geochemical study: Journal
Hergt, J.M., Peate, D.W., and Hawkesworth, C.J., 1991, The petrogenesis of Petrology, v. 45, p. 793−834.
of Mesozoic Gondwana low-Ti flood basalts: Earth and Planetary Science Konc, Z., Marchesi, C., Hidas, K., Garrido, C.J., Szabó, C., and Sharygin,
Letters, v. 105, p. 134−148. V.V., 2012, Structure and composition of the subcontinental lithospheric
Herzberg, C., 1999, Phase equilibrium constraints on the formation of cra- mantle beneath the Sangilen plateau (Tuva, southern Siberia, Russia): Evi-
tonic mantle, in Fei, Y., Bertka, C.M., and Mysen, B.O., eds., Mantle petrol- dence from lamprophyre-hosted spinel peridotite xenoliths: Lithos, v. 146,
ogy: Field observations and high-pressure experimentation: Houston, TX, p. 253−263.
Geochemical Society, p. 13−46. Lassiter, J.C., and DePaolo, D.J., 1997, Plume/lithosphere interaction in the
Hofmann, A.W., 1988, Chemical differentiation of the Earth: The relation- generation of continental and oceanic flood basalts: Chemical and isotopic
ship between mantle, continental crust, and oceanic crust: Earth and Plan- constraints, in Mahoney, J.J., and Coffin, M.F., eds., Large igneous prov-
etary Science Letters, v. 90, p. 297−314. inces: Continental, oceanic and planetary flood volcanism: Washington,
Hofmann, A.W., and White, W.M., 1982, Mantle plumes from ancient oce- American Geophysical Union, p. 335−356.
anic crust: Earth and Planetary Science Letters, v. 57, p. 421−436. Lee, C.-T.A., 2002, Platinum-group element geochemistry of peridotite xen-
Hong, D., Wang, S., Xie, X., Zhang, J., and Wang, T., 2002, Metallogenic oliths from the Sierra Nevada and the Basin and Range, California: Geo-
province derived from mantle sources: Nd, Sr, S and Pb isotope evidence chimica et Cosmochica Acta, v. 66, p. 3987–4005.
from the Central Asian orogenic belt: Gondwana Research, v. 6, p. 711–728. Lehmann, J., Arndt, N.T., Windley, B., Zhou, M.F., Wang, C.Y., and Harris,
Hronsky, J.M.A., Groves, D.I., Loucks, R.R., and Begg, G.C., 2012, A unified C., 2007, Field relationships and geochemical constraints on the emplace-
model for gold mineralisation in accretionary orogens and implications for ment of the Jinchuan intrusion and its Ni-Cu-PGE sulfide deposit, Gansu,
regional-scale exploration targeting methods: Mineralium Deposita, v. 47, China: Economic Geology, v. 102, p. 75−94.
p. 339−358. Lesher, C.E., and Burnham, O.M., 1999, Mass balance and mixing in dynam-
Ionov, D., 2010, Petrology of mantle wedge lithosphere: New data on icore-forming magmatic systems, in Keays, R.R., Lesher, C.E., Lightfoot,
supra-subduction zone peridotite xenoliths from the andesitic Avachavol- P.C., and Farrow, C.E.G., eds., Dynamic processes in magmatic ore depos-
cano, Kamchatka: Journal of Petrology, v. 51, p. 327−362. its and their application to mineral exploration: Ottawa, Geological Associa-
Ionov, D., and Harmer, R.E., 2002, Trace element distribution in calcite- tion of Canada, Short Course, p. 413−449.
dolomite carbonatites from Spitskop: Inferences for differentiation of car- Lesher, C.M., and Keays, R.R., 2002, Komatiite-associated Ni-Cu-PGE
bonatite magmas and the origin of carbonates in mantle xenoliths: Earth deposits: Canadian Institute of Mining, Metallurgy Petroleum Special Vol-
and Planetary Science Letters, v. 198, p. 495−510. ume 54, p. 579–617.
LITHOSPHERIC MANTLE: FORMATION OF ORTHOMAGMATIC ORE DEPOSITS 1969

Li, C., and Naldrett, A.J., 1999, Geology and petrology of the Voisey’s Bay Menzies, M.A., and Hawkesworth, C.J., 1987, Mantle metasomatism: Lon-
intrusion: Reaction of olivine with sulfide and silicate liquids: Lithos, v. 47, don, Academic Press, 472 p.
p. 1−31. Mitchell, R.H., and Keays, R.R., 1981, Abundance and distribution of gold,
Li, C., Ripley, E.M., and Naldrett, A.J., 2009, A new genetic model for the palladium and iridium in some spinel and garnet lherzolites: Implications
giant Ni-Cu-PGE sulfide deposits associated with the Siberian flood basalts: for the nature and origin of precious metal, sulphur-rich intragranular com-
Economic Geology, v. 104, p. 291−301. ponents in the Upper Mantle: Geochimica et Cosmochimica Acta, v. 45, p.
Lightfoot, P.C., and Hawkesworth, C.J., 1997, Flood basalts and magmatic 2425−2442.
Ni, Cu, and PGE sulphide mineralization: Comparative geochemistry of Molzahn, M., Resiberg, L., and Wörner, G., 1996, Os, Sr, Nd, Pb, O isotope
the Noril’sk (Sinerian Trap s) and West Greenland sequences, in Mahoney, and trace element data from the Ferrar flood basalts, Antarctica: Evidence
J.J., and Coffin, M.F., eds., Large igneous provinces: Continental, oceanic for an enriched subcontinental mantle: Earth and Planetary Science Let-
and planetary flood volcanism: Washington, American Geophysical Union, ters, v. 144, p. 529−546.
p. 357−380. Mungall, J.E., 2002, Roasting the mantle: Slab melting and the genesis of
Lightfoot, P.C., and Naldrett, A.J., 1994, Proceedings of the Sudbury-Noril’sk major Au and Au-rich Cu deposits: Geology, v. 30, p. 915−918.
Symposium: Sudbury, Ministry of Northern Development and Mines, p. ——2007, Crustal contamination of picritic magmas during transport through
421. dikes: the Expo Intrusive Suite, Cape Smith fold belt, New Quebec: Journal
Lightfoot, P.C., Naldrett, A.J., Gorbachev, N.S., and Fedorenko, V.A., 1990, of Petrology, p. 1021−1039.
Geochemistry of the Siberian Trap of the Noril’sk area, USSR, with impli- Naldrett, A.J., 1992, A model for the Ni-Cu-PGE ores of the Noril’sk region
cations for the relative contributions of crust and mantle to flood basalt and its application to other areas of flood basalt: Economic Geology, v.
magmatism: Contributions to Mineralogy and Petrology,v. 104, p. 631−644. 87, p. 1945−1962.
Lightfoot, P.C., Hawkesworth, C.J., Hergt, J., Naldrett, A.J., Gorbachev, N.S., ——1999, World class Ni-Cu-PGE deposits: Key factors in their genesis:
Fedorenko, V.A., and Doherty, W., 1993, Remobilization of the continen- Mineralium Deposita, v. 34, p. 227–240.
tal lithosphere by mantle plumes: Major-, trace-element and Sr-, Nd-, and ——2004, Magmatic sulfide deposits: Geology, geochemistry and explora-
Pb-isotope evidence from picritic and tholeiitic lavas of the Noril’sk district tion: Heidelberg, Berlin, Springer Verlag, 728 p.
Siberian Trap, Russia: Contributions to Mineralogy and Petrology, v. 114, Pearce, J.A., Stern, R.J., Bloomer, S.H., and Fryer, P., 2005, Geochemical
p. 171−188. mapping of the Mariana arc-basin system: Implications for the nature and
Lightfoot, P.C., Naldrett, A.J., Gorbachev, N.S., Fedorenko, V.A., Hawkes- distribution of subduction components: Geochemistry, geophysics, geosys-
worth, C.J., Hergt, J., and Doherty, W., 1994, Chemostratigraphy of Sibe- tems, v. 6, no. 7, Q07006, doi: 10.1029/2004GC000895.
rian Trap lavas, Noril’sk district, Russia: Implications for the source of flood Pearson, D.G., and Nowell, G.M., 2002, The continental lithospheric mantle:
basalt magmas and their associated Ni-Cu mineralization, in Naldrett, A.J., Characteristics and significance as a mantle reservoir: Proceedings of the
and Lightfoot, P.C., eds., Proceedings of the Sudbury-Noril’sk Symposium: Royal Society, Series A., v. 360, p. 2383–2410.
Sudbury, Ontario Geological Survey, p. 283−312. Pearson, D.G., Canil, D., and Shirey, S.B., 2003, Mantle samples included
Lorand, J.P., Alard, O., Luguet, A., and Keays, R.R., 2003, Sulfur and sele- in volcanic rocks: Xenoliths and diamonds: Treatise on Geochemistry, v. 2,
nium systematics of the subcontinental lithospheric mantle: Inferences
p. 171–275.
from the Massif Central xenolith suite (France): Geochimica et Cosmo-
Pearson, D.G., Irvine, G.J., Ionov, D.A., Boyd, F.R., and Dreibus, G.E., 2004,
chimica Acta, v. 67, p. 4137−4151.
Re-Os isotope systematics and platinum group element fractionation dur-
Maaloe, S., and Aoki, K.-I., 1977, The major element composition of the
ing mantle melt extraction : A study of massif and xenolith peridotite suites:
upper mantle estimated from the composition of lherzolites: Contributions
Chemical Geology, v. 208, p. 29−59.
to Mineralogy and Petrology, v. 63, p. 161−173.
Quick, J.E., Sinigoi, S., Peressini, G., Demarchi, G., Wooden, J.L., and Sbisà,
MacGregor, I.D., and Manton, W.I., 1981, The Roberts Victor eclog-
A., 2009, Magmatic plumbing of a large Permian caldera exposed to a depth
ites: Ancient oceanic crust: Journal of Geophysical Research, v. 91, p.
of 25 km: Geology, v. 37, p. 603−606.
14,063−14,079.
Maier, W.D., and Barnes, S.-J., 2004, Pt/Pd and Pd/Ir ratios in mantle-derived Rehkämper, M., Halliday, A.N., Fitton, J.G., Lee, D.-C., Wieneke, M., and
magmas: A possible role for mantle metasomatism: South African Journal of Arndt, N.T., 1999, Ir, Ru, Pt, and Pd in basalts and komatiites: New con-
Geology, v. 107, p. 333−340. straints for the geochemical behavior of the platinum-group elements in the
Maier, W.D., Arndt, N.T., and Curl, E.A., 2000, Progressive crustal contami- mantle: Geochimica et Cosmochimica Acta, v. 63, p. 3915−3934.
nation of the Bushveld Complex: Evidence from Nd isotopic analyses of Reisberg, L., and Zindler, A., 1987, Extreme isotopic variations in the upper
the cumulate rocks: Contributions to Mineralogy and Petrology, v. 140, p. mantle: Evidence from Ronda: Earth and Planetary Science Letters, v. 81,
328−343. p. 29−45.
Maier, W.D., Peltonen, P., McDonald, L., Barnes, S.-J., Hatton, C., and Vil- Richardson, S.H., and Shirey, S.B., 2008, Continental mantle signature of
joen, F., 2012, The concentration of platinum-group elements and gold in Bushveld magmas and coeval diamonds: Nature, v. 453, doi:10.1038/
southern African and Karelian kimberlite-hosted mantle xenoliths: Implica- nature07073.
tions for the noble metal content of the Earth’s mantle: Chemical Geology, Rock, N.M.S., 1991, Lamprophyres: Glasgow, Blackie, 285 p.
v. 302, p. 119–125. Rudnick, R.L., and Fountain, D.M., 1995, Nature and composition of the
Mavrogenes, J.A., and O’Neill, C., 1999, The relative effects of pressure, tem- continental crust: A lower crust perspective: Reviews in Geophysics, v. 33,
perature and oxygen fugacity on the solubility of sulfide in mafic magmas: p. 267−309.
Geochimica et Cosmochimica Acta, v. 63, p. 1173−1180. Saunders, A.D., England, R.W., Reichow, M.K., and White, R.V., 2005, A
McDonough, W.F., 1990, Constraints on the composition of the continental mantle plume origin for the Siberian Traps: Uplift and extension in the
lithospheric mantle: Earth and Planetary Science Letters, v. 101, p. 1−18. West Siberian basin, Russia: Lithos, v. 79, p. 407−424.
McInnes, B.I.A., McBride, J.S., Evans, N.J., Lambert, D.D., and Andrew, Sharma, M., Basu, A.R., and Nesterenko, G.V., 1992, Temporal Sr-, Nd- and
A.S., 1999, Osmium isotope constraints on ore metal recycling in subduc- Pb-isotopic variations in the Siberian flood basalts: Implications for the
tion zones: Science, v. 286, DOI: 10.1126/science.286.5439.512. plume-source characteristics: Earth and Planetary Science Letters, v. 113,
McKenzie, D.P., 1989, Some remarks on the movement of small melt frac- p. 365−381.
tions in the mantle: Earth and Planetary Science Letters, v. 95, p. 53−72. Sharpe, M.R., 1982, Noble metals in the marginal rocks of the Bushveld
Menzies, M.A., 1990a, Archaean, Proterozoic and Phanerozoic lithospheres, Complex: Economic Geology, v. 77, p. 1286−1295.
in Menzies, M.A., ed., Continental mantle: Oxford, Clarendon, p. 67−86. Shirey, S.B., 1997, Re-Os isotopic composition of Mid-continent Rift Sys-
——1990b, Continental mantle: Oxford, Oxford University Press, 184 p. tem picrites and tholeiites: Implications for plume lithosphere interaction
——1992a, The lower lithosphere as a major source for continental flood and enriched mantle sources: Canadian Journal of Earth Sciences, v. 34, p.
basalts: A re-appraisal: London, Geological Society Special Publication 68, 489−503.
p. 31−39. Simon, N.S.C., Carlson, R.W., Pearson, D.G., and Davies, G.R., 2007, The
——1992b, The lower lithosphere as a major source for continental flood origin and evolution of the Kaapvaal cratonic lithospheric mantle: Journal
basalts: A re-appraisal, in Storey, B.C., Alabaster, T., and Pankhurst, R.J., of Petrology, v. 48, p. 589−625.
eds., Magmatism and the causes of continental break-up: London, Geologi- Sinigoi, S., Antonini, P., Demarchi, G., Longinelli, A., Mazzuchelli, M.,
cal Society, p. 33−41. Negrini, L., and Rivalenti, G., 1991, Interactions of mantle and crustal
1970 NICHOLAS ARNDT

magmas in the southern part of the Ivrea zone (Italy): Contributions to White, W.M., 1985, Sources of oceanic basalts: Radiogenic isotopic evidence:
Mineralogy and Petrology, v. 108, p. 385−395. Geology, v. 13, p. 115−118.
Sinigoi, S., Quick, J.E., Clemens-Knott, D., Mayer, A., Demarchi, G., Maz- ——2010, Oceanic island basalts and mantle plumes: The geochemical per-
zuchelli, M., Negrini, L., and Rivalenti, G., 1994, Chemical evolution of a spective: Annual Review of Earth and Planetary Sciences, v. 38, p. 133−160.
large mafic intrusion in the lower crust, Ivrea-Verbano zone, northern Italy: Williams, D.A., Kerr, R.C., and Lesher, C.M., 2011, Mathematical model-
Journal of Geophysical Research, v. 99, p. 21575−21590. ing of thermo-mechanical erosion beneath Proterozoic komatiitic basaltic
Sobolev, A.V., Hofmann, A.W., Kuzmin, D.V., Yaxley, G.M., Anderson, A.T., sinuous rilles in the Cape Smith belt, New Québec: Mineralium Deposita,
Arndt, N.T., Chung, S.-L., Garcia, M.O., Gurenko, A.A., Danyushevsky, v. 46, p. 943−958.
L.V., Elliott, T., Frey, F.A., Kamenetsky, V.S., Kerr, A.C., Krivolutskaya, Wilson, A., 2012, A chill sequence to the Bushveld Complex: Insight into
N.A., Matvienkov, V.V., Nikogosian, I.K., Rocholl, A., Suschevskaya, N.M., the first stage of emplacement and implications for the parental magmas:
and Teklay, M., 2007, Estimating the amount of recycled crust in sources of Journal of Petrology, doi: 10.1093/petrology/egs011.
mantle-derived melts: Science. Wittig, N., Webb, M., Pearson, D.G., Dale, C.W., Ottley, C.J., Hutchinson,
Sobolev, S.V., Sobolev, A.V., Kuzmin, D.V., Krivolutskaya, N.A., Petrunin, M., Jensen, S.M., and Luguet, A., 2010, Formation of the North Atlantic
A.G., Arndt, N.T., Radko, V.A., and Vasiliev, Y.R., 2011, Linking mantle craton: Timing and mechanisms constrained from Re-Os isotope and PGE
plumes, large igneous provinces and environmental catastrophes: Nature, data of peridotite xenoliths from S.W.Greenland: Chemical Geology, v. 276,
v. 477, p. 312−316. p. 166−187.
Sun, S.-S., Wallace, D.A., Hoatson, D.M., Gilkson, A.Y., and Keays, R.R., Wooden, J.L., Czamanske, G.K., Fedorenko, V.A., Arndt, N.T., Chauvel, C.,
1991, Use of geochemistry as a guide to platinum group element poten- Bouse, R.M., King, B.W., Knight, R.J., and Siems, D.F., 1993, Isotopic and
tial of mafic-ultramafic rocks: Examples from the west Pilbara block and trace-element constraints on mantle and crustal contributions to Siberian
Halls Creek mobile zone, Western Australia: Precambrian Research, v. 50, continental flood basalts, Noril’sk area, Siberia: Geochimica et Cosmochi-
p. 1−33. mica Acta, v. 57, p. 3677−3704.
Sweeney, R.J., and Watkeys, M.K., 1990, A possible link between Mesozoic Woodland, S.J., Pearson, D.G., and Thirlwall, M.F., 2002, A platinum group
lithospheric architecture and Gondwana basalts: Journal of African Earth element and Re-Os isotope investigation of siderophile element recycling
Science, v. 10, p. 707−716. in subduction zones: Journal of Petrology, v. 43, p. 171−198.
Tang, Z.L., 1993, Genetic model of the Jinchuan nickel-copper deposit: Yaxley, G.M., Crawford, A.J., and Green, D.H., 1991, Evidence for carbon-
Ottawa, Geological Association of Canada Special Paper 40, p. 398−401. atite metasomatism in spinel peridotite xenoliths from western Victoria,
Thompson, R.N., Morrison, M.A., Hendry, G.L., and Parry, S.J., 1984, An Australia: Earth and Planetary Science Letters, v. 107, p. 305−317.
assessment of the relative roles of crust and mantle in magma genesis: An Yaxley, G.M., Green, D.H., and Kamenetsky, V., 1998, Carbonatite metaso-
elemental approach: Philosophical Transactions of the Royal Society of matism in the southeastern Australian lithosphere: Journal of Petrology, p.
London, v. A310, p. 549–590. 1917−1930.
Voshage, H., Hofmann, A.W., Mazzucchelli, M., Rivalenti, G., Sinigoi, S., Zhang, M., O’Reilly, S.Y., Wang, K., Hronsky, J., and Griffin, W.L., 2008,
Raczek, I., and Demarchi, G., 1990, Isotopic evidence from the Ivrea zone Flood basalts and metallogeny: The lithospheric mantle connection: Earth
for a hybrid lower crust formed by magmatic underplating: Nature, v. 347, Science Reviews, v. 86, p. 145−174.
p. 731−736. Zhang, M.J., Niu, Y.L., and Hu, P.Q., 2009, Volatiles in the mantle litho-
Webb, S.J., Nguuri, T., Cawthorn, R.G., and James, D.E., 2004, Gravity sphere: modes of occurrence and chemical compositions, in Anderson,
modeling of Bushveld Complex connectivity supported by southern Africa J.E., and Coates, R.W., eds., The Lithosphere: Geochemistry, geology and
seismic experiment results: South African Journal of Geology, v. 107, p. geophysics: New York, Nova Science Publishers, p. 171−272.
207−220. Zindler, A., and Hart, S., 1986, Chemical geodynamics: Annual Review of
Wendlandt, R.F., 1982, Sulfide saturation of basalt and andesite melts at high Earth and Planetary Science, v. 14, p. 493−571.
pressures and temperatures: American Mineralogist, v. 67, p. 877−885.

Das könnte Ihnen auch gefallen