Earth and Planetary Science Letters

advertisement
Earth and Planetary Science Letters 389 (2014) 74–85
Contents lists available at ScienceDirect
Earth and Planetary Science Letters
www.elsevier.com/locate/epsl
Pyroxene megacrysts in Proterozoic anorthosites: Implications for
tectonic setting, magma source and magmatic processes at the Moho
G.M. Bybee a,b,∗ , L.D. Ashwal a , S.B. Shirey b , M. Horan b , T. Mock b , T.B. Andersen c
a
b
c
School of Geosciences, University of the Witwatersrand, Private Bag 3, Wits, 2050, South Africa
Department of Terrestrial Magnetism, Carnegie Institute for Science, 5142 Broad Branch Road NW, Washington, DC 20015, USA
Center of Earth Evolution and Dynamics (CEED), University of Oslo, P.O. Box 1047, Blindern, 0316, Oslo, Norway
a r t i c l e
i n f o
Article history:
Received 1 September 2013
Received in revised form 6 December 2013
Accepted 11 December 2013
Available online xxxx
Editor: T.M. Harrison
Keywords:
Proterozoic anorthosites
megacryst
crustal differentiation
magma ponding
magmatic processes
a b s t r a c t
Proterozoic anorthosites from the 1630–1650 Ma Mealy Mountains Intrusive Suite (Grenville Province,
Canada), the 1289–1363 Ma Nain Plutonic Suite (Nain–Churchill Provinces, Canada) and the 920–949 Ma
Rogaland Anorthosite Province (Sveconorwegian Province, Norway), all entrain comagmatic, cumulate,
high-alumina orthopyroxene megacrysts (HAOMs). The orthopyroxene megacrysts range in size from 0.2
to 1 m and all contain exsolution lamellae of plagioclase that indicate the incorporation of an excess Ca–
Al component inherited from the host magma at pressures in excess of 10 kbar at or near Moho depths
(>30–40 km). Suites of HAOMs from each intrusion display a large range in 147 Sm/144 Nd (0.10 to 0.34)
making them amenable for precise age dating with the Sm–Nd system. Sm–Nd isochrons for HAOMs
give ages of 1765 ± 12 Ma (Mealy Mountains), 1041 ± 17 Ma (Rogaland) and 1444 ± 100 Ma (Nain), all
of them older by about 80 to 120 m.y. than the respective 1630–1650, 920–949 and 1289–1363 Ma
crystallization ages of their host anorthosites. Internal mineral Sm–Nd isochrons between plagioclase
exsolution lamellae and the orthopyroxene host for HAOMs from the Rogaland and Nain complexes yield
ages of 968 ± 43 and 1347 ± 6 Ma, respectively – identical within error to the ages of the anorthosites
themselves. This age concordance establishes that decompression exsolution in the HAOM was coincident
with magmatic emplacement of the anorthosites, ∼100 m.y. after HAOMs crystallization at the Moho.
Correspondence of Pb isotope ages (206 Pb/204 Pb vs. 207 Pb/204 Pb) with Sm–Nd ages and other strong
lines of evidence indicate that the older megacryst ages represent true crystallization ages and not the
effects of time-integrated mixing processes in the magmas. Nd isotopic evolution curves, AFC/mixing
calculations and the age relations between the HOAMs and their anorthosite hosts show that the HAOMs
are much less contaminated with crustal components and are an older part of the same magmatic system
from which the anorthosites are derived. Modeling of these anorthositic magmas with MELTS indicates
that their ultramafic cumulates would have sunk in the magma and been sequestered at the Moho,
where they may have sunk deeper into the mantle resulting in large-scale compositional differentiation.
The HAOMs thus represent a rare example of part of a cumulate assemblage that was carried to the
upper crust during anorthosite emplacement and, together with the anorthosites, illustrate the dramatic
influence that magma ponding and differentiation at the Moho has on residual magmas traveling towards
the surface. The new geochronologic and isotopic data indicate that the magmas were derived by melting
of the mantle, forming magmatic systems that could have been long-lived (e.g. 80–100 m.y.). A geologic
setting that would fit these temporal constraints is a long-lived Andean-type margin.
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
Near monomineralic plagioclase-rich intrusions, up to 18 000
km2 in areal extent, known as Proterozoic massif-type anorthosites,
have been studied for decades, yet many questions remain regard-
*
Corresponding author at: School of Geosciences, University of the Witwatersrand, Private Bag 3, Wits, 2050, South Africa. Tel.: +27 11 717 6633.
E-mail address: grant.bybee@wits.ac.za (G.M. Bybee).
0012-821X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.epsl.2013.12.015
ing their tectonic setting, magma source and temporal restriction
(1800–1100 Ma; Ashwal, 1993; Morse, 1982). These batholiths
were emplaced at upper crustal levels as crystal-laden magmas carrying with them giant (up to 1 m in length), cogenetic,
high pressure (10–15 kbar), high-Al orthopyroxene megacrysts
(HAOMs; 3–9 wt% Al2 O3 ; Fig. 1; Emslie, 1975; Emslie et al.,
1994). Proterozoic anorthosites are known to have crystallized
from high-Al basaltic magmas and are so plagioclase-rich that
they are widely believed to be missing 30–40% mafic minerals
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
75
1.1. The enigmatic petrogenesis of Proterozoic anorthosites
Fig. 1. (Color on web.) The field setting and microscopy of HAOMs. a. One large
(∼30 cm), euhedral HAOM crystal forming part of aggregate/pod from the MMIS.
b. Equant and subhedral megacryst aggregation from the NPS (photo courtesy B.
Ryan). c. Rounded HAOM aggregate from RAP. d. Elongate (∼40 cm) orthopyroxene megacryst from RAP possibly fractured during magma ascent/transport.
e. Two HAOMs from RAP showing interstitial relationships with plagioclase grains.
f. Crossed-polarized photomicrographs of HAOM showing plagioclase exsolution
lamellae (RAP). g. Plane polarized microphotographs of HOAMs (RAP).
(Charlier et al., 2010; Emslie, 1990; Emslie et al., 1994; Fram
and Longhi, 1992; Heinonen et al., 2010; Mitchell et al., 1995).
The comagmatic HAOMs, which may represent a small portion of
this missing component, are trapped fortuitously by high-viscosity
anorthositic magma mushes and transported to upper crustal levels. HAOMs thus provide a rare look at magmatic processes at
the crust–mantle interface, where evidence is usually sequestered
30–40 km below the surface, and can help explain the processes
that operate at the Moho.
We explore possible constraints on crustal differentiation processes as revealed by the HAOMs and associated anorthosites from
three classic Proterozoic anorthosite massifs: the Mealy Mountains
Intrusive Suite (MMIS), Nain Plutonic Suite (NPS, both in Labrador,
Canada, Fig. 2a) and Rogaland Anorthosite Province (RAP, Norway, Fig. 2b). Direct petrologic and geochemical evidence for the
ponding of upwelling magmas at the Moho is documented using the Nd and Pb isotopic systems. At Moho depths cumulates
can form, be sequestered, and sink across the Moho due to their
high density. The geochemical and geochronological information
for these HAOMs and their host anorthosites indicates a mantle source and Andean-type arc setting for Proterozoic massif-type
anorthosites.
Proterozoic massif-type anorthosites remain an enigmatic mode
of anorthosite occurrence. Ongoing debate surrounds the tectonic setting, parental melt compositions, source and emplacement
mechanisms of these predominantly felsic (80–90% plagioclase;
An contents = 50 ± 10), temporally restricted, batholithic intrusions. Competing hypotheses describing the source of anorthositeforming magmas in the Proterozoic have existed for decades
(e.g. Berg, 1969; Morse, 1969; Philpotts, 1969) and in the simplest sense the two schools of thought argue that the melts
were derived either from the mantle or the lower crust. Geochemical and petrologic studies investigating not only Proterozoic anorthosites, but also associated jotunites/monzonites and
marginal basaltic/high-Al gabbroic rocks, pointed to several petrologic and geochemical lines of evidence indicating that the magmas were derived from the mantle (Ashwal and Wooden, 1983;
Emslie, 1978; Icenhower et al., 1998; Mitchell et al., 1995, 1996;
Morse, 1982; Olson, 1992; Wiebe, 1990). Key to many of the
mantle-derivation proposals is the recognition that crustal assimilation played a role in developing the petrologic and geochemical composition of the magmas. Stable and radiogenic isotope data support the proposals that the magmas are derived
from melting of the depleted mantle combined with all-important
crustal contamination (Emslie et al., 1994; Peck and Valley, 2000;
Peck et al., 2010). In recent years lower crustal melting, through
underthrusting of lower crustal material into the mantle, has become popular based on several experimental investigations, geochemical arguments and field/geophysical observations (Duchesne,
2001; Duchesne et al., 1999; Longhi, 2005; Longhi et al., 1999;
Sauer et al., 2013; Schiellerup et al., 2000; Taylor et al., 1984).
A number of potential tectonic settings have, over the years,
been proposed for Proterozoic anorthosites such as meteorite impacts, extensional regimes, convergent margins and even anorogenic settings (Anderson, 1975; Ashwal, 1993; Berg, 1977; Corrigan
and Hanmer, 1997; Dewey and Burke, 1973; Emslie, 1978; Hoffman, 1989; Morse, 1982; Scoates and Chamberlain, 1997; Vigneresse, 2005). Proposals that anorthosites may represent part of the
deep roots of Andean-arc systems were favored for a time because
these settings could account for long, linear arrays of anorthosites
(e.g. Grenville Province, Eastern Ghats Belt; Ashwal, 1993). With
increasing geochronological evidence linking anorthosite emplacement to the diminishing stages of collisional orogeny, consensus
on the tectonic setting of Proterozoic anorthosites has shifted to
that of a late- to post-collisional orogenic setting (Martignole and
Schrijver, 1970; McLelland et al., 2010).
1.2. High-Al orthopyroxene megacryst geobarometry
Igneous relationships between HAOMs and surrounding megacrystic plagioclase (Emslie, 1975) and geochemical modeling
(Charlier et al., 2010) indicate a cogenetic relationship with their
host anorthosites (Fig. 1). Four independent studies using general
petrography/geochemistry (Wiebe, 1986), Al-in-orthopyroxene geobarometry (Emslie, 1975), experimental work (Longhi et al., 1993;
Fram and Longhi, 1992) and major element modeling (Charlier
et al., 2010), show that HAOMs crystallized from fractionating magmas, beginning at pressures of 10–15 kbar (30–40 km).
These crystallization depths are at, or near, the Moho, an interval seldom sampled by other geological processes, and are far
deeper than the emplacement depths of the host anorthosites
(<10–20 km; Berg, 1977; Valley and O’Neil, 1982). Although Al
content in many mafic systems is not the sole indicator of depth
of crystallization, detailed calibration and studies on pyroxenes
from the anorthositic system show that Al content is a wellcalibrated geobarometer (Emslie, 1975; Fram and Longhi, 1992;
76
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
Fig. 2. a. The geological setting of the Nain Plutonic Suite (NPS) and Mealy Mountains Intrusive Suite (MMIS) in Labrador, the easternmost province of Canada. b. The Rogalnd
Anorthosite Province is located on the southwestern coast of Norway and intrudes into rocks of the Sveconorwegian Orogen.
Longhi et al., 1993 and Charlier et al., 2010). Dymek and Gromet
(1984) and Morse (1975) have alternatively suggested that these
HAOMs were products of rapid in-situ crystallization and an increased incorporation of Al2 O3 into orthopyroxene at the liquid–
crystal interface as a result of sluggish diffusion of Al2 O3 into the
liquid. Longhi et al. (1993) showed through experiments and calculations that compatible elements like Cr2 O3 would have been
depleted in the crystals if rapid, low pressure crystallization had
taken place. However, a natural, positive correlation of Cr2 O3 and
Al2 O3 is observed (Longhi et al., 1993). Consensus has therefore
settled on the fact that these HAOMs are products of crystallization of parental magma to anorthosites at upper mantle to lower
crustal depths, providing a window into primitive magmatic processes operating near the source of these magmas.
1.3. Regional geology of a trio of Proterozoic anorthosite intrusions
The Mealy Mountains Intrusive Suite (MMIS), Nain Plutonic
Suite (NPS), and Rogaland Anorthosite Province (RAP) intruded into
pre- and early-Labradorian orthogneisses of the Grenville Province,
Archaean ortho- and paragneisses from the Torngat Orogen and
gneisses/amphibolites of the Sveconorwegian Orogen, respectively,
are all classic examples of Proterozoic massif-type anorthosites
(Fig. 2).
The Mealy Mountains Intrusive Suite (MMIS; Fig. 2, S1a), which
forms part of a deep-level, thrust-stacked block in the exterior thrust belt of the Grenville Province known as the Mealy
Mountains Terrane, is the largest of several anorthosite massifs in the easternmost section of the Eastern Grenville Province
(Emslie, 1976; Emslie et al., 1983; Gower et al., 2008a, 2008b).
The MMIS lies approximately 200 km southeast of the Grenville
Front and has been emplaced into relatively juvenile and immature, pre-Labradorian (1800–1770 Ma) and early-Labradorian
(1710–1655 Ma) orthogneisses and paragneisses that make up
a significant portion of the NE sector of the Grenville Province
(Emslie and Hegner, 1993; Gower et al., 2008b; Hegner et al.,
2010). Two distinct massifs make up the anorthositic member of the suite: leuconorites and pyroxene-bearing anorthosites
of the Etagaulet massif and leucotroctolites and olivine-bearing
anorthosites of the Kenemich massif (Emslie and Bonardi, 1979).
Geochemically, the pre-Labradorian and early-Labradorian gneisses
of the Mealy Mountains Terrane, into which the anorthositic
suite has intruded, are juvenile with 143 Nd/144 NdI at 1640 Ma =
0.51040–0.51051.
The Nain Plutonic Suite is a mid- to upper-crustal, coalesced
batholithic assemblage of more than 12 anorthositic, granitic, troctolitic and gabbronoritic intrusions that were emplaced in two
distinct sequences spanning 1363–1289 Ma (Fig. 2, S1b; Myers et
al., 2008; Ryan, 2000; Xue and Morse, 1993). A roughly north–
south trending tectonic junction, the Torngat Orogen, plays host
to the NPS and also marks a zone of oblique collision, transcurrent
shear and subsequent uplift between the Archaean Nain Province
and the Palaeoproterozoic Churchill/South Eastern Rae Province between 1860 and 1740 Ma (Ryan, 2000; Van Kranendonk, 1996).
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
The anorthositic lithologies, of which leuconorites and troctolites
are dominant, separate granitoid lithologies to the west and Archaean gneisses of the Nain Province to the east (Ryan, 2000).
Small plutons, dykes and sheets of ferrodiorite are also common
in the NPS as are layered troctolitic intrusions (Ryan, 2000).
The Rogaland Anorthosite Province (Fig. 2) formed during the
Meso- to Neoproterozoic within the Sveconorwegian Orogen on
the western margin of the Fennoscandian plate (within the larger
Baltica or East European Craton; Bingen et al., 2008; Bogdanova
and et al., 2008; Rivers and Corrigan, 2000). Three separate massifs,
the Egersund-Ogna, Håland-Helleren and Åna-Sira bodies, make
up the anorthositic end member of the RAP. The Egersund-Ogna
massif, from which the majority of the HAOMs in this study are
derived, is a mantled dome with margins consisting predominantly of deformed leuconoritic lithologies (Fig. S1c). The central
regions of the body are composed of a leuconoritic–noritic facies
known as the anorthositic–noritic complex that displays granulated, 1–3 cm plagioclase grains and hosts substantial concentrations of sub-ophitic pyroxene and plagioclase megacryst aggregates
(Charlier et al., 2010). These lithologies are emplaced into granulite
facies, migmatitic gneisses that represent metamorphosed granitoids and volcanic lithologies that formed at 1.5 Ga during one
of the first recorded magmatic, crust-forming events in the Sveconorwegian Orogen (Slagstad et al., 2013; Vander Auwera et al.,
2011). These gneisses show similar juvenile characteristics to those
in the Mealy Mountains Terrane with 143 Nd/144 NdI at 930 Ma =
0.51127–0.51142 compositions (Menuge, 1988).
2. Analytical methods
Major element data were obtained from the School of Geosciences Earth Lab at the University of Witwatersrand using XRF.
A full suite of trace elements was analyzed at the Department
of Geological Sciences at the University of Cape Town. Radiogenic isotopic abundances and concentration data for Sm, Nd and
Pb were determined at the Department of Terrestrial Magnetism
(DTM) of the Carnegie Institute of Washington using traditional
isotope dilution, ion-exchange chromatography and a combination
of Thermal Ionisation Mass Spectrometry (TIMS; for Nd) and MultiCollector Inductively Coupled Mass Spectrometry (MC-ICPMS; for
Sm and Pb). A detailed analytical procedure can be found in Section 1.3 of the Supplementary Information.
3. Results
3.1. Megacryst and anorthosite petrology, geochemistry, and isotopic
compositions
HAOMs are ubiquitous in Proterozoic massif-type anorthosites
and in both the MMIS and RAP, these megacrysts either occur
as single, euhedral to subhedral crystals or form part of aggregates of several subhedral megacrysts (Fig. 1a–e). HAOM habit in
the NPS is more varied and in addition to occurring as single,
curved crystals or giant aggregate pods, these phases also occur
as single, angular pyroxene crystals with an intercumulus relationship to surrounding plagioclase. In the MMIS, field observations, in addition to collations of previous field workers’ observations, shows that the orthopyroxene megacrysts are restricted
to leuconorite-bearing lithologies and appear to be absent in
the olivine-bearing Kenemich massif. Similar correlations between
orthopyroxene-bearing lithologies and megacrysts are also noted
in the leuconoritic Egersund-Ogna massif, which hosts the most
HAOMs of the three intrusive suites making up the RAP. HAOMs
studied in this contribution show the typical microscopic features
such as plagioclase exsolution lamellae and pervasive inclusions
of Fe-oxide material. HAOMs from the MMIS are undeformed and
77
Fig. 3. Chondrite-normalized (Anders and Grevesse, 1989) REE element diagram
comparing HAOMs and anorthosites from the three localities sampled in this study.
Numbers adjacent to HAOMs from the MMIS represent Al2 O3 contents (wt.%).
pristine and, while the outer margins of the Egersund-Ogna massif
do preserve deformed and recrystallized megacrysts, the only samples collected from the RAP were pristine and showed no signs of
post-intrusion alteration. On the other hand, many of the samples
from the NPS show signs of recrystallization and some show kinkbanding caused by deformation.
Microscopically, these HAOMs are unzoned crystals of orthopyroxene with abundant, thin, planar, calcic plagioclase exsolution
lamellae (average An80 ) occasionally incorporating olivine, ilmenite
and biotite (Fig. 1f–g; Emslie, 1975). Light rare earth element enrichment (LREE) and Eu anomalies vary substantially within and
between sample sets with the highest Al, most LREE depleted samples showing no or very small Eu anomalies, whilst more LREE
enriched samples display prominent negative Eu anomalies (Fig. 3
and S5). Based on experimental data (Fram and Longhi, 1992;
Longhi et al., 1993), an Al-in-orthopyroxene geobarometer for
Proterozoic anorthosites calibrated by Emslie et al. (1994), suggests that our sampled HAOMs formed at pressures between
12.14–5.85 ± 0.4 kbars (RAP), 9.67–4.52 ± 0.4 kbars (MMIS), and
9.36–7.18 ± 0.4 kbars (NPS; Fig. S2).
The isotopic compositions of the different suites are an important first order observation (Supplementary Data Tables). The
MMIS, NPS and RAP display remarkably different isotopic compositions with the MMIS and RAP showing positive εNd,T compositions
(2.3 to 3.6 and 2.2 to 4.4, respectively) whilst the NPS presents a
range of negative values (−1.6 to −8.5; Fig. 4a). The systematic
differences and the precision of the isotopic analyses in the MMIS
allow us to differentiate different rock types isotopically (Fig. 4a
inset). The anorthosites (sensu stricto) have the highest εNd,T compositions, followed by leucotroctolites and then leuconorites (although the compositional ranges of the leuconorites and leucotroctolites overlap substantially when plotted with 2 sigma errors). On
a εNd -208 Pb plot, suites of HAOMs have approximately constant εNd
compositions, but show variations in 208 Pb (Fig. 4b). In the RAP,
lower-Al megacrysts are displaced, in varying degrees, to lower εNd
and 208 Pb compositions on a vector trending towards upper crustal
compositions. Anorthositic lithologies from the MMIS are displaced
to slightly lower εNd than the HAOMs although many lithologies
overlap. HAOMs from the NPS have been displaced to much lower
εNd and slightly lower 208 Pb compositions than the other HAOM
suites, while the host anorthosites show similar 208 Pb but lower
εNd compositions.
78
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
Fig. 4. a. A depleted mantle origin for HAOMs and Proterozoic anorthosites. εNd
vs. time illustrating how emplacement of anorthositic mushes into terranes of different ages dramatically influences the isotopic composition of the anorthosites. In
young terranes, where crust is relatively juvenille, the initial megacrystic isotopic
compositions (large triangles, derived from isochron regression) are consistently
only slightly displaced from depleted mantle evolution, indicating magma derivation
from depleted mantle. Linear fields indicate evolution of potential crustal contaminants (potential contaminants for the NPS include both Churchill and Nain Province
gneisses, due to intrusion into the province-suturing Torngat Orogen). (Nägler and
Kramers, 1998.) b. εNd vs. (208 Pb/204 Pb)i plot illustrating isotopic variation between
megacrysts and host anorthosite. The 232 Th–208 Pb system is chosen as a better
tracer of contamination because of increased concentrations of Th in the continental crust. Depleted mantle composition determined from Rehkämper and Hofmann
(1997) and the representative lower crustal composition is taken from Ben-Othman
et al. (1984). HAOM and anorthosite compositions are plotted at the ages shown in
Table 1.
Fig. 5. a. 143 Nd/144 Nd vs. 147 Sm/144 Nd illustrating the isochronous relationships
between high-Al megacrysts from three different Proterozoic anorthosite massifs.
Each isochron defines an age approximately 110–130 m.y. older than the recognized age range of the anorthosite intrusions. 2σ errors are all smaller than
the symbol size. Isochron regressions produce εNd values for MMIS: εNd;1765 Ma =
+3.17 ± 0.31, RAP: εNd;1041 Ma = +4.83 ± 0.61 and NPS: εNd;1444 Ma = −3.16 ± 2.55.
The Churchill and Nain Province Gneisses cluster around the following isotopic compositions not shown on the graph: 143 Nd/144 Nd: 0.511421–0.510451; 147 Sm/144 Nd:
0.107025–0.085978 (Supplementary Data Tables). MSWD: mean square of the
weighted deviates. b. Four point isochrons (including repeat analyses of wholerock megacryst) between plagioclase lamellae, host orthopyroxene and whole-rock
megacrysts indicating that the ascent of HAOMs in plagioclase-rich mushes to final
emplacement levels in the upper crust requires significant lengths of time – similar
to the age differences between HAOMs and the host anorthosites.
3.2. High-Al orthopyroxene megacryst geochronology
than the anorthosite crystallization age as given by the U–Pb zircon, baddeleyite, and apatite ages of 1289 to 1363 Ma on the NPS
anorthosites directly and the very precise megacryst exsolution age
of 1347 ± 6 Ma (see summary in Myers, 2008; this study). Similarly, the Sm–Nd isochron age for the HAOMs from the RAP with
>8 wt% Al2 O3 of 1041 ± 17 Ma is 92 to 121 m.y. older than the inferred anorthosite crystallization age of 920 to 949 Ma as given by
U–Pb ages of zircon and baddeleyite separated from the cogenetic,
high-Al orthopyroxene megacrysts (Andersen and Griffin, 2004;
Sauer et al., 2013; Schärer et al., 1996; this study). That three
different anorthosite massifs, formed at different times, show consistent relative age relations tied to identical mineralogy can only
Isochronous relationships amongst HAOMs using the Sm–Nd
system demonstrate that the highest pressure HAOMs, in each
suite, are 110–130 m.y. older than their host, comagmatic anorthosites (Fig. 5a; Table 1). The Sm–Nd isochron age for the HAOMs
from the MMIS is very precise at 1765 ± 12 Ma and statistically
significantly older by 115–130 m.y. than the anorthosite crystallization age as given by the precise U–Pb zircon and baddeleyite ages on the MMIS anorthosites themselves (Emslie, 1990;
Gower et al., 2008b; this study). The Sm–Nd isochron age for the
HAOMs from the NPS is 1444 ± 100 Ma, older by 81–155 m.y.
79
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
Table 1
Comparative table of megacryst isochron age (this study) vs. emplacement age of the host anorthosite (U–Pb zircon/baddeleyite; see geochronological sources below) and
Sm–Nd megacryst exsolution ages (this study). Point values in the parentheses indicate the number of points creating the isochron.
Intrusion
MMIS
NPS
RAP (>8 wt% Al2 O3 )
Megacryst isochron age
(Ma)
1765 ± 12
(n = 6 this study)
1444 ± 100
(n = 9, this study)
1041 ± 17
(n = 5; >8 wt% Al2 O3 , this study)
Megacryst exsolution age
(Ma; see Fig. 4)
–*
1346.9 ± 5.9
(n = 4, this study)
968 ± 43
(n = 4, this study)
Anorthosite emplacement
age (Ma)
1650–16301
1363–12892
949–9203
1
2
3
*
U–Pb zircon/baddeleyite ages from anorthosites and granitoids (Gower et al., 2008b; Emslie, 1990).
U–Pb zircon/baddeleyite/apatite, Ar–Ar ages from multiple sources summarized in (Myers et al., 2008).
U–Pb zircon/baddeleyite in high-Al megacrysts (Sauer et al., 2013; Schärer et al., 1996) and Storgangen orebody of the RAP (Andersen and Griffin, 2004).
Insufficient plagioclase could be separated from megacrysts in the MMIS to perform exsolution geochronology.
be explained by a process general to all three massifs and perhaps
to all anorthosites. Fig. S2 in the Supplementary Information indicates the calculated pressure of crystallization for the HAOMs that
form these isochrons and Section 1.4 highlights more detailed discussions surrounding the geochronology.
Previous U–Pb geochronology and detailed field studies reveal
that these comagmatic, coalesced anorthosite massifs were emplaced over at least a 12–80 m.y. period (Gower et al., 2008a;
Myers et al., 2008; Schärer et al., 1996). HAOMs from RAP with
lower Al2 O3 contents, as well as intercumulus orthopyroxene display various degrees of isotopic disequilibrium with the HAOMs
(Fig. 5a). Recrystallization has affected, and is unavoidable, in many
of the samples from the Nain Plutonic Suite, creating a larger
isochron age error. Age data from Pb–Pb isochrons are of poorer
quality, perhaps due to the ease with which Pb remobilization occurs. In one case (RAP), however, data trends are preserved in the
U–Pb system, and although not a true isochron, this array produces
a comparable age to the isochron in the Sm–Nd system (Fig. S4).
Four-point Sm–Nd isochrons between plagioclase lamellae, host
orthopyroxene and whole-rock HAOM compositions from two
HAOMs in the NPS and RAP (Fig. 5b, Table 1) show that decompression exsolution occurred at 1346.9 ± 5.9 Ma and 968 ±
43 Ma, respectively – ages that correspond (within error) with
anorthosite crystallization age ranges in both massifs. Plagioclase exsolved from these HAOMs 70–100 m.y. after crystallization, suggesting that significant timescales are required for the
20–30 km ascent from crystallization depth to plagioclase exsolution depth. The low Al2 O3 contents (less than 3 wt%) and lack
of plagioclase exsolution lamellae in intercumulus orthopyroxene of the host anorthosites indicate that these lithologies must
have crystallized at shallow, upper crustal depths (Berg, 1977;
Valley and O’Neil, 1982) over the time spans indicated above and
110–130 m.y. after HAOM formation. Interestingly, Sm–Nd plagioclase exsolution ages in HAOMs from the RAP overlap with
U–Pb ages of zircon and baddeleyite separated from the cogenetic,
HAOMs (Table 1). These U–Pb ages are interpreted as the emplacement age of the anorthosites in the RAP by Andersen and Griffin
(2004), Sauer et al. (2013) and Schärer et al. (1996) and the implications of this seemingly contradictory overlap are discussed in
Section 4.3.
Viewed in isolation, the ∼100 m.y. differences between HAOMs
and anorthosite ages could be interpreted as a result of differences
in closure temperature between the Sm–Nd system in the primitive megacrysts and U–Pb in anorthositic zircon that crystallized
at a late-stage (or even at sub-solidus conditions) in zirconium depleted magmas. However, associated with Proterozoic anorthosites
are suites of coeval granitoids (Ashwal, 1993), which are more
likely to be saturated in zircon, confirming that 100 m.y. differences between megacryst and younger anorthosites and granitoids
cannot be the effect of different closing temperatures and slow
cooling.
3.3. True isochrons vs. pseudo-isochronous mixing lines?
One interpretation of the isotopic arrays created by the HAOMs
(Fig. 5a) may be as time-integrated mixing lines created by initial I Nd variations due to either mixing or assimilation-fractionalcrystallization (AFC) with crustal material during magma crystallization (Davidson et al., 2005; DePaolo, 1981). In theory, while
this hypothesis could explain the data, several lines of evidence
(described below) disprove the “mixing” hypothesis and indicate
that these arrays are true isochrons, ∼100 m.y. older than the host
anorthosites.
3.3.1. Isotopic disequilibrium in lower-Al2 O3 megacrysts and in
intercumulus orthopyroxene
In the RAP, HAOMs with Al2 O3 content >8 wt% form a ca.
1041 Ma isochron, but lower Al2 O3 megacrysts (3–7 wt%) are in
disequilibrium with this isochron and are displaced to lower Nd
isotopic compositions (Fig. 5a). Regression analysis reveals that this
lower-Al suite forms an errorchron with a shallower slope in Sm–
Nd space (Section 1.4, Supplementary Information). Isotopic disequilibrium between the higher- and lower-Al compositions suggests that magmatic differentiation processes, like AFC, affected
only the lower-Al megacrysts. Had the mixing between two separate components occurred during an AFC process that affected
all the megacrysts, it would be expected that they all plot on
an isochron. This is not the case. The fact that a distinct suite
of higher Al2 O3 content megacrysts retains an older isochron age
suggests that this age dates the crystallization of these minerals,
otherwise they should not yield an isochron either.
3.3.2. Coincidence in Pb–Pb isochron age
Although much of the Pb–Pb isotopic data appears reset, due
perhaps to later orogenic activity or secondary effects, HAOMs
from the RAP form a linear array on a 207 Pb–206 Pb plot with an age
of 1062 ± 330 Ma (MSWD = 258; Fig. S4). Although not strictly an
isochron, perhaps due to Pb remobilization, the age produced provides a useful comparison to the Sm–Nd isochrons (Fig. 5a). Given
the differences in fractionation behaviors and half-lives between
the Sm–Nd and U–Pb isotope systems, it is highly unlikely that
similar age differences between the HAOMs and host anorthosites
would result from initial crustal mixing or assimilation (Davidson
et al., 2005). The close correspondence between Sm–Nd and Pb–Pb
isochron ages, in at least one intrusion, disproves the hypothesis
that the isochrons result from variation in the initial isotopic ratio
of the magma.
3.3.3. Anorthosite isochrons and U–Pb ages
Sm–Nd isotopic arrays for anorthosite whole-rock compositions
from the MMIS, although not isochronous, produce an age equivalent to the U–Pb zircon/baddeleyite ages determined for these
80
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
intrusions (1653 ± 150 Ma vs. 1650–1630 Ma, respectively). If initial contamination had affected the magmas and produced old
isochrons in the megacrysts, one would expect the anorthosites,
which crystallize from the same, if not more contaminated magma,
to display similarly old ages relative to the U–Pb ages. This is not
the case, providing further evidence against the HAOM arrays representing pseudoisochrons. Anorthosites from the NPS do not form
arrays with any meaningful age and so no comparison can be made
in this case.
3.3.4. Consistent isochron ages between intrusions
The RAP, MMIS and NPS intrusions are each of significantly
different age (ca. 930 Ma, 1650–1630 Ma, 1363–1289 Ma, respectively; see Table 1) and intrude into crust with significantly different isotopic compositions and age. If the arrays do represent mixing or AFC trends between melt and a crustal assimilant, the diversity of age and crustal compositions would be highly unlikely to
produce the observed, self-consistent age differences between the
megacrysts and the comagmatic host anorthosites (100–144 m.y.,
see Table 1).
3.3.5. Time-integrated AFC/Mixing forward models
Using megacrysts from the RAP, where a range of Al2 O3 contents are preserved in the megacrysts and reasonable constraints
on the isotopic composition of the lower continental crust exist
(Othman et al., 1984), we forward model the effect of variations
in I Nd ratio due to mixing or assimilation on aged isotope data
(Section 1.5 of Supplementary Information). All of the “isochron”
ages produced by AFC and pure mixing using constraints appropriate to the RAP are between 285 m.y. and over a billion years
older than the anorthosite, illustrating that it is not possible to
create isochrons only ∼100 m.y. older than the anorthosites. Only
in a simulation where no fractionation occurs (r = 1), representing
pure mixing, do array ages approach those of the actual megacryst
isochrons (1215 Ma cf. 1041 Ma; Fig. S3c). A scenario involving
pure mixing is highly unlikely (DePaolo, 1981) and not applicable in the case of the RAP as variations in Eu anomalies amongst
the isochronous HAOMs (<8 wt% Al2 O3 ) indicate that the magma
was fractionating orthopyroxene and began fractionating plagioclase during HAOM crystallization (Fig. S5).
These various and diverse lines of evidence all disprove the
hypothesis that the isochronous Sm–Nd data arrays could be produced by time-integrated initial I Nd variations due to mixing or
AFC processes in a magma chamber. The more likely scenario for
the creation of isochronous data arrays in all three anorthositehosted megacryst suites is directly related to magmatism near the
Moho.
4. Discussion
4.1. Direct geochemical evidence for magma ponding at the Moho
Given our evidence that HAOM isochron ages represent true
crystallization ages, we can now place HAOM crystallization in
a plausible sequence of events to better understand Proterozoic
anorthosite petrogenesis and general magmatic processes operating in the upper mantle and lower crust.
Broadly basaltic high-Al, tholeiitic melts, known to be the
parental magmas of Proterozoic anorthosites (Charlier et al.,
2010; Emslie, 1990; Emslie et al., 1994; Fram and Longhi, 1992;
Heinonen et al., 2010; Mitchell et al., 1995), initially crystallize
olivine and clinopyroxene while rising towards the base of the
crust (Müntener and Kelemen, 2001; Müntener and Ulmer, 2006).
Interactions between this magma and partial melts of the mafic
lower crust would induce a limited pulse of more silicic lower
crustal contamination, not only inducing orthopyroxene formation
due to peritectic olivine reaction (Müntener and Kelemen, 2001),
but lowering and shifting the Nd and Pb isotopic composition of
the high-Al orthopyroxene cumulates, as we observe (Fig. 4a, b).
Although most Proterozoic anorthosite massifs commonly preserve
orthopyroxene megacrysts, rare examples of the occurrence of
clinopyroxene megacrysts together with HAOMs have been noted
in the Labrieville and Grenville Township massifs (Quebec; Ashwal,
1993; Owens and Dymek, 1995; Philpotts, 1969). These observations illustrate that most clinopyroxene (and olivine) is likely to
have crystallized before, and in rare cases, with orthopyroxene
and plagioclase at high pressures. The paucity of clinopyroxene
(and olivine) megacrysts in anorthosite massifs suggests that these
phases formed (and perhaps sank in the magma chamber) before
sufficient, buoyant plagioclase mush had formed – a vital ingredient for delivering some HAOMs to upper crustal depths with the
anorthosite. The isochronous isotopic composition of the highestAl megacrysts (Fig. 5a) indicates that they were in equilibrium
with a fractionating magma that received compositionally indistinct lower crustal contamination and/or magmatic recharge at
upper mantle or lower crustal pressures. Ponding and subsequent
crystallization of a rising basaltic magma at a suitable interface in
the lithosphere allows creation of such a locus for such crystallization. The Moho provides the required rheological contrast to
stall an upwelling mafic magma at the calculated depths of HAOM
crystallization on modern-day Earth (Artemieva, 2011). Depths to
the Moho in arc/orogenic environments range between 27–60 km
(with the maximum value typical of continent–continent collisional zones; Artemieva, 2011). Ponding of magma at the Moho is
a commonly proposed, but poorly documented phenomenon when
describing magma ascent from source to surface through the continental lithosphere. Our new data for HAOMs, however, provide
direct, geochemical and petrological evidence for these processes
at the Moho. Our findings are supported by recent seismic refraction studies by Richards et al. (2013) imaging large ultramafic
bodies (V p ∼ 7.4–8.0 km/s) at or above the Moho below older
oceanic hotspots formed beneath thick oceanic lithosphere. These
findings, together with our results, illustrate the importance of cumulate formation in magmas that pond at the Moho. Implications
of these processes are discussed in Section 4.2.
The scarcity of negative Eu anomalies in some of the highest-Al
megacrysts, in addition to their high Al content, suggest that plagioclase was not yet part of the crystallizing assemblage at these
pressures (Fig. 3, Fig. S5). These magmas are so plagioclase rich
that if the HAOMs were crystallizing in equilibrium with them,
they would be expected to have prominent negative Eu anomalies as are seen in lower-Al megacrysts. Positive-buoyancy instabilities develop once sufficient plagioclase has formed, allowing
plagioclase-rich mushes (defined as magmas with volume fraction crystallinity of 0.25–0.55) to begin rising through the crust
(Kushiro and Fuji, 1977). Both lower-Al megacrysts and intercumulus orthopyroxene in the host leuconorites are in isotopic disequilibrium with the highest-pressure megacrysts, indicating that polybaric assimilation and fractional crystallization characterize the
remainder of the ascent of these buoyant plagioclase mushes to
their final level of emplacement (Fig. 5a). Later pulses of magma
that crystallize olivine and plagioclase may also pond at the Moho
and create younger troctolitic mushes (as observed; Gleißner et
al., 2011; Morse, 2006) that rise along similar pathways to the
final levels of emplacement. These magmas are not significantly
contaminated by crustal material (yielding the most radiogenic,
mantle-like compositions; Fig. 4a inset) because earlier anatexis
depleted and sealed the country rock. This explains the restriction of HAOMs to orthopyroxene-bearing massifs, as we observe,
particularly in the MMIS.
In the model presented here, HOAMs form in a ponded magma
at the Moho ∼100 m.y. before anorthosite emplacement in the up-
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
per crust, but we do not envisage these crystals remaining at the
Moho for 100 m.y., and suggest that part of this time period was
spent in a convecting, recharging magma at the Moho with the remainder of the period involving the ascent of the magma mush
to mid-crustal levels. The ascent of magmas, sufficiently saturated
with plagioclase to be positively buoyant, may take significant
lengths of time – longer than predicted for less-viscous basaltic
magmas (Kushiro, 1980).
4.2. Implications for magma chamber processes in the deep crust and
mantle
Recent work has shown that equilibrium crystallization of
basaltic magmas produces, at most, 50–60% of plagioclase (Müntener and Ulmer, 2006) in the crystallizing assemblage. Based
on our maximum estimates, any given anorthositic massif contains 10–15% of mafic phases (olivine, orthopyroxene, clinopyroxene and magnetite/ilmenite), suggesting that 35–40% of the mafic
mineralogy is unaccounted for. Considering our evidence for cumulate formation in ponded magmas at the Moho, we propose
that ultramafic crystallization and sequestration in ponded magmas at these depths could account for the missing component
in Proterozoic anorthosites. Calculations show that, in arc environments, ∼21% ultramafic cumulate formation from primitive
mantle-derived magmas occurs below the Moho (Conrad and Kay,
1984) and the observation of high-pressure ultramafic xenoliths
in lavas from the Aleutian arc (Conrad and Kay, 1984) and the
Nógrád-Gömör Volcanic Field in northern Hungary (Zajacz et al.,
2007) provide corroboratory evidence that upper mantle cumulates
do form in other environs, but are rarely brought to the surface.
HAOM compositions also compare favorably with fields of experimentally determined, high-pressure (12 kbar) basaltic arc cumulates (Fig. S6; Müntener and Kelemen, 2001). HAOMs are slightly
more evolved than these experimental cumulates, suggesting that
olivine fractionation did occur prior to HAOM crystallization. These
lines of evidence point to similar ultramafic cumulates forming
in major continental crust generation zones where basaltic magmas pond at the Moho. HAOMs, therefore, are not peculiarities
restricted to Proterozoic anorthosites, but instead are vestiges of
ubiquitous and significant magmatic processes operating at the
Moho, fortuitously brought to the surface as a result of high viscosity plagioclase-rich mushes which form Proterozoic anorthosites.
Arndt and Goldstein (1989), working on the assumption that
mafic magmas pond, crystallize and assimilate lower crust at
the Moho (as we have demonstrated), show that olivine and
pyroxene cumulates would sink back into the mantle due to
negative density contrasts with the surrounding mantle. MELTS
(Ghiorso and Sack, 1995) modeling was performed to quantify
the density contrasts between cumulates, magmas and the mantle at the Moho using starting conditions expected for Proterozoic anorthosite formation. An experimentally produced highAl basalt was used as the parental magma (Fram and Longhi,
1992), with close similarities to natural samples found in various anorthositic massifs and reported in Charlier et al. (2010).
Models were run under equilibrium and fractional crystallization
scenarios at 10 and 15 kbar. Pyroxenes formed between 10 and
15 kbar will be denser than both the crystallizing magma and the
surrounding mantle (e.g. #ρmagma-pyroxene [10 kbar] = −0.62 g cm−3 ;
#ρPREMmantle-pyroxene [10 kbar] = −0.07 g cm−3 ) and will sink, not
only in the ponded magma, but also into the mantle (Fig. 6). The
density contrasts between the cumulate pyroxene and mantle are
less pronounced at 15 kbar, although still exhibit relatively large
density contrasts with the magma. The small negative density contrast between pyroxene cumulates and the surrounding mantle at
both pressures will be exacerbated as the cumulates transition to
81
eclogite facies. These results demonstrate that ultramafic material formed at the Moho, under conditions and using chemistries
likely to crystallize anorthosites, would be denser than the magmas and the surrounding mantle and would consequently sink into
the mantle, or at least be sequestered beneath the Moho. These
MELTS models (particularly at 10 kb) also produce positively buoyant plagioclase (with geologically reasonable An contents; An±50 )
that will accumulate at the top of a ponded magma and eventually rise diapirically through the crust, as predicted in models
of anorthosite petrogenesis (Ashwal, 1993; Charlier et al., 2010;
Emslie et al., 1994). Jull and Keleman (2001) modeled and quantified density contrasts between cumulates/lower crust at the Moho,
using different methods, and demonstrated that ultramafic material (olivine clinopyroxenites) present at Moho conditions (on
arc-type geotherms) would have density contrasts with the mantle between 0.05–0.1 g/cm3 and would sink into the mantle on
timescales of 10 million years, supporting the MELTS models created in this study.
The fate of these ultramafic cumulates is crucial. Although it
is not known whether bulk crustal differentiation occurs by sinking of mantle-derived cumulates, or foundering of lower crustal
material into the mantle (Kay and Kay, 1991, 1993; Lee et al.,
2006), our results indicate that cumulate formation and foundering at the Moho is a very real and significant mechanism for
generating discrepancies between single-stage melts and observed
magmatic products (Fig. 7). Similar processes of cumulate formation and delamination are inferred to have occurred in other arc
environments from studies in exposed arc sections (e.g. Kohistan Arc, N. Pakistan) and represent analogs to the processes we
have observed in Proterozoic anorthosites (Jagoutz et al., 2006;
Jagoutz and Schmidt, 2012). As shown in Fig. 7, the effects of ultramafic cumulate sequestration in ponded magmas at the Moho
plays a major role in developing the bulk, felsic composition of
Proterozoic anorthosites and we suggest that the same process is
crucial for development of the intermediate composition of the
bulk continental crust. The ultimate fate of these cumulates may
be remelting and creation of chemical heterogeneities elsewhere
in the convecting mantle, followed by subsequent recycling into
the sources of MORB and OIB (Tatsumi, 2000).
4.3. The tectonic setting and source of Proterozoic anorthosites
A variety of tectonic settings have been proposed for Proterozoic anorthosites, with opinion divided between intraplate,
divergent and convergent plate tectonic settings (McLelland et
al., 2010). Anorthositic magma production, in terranes such as
the Grenville–Labrador region (over 1600 km long), spans about
500 m.y. and it is difficult to envisage even the largest continental rifts supplying magma to a region for this length of
time without initiating continental break-up (Ashwal, 1993). In
some tectonic models of anorthosite formation (Martignole et
al., 1993; McLelland et al., 2010), the magmas are produced by
post-collisional lithospheric delamination followed by ascent and
partial melting of upwelling asthenospheric mantle. Such models
are possible for some anorthosite-bearing terranes (e.g. Grenville
Province), but fail for others (e.g. Nain), where no collisional events
of the appropriate ages have been recognized. Furthermore, postcollisional delamination models imply short-lived magmatism, in
contrast to our results, which suggest longer-lived magmatic systems on the order of ∼100 m.y.
We suggest that one geodynamic setting capable of delivering
sufficient, geographically-focused quantities of magma (and heat)
to the base of the crust for ∼100 m.y., or more, is a convergent arc
environment. Well-known convergent systems such as the Californian Arc were in operation for >150 m.y., and produced volcanic
and plutonic magmatism in only two short episodes (10–20 m.y.
82
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
Fig. 6. Extracted magma and cumulate densities from MELTS modeling of high-Al basalt crystallization at Moho pressures. The starting composition is the same in each model
permutation. a. Model fractionating solids at 1 GPa (10 kbars). b. Model not fractionating solids at 1 GPa (10 kbars). c. Model not fractionating solids at 1.5 GPa (15 kbars).
d. Model fractionating solids at 1.5 GPa (15 kbars).
in duration; Ducea, 2001). Subduction-related plutonism is also
recorded for a period of over 150 m.y. in Late Jurassic to Neogene granitoids constituting the South Patagonian Batholith (Hervé
et al., 2007). In the RAP, the formation of magmas in an arc environment is supported by the overlap in HAOM crystallization age
and the onset of calc-alkaline magmatism, forming the Sirdal Magmatic Belt in the Sveconorwegian Orogen, both ∼100 m.y. before
anorthosite emplacement (Bybee et al., in preparation; Slagstad et
al., 2013). As shown in this paper, Sm–Nd plagioclase exsolution
ages in HAOMs record the times of anorthosite massif emplacement. This final anorthosite emplacement in the mid- to upper
crust, ∼100 m.y. after magma generation, may be facilitated by
extensional regimes common to post-collisional tectonics settings
by, for instance, weakening the crust through orogenic collapse, or
warming the collapsed crust to near-solidus temperatures, thereby
facilitating ascent of viscous magmas. Correspondence of Sm–Nd
plagioclase exsolution ages and U–Pb zircon-in-megacryst ages
may be due to zircon, like plagioclase, being an exsolution product in the megacrysts and consequently dating an exsolution and
mid-crustal emplacement event.
Batholithic bodies of massif-type anorthosite have not yet been
found in Phanerozoic or modern-day arc-related terranes. It has
been suggested that this might be due to insufficient levels of erosion in young arc terranes (Hamilton, 1981); perhaps a search in
poorly explored plutonic parts of young arcs, such as in southern
Patagonia, may reveal new anorthosite occurrences. However, there
are minor anorthosites that occur as cognate xenoliths in volcanics
from several magmatic arcs (Ashwal, 1993), although the plagioclase in these is far more calcic (An70–100 ) than the typical An45–60
found in Proterozoic massifs. If these occurrences are representative of a general anorthosite-forming process, then this might
imply a difference in the compositional details between Proterozoic and modern magmatic systems, perhaps involving volatiles.
For instance, dry magmatism in Proterozoic arcs would enhance
plagioclase production, promote tholeiitic differentiation trends
and Fe enrichment, whereas higher water content in Phanerozoic
arcs would suppress plagioclase crystallization, possibly leading to
smaller anorthosite bodies, with more calcic plagioclase. We do
recognize the petrological and compositional differences between
our proposed anorthosite-producing continental arcs and modern
arcs typified by calc-alkaline magmas, but we call attention to
the clear diversity in the magmatic products in modern arc environments, which include rhyolitic volcanics and equivalent granitoid intrusions as well as basaltic and gabbroic bodies, spanning
four main magma series including low-K, tholeiitic compositions
(Wilson, 2007). It is this part of the arc, characterized by low-K,
tholeiitic compositions that could also conceivably produce Proterozoic anorthosites and their parental magmas.
Magmas forming Proterozoic anorthosites and their cogenetic
HAOMs are considered to have been derived from melting of ei-
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
83
external source of heat from upwelling magmas (for which the
concomitant mantle-derived magmatism is not observed; Morse,
1991). We suggest, therefore, that Proterozoic anorthosites formed
from melting of the depleted mantle in long-lived continental-arc
environments.
5. Conclusion
Fig. 7. a. MgO vs. SiO2 illustrating the compositional discrepancy between singlestage melts (grey-shaded box) of peridotitic mantle [Stage 1] in arc environments
and anorthosites and the bulk continental crust (CC; Rudnick and Gao, 2003 and
references therein). Fractionation [Stage 2] of these peridotitic melts at the Moho
and associated sequestration of dense ultramafic cumulates (real megacryst compositions, experimental cumulates and modeled MELTS phases) below the Moho,
drives the composition of the residual magma to more intermediate compositions
– a process operating at the Moho that is fundamental to generating, not only the
character of Proterozoic anorthosites, but all magmatic products that rise from this
depth. Primitive mantle (PM) compositions are taken from (Palme and O’Neill, 2003)
and references therein. Yellow pentagon represents the proposed parental magma of
Proterozoic anorthosites (Charlier et al., 2010). Density contrast calculations (#ρ )
between magma, mantle and cumulates derived from MELTS modeling. b. We show
that ultramafic cumulates formed in fractionating magmas ponded at the Moho are
denser than the melt and the surrounding mantle and may founder into the deeper
mantle. This process is crucial to the development of anorthositic mushes in arc
environments as we propose. Furthermore, the effect of cumulate sequestration at
this depth cannot be discounted in models of continental crust differentiation and
is likely to play an important role in developing the intermediate composition of
the bulk crust.
ther the depleted mantle (Emslie et al., 1994) or of the lower
crust (Schiellerup et al., 2000). Data in this study support previous workers’ conclusions (Ashwal et al., 1986) that the age and
nature of the continental crust into which anorthositic magmas intrude dramatically affects their isotopic composition, changing the
convecting or depleted mantle isotopic signature of these rocks
toward more evolved compositions (Fig. 4a, b). The NPS, which intrudes into old, isotopically evolved crust of the Nain and Churchill
Provinces (Archaean–Palaeoproterozoic) has Nd isotopic compositions up to 10 ε -units lower than the MMIS and RAP, which
intrude into lithosphere of the Palaeoproterozoic Grenvillian and
Mesoproterozoic Sveconorwegian provinces (Ashwal, 1993; Ashwal
et al., 1986). In these younger terranes, the initial isotopic compositions of the magmas from which the highest-pressure megacrysts
crystallized has only been slightly displaced from depleted mantle, indicating that the parental magmas forming both the HAOMs
and anorthosites were derived from melting of the depleted mantle. Melting of the lower crust alone to produce the necessary
mafic melts requires unreasonably high degrees of melting and an
Nd and Pb isotopic geochronology and geochemistry of Proterozoic anorthosites and ubiquitous, cogenetic high-pressure orthopyroxene megacrysts suggests that these intrusions are derived from
melting of the depleted mantle in long-lived Andean-type arc systems. The ca. 100 m.y. time differences between HAOM crystallization and anorthosite emplacement, as well as isochronous HAOM
compositions are evidence that these phases were in equilibrium
with a fractionating, ponded magma that received compositionally
indistinct lower crustal contamination and/or magmatic recharge
at upper mantle or lower crustal pressures. Although products of
cumulate formation in ponded magmas at Moho are rarely brought
to surface, the viscous plagioclase-rich magmas forming Proterozoic anorthosites entrain evidence from the crust–mantle boundary
in the form of high-Al megacrysts, thereby providing an accessible proxy for perhaps more widespread, but less visible crustal
differentiation that occurs in silicic systems. Our results indicate
that crustal differentiation by cumulate formation and sequestration at the Moho is an important mechanism that plays a significant role in creation of the distinctive composition of Proterozoic
anorthosite massifs, but that could also explain the intermediate
(SiO2 > 60%) compositions of the bulk continental crust. These
magmatic processes operating at the Moho should be taken into
account in future studies on the evolution and differentiation of
the crust, along with more popular processes like crustal delamination.
Acknowledgements
G.M.B. acknowledges the School of Geosciences and Faculty of
Science at the Univ. of Witwatersrand for financial support as well
as the staff at the Department of Terrestrial Magnetism (Carnegie
Institute) for technical training and support during a pre-doctoral
fellowship. C.F. Gower and B. Ryan are thanked for field expedition
assistance. L.D.A. acknowledges funding from the South African National Research Foundation. A Center of Excellence grant from the
Norwegian Research Council to PGP funded T.B.A.
Appendix A. Supplementary information
Supplementary material related to this article can be found online at http://dx.doi.org/10.1016/j.epsl.2013.12.015.
References
Anders, E., Grevesse, N., 1989. Abundances of the elements: Meteoritic and solar.
Geochim. Cosmochim. Acta 53, 197–214.
Andersen, T., Griffin, W.L., 2004. Lu–Hf and U–Pb isotope systematics of zircons
from the Storgangen intrusion, Rogaland Intrusive Complex, SW Norway: implications for the composition and evolution of Precambrian lower crust in the
Baltic Shield. Lithos 73, 271–288.
Anderson, D.L., 1975. Chemical plumes in the mantle. Geol. Soc. Am. Bull. 86,
1593–1600.
Arndt, N.T., Goldstein, S.L., 1989. An open boundary between lower continental crust
and mantle: its role in crust formation and recycling. In: Ashwal, L.D. (Ed.),
Growth of the Continental Crust. Tectonophysics 161, 201–212.
Artemieva, I., 2011. The Lithosphere: An Interdisciplinary Approach. Cambridge University Press.
Ashwal, L., 1993. Anorthosites. Springer-Verlag.
Ashwal, L.D., Wooden, J.L., 1983. Isotopic evidence from the eastern Canadian shield
for geochemical discontinuity in the Proterozoic mantle. Nature 306, 679–680.
84
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
Ashwal, L., Wooden, J., Emslie, R., 1986. Sr, Nd and Pb isotopes in Proterozoic intrusives astride the Grenville Front in Labrador: Implications for crustal contamination and basement mapping. Geochim. Cosmochim. Acta 50, 2571–2585.
Berg, J.H., 1969. Petrology of anorthosites of the Bitterroot Range, Montana. In:
Isachsen, Y.W. (Ed.), Origin of Anorthosite and Related Rocks, Memoir 18, New
York State Museum and Science Service, pp. 387–398.
Berg, J.H., 1977. Regional Geobarometry in the Contact Aureoles of the Anorthositic
Nain Complex, Labrador. J. Petrol. 18, 399–430.
Bingen, B., Nordgulen, Ø., Viola, G., 2008. A four-phase model for the Sveconorwegian orogeny, SW Scandinavia. Norwegian J. Geol. 88, 43–72.
Bogdanova, S.V., et al., 2008. The East European Craton (Baltica) before and during
the assembly of Rodinia. Precambrian Res. 160, 23–45.
Bybee, G.M., Ashwal, L.D., Andersen, T.B., in preparation. New evidence for an accretionary Sveconorwegian orogen using high-pressure megacrysts from Rogaland
anorthosites.
Charlier, B., Duchesne, J.C., Vander Auwera, J., Storme, J.Y., Maquil, R., Longhi,
J., 2010. Polybaric fractional crystallization of high-alumina basalt parental
magmas in the egersund-ogna massif-type anorthosite (Rogaland, SW Norway) constrained by plagioclase and high-alumina orthopyroxene megacrysts.
J. Petrol. 51, 2515–2546.
Conrad, W.K., Kay, R.W., 1984. Ultramafic and mafic inclusions from adak island:
crystallization history, and implications for the nature of primary magmas and
crustal evolution in the Aleutian Arc. J. Petrol. 25, 88–125.
Corrigan, D., Hanmer, S., 1997. Anorthosites and related granitoids in the Grenville
orogen: A product of convective thinning of the lithosphere? Geology 25, 61–64.
Davidson, J., Charlier, B., Hora, J.M., Perlroth, R., 2005. Mineral isochrons and isotopic
fingerprinting: Pitfalls and promises. Geology 33, 29–32.
DePaolo, D., 1981. Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization. Earth Planet. Sci. Lett. 53, 189–202.
Dewey, J.F., Burke, K.C.A., 1973. Tibetan, Variscan, and Precambrian basement reactivation: products of continental collision. J. Geol. 81, 683–692.
Ducea, M., 2001. The California arc: Thick granitic batholiths, eclogitic residues,
lithospheric-scale thrusting, and magmatic flare-ups. GSA Today 11, 4–10.
Duchesne, J.-C., 2001. Anorthosites: the lower crustal connection. EUROPROBE, 1–4
November 2011, St Petersburg.
Duchesne, J.C., Liégeois, J.P., Vander Auwera, J., Longhi, J., 1999. The crustal tongue
melting model and the origin of massive anorthosites. Terra Nova 11, 100–105.
Dymek, R., Gromet, L., 1984. Nature and origin of orthopyroxene megacrysts from
the St-Urbain anorthosite massif Quebec. Can. Mineral. 22, 297–326.
Emslie, R.F., 1975. Pyroxene megacrysts from anorthositic rocks: New clues to the
sources and evolution of the parent magmas. Can. Mineral. 13, 138–145.
Emslie, R.F., 1976. Mealy Mountains complex, Grenville Province, southern Labrador.
In: Current Research, Part A. Geol. Survey Canada Paper 76-1A, 165–170.
Emslie, R.F., 1978. Anorthosite massifs, rapakivi granites, and late proterozoic rifting
of North America. Precambrian Res. 7, 61–98.
Emslie, R.F., 1990. Ages and petrogenetic significance of igneous mangeritecharnockite suites associated with massif anorthosites, Grenville Province. J.
Geol. 98, 213–231.
Emslie, R.F., Bonardi, M., 1979. Anorthositic Massifs and Associated Rocks of the
Mealy Mountains Complex, Grenville Province, Southern Labrador. GACMAC.
Emslie, R.F., Hegner, E., 1993. Reconnaissance isotopic geochemistry of anorthosite–
mangerite–charnockite–granite (AMCG) complexes, Grenville Province, Canada.
Chem. Geol. 106, 279–298.
Emslie, R.F., Loveridge, W.D., Stevens, R.D., Sullivan, R.W., 1983. Igneous and
tectonothermal evolution, Mealy Mountains, Labrador. Program Abstr. - Geol.
Assoc. Can. 8, A-20.
Emslie, R., Hamilton, M., Theriault, R., 1994. Petrogenesis of a mid-Proterozoic
Anorthosite–Mangerite–Charnockite–Granite (AMCG) complex: Isotopic and
chemical evidence from the nain plutonic suite. J. Geol. 102, 539–558.
Fram, M., Longhi, J., 1992. Phase-equilibria of dikes associated with proterozoic
anorthosite complexes. Am. Mineral. 77, 605–616.
Ghiorso, M.S., Sack, R.O., 1995. Chemical mass transfer in magmatic processes IV.
A revised and internally consistent thermodynamic model for the interpolation and extrapolation of liquid–solid equilibria in magmatic systems at elevated
temperatures and pressures. Contrib. Mineral. Petrol. 119, 197–212.
Gleißner, P., Drueppel, K., Romer, R.L., 2011. The role of crustal contamination in
massif-type anorthosites, new evidence from Sr–Nd–Pb isotopic composition of
the Kunene Intrusive Complex, NW Namibia. Precambrian Res. 185, 18–36.
Gower, C.F., Kamo, S., Krogh, T.E., 2008a. Indentor tectonism in the eastern Grenville
Province. Precambrian Res. 167, 201–212.
Gower, C.F., Kamo, S.L., Kwok, K., Krogh, T.E., 2008b. Proterozoic southward accretion and Grenvillian orogenesis in the interior Grenville Province in eastern Labrador: Evidence from U–Pb geochronological investigations. Precambrian
Res. 165, 61–95.
Hamilton, W.B., 1981. Crustal evolution by arc magmatism. In: Moorbath, S., Windley, B.F. (Eds.), The Origin and Evolution of Earth’s Continental Crust. In: Philos.
Trans. R. Soc. Lond., vol. 301, pp. 279–291.
Hegner, E., Emslie, R.F., Iaccheri, L.M., Hamilton, M.A., 2010. Sources of the
Mealy Mountains and Atikonak River anorthosite–granitoid complexes, Grenville
Province, Canada. Can. Mineral. 48, 787–808.
Heinonen, A., et al., 2010. Formation and fractionation of high-Al tholeiitic magmas
in the Ahvenisto Rapakivi Granite Massif-type anorthosite complex, SE Finland.
Can. Mineral. 48, 969–990.
Hervé, F., Pankhurst, R.J., Fanning, C.M., Calderón, M., Yaxley, G.M., 2007. The
South Patagonian batholith: 150 my of granite magmatism on a plate margin.
Lithos 97, 373–394.
Hoffman, P., 1989. Speculations on Laurentia’s first gigayear (2.0 to 1.0 Ga). Geology 17, 135–138.
Icenhower, J., Dymek, R., Weaver, B., 1998. Evidence for an enriched mantle
source for jotunite (orthopyroxene monzodiorite) associated with the St. Urbain
anorthosite, Quebec. Lithos 42, 191–212.
Jagoutz, O., Schmidt, M.W., 2012. The formation and bulk composition of modern
juvenile continental crust: The Kohistan arc. Chem. Geol. 298–299, 79–96.
Jagoutz, O., Müntener, O., Burg, J., Ulmer, P., Jagoutz, E., 2006. Lower continental
crust formation through focused flow in km-scale melt conduits: The zoned ultramafic bodies of the Chilas Complex in the Kohistan Island arc (NW Pakistan).
Earth Planet. Sci. Lett. 242, 320–342.
Jull, M., Keleman, P.B., 2001. On the conditions for lower crustal convective instability. J. Geophys. Res. 106, 6423–6446.
Kay, R.W., Kay, S.M., 1991. Creation and destruction of lower continental crust. Geol.
Rundsch. 80, 259–278.
Kay, R.W., Kay, S.M., 1993. Delamination and delamination magmatism. Tectonophysics 219, 177–189.
Kushiro, I., 1980. Viscosity, density, and structure of silicate melts at high pressures,
and their petrological applications. In: Hargraves, R.B. (Ed.), Physics of Magmatic
Processes. Princeton University Press, Princeton, NJ, pp. 93–120.
Kushiro, I., Fuji, T., 1977. Floatation of plagioclase in magma at high pressures and
its bearing on the origin of anorthosite. Proc. Jpn. Acad. Sci. 53, 262–266.
Lee, C.-T.A., Cheng, X., Horodyskyj, U., 2006. The development and refinement of
continental arcs by primary basaltic magmatism, garnet pyroxenite accumulation, basaltic recharge and delamination: insights from the Sierra Nevada, California. Contrib. Mineral. Petrol. 151, 222–242.
Longhi, J., 2005. A mantle or mafic crustal source for Proterozoic anorthosites?.
Lithos 83, 183–198.
Longhi, J., Fram, M., Vander Auwera, J., Montieth, J., 1993. Pressure effects, kinetics,
and rheology of anorthositic and related magmas. Am. Mineral. 78, 1016–1030.
Longhi, J., Vander Auwera, J., Fram, M., 1999. Some phase equilibrium constraints
on the origin of Proterozoic (Massif) anorthosites and related rocks. J. Petrol. 40,
339–362.
Martignole, J., Schrijver, K., 1970. Tectonic setting and evolution of the Morin
anorthosite, Grenville Province, Quebec. Bull. Geol. Soc. Finl. 42, 165–209.
Martignole, J., Machado, N., Nantel, S., 1993. Timing of intrusion and deformation of
the Rivière–Pentecôte anorthosite (Grenville Province). J. Geol. 101, 652–658.
McLelland, J.M., Selleck, B.W., Hamilton, M.A., Bickford, M.E., 2010. Late- to postTectonic setting of some major Proterozoic Anorthosite–Mangerite–Charnockite,
Granite (AMCG) Suites. Can. Mineral. 48, 729–750.
Menuge, J.F., 1988. The petrogenesis of massif anorthosites: a Nd and Sr isotopic
investigation of the Proterozoic of Rogaland/Vest-Agder, SW Norway. Contrib.
Mineral. Petrol. 98, 363–373.
Mitchell, J., Scoates, J., Frost, C., 1995. High-Al gabbros in the Laramie Anorthosite
Complex, Wyoming – Implications for the composition of melts parental to Proterozoic anorthosite. Contrib. Mineral. Petrol. 119, 166–180.
Mitchell, J., Scoates, J., Frost, C., Kolker, A., 1996. The geochemical evolution of
anorthosite residual magmas in the Laramie Anorthosite Complex, Wyoming.
J. Petrol. 37, 637–660.
Morse, S.A., 1969. Layered intrusions and anorthosite genesis. In: Isachsen, Y.W.
(Ed.), Origin of Anorthosite and Related Rocks, Memoir 18, New York State Museum and Science Service, pp. 175–188.
Morse, S., 1975. Plagioclase lamellae in hypersthene, Tikkoatokhakh Bay, Labrador.
Earth Planet. Sci. Lett. 26, 331–336.
Morse, S., 1991. Basaltic magma from the crust is not a free option. Eos Trans.
AGU 72, 161.
Morse, S., 2006. Labrador massif anorthosites: Chasing the liquids and their sources.
Lithos 89, 202–221.
Morse, S.A., 1982. A partisan review of Proterozoic anorthosites. Am. Mineral. 67,
1087–1100.
Müntener, O., Kelemen, P.B., 2001. The role of H2 O during crystallisation of primitive arc magmas under uppermost mantle conditions and genesis of igneous
pyroxenites: an experimental study. Contrib. Mineral. Petrol. 141, 643–658.
Müntener, O., Ulmer, P., 2006. Experimentally derived high-pressure cumulates from
hydrous arc magmas and consequences for the seismic velocity structure of
lower arc crust. Geophys. Res. Lett. 33, L21308.
Myers, J.S., Voordouw, R.J., Tettelaar, T.A., 2008. Proterozoic anorthosite–granite Nain
batholith: structure and intrusion processes in an active lithosphere-scale fault
zone, northern Labrador. Can. J. Earth Sci. 45, 909–934.
Nägler, Th.F., Kramers, J.D., 1998. Nd isotopic evidence of the upper mantle during
the Precambrian: models, data and the uncertainty of both. Precambrian Res. 91,
233–252.
Olson, K., 1992. The Petrology and geochemistry of mafic igneous rocks in the
anorthosite-bearing Adirondack Highlands, New York. J. Petrol. 33, 471–502.
G.M. Bybee et al. / Earth and Planetary Science Letters 389 (2014) 74–85
Othman, D.B., Polvé, M., Allègre, C.J., 1984. Nd–Sr isotopic composition of granulites
and constraints on the evolution of the lower continental crust. Nature 307,
510–515.
Owens, B., Dymek, R., 1995. Significance of pyroxene megacrysts for massif
anorthosite petrogenesis – constraints from the Labrieville, Quebec, Pluton. Am.
Mineral. 80, 144–161.
Palme, H., O’Neill, H.S.C., 2003. Cosmochemical estimates of mantle composition.
Treat. Geochem. 2, 1–38.
Peck, W., Valley, J., 2000. Large crustal input to high delta O-18 anorthosite massifs of the southern Grenville Province: new evidence from the Morin Complex,
Quebec. Contrib. Mineral. Petrol. 139, 402–417.
Peck, W.H., Clechenko, C.C., Hamilton, M.A., Valley, J.W., 2010. Oxygen isotopes in
the Grenville and Nain AMCG suites: Regional aspects of the crustal component
in massif anorthosites. Can. Mineral. 48, 763–786.
Philpotts, A.R., 1969. Parental magma of the anorthosite-mangerite suite. In: Isachsen, Y.W. (Ed.), Origin of Anorthosite and Related Rocks, Memoir 18, New York
State Museum and Science Service, pp. 207–212.
Rehkämper, M., Hofmann, A., 1997. Recycled ocean crust and sediment in Indian
Ocean MORB. Earth Planet. Sci. Lett. 147, 93–106.
Richards, M., Contreras-Reyes, E., Lithgow-Bertelloni, C., Ghiorso, M., Stixrude ,
L., 2013. Petrological interpretation of deep crustal intrusive bodies beneath
oceanic hotspot provinces. Geochem. Geophys. Geosyst. 14. http://dx.doi.org/
10.1029/2012GC004448.
Rivers, T., Corrigan, D., 2000. Convergent margin on southeatern Laurentia during
the Mesoproterozoic: tectonic implications. Can. J. Earth Sci. 37, 359–383.
Rudnick, R., Gao, S., 2003. Composition of the continental crust. Treat. Geochem. 3,
1–64.
Ryan, B., 2000. The Nain-Churchill boundary and the Nain Plutonic Suite: A regional
perspective on the geologic setting of the Voisey’s Bay Ni–Cu–Co deposit. Econ.
Geol. 95, 703–724.
Sauer, S., Slagstad, T., Andersen, T., Kirkland, C.L., 2013. Zircon Lu–Hf isotopes
in high-alumina orthopyroxene megacrysts from the Neoproterozoic Rogaland
Anorthosite Province, SW Norway: A window into the Sveconorwegian lower
crust. Geophys. Res. Abstr. 15, 1.
Schärer, U., Wilmart, E., Duchesne, J., 1996. The short duration and anorogenic character of anorthosite magmatism: U–Pb dating of the Rogaland Complex, Norway.
Earth Planet. Sci. Lett. 139, 335–350.
85
Schiellerup, H., Lambert, R., Prestvik, T., Robins, B., McBride, J., Larsenk, R., 2000.
Re–Os isotopic evidence for a lower crustal origin of massif-type anorthosites.
Nature 405, 781–784.
Scoates, J.S., Chamberlain, K.R., 1997. Orogenic to Post-Orogenic Origin For the
1.76 Ga Horse Creek anorthosite complex. J. Geol. 105, 331–344.
Slagstad, T., Marker, M., Røhr, T.S., Schiellerup, H., 2013. A non-collisional, accretionary Sveconorwegian orogen. Terra Nova 25 (1), 30–37. http://dx.doi.org/
10.1111/ter.12001.
Tatsumi, Y., 2000. Continental crust formation by crustal delamination in subduction
zones and complementary accumulation of the enriched mantle I component in
the mantle. Geochem. Geophys. Geosyst. 1.
Taylor, S., et al., 1984. A lower crustal origin for massif-type anorthosites. Nature 311, 372–374.
Valley, J., O’Neil, J.R., 1982. Oxygen isotope evidence for shallow emplacement of
Adirondack anorthosite. Nature 300, 497–500.
Van Kranendonk, M.J., 1996. Tectonic evolution of the Palaeoproterozoic Torngat
orogen: Evidence from pressure–temperature–time–deformation paths in the
North River map area, Labrador. Tectonics 15, 843–869.
Vander Auwera, J., Bolle, O., Bingen, B., Liégeois, J.P., Bogaerts, M., Duchesne, J.C.,
De Waele, B., Longhi, J., 2011. Sveconorwegian massif-type anorthosites and related granitoids result from post-collisional melting of a continental arc root.
Earth-Sci. Rev. 107, 375–397.
Vigneresse, J.L., 2005. The specific case of the Mid-Proterozoic rapakivi granites and
associated suite within the context of the Columbia supercontinent. Precambrian Res. 137, 1–34.
Wiebe, R.A., 1986. Lower crustal cumulate nodules in Proterozoic dikes of the
Nain Complex: Evidence for the origin of Proterozoic anorthosites. J. Petrol. 27,
1253–1275.
Wiebe, R.A., 1990. Evidence for unusually feldspathic liquids in the Nain Complex,
Labrador. Am. Mineral. 75, 1–12.
Wilson, M., 2007. Igneous Processes – A Global Tectonic Approach. Springer. 466 pp.
Xue, S., Morse, S.A., 1993. Geochemistry of the Nain Massif Anorthosite,
Labrador: Magma diversity in five intrusions. Geochim. Cosmochim. Acta 57,
3925–3948.
Zajacz, Z., Kovacs, I., Szabo, C., Halter, W., Pettke, T., 2007. Evolution of mafic alkaline melts crystallized in the uppermost lithospheric mantle: a melt inclusion study of olivine–clinopyroxenite xenoliths, Northern Hungary. J. Petrol. 48,
853–883.
Download