Received: 9 August 2018 Revised: 24 February 2019 Accepted: 28 February 2019
DOI: 10.1002/gj.3512R E S E A R CH AR T I C L EUltramafic–mafic and granitoids supra‐subduction magmatism
in the southern Ashanti volcanic belt, Ghana: Evidence from
geochemistry and Nd isotopesSamuel B. Dampare1,2 | Tsugio Shibata3 | Daniel K. Asiedu4 | Osamu Okano3 |
Shiloh K.D. Osae1,2 | David Atta‐Peters4 | Patrick A. Sakyi41School of Nuclear and Allied Sciences,
University of Ghana‐Atomic, Legon‐Accra,
Ghana
2National Nuclear Research Institute, Ghana
Atomic Energy Commission, Legon‐Accra,
Ghana
3Department of Earth Sciences, Okayama
University, Okayama, Japan
4Department of Earth Science, University of
Ghana, Legon‐Accra, Ghana
Correspondence
Samuel B. Dampare, School of Nuclear and
Allied Sciences, University of Ghana‐Atomic,
P.O. Box LG 80, Legon‐Accra, Ghana.
Email: sbdampare@ug.edu.gh
Funding information
Japanese Government (Monbukagakusho:
MEXT) and Okayama University; National
Nuclear Research Institute (NNRI) of the
Ghana Atomic Energy Commission (GAEC)2.75 Ga. The pyroxenitic and gabbroic rocks were most likely originated from a
Handling Editor: Y. LiuGeological Journal. 2019;1–37.Geochemical and Nd isotope data are presented for Palaeoproterozoic ultramafic to
mafic rocks and granitoids which are associated with volcanic rocks in the southern
part of the Ashanti greenstone belts of Ghana. The Prince's Town granitoids display
subduction‐zone geochemical features, with some showing signatures similar to
high‐SiO2 adakites (HSA). Their initial εNd (2.1 Ga) values range from −1.01 to
+2.92, and they have TDM2 of 2.17–2.51 Ga. The gabbros show slightly LREE‐
depleted and ‐enriched patterns, Th–U troughs, negative Nb, Zr–Hf and Ti anomalies,
spikes in Sr and Pb, have εNd (2.1 Ga) values ranging from −1.23 to +5.23, and TDM2
values from 2.08 to 2.57 Ga. The Ahama ultramafic body is characterized by two
groups of pyroxenites. Both groups show LREE enrichment patterns, display Th–U
and Nb–Ta troughs, minor or pronounced negative Zr–Hf and Ti anomalies. However,
the Group I pyroxenites exhibit minor Ce anomalies and lower total REE contents
(28.1–32.8 ppm) whereas Group II pyroxenites show significant negative Ce anoma-
lies and relatively higher total REE contents of 57.1–124 ppm. The pyroxenites show
negative initial εNd (2.1 Ga) values (−0.58 to −5.68) and have TDM2 values of 2.52 to
subduction‐related lithospheric mantle. The Aketakyi ultramafic–mafic complex is
made up of mainly cumulate dunite, harzburgite, pyroxenites, and gabbro. The rocks
display LREE‐depleted to flat patterns with low total REE contents (1.71 to
22.3 ppm) and demonstrate a supra‐subduction affinity. They show high positive ini-
tial εNd (2.1 Ga) values (+3.69 to +4.93), and TDM2 values of 1.99 to 2.04 Ga, which
suggest that they were derived from depleted mantle magmas and were juvenile at
their time of formation. The Nd isotopic data provides evidence for a possible con-
tamination of the juvenile Birimian crust of the southern Ashanti belt by some
amount of a pre‐Birimian (or Archaean?) crustal material. The trace element and iso-
topic signatures as well as the field relations of the studied granitoids and mafic–
ultramafic rocks show they were derived through supra‐subduction‐related
magmatism during the Palaeoproterozoic.
KEYWORDS
Birimian, geochemistry, mafic–ultramafic rocks, Nd isotopes, Palaeoproterozoicwileyonlinelibrary.com/journal/gj © 2019 John Wiley & Sons, Ltd. 1
2 DAMPARE ET AL.1 | INTRODUCTION
Palaeoproterozoic (Birimian) rocks form a significant part of the West
African Craton (WAC; Figure 1). The Palaeoproterozoic and Archaean
terranes of the WAC have been recognized as a major zone of crustal
growth (e.g., Abouchami, Boher, Michard, & Albarede, 1990; Boher,
Abouchami, Michard, Albarède, & Arndt, 1992; Taylor, Moorbath,
Leube, & Hirdes, 1992). It has been shown from a large‐scale isotopic
studies that the main Palaeoproterozoic crustal growth event (or
Eburnean orogeny) in the WAC represents major juvenile crust‐
forming process with involvement of only a negligible Archaean
crustal components (e.g., Abouchami et al., 1990; Boher et al., 1992;
Liégeois, Claessens, Camara, & Klerkx, 1991; Pawlig et al., 2006;
Taylor et al., 1992). The Palaeoproterozoic terranes of theWAC greatly
resemble the Archaean greenstone terranes, particularly in their struc-
tural style and composition, and like the Archaean, are surrounded
with controversies regarding the tectonic processes that formed them.
In terms of the tectonic setting in which the rocks were formed, inter-
pretations vary between generation of the Birimian juvenile crust in
arc environments (e.g., Asiedu et al., 2004; Dampare et al., 2008; Mor-
timer, 1992; Pohl & Carlson, 1993; Sylvester & Attoh, 1992) or from
plume‐related magmatism (e.g., Abouchami et al., 1990). Most of theseFIGURE 1 Geologic sketch of the Leo‐Man Shield of the West Africa Cra
belts and the location of the study area (adopted after Attoh, Evans, & Bicmodels were based on isotopic and geochemical compositions of volca-
nic rocks from these greenstone belts, with some contributions from
the sedimentary units and the associated granitoid intrusions, whereas
the associated mafic–ultramafic intrusives have received little atten-
tion. Although voluminously minor, the mafic–ultramafic rocks may
provide useful information to enable us to understand the geodynamic
evolution of the Palaeoproterozoic greenstone belts. The mafic–
ultramafic rocks and the granitoids are believed to provide evidence
for juvenile crust‐forming events near the Archaean–Proterozoic
boundary (Attoh & Ekwueme, 1997; Hirdes & Davis, 2002; Taylor
et al., 1992).
Several ultramafic–mafic intrusives are associated with the volca-
nic rocks and granitoids in the Palaeoproterozoic southern Ashanti
greenstone belt of Ghana. These include the Aketakyi ultramafic com-
plex (the largest ultramafic–mafic body in Ghana), the Ahama ultra-
mafic body, and the Axim, Kegyina, and Aketakyi mafic bodies. This
study has focused on the intrusive rocks in the Axim–Aketakyi area
of the southern Ashanti volcanic belt. Consequently, the Prince'sTown
pluton is considered here together with the mafic–ultramafic rocks to
enhance our understanding of the geodynamic setting of the green-
stone belt. Petrographic and geochemical studies of some of the rep-
resentative intrusions are available in Loh and Hirdes (1999) andton (WAC; inset) showing the Palaeoproterozoic (Birimian) greenstone
kford, 2006)
DAMPARE ET AL. 3Attoh et al. (2006) and provide important background information for
this study. The objectives of this paper are (a) to present chemical data
of Palaeoproterozoic plutonic rocks exposed in the southern Ashanti
greenstone terrane in south‐western Ghana and (b) to discuss the pet-
rogenesis of the plutonic rocks and the tectonic setting during their
emplacement.2 | GEOLOGICAL SETTING
Palaeoproterozoic rocks of the Birimian terrane of the Leo‐Man Shield
(Figure 1) form a significant portion of the WAC (Bessoles, 1977) to
the east and north of the Archaean Liberian cratonic nucleus. The
Palaeoproterozoic terrane, characterized by narrow sedimentary
basins and linear, arcuate volcanic belts, is intruded by several gener-
ations of granitoids (Doumbia et al., 1998; Hirdes, Davis, Lüdtke, &
Konan, 1996; Leube, Hirdes, Mauer, & Kesse, 1990) and corresponds
to a period of accretion during the 2.1–2.0 Ga Eburnean orogeny
(e.g., Abouchami et al., 1990). The Eburnean tectono‐thermal event
was not only accompanied by the deformation and the emplacement
of syn‐ to post‐orogenic tonalite–trondhjemite–granodiorite (TTG)
and granite plutons along fractures (e.g., Feybesse et al., 2006; Leube
et al., 1990; Vidal & Alric, 1994) but also resulted in widespread meta-
morphism mostly under greenschist‐facies conditions (e.g., Oberthür,
Vetter, Davis, & Amanor, 1998; Béziat et al., 2000). Also, evidence
of medium amphibolite–facies metamorphism may be observed in
the vicinity of granitoid plutons (e.g., Debat et al., 2003; Eisenlohr &
Hirdes, 1992; Junner, 1940). Recently, John, Klemd, Hirdes, and Loh
(1999), on the basis of mineral chemistry, have suggested that the
entire Ashanti belt of south‐eastern Ghana underwent epidote–
amphibolite–facies metamorphism (T = 500–650°C and P = 5–6 kbar)
before experiencing retrograde metamorphism under the greenschist‐
facies conditions.2.1 | Lithostratigraphy
The lithostratigraphic succession of these Palaeoproterozoic rocks (i.e.,
the volcano‐sedimentary and bimodal volcanics) has over the years
been contentious with respect to whether the sedimentary unit lies
below (e.g., Feybesse & Milési, 1994; Junner, 1940; Milési, Ledru,
Feybesse, Dommanget, & Marcoux, 1992) or above (e.g., Asiedu
et al., 2004; Hirdes et al., 1996; Tagini, 1971) the volcanic unit. It is
now widely accepted that the two units formed quasi‐
contemporaneously, as lateral facies equivalents (e.g., Leube et al.,
1990). According to Béziat et al. (2008), the lithostratigraphic succes-
sion of the Birimian crust sensu lato could be established from bottom
to top as (a) a thick sequence of mafic rocks, including basalt, locally
pillowed, as well as dolerite and gabbro, all of tholeiitic composition,
locally interlayered with immature detrital sediments and limestone;
(b) a detrital sedimentary pile (volcanics, turbidite, mudstone and car-
bonate) including interbedded calc‐alkaline volcanics; and (c) a coarse
clastic sedimentary sequence belonging to the Tarkwaian Group.2.1.1 | Sediments and volcanics
The Palaeoproterozoic supracrustal rocks of Ghana are subdivided
into the Birimian and Tarkwaian. The Birimian rocks comprise a series
of subparallel, roughly equal‐spaced north‐east trending volcanic belts
of volcanic rocks separated by an assemblage of volcano‐sedimentary
rocks. Traditionally, the Birimian rocks were classified into a two‐fold
lithostratigraphic and chronological system, consisting of an older
metasedimentary sequence (referred to as Lower Birimian) and a
younger volcanic sequence of predominantly tholeiitic basalt volcanic
and pyroclastic rocks (referred to as Upper Birimian; e.g., Junner,
1940). Leube et al. (1990) have indicated that the volcanic and sedi-
mentary rock sequences formed contemporaneously as lateral facies
equivalents. The Birimian rocks are overlain by the detrital Tarkwaian
sedimentary rocks in almost all the prominent volcanic belts, and both
formations were subjected to similar deformation events, which
involved compression along a southeast ‐ northwest‐trending axis,
resulting in folding and thrusting with subsequent flattening and local-
ized oblique‐slip shearing (e.g., Allibone et al., 2002; Blenkinsop,
Schmidt‐Mumm, Kumi, & Sangmor, 1994; Eisenlohr, 1989; Eisenlohr
& Hirdes, 1992). The Tarkwaian rocks are made up of conglomerates,
sandstones, and subordinate shale, with much of the clastic material
in the sediments derived from the adjacent Birimian units (Tunks,
Selley, Rogers, & Brabham, 2004).
Radiometric dating on the Birimian of Ghana indicates that volca-
nic belts were deposited between 2,240 and 2,186 Ma whereas the
sediments were deposited 100–60 Ma later into basins (Davis, Hirdes,
Schaltegger, & Nunoo, 1994; Leube et al., 1990; Oberthür et al., 1998;
Pigois, Groves, Fletcher, McNaughton, & Snee, 2003). Recently, on the
basis of SHRIMP U–Pb analyses of detrital zircons, Pigois et al. (2003)
have constrained the maximum age of sedimentation of theTarkwaian
to 2,133 ± 4 Ma. The Birimian and Tarkwaian rocks host the major
gold deposits in Ghana, with most of the gold deposits concentrated
along the western flank of the Ashanti belt. The major epigenetic
lode‐gold event occurred late in the Eburnean orogeny, after the peak
of metamorphism was reached (John et al., 1999; Oberthür et al.,
1998; Pigois et al., 2003; Yao et al., 2001). Using hydrothermal rutile,
Oberthür et al. (1998) have obtained the age of hydrothermal alter-
ation to be 2,092 ± 3 and 2,086 ± 4 Ma. Recently, Pigois et al.
(2003) have determined the most robust age of 2,063 ± 9 Ma for gold
mineralization in Ghana from SHRIMP II U–Pb analyses of hydrother-
mal xenotime.
2.1.2 | Granites
The Palaeoproterozoic granitoids have been classified into two main
groups according to the presence or absence of amphibole (Lompo,
2009 and references therein): (a) the amphibole‐bearing granitoids
(PAG) and (b) the biotite ± muscovite‐bearing granitoids without
amphibole (PBG). The PAG granitoids are syntectonic intrusives in
the greenstone terranes and are older than the PBG granitoids, which
intrude the greenstone belts or the PAG granitoids (Lompo, 2009). On
the basis of 207Pb/206Pb and U–Pb zircon geochronological data, two
4 DAMPARE ET AL.main coeval plutonism and volcanism have been recognized and con-
sidered to mark two successive major episodes (i.e., Birimian sensu
stricto [2,185–2,150 Ma] and Bandamian [2,115–2,100 Ma]) of
Palaeoproterozoic crust formation in the West African Craton
(Doumbia et al., 1998; Hirdes et al., 1996; Hirdes & Davis, 2002).
However, the occurrence of some pre‐Birimian ages in detrital zircons
(2,245–2,266 Ma) from the Tarkwaian sediments (Davis et al., 1994;
Loh & Hirdes, 1999), in two zircon grains (2,208 and 2,220 Ma) from
theTafolo tonalite (Doumbia et al., 1998), in zircon cores from the Issia
granite (2,212–2,305 Ma) in south‐western Cote d'Ivoire (Kouamelan,
1996; Kouamelan, Delor, & Peucat, 1997), and in zircon cores from the
Dabakala tonalite (2,312 ± 17 Ma; Gasquet, Barbey, Adou, & Paquette,
2003) could probably suggest a pre‐Birimian crustal growth episode.
The Birimian terrane of Ghana is intruded by two main suites of
granitoids, namely, the Dixcove‐ (or belt‐type) and Cape Coast‐type
(or basin‐type) granitoids. The metaluminous, relatively Na‐rich,
Dixcove‐type granitoids, which are dominantly hornblende‐ to
biotite‐bearing granodiorite to diorite, monzonite and syenite, intrude
the Birimian volcanic belts and they may be coeval with the volcanic
rocks (Eisenlohr & Hirdes, 1992). On the other hand, the Cape
Coast‐type granitoids, which are predominantly peraluminous, two‐
mica granodiorites, with lesser hornblende‐ and biotite‐bearing grano-
diorites, are emplaced within the Birimian sedimentary basins. The
Dixcove granitoids have been dated at 2,172 ± 1.4 Ma in the Ashanti
Belt (Hirdes, Davis, & Eisenlohr, 1992) and the Cape Coast type
between 2,104 ± 2 and 2,123 ± 3 Ma in the Kumasi Basin (Oberthür
et al., 1998). Other types of granitoids include the localized, biotite
and hornblende Winneba granitoids, and the K‐rich Bongo granitoids
which cut the Tarkwaian rocks in the volcanic belts. The Cape
Coast‐, Kumasi‐, Dixcove‐, and Bongo‐type granitoids have strong
mantle affinities whereas the Winneba type has an Archaean sialic
precursor (Sm–Nd model age of ~2.6 Ga; Taylor et al., 1992). The
Winneba‐type granitoids are believed to be the only rock suite in
Ghana which show evidence for a significant magmatic contribution
from an Archaean continental crust in their genesis.
2.1.3 | Mafic and ultramafic rocks
The Birimian greenstone terrane is often characterized by
metavolcanic and metasedimentary rocks, which have been intruded
by granitoids and other mafic intrusive rocks. According to Béziat
et al. (2000), the Birimian mafic suite comprises both volcanic rocks
and plutonic complexes (ultrabasites and gabbros). Some of these plu-
tonic complexes appear to outcrop over wide areas; these include the
ultrabasic bodies at the eastern border of the Mako series, Eastern
Senegal and those from the Ashanti belt of Ghana.
As intimated earlier, discussions on the processes of growth of the
Proterozoic continental crust have mainly been driven by isotopic and
geochemical compositions of volcanic rocks from these greenstone
belts, with some contributions from the sedimentary units and the
associated granitoid intrusions. In spite of the fact that the mafic–
ultramafic rocks are voluminously minor, they could still provide useful
information to enable us to understand the geodynamic evolution ofthe Palaeoproterozoic greenstone belts. For example, Béziat et al.
(2000) have interpreted the Loraboué Birimian ultramafic–mafic
assemblage, outcropping in the Boromo greenstone belt of Burkina
Faso as the remains of a magma chamber that crystallized at the base
of an island arc. Also, Attoh et al. (2006) have suggested that the
rodingite‐bearing ultramafic complex and associated pillowed basalts
located in the Palaeoproterozoic (Birimian) Dixcove greenstone belt
in south‐western Ghana may represent fragments of a
Palaeoproterozoic oceanic lithosphere.
In the study area, ultramafic to mafic complexes as well as isolated
mafic bodies are associated with this greenstone–granitoid belt. The
mafic and ultramafic rocks form a minor part of the entire intrusive
bodies in the belt. The mafic rocks include the Axim gabbro, Kegyina
gabbro and gabbroic rocks from the Aketakyi area. These mafic bodies
appear to have been emplaced at different times (Loh & Hirdes, 1999).
Ultramafic rocks occur as two discrete bodies, namely, the
Aketakyi ultramafic complex and the Ahama ultramafic body. Both
bodies were emplaced in volcanic rocks and are separated by the
Prince's Town granitoid body with the Ahama ultramafic body located
to the west and the Aketakyi ultramafic complex to the east of the
granitoid body.2.2 | Tectonic setting
2.2.1 | Models
The context of crustal accretion and overall tectonic regime of the
Palaeoproterozoic (Birimian) rocks has been a subject of contrasted
interpretations. In terms of the tectonic setting in which the rocks
were formed, interpretations vary between generation of the Birimian
juvenile crust in arc environments (e.g., Asiedu et al., 2004; Dampare,
Shibata, Asiedu, & Osae, 2005; Mortimer, 1992; Pohl & Carlson, 1993;
Sylvester & Attoh, 1992) and from plume‐related magmatism (e.g.,
Abouchami et al., 1990; Lompo, 2009). Two main models have been
proposed to account for the tectonic processes involved in the forma-
tion of the rocks: accretionary orogeny versus transcurrent tectonics
models. In the first model, the Eburnean event is seen as the accre-
tionary orogeny that resulted from the collisions of island arcs and
oceanic plateaus with an Archaean craton. In this case, the Eburnean
is considered to be part of a much larger accretionary event that
involved fragments of Precambrian crust now preserved over large
parts of Africa and South America (Abouchami et al., 1990; Boher
et al., 1992; Davis et al., 1994; Feybesse & Milési, 1994; Hirdes
et al., 1996; Hirdes & Davis, 2002; Ledru, Johan, Milési, & Tegyey,
1994). The proponents of the second model intimate that the
Palaeoproterozoic terranes are largely the product of transcurrent tec-
tonics either in an oceanic domain that differentiated to produce con-
tinental crust or in association with Archaean basement but lacked any
clear evidence of collisional processes (e.g., Bassot, 1987; Doumbia
et al., 1998; Kouamelan, 1996). Apart from these two main models,
there are some workers who favour the involvement of subduction
processes in the generation of the Palaeoproterozoic terranes but
DAMPARE ET AL. 5prefer small‐scale extensional and accretionary processes to a large‐
scale collision (Ama‐Salah, 1996; Vidal & Alric, 1994).
The tectonic setting in which the Palaeoproterozoic rocks of
Ghana formed is contentious. One of the popular views is the tectonic
model proposed by Leube et al. (1990), where an evolution of the tec-
tonic setting of the Birimian/Eburnean in Ghana from an intracratonic‐
rift to oceanic‐spreading and finally to an accretion–collision‐related
setting is envisaged. The Birimian volcanic belts originated as chains
of volcanic islands developed during intracontinental rifting. The pyro-
clastic rocks which erupted from the volcanoes and epiclastic erosion
products of the volcanic belts were deposited synchronously in adja-
cent subsiding basins to form the Birimian sediments. The geochrono-
logical and isotopic data suggest that the Birimian of Ghana represents
a major Early Proterozoic magmatic crust‐forming event around 2.3–
2.0 Ga by differentiation from a slightly depleted mantle source (Tay-
lor et al., 1992). Accordingly, the Birimian of Ghana is considered as
part of a major Proterozoic (Eburnean) episode of juvenile crustal
accretion which has been recognized in other surrounding areas in
West Africa, where it has been dated at 2.2–2.1 Ga by Abouchami
et al. (1990). On the basis of isotopic data and occasional presence
of calc‐alkaline andesite‐dacite‐rhyolite sequences, the Birimian volca-
nic belts mostly represent island arc complexes and some are external
back‐arc basins (Pohl & Carlson, 1993). The authors have indicated
that the crustal shortening period of the Birimian involved a closure
of an oceanic basin (located between the Man Shield and the Nigeria
Craton) where multiple arc–arc collisions or, in the west, arc‐
continental platform collisions took place. It is believed that the mod-
ern paradigm of the Birimian situation is the compression of the
Indonesian‐Melanesian Archipelago and the intervening basins by
the northward migration of the Australian Plate towards the Asian
continent (Pohl & Carlson, 1993). On the basis of trace element geo-
chemistry of the Birimian volcanic rocks and field observations, the
Birimian greenstone belts were believed to have formed in oceanic
island arc environments similar to those inferred for the Late Archaean
Superior Craton (Sylvester & Attoh, 1992). Thus, the Birimian volcanic
rocks probably originated as immature island arcs built on oceanic
crust. It has also been observed that there exist similarities in regional
geological patterns and relative internal age relationships between the
Palaeoproterozoic Eburnean orogen of Ghana and the late Archaean
Kenoran orogen of the southern Superior Province in Canada (Davis
et al., 1994). In the southern Superior Province of Canada, deposition
of volcanogenic‐turbiditic basin sediments is suggested to be synchro-
nous with accretion of arcs and micro‐continental fragments against a
growing continental mass. The authors have, therefore, proposed that
the Birimian sedimentary basins resulted from accretion of arcs and
oceanic plateaus now represented by allochthonous Birimian volcanic
belts. Loh and Hirdes (1999), largely based on field observation, low
TiO2 contents as well as a calc‐alkaline differentiation trend for basalts
from the southern Ashanti belt, agree with Sylvester and Attoh (1992)
that the Birimian rocks erupted in a primitive island‐arc environment.
When metagreywackes and metapelites from the Birim diamondifer-
ous field were studied, they showed geochemical signatures similar
to their Archaean counterparts (Asiedu et al., 2004). The chemical dataof the rocks suggest that the rocks were derived from the basaltic to
dacitic volcanic rocks and granItoids within the Birimian greenstone
belts and were deposited in a tectonic setting comparable to modern
island arcs (Asiedu et al., 2004). Furthermore, the geochemical signa-
tures of the I‐type granitoids from the southern Ashanti belt are con-
sistent with those of granitoids emplaced in volcanic‐arc settings
(Dampare et al., 2005). Considering the fact that the belt granitoids
and volcanic rocks of the Birimian terrane of Ghana are coeval (Davis
et al., 1994; Hirdes et al., 1992), an island arc tectonic setting is
inferred for the Birimian belts (Dampare et al., 2005). On the basis
of metallogenesis model for the occurrence of gold deposits in
Ghanam, a combination of continental margin, juvenile magmatism
and convergence and collision between an old continent and a juvenile
crust has been implied for the tectonic environments in which the
Birimian greenstone belts were generated (Feybesse et al., 2006).
The authors have stressed that the Ashanti volcanic belt marks the
boundary between an Archaean continental domain and a Birimian
oceanic domain. However, the results of numerical modelling using
thermal parameters such as thermal conductivity values and heat‐
production rates led Harcouët, Guillou‐Frottier, Bonneville, Bouchot,
and Milesi (2007) to suggest that the basement in the Ashanti area
is likely to be of continental rather than of oceanic type. Using field
relationships, geochemistry, and the association of rodingite with the
mafic–ultramafic complex in the southern Ashanti belt, the complex
has recently been re‐interpreted as a fragment of Palaeoproterozoic
ophiolitic complex emplaced in a supra‐subduction zone (SSZ) tectonic
setting (Attoh et al., 2006). Also, an intra‐oceanic island arc–fore‐arc–
back‐arc setting is inferred for the Palaeoproterozoic metavolcanic
rocks associated with the mafic–ultramafic complex (Dampare,
Shibata, Asiedu, Osae, & Banoeng‐Yakubo, 2008). Thus, interpreta-
tions of tectonic models for the Birimian terrane of Ghana basically
varies between the intracratonic rift setting and island arc complexes.
The southern Ashanti volcanic belt follows the general geological
disposition of the Ashanti belt, which is characterized by alternating
NE–SW trending volcanic rocks and synvolcanic granitoids. The over-
all geological structural of the Ashanti belt is that of a synform,
whereby the overlying Tarkwaian rocks occupy the centre of the belt
and the Birimian volcanic rocks occur along the margins of the belt.
The north‐western margin of the Ashanti belt is strongly tectonized,
represented by the Prestea–Obuasi–Konongo high‐strain zone. The
southern portion of the Ashanti volcanic belt forms three branches
referred to as the Axim branch, Cape Three Point branch, and Butre
branch, with three plutons (locally termed as Prince's Town, Dixcove,
and Ketan pluton) occupying positions between these branches (Loh
& Hirdes, 1999). The volcanic branches are composed of basaltic and
andesitic lavas and pyroclastic rocks. The bulk of these andesitic rocks
occur in the Axim volcanic branch where they probably occupy a
stratigraphically upper position in the volcanic sequence of the south-
ern Ashanti belt (Loh & Hirdes, 1999). The volcanic branches and the
synvolcanic plutons are generally separated by normal faults/thrust
zones. Several mafic–ultramafic rocks are associated with the volcanic
rocks in the southern Ashanti volcanic belt (Figure 2). The age of the
metavolcanic rocks is mainly constrained by the ages of the granitoids.
6 DAMPARE ET AL.
FIGURE 2 Geological map of the southern Ashanti volcanic belt of Ghana showing the Axim (left) and Cape Three Points (right) volcanic
branches (after Loh & Hirdes, 1999)The precise TIMS U–Pb zircon ages obtained for the migmatites and
granitoids from the Ketan pluton, a tonalite from the Dixcove pluton,
and a granodiorite from the Prince's Town pluton are 2,172 ± 4 Ma
(Opare‐Addo, John, Mukasa, & Browning, 1993), 2,174 ± 2 Ma (Hirdes
et al., 1992) and 2,159 ± 4 Ma (Attoh et al., 2006), respectively. The
age of the Prince's Town pluton has constrained a minimum age for
the Aketakyi ophiolitic complex, as it intrudes the western flank of
the complex.
2.2.2 | Birimian and pre‐Birimian crust
Existing geochronological data to date appear to suggest two succes-
sive, major plutonism and volcanism episodes of Palaeoproterozoic
crust formation in the West African Craton. These include the so‐
called Birimian sensu stricto (2,185–2,150 Ma, occurring in Ghana,
eastern Cote d'Ivoire, and eastern part of Burkina Faso) and the so‐
called Bandamian (2,115–2,100 Ma, covering central and western
Cote d'Ivoire, western Burkina Faso, Guinea, and Mali). Gasquet
et al. (2003) obtained U–Pb ID‐TIMS and SIMS zircon ages in the
range of 2,103–2,162 Ma) and Nd model ages of 2,039–2,148 Ma
for rocks from the Dabakala area of Cote d'Ivoire. An age of
2,312 ± 17 Ma was, however, obtained for zircon cores from the
Dabakala tonalite, which also yielded an older Nd model age of
2,471 Ma. The authors interpreted this model age in the context of
recycling of old 2.3 Ga material rather than the contribution of an
ancient Archaean material due to the presence of 2,312 Ma old zirconcores and the low 87Sr/86Sr initial ratio of the rocks. Gasquet et al.
(2003) suggested that the Birimian and Bandamian events were prob-
ably preceded by an early Palaeoproterozoic crustal growth event at
~2.3 Ga (Gasquet et al., 2003). The existence of a pre‐Birimian event
in the West African Craton is suggested by abundant geochronological
and isotope geochemical data (e.g., Davis et al., 1994; Gasquet et al.,
2003; Lemoine, 1988; Loh & Hirdes, 1999; Pawlig et al., 2006). Pawlig
et al. (2006) demonstrated that this Early Birimian phase also occurred
in the Kedougou‐Kenieba‐Inlier (KKI) of Eastern Senegal, which had
previously been suggested to be among the youngest parts of the
Birimian terrane (Hirdes et al., 1996; Hirdes & Davis, 2002). Thus, such
an event is not limited to a particular region within the
Palaeoproterozoic terrane of the WAC. Isotopic data also appear to
suggest that the Eburnean orogeny represents a major juvenile
crust‐forming process with little or negligible involvement of an
Archaean continental crust. The Winneba area and the vicinity of
the Man nucleus have often been cited as the exceptional case, where
a stronger influence of recycled Archaean basement is observed (e.g.,
Kouamelan, Peucat, & Delor, 1997; Taylor et al., 1992).
Taylor et al. (1992) carried out Rb–Sr, Pb/Pb, and Sm–Nd isotopic
studies on selected Birimian metavolcanics, metasedimentary rocks
and granitoids from Ghana. The absence of contamination in the igne-
ous rocks, as indicated by the Sm/Nd ratios, led these authors to sug-
gest that there was little evidence for the involvement of significantly
older crust in the genesis of the volcanic‐plutonic rocks. Taylor et al.
(1992), therefore, discounted the possibility of an Archaean crust in
DAMPARE ET AL. 7Ghana. Feybesse et al. (2006) held a contrary view; that is, the basic
nature of the Palaeoproterozoic magmatism suggests a deep (i.e., man-
tle) origin and a very rapid rise, which probably accounted for the
absence of contamination. The isotopic signature of the Winneba
granitoids in south‐eastern Ghana provides evidence of magmatic
contribution from an older Archaean crust, which suggests that the
Winneba area is probably underlain by an Archaean basement (Taylor
et al., 1992). Other lines of evidence for the presence of an Archaean
basement in Ghana include (a) the recent discovery of diamondiferous
kimberlite in the Sunyani/Cape Coast Basin or the Akwatia and (b) the
presence of end‐Archaean to Siderian BIF‐bearing supracrustal rocks
in south‐eastern Ghana (Feybesse et al., 2006). The authors sug-
gested, therefore, that an Archaean basement probably exists below
the Sunyani Basin and south‐eastern Ghana, and that the Ashanti Belt
could represent the boundary between an Archaean continental
domain and a Birimian oceanic domain. The results of the numerical
modelling, using thermal parameters such as thermal conductivity
values and heat‐production rates, suggest that the basement in the
Ashanti area is likely to be of continental rather than of oceanic type
(Harcouët et al., 2007). These authors supported the view that the lack
of evidence of significantly older materials in Birimian rocks was prob-
ably due to the rapid nature of the ascending magma (e.g., Feybesse
et al., 2006) rather than absence of continental basement, as sug-
gested by Taylor et al. (1992).3 | FIELD RELATIONS AND PETROGRAPHY
The field relations among the various rock types of the southern
Ashanti volcanic belt are described in detail by Loh and Hirdes
(1999), who indicated that the volcanic rocks form three major NE–
SW trending lobes/branches referred to from east to west as the
Butre branch, Cape Three Points branch, and the Axim branch. The
volcanic branches are composed of basaltic and andesitic lavas, and
pyroclastic rocks. The bulk of these andesitic rocks occur in the Axim
volcanic branch where they probably occupy a stratigraphically upper
position in the volcanic sequence of the southern Ashanti belt (Loh &
Hirdes, 1999). Three granitoid intrusives, locally called the Prince's
Town, Dixcove, and Ketan pluton, occur at intervening positions
between the three major lobes/branches. Ultramafic to mafic com-
plexes as well as isolated mafic bodies are associated with this
greenstone–granitoid belt. The mafic and ultramafic rocks form a
minor part of the entire intrusive bodies in the belt.
The study area is characterized by a major structural element,
which is the high‐strain zone to the north‐west of Axim town in the
belt/basin transition zone on the western flank of the Ashanti belt
(Loh & Hirdes, 1999). This Axim high‐strain zone is the southern con-
tinuation of the Prestea–Obuasi–Konongo high‐strain zone, which
hosts several major gold mines in Ghana. In this zone, sediments of
the Kumasi Basin are thrust over the rocks of the Ashanti belt, includ-
ing the Tarkwaian, in an oblique manner. Apart from the NE/NNE
striking regional foliation, the several kilometre‐wide Axim high‐strain
zone is characterized by numerous N–S shears, which are notprominently developed elsewhere in the belt (Loh & Hirdes, 1999).
The effect of deformation in this high‐strain zone is mainly observed
in the incompetent argillitic sediments of the Kumasi Basin, whereas
rocks within the Ashanti belt itself appear relatively undeformed
(Loh & Hirdes, 1999).
3.1 | Prince's Town pluton
The Prince's Town granitoids have previously been studied by a num-
ber of workers including Loh and Hirdes (1999) and Dampare et al.
(2005). The pluton consists of granitic to dioritic rocks, which are gen-
erally massive but occasionally display an alignment of ferromagnesian
minerals. They are mostly greenish with pinkish tints and are typically
medium‐grained, but some of the granodiorites and the tonalites are
coarse‐grained. The mafic members (i.e., quartz diorites and diorites)
tend to characterize the margins of the pluton, and the Prince's Town
pluton is visualized as a single zoned magmatic intrusion (Loh &
Hirdes, 1999). Thin section analysis indicates that the granitoids con-
tain mainly plagioclase, K‐feldspar, quartz, amphibole, biotite, and
opaques. Accessory phases include zircon, apatite, and sphene. Plagio-
clase occurs as euhedral grains which display a variety of textures,
including albite twinning and compositional zoning. Plagioclase is
often sericitized and carbonated, and/or saussuritized, with the alter-
ation being intense at the core. Fresh plagioclase crystals also occur
in the rocks. K‐feldspar partly displays a microperthitic to perthitic tex-
ture and is normally wrapped by euhedral plagioclase grains. Micro-
cline is sericitized, carbonated, and/or saussuritized; fresh types are
also present. In some of the rocks, feldspar exhibits a myrmekitic tex-
ture. The amphiboles are brownish‐green in colour and have prismatic
shapes and are partially altered to epidote and chlorite (Figure 3a).
Biotite, mostly flaky, is often reddish‐brown and is commonly altered
to epidote and chlorite. Quartz is anhedral and displays undulatory
extinction as well as sutured grained boundaries. Some of the quartz
grains have been recrystallized and form fine‐grained mosaics in the
rocks. In the quartz diorites, quartz frequently occurs as interstitial
grains between plagioclase crystals. Apatite mostly occurs as stubby‐
prismatic grains. Sphene and zircon are present as subhedral and
euhedral crystals, respectively.
3.2 | Ultramafic rocks
The ultramafic–mafic rocks in the study area occur as two discrete
bodies, that is, Aketakyi ultramafic complex (UMC) and Ahama ultra-
mafic body.
3.2.1 | Aketakyi ultramafic complex
The Aketakyi UMC, which was previously mapped by Bartholomew
(1961) and Loh and Hirdes (1999), has recently been re‐interpreted
as a fragment of Palaeoproterozoic ophiolitic complex by Attoh et al.
(2006) based on geochemistry and the association of rodingite with
the UMC. The rocks are well exposed on the coast and form a jagged
coastline from Aketakyi to Cape Three Points, where they represent
8 DAMPARE ET AL.
FIGURE 3 Photomicrographs of selected plutonic rocks from the study area: (a) Prince's Town granodiorite showing the occurrence of green
amphibole (amp) in association with saussuritized plagioclase (plag); chlorite (chl) occurs as an alteration product of amphibole; Plane‐Polarized
Light. (b) Serpentinized peridotite of the Aketakyi ultramafic complex (UMC) showing relict olivine (ol) and orthopyroxene (opx) enclosed in a large
clinopyroxene (cpx) crystal; Olivine is partly replaced by serpentine (ser). Plane‐Polarized Light. Mt = magnetite. (c) Serpentinized peridotite of
Aketakyi UMC showing brown amphibole (amp) oikocryst poikilitically enclosing serpentinized olivine (s.ol); Plane‐Polarized Light. (d) Pyroxenite
from the Ahama ultramafic‐mafic body showing a large clinopyroxene crystal being replaced by chlorite; Plane‐Polarized Light. (e) Axim gabbro
showing relict clinopyroxene (cpx) which commonly forms the core of amphibole (amp) crystals; hornblende is partly replaced by chlorite. Plane‐
Polarized Light. (f) A typical large clinopyroxene (cpx) crystal in the Kegyina gabbro. Cross‐Polarized Light [Colour figure can be viewed at
wileyonlinelibrary.com]
DAMPARE ET AL. 9the east central synclinal flank of the southern Ashanti volcanic belt.
The western contact of the Aketakyi UMC is marked by fault and con-
sidered to represent the lower section of the complex, where it is
made up of ~2.0‐km‐thick zone of coarse‐grained peridotites, includ-
ing harzburgite and dunite (Attoh et al., 2006). The rocks are variably
serpentinized and altered to various degrees. Layering in the ultra-
mafic complex appears to follow the regional north‐east strike of the
volcanic belt. This layering is more pronounced in the gabbroic rocks,
which occur in the eastern contact zone. Rodingite occurs in the ultra-
mafic rocks. Attoh et al. (2006) recognized the similarities between the
Birimian rodingite occurrence in the Aketakyi UMC and those from
ultramafic rocks of Archaean, including the Abitibi and Barberton
greenstone belts, as well as the Phanerozoic rodingite occurrence in
the ultramafic zones of ophiolitic complexes or modern oceanic crust.
The protolith of the rodingites is uncertain.
The gabbroic zone of the ultramafic complex is stratigraphically
overlain by a sheeted‐dyke complex, which consists of basaltic
hyaloclastites intercalated with fine‐grained sediments including chert
and disrupted basaltic‐andesitic sills. A detailed description of the
Aketakyi UMC is provided by Attoh et al. (2006). The ultramafic com-
plex is associated with volcanic rocks. The volcanic pile consists of pil-
low breccias and hyaloclastites at the base, pillowed and massive lava
flows in the middle, and aquagene tuffs interbedded with massive
flows at the top, capped by a manganese‐oxide and manganese‐
carbonate deposit (Sylvester & Attoh, 1992). The ultramafic complex
with the overlying sheeted‐dyke complex and associated volcanics
has been suggested to represent a cross‐section of a
Palaeoproterozoic oceanic crust (Attoh et al., 2006).
Primary minerals in the Aketakyi UMC include olivine, pyroxene,
and amphibole (Figure 3b). Olivine occurs as oval and subrounded
crystals, which show a wide range of sizes. Olivine is commonly
serpentinized. Few grains are preserved. Both brown and green
amphiboles occur in the rocks. Brown oikocrysts occur in almost all
the samples as a postcumulus phase, enclosing serpentinized olivine
(Figure 3c) and sometimes pyroxene. Secondary minerals include
actinolite, chlorite, and tremolite. Actinolite replaces pyroxene and
chlorite partly replaces amphibole. Chromite and magnetite commonly
occur in the rocks. Ilmenite occurs as anhedral minerals in some of the
rocks.
3.2.2 | Ahama ultramafic body
The Ahama ultramafic body was first discovered and mapped by Loh
and Hirdes (1999). It is located in the area of the headwaters of the
Ahama stream in the Axim volcanic branch. The rocks are poorly
exposed. Like the Aketakyi ultramafic complex, the emplacement of
the Ahama ultramafic body appears to predate shearing which
affected the Birimian terrane of Ghana around 2.1 Ga (Loh & Hirdes,
1999). The Ahama rocks are massive, coarse‐grained, dark green, and
are mostly weathered and show various degrees of serpentinization.
The Ahama rocks are predominantly pyroxenites. Primary minerals
are poorly preserved. Only few pyroxene grains were identified.
Pyroxene shows sector zoning, and it is mostly replaced by talc.Actinolite occurs as a secondary mineral. Opaques are also abundant
in them. Epidote occurs as single euhedral crystals and as a replace-
ment product of plagioclase. Pyroxene is replaced by hornblende. In
some of the rocks, pyroxene is replaced by chlorite (Figure 3d).
3.3 | Mafic rocks
The gabbroic rocks occur as minor bodies in the study area, and they
include the Axim gabbro, Kegyina gabbro, and gabbroic rocks from
the Aketakyi area. Although the gabbroic rocks have not been dated,
they appear to have different origins as well as emplacement records.
The Axim body is believed to have formed later than the granitoids,
whereas the Kegyina gabbro could be coeval with the granitoids.
The Aketakyi gabbro is inferred to have been emplaced earlier than
the Prince's Town pluton, as the latter intrudes the western flank of
the Aketakyi ultramafic complex, of which the Aketakyi gabbro is con-
sidered to be part. According to Loh and Hirdes (1999), most of the
gabbroic rocks appear to predate the Eburnean tectono‐thermal
event.
3.3.1 | Axim gabbro
The Axim gabbro is associated with argillite and andesitic lavas in the
transition zone between the Kumasi basin and the Ashanti belt. The
rocks are massive and show various degrees of alteration. The Axim
gabbroic rocks are equigranular, medium‐grained, and consist of pla-
gioclase, pyroxene, and hornblende. Plagioclase is commonly
saussuritized. Relict pyroxenes usually form the cores of hornblende
crystals (Figure 3e). This feature is considered to have developed dur-
ing the late magmatic stage and is not a product of post‐magmatic
replacement. Hornblende is partly replaced by chlorite. Sphene coex-
ists with an opaque mineral suspected to be ilmenite. Sulphides also
occur in the rocks.
3.3.2 | Aketakyi and Kegyina gabbroic rocks
The gabbroic body, which intrudes the area north of Aketakyi town, is
considered as an equivalent of the Kegyina gabbro. This body is made
up of fine‐ to coarse‐grained types. The Aketakyi gabbroic rocks are
associated with the Aketakyi ultramafic complex, and they commonly
occur as strongly foliated rocks at the outer flanks of the ultramafic
complex. The ultrabasic nature of some analysed gabbroic rocks led
Loh and Hirdes (1999) to suggest that the gabbroic rocks could be part
of the main Aketakyi ultramafic complex. For Attoh et al. (2006), the
gabbro rocks represent the upper section of the ultramafic complex.
The Kegyina gabbro is poorly exposed and deeply weathered. It is
located within the Prince's Town pluton towards its margin and prob-
ably represents the mafic boarder of the Prince's Town pluton (Loh &
Hirdes, 1999). The rocks are coarse‐grained and massive.
In both the Kegyina and Aketakyi gabbros, green amphiboles coex-
ist with pyroxenes (Figure 3f). Pyroxene is often replaced by amphi-
bole. Plagioclase is commonly saussuritized. Epidote occurs in some
of the samples. Chlorite also occurs as an alteration product of
10 DAMPARE ET AL.
TABLE 1 Major element abundances in the representative samples from the Prince's Town granitoid suite
Rock type Granodiorite Granodiorite Tonalite Quartz diorite Quartz diorite Quartz diorite Quartz diorite Granodiorite Granodiorite
Sample No. D9013A D9013B D9021 D9023A D9024A D9024B D9024C D9025 D9038
SiO2 (wt.%) 66.26 67.09 65.69 61.61 63.74 63.78 62.17 69.62 67.11
TiO2 0.34 0.34 0.33 0.36 0.36 0.37 0.37 0.27 0.37
Al2O3 15.25 15.42 15.82 15.98 15.91 15.75 15.66 15.53 14.25
Fe2O
a
3 3.55 3.63 3.86 5.63 5.51 5.57 5.68 3.06 4.16
MnO 0.07 0.08 0.08 0.11 0.11 0.11 0.11 0.06 0.08
MgO 2.03 2.01 2.02 2.67 2.68 2.80 2.83 1.51 2.11
CaO 3.70 3.64 3.84 5.51 5.34 5.20 5.62 3.26 3.42
Na2O 4.43 4.19 4.67 3.66 3.87 3.67 3.53 4.69 3.88
K2O 2.42 2.37 1.95 1.60 1.43 1.48 1.37 1.85 2.90
P2O5 0.12 0.12 0.13 0.11 0.11 0.12 0.12 0.10 0.11
LOI 1.58 1.59 1.46 2.12 2.08 2.24 2.13 1.25 1.75
Total 99.74 100.49 99.84 99.35 101.14 101.08 99.58 101.20 100.13
Mg# 53.08 52.31 50.92 48.39 49.03 49.92 49.68 49.39 50.06
A/CNK 0.92 0.96 0.95 0.9 0.91 0.92 0.89 0.99 0.91
V (ppm) 73 74 49 76
Cr 200 200 160 170
Co 13 13 27 13
Ni 90 80 80 70
Cu 40 20 20 40
Zn 60 50 50 50
Ga 20 16 20 16
Rb 55 77 46 75
Sr 667 380 655 366
Y 9.3 23.9 9.7 24.3
Zr 104 118 98 130
Nb 3.2 4 2.9 4
Cs 1.1 1.4 1.1 1.4
Ba 712 837 681 845
La 15.8 32.7 17.7 32.7
Ce 33.8 42.4 33.2 41.5
Pr 4 6.18 4.18 6.16
Nd 14.9 20.9 14.9 21.6
Sm 2.76 3.97 2.76 3.95
Eu 0.812 1.09 0.775 1.08
Gd 2.16 3.85 2.16 3.82
Tb 0.31 0.59 0.29 0.6
Dy 1.65 3.31 1.5 3.33
Ho 0.3 0.68 0.28 0.67
Er 0.9 2.07 0.8 2.08
Tm 0.134 0.303 0.121 0.306
Yb 0.95 1.95 0.83 1.96
Lu 0.148 0.303 0.13 0.31
Hf 2.8 3.3 2.6 3.5
(Continues)
DAMPARE ET AL. 11
TABLE 1 (Continued)
Rock type Granodiorite Granodiorite Tonalite Quartz diorite Quartz diorite Quartz diorite Quartz diorite Granodiorite Granodiorite
Sample No. D9013A D9013B D9021 D9023A D9024A D9024B D9024C D9025 D9038
Ta 0.29 0.35 2.66 0.4
Pb 12 9 9 11
Th 2.9 4.01 3.77 4.03
U 0.95 1.15 0.89 1.1
Note. Mg# = 100*(MgO/40.32)/({Fe2O3/79.85} + MgO/40.32); A/CNK = Molecular Al2O3/(CaO + Na2O).
aTotal iron as Fe2O3.amphibole. Accessory minerals include apatite, ilmenite, and magne-
tite. Some varieties contain hornblende and plagioclase which are
commonly replaced by chlorite and epidote, respectively. Sulphides
are, however, abundant in the Aketakyi gabbros.4 | WHOLE‐ROCK COMPOSITIONS AND
ISOTOPE GEOCHEMISTRY
The analytical techniques used for the major, trace, and Sr–Nd isotope
analyses of samples collected from mafic, ultramafic, and granitoid
outcrops associated with metavolcanic rocks in the Axim and Cape
Three Points volcanic branches of the southern Ashanti greenstone
belt have been provided in Appendix A.4.1 | Major and trace element geochemistry
4.1.1 | Prince's Town pluton
The chemical compositions of the Prince'sTown granitoids are given in
Table 1. Although some of the analysed samples show variable effects
of alteration, this is not pronounced and is not evident in the chemical
data. In particular, the trace elements show systematic patterns
reflecting petrogenetic processes, although we cannot rule out that
some of the variations in K2O, Na2O, Ba, Rb, and Sr may be due to
minor post‐crystallization changes. The major element characteristics
of the analysed granitoids have been discussed by Dampare et al.
(2005). Accordingly, only a summary of these features is reported.
All the analysed samples from the Prince's Town pluton have SiO2
content of 63.0–70.5 wt.%; TiO2 of 0.27–0.38 wt.%; Al2O3 of 14.50–
16.33 wt.%; total iron, as Fe2O3 of 3.10–5.80 wt.%; MnO of 0.06–
0.11 wt.%; MgO of 1.53–2.89 wt.%; CaO of 3.30–5.74 wt.%;
(Na2O + K2O) content of 5.01–6.96 wt.% and Na2O/K2O ratios from
1.34 to 2.70; and P2O5 of 0.10–0.13 wt.%. The Mg# of the rocks
ranges from 53 to 48. The rocks may be classified as meta‐aluminous,
medium‐K, calc‐alkaline I‐type granitoids (Dampare et al., 2005; this
study). This confirms the findings of Loh and Hirdes (1999) that the
Prince's Town pluton is largely meta‐aluminous and calc‐alkaline in
character. Loh and Hirdes (1999) have, however, indicated that the
Prince's Town pluton is zoned with a mafic rim, and these mafic rocks
(i.e., quartz diorite and diorite) show some deviation from the calc‐alkaline trend defined by the other rock types such as the granites,
granodiorites, and tonalites. Although there appears to be some differ-
ences in the chemistry of the analysed quartz diorite, granodiorite,
and tonalite (Table 1), we are unable to confirm Loh and Hirdes'
(1999) assertion, probably due to relatively limited number of samples
examined in this study.
The analysed granitoids have low REE abundances (ΣREE = 79–
120 ppm). The chondrite‐normalized REE patterns have concave‐
upward shapes with slight LREE enrichment (LaN/SmN = 3.70–5.34;
LaN/YbN = 11.93–15.30) and weak negative to no Eu anomalies (Eu/
Eu* = 0.85–1.02; Figure 4a). The granitoids show LREE enrichment
patterns with minor negative and positive Eu anomalies and demon-
strate enrichment in LILE relative to HFSE and in LREE relative to
HREE, with Th–U trough, negative Nb (and Ta), and pronounced Ti
anomalies, and spikes in Pb (and Sr; Figure 5a).4.1.2 | Ultramafic rocks
The ultramafic rocks in the study area include the Aketakyi ultramafic
complex and the Ahama ultramafic body. The chemical compositions
of the ultramafic rocks (Tables 2 and 3) as well as the mafic rocks have
been plotted on the SiO2 versus FeO*/MgO diagram (Figure 6).
Figure 6 also includes the chemical data of the mafic and ultramafic
rocks provided in Loh and Hirdes (1999), and our data are comparable
to that of Loh and Hirdes (1999). The rocks are mostly subalkaline and
fall on both sides of the tholeiitic/calc‐alkaline divide defined by
Miyashiro (1974). Variations in the contents of major elements and
selected trace elements with MgO are also shown in Figure 7.
Aketakyi ultramafic complex
Major and trace element data for the Aketakyi UMC are listed in
Table 2. The analysed ultramafic–mafic rocks have SiO2, (Na2O + K2O),
MgO, and TiO2 contents of 42.3–53.7 wt.%, 0.0–2.14 wt.%, 15.1–
36.5 wt.%, and 0.09–0.61 wt.%, respectively. The Al2O3 (2.75–
11.8 wt.%) and CaO (2.20–13.5 wt.%) contents in the rocks range
from low to high values with the high values observed in the pyroxe-
nitic and gabbroic rocks. The ultramafic–mafic rocks are characterized
by moderate to high Mg numbers ranging from 72 to 84 with Ni con-
tents of 710–1,650 ppm and Cr of 1,590–3,940 ppm (Table 2).
12 DAMPARE ET AL.
FIGURE 4 Chondrite‐normalized patterns of
plutonic rocks from the southern Ashanti
greenstone belt: (a) Prince's Town granitoids,
(b) Gabbroic rocks, (c) Ahama pyroxenites, and
(d) Aketakyi ultramafic complex.
Normalization values after Sun and
McDonough (1989)The high loss‐on‐ignition observed in these rocks resulted from
serpentinization. Compositional variations in the ultramafic–mafic
rocks are represented by a plot of major and trace elements versus
MgO (Figure 7). The chemical compositions of the ultramafic–maficrocks are plotted together with the previous chemical data of Loh
and Hirdes (1999) in the Al2O3–CaO–MgO diagram of Coleman
(1977; Figure 8). The analysed rocks fall in the fields of ophiolitic
cumulate rocks.
DAMPARE ET AL. 13
FIGURE 5 Primitive‐mantle‐normalized
multi‐element patterns of plutonic rocks from
the southern Ashanti greenstone belt: (a)
Prince's Town granitoids, (b) Gabbroic rocks,
(c) Ahama pyroxenites, and (d) Aketakyi
ultramafic complex. Normalization values,
OIB, and N‐MORB are from Sun and
McDonough (1989)The ultramafic–mafic rocks display LREE‐depleted to flat,
chondrite‐normalized REE patterns ((La/Sm)N = 0.32–1.16; (La/Yb)
N = 0.16–1.16) with no to minor positive Eu anomalies (Eu/Eu* = 1.00–
1.41) and with total REE contents ranging from 1.71 to 22.3 ppm
(Figure 4d). The positive Eu anomalies observed in some of the rockssuggest that plagioclase was a cumulative phase in the formation of
those rocks. On the primitive‐mantle‐normalized, trace element dia-
gram, they show enrichment in Cs, slightly negative Th and Nb anom-
alies, and variable but minor positive and negative Sr, Hf, and Ti
anomalies (Figure 5d).
14 DAMPARE ET AL.
TABLE 2 Representative whole‐rock major and trace element chemical analyses of the Aketakyi ultramafic complex from the southern Ashanti
volcanic belt
AKT9 AKT10 D90410 D90413 D90416 D90411A D90411B AKT2A D9046 D90412B D90415 D90417 D90412
Hz Hz Hz Hz Hz D D D Py Py Py Py G
SiO2 41.35 40.09 39.60 39.19 40.52 38.73 38.10 37.77 48.44 51.13 44.11 46.63 49.63
(wt.%)
TiO2 0.22 0.20 0.26 0.09 0.19 0.15 0.18 0.11 0.54 0.18 0.08 0.57 0.15
Al2O3 5.30 4.41 3.42 2.46 3.74 2.70 2.84 2.98 5.29 3.22 5.93 10.03 11.39
Fe2O
a
3 12.57 12.88 13.58 13.19 13.12 13.88 13.82 13.77 10.50 8.62 11.50 12.29 6.59
MnO 0.18 0.18 0.18 0.19 0.20 0.19 0.18 0.20 0.13 0.18 0.17 0.17 0.13
MgO 26.71 28.41 29.83 31.03 28.72 32.79 31.82 32.11 21.61 22.29 23.55 15.62 14.52
CaO 4.31 3.55 2.93 1.97 2.99 2.18 2.34 2.31 6.76 9.12 7.02 7.02 12.96
Na2O 0.35 0.25 0.21 0.06 0.18 0.08 0.13 0.02 0.43 0.21 0.03 2.07 0.82
K2O 0.03 0.02 0.01 0.00 0.01 0.00 0.00 0.00 0.03 0.00 0.00 0.07 0.01
P2O5 0.02 0.03 0.02 0.01 0.02 0.02 0.02 0.01 0.03 0.01 0.01 0.06 0.01
LOI 8.74 9.39 9.97 10.46 10.02 10.23 10.14 10.78 6.02 4.68 7.10 5.23 3.60
Total 99.79 99.43 100.01 98.65 99.71 100.95 99.57 100.07 99.77 99.65 99.50 99.77 99.82
Mg# 80.80 81.37 81.30 82.32 81.26 82.39 82.02 82.20 80.30 83.67 80.21 71.57 81.34
V (ppm) 105 61 86 80 129 81 223
Cr 3,550 3,410 3,190 3,940 3,170 2,560 1,590
Co 127 126 124 140 76 113 84
Ni 1,510 1,500 1,450 1,650 710 1,170 740
Cu 50 10 30 30 20 <10 100
Zn 70 60 70 70 50 40 80
Ga 5 4 6 4 5 6 12
Rb 1 <1 1 <1 <1 <1 <1
Sr 22 10 17 10 13 7 24
Y 6.8 1.6 3.9 3.9 5.8 2 14
Zr 13 2 9 10 4 <1 34
Nb 0.5 <0.2 0.3 0.3 <0.2 <0.2 1.5
Cs 0.8 0.6 0.6 0.5 <0.1 <0.1 0.5
Ba 9 <3 6 14 4 <3 23
La 0.86 0.16 0.58 0.57 0.63 0.05 2.09
Ce 2.43 0.41 1.53 1.58 1.22 0.15 5.44
Pr 0.39 0.06 0.24 0.24 0.18 0.03 0.82
Nd 2.17 0.32 1.17 1.28 1.02 0.24 4.14
Sm 0.7 0.11 0.36 0.4 0.35 0.1 1.29
Eu 0.267 0.061 0.154 0.158 0.149 0.051 0.542
Gd 0.96 0.16 0.47 0.51 0.57 0.17 1.72
Tb 0.18 0.04 0.09 0.1 0.11 0.04 0.33
Dy 1.17 0.27 0.62 0.69 0.8 0.3 2.19
Ho 0.25 0.06 0.13 0.14 0.18 0.07 0.47
Er 0.75 0.18 0.41 0.43 0.56 0.22 1.41
Tm 0.108 0.029 0.066 0.065 0.088 0.034 0.213
Yb 0.69 0.21 0.44 0.43 0.6 0.22 1.38
Lu 0.106 0.036 0.069 0.07 0.095 0.034 0.212
(Continues)
DAMPARE ET AL. 15
TABLE 2 (Continued)
AKT9 AKT10 D90410 D90413 D90416 D90411A D90411B AKT2A D9046 D90412B D90415 D90417 D90412
Hz Hz Hz Hz Hz D D D Py Py Py Py G
Hf 0.4 <0.1 0.3 0.3 0.1 <0.1 0.9
Ta 0.03 <0.01 0.02 0.02 <0.01 <0.01 0.08
Pb <5 <5 <5 <5 <5 <5 <5
Th 0.07 <0.05 <0.05 <0.05 <0.05 <0.05 0.18
U 0.04 0.02 <0.01 0.01 <0.01 0.06 0.05
Note. Hz: Harzburgitite; D: Dunite; Py: Pyroxenite; G: Gabbro. Magnesium number (Mg#) = 100 × molar Mg2+/(Mg2+ + Fe2+).
aTotal iron as Fe2O3.Ahama ultramafic body (pyroxenites)
The chemical compositions of the Ahama pyroxenites are listed in
Table 3. The LOI values of the ultramafic rocks in the study area
(Ahama and Aketakyi rocks) are high due to alteration; microscopic
studies also indicate that these are variably altered and serpentinized.
In what follows, the major oxides of the rocks in data plots are pre-
sented on a volatile‐free basis to reduce the effect of variable dilution
resulting from serpentinization.
The Ahama pyroxenites have SiO2, (Na2O + K2O), MgO, and TiO2
contents of 48.7–50.5 wt.%, 1.30–2.81 wt.%, 7.98–12.6 wt.%, and
0.54–0.78 wt.%, respectively, and are characterized by low Mg num-
bers ranging from 63 to 74 with Ni contents of 170–520 ppm, and
Cr of 630–1,580 ppm. The trace element geochemistry reveals two
types of pyroxenites. The Group I pyroxenites show LREE enrichment
patterns with (La/Sm)N = 1.16–1.70, (La/Yb)N = 1.94–2.34, minor pos-
itive Eu anomalies (Eu/Eu* = 1.09–1.17), minor Ce anomalies of 0.98–
1.03, and ∑REE contents of 28.1–32.8 ppm (Figure 4c). The Group II
pyroxenites also show fractionated REE patterns with (La/Sm)
N = 1.20–1.47, (La/Yb)N = 1.06–2.55, Eu anomalies of 0.99–1.03, sig-
nificant negative Ce anomalies of 0.53–0.88, and relatively higher
∑REE contents of 57.1–124 ppm (Figure 4c). Both types of pyroxe-
nites are characterized by enrichment in LILE relative to HFSE and in
LREE relative to HREE. The Group I pyroxenites show Th–U and
Nb–Ta troughs, positive Sr anomalies and minor negative Zr–Hf and
Ti anomalies, whereas the Group II pyroxenites exhibit negative Th
anomalies, pronounced negative Nb, Zr–Hf and Ti anomalies, and
spikes in Sr (Figure 5c).4.1.3 | Mafic rocks
Whole‐rock compositions of the gabbroic intrusives in the southern
Ashanti greenstone belt are presented in Table 4. The chemical com-
positions of the mafic and ultramafic rocks have been plotted on the
SiO2 versus FeO*/MgO diagram (Figure 6). Figure 6 also includes
the chemical data of the mafic and ultramafic rocks provided in Loh
and Hirdes (1999), and our data are comparable to that of Loh and
Hirdes (1999). The rocks are mostly subalkaline and fall on both sides
of the tholeiitic/calc‐alkaline divide defined by Miyashiro (1974). Var-
iations in the contents of major elements and selected trace elements
with MgO are also shown in Figure 7.Axim gabbro
The analysed Axim gabbros have SiO2, total alkalis (Na2O + K2O), high
MgO, and low TiO2 contents of 50.3–51.8 wt.%, 4.38–4.39 wt.%,
7.02–10.5 wt.%, and 0.65–0.82 wt.%, respectively. Their Mg numbers
range from 54 to 68, and they have Ni and Cr contents of 140–250
and 340–880 ppm, respectively. The Axim gabbros exhibit slightly
LREE‐enriched patterns, (La/Sm)N = 2.16–2.46 and (La/Yb)N = 4.35–
5.76, with minor negative and positive Eu anomalies (Eu/Eu* = 0.94–
1.14; Figure 4b). Primitive‐mantle‐normalized, trace element patterns
show that the gabbros have geochemical patterns characterized by
enrichment in LILE (e.g., Cs, Rb, and Ba) relative to HFSE (e.g., Nb,
Zr, Hf, Ta, Ti, and Y) and in LREE relative to HREE, with Th–U trough,
pronounced negative Nb anomalies, negative Zr–Hf and Ti anomalies,
and spikes in Pb and Sr (Figure 5b).
Kegyina gabbro
The Kegyina gabbroic rocks have SiO2, (Na2O + K2O), MgO, and TiO2
contents of 48.1–52.6 wt.%, 0.83–4.78 wt.%, 4.88–15.4 wt.%, and
0.21–1.10 wt.%, respectively, and Mg numbers of 63–74, Ni contents
of 170–520 ppm and Cr of 630–1,580 ppm. The rocks commonly
show LREE enrichment patterns with (La/Sm)N = 1.03–2.64, (La/Yb)
N = 2.43–5.70, and minor negative to strong positive Eu anomalies
(Eu/Eu* = 0.87–1.87). Sample D9039 (isolated massive gabbro from
Kegyina) shows a convex‐upward REE pattern which is typical of
pyroxene‐dominated cumulates (McDonough & Frey, 1989) with the
apex at Nd (McDonough & Frey, 1989; Figure 4b). On the primitive‐
mantle‐normalized, trace element diagram, the gabbroic rocks show
enrichment in LILE relative to HFSE and in LREE relative to HREE with
negativeTh (and U), Nb (and Ta), Zr–Hf and Ti anomalies, and negative
and positive Sr anomalies (Figure 5b)
Aketakyi gabbroic rocks
The analysed Aketakyi gabbroic rocks (i.e., those from the northern part
of Aketakyi town) have SiO2, (Na2O + K2O), MgO, and TiO2 contents of
45.1–50.1 wt.%, 0.50–3.79 wt.%, 6.81–5.81 wt.%, and 0.49–0.99 wt.%,
respectively. The Mg numbers range from 56 to 73, and their Ni and Cr
contents are 160–430 and 270–440 ppm, respectively. They show
LREE‐depleted patterns with (La/Sm)N = 0.43–0.63 and (La/Yb)
N = 0.25–0.63 (Figure 4b). The gabbroic rocks are characterized by
enrichment in LILE relative to HFSE, negative Zr–Hf and Ti anomalies,
pronounced negative Nb and Ta anomalies and spikes in Sr (Figure 5b).
16 DAMPARE ET AL.
TABLE 3 Whole‐rock major and trace element chemical analyses of the Ahama pyroxenitic rocks from the southern Ashanti volcanic belt
D8301A D8301B D8301C D8302 D90211A D90211B D90211C D90212A D90212B D90212C
SiO2 (wt.%) 47.11 47.44 47.30 47.80 46.63 46.56 47.36 47.07 46.35 47.67
TiO2 0.54 0.51 0.55 0.52 0.67 0.69 0.74 0.64 0.68 0.68
Al2O3 9.78 9.23 9.83 10.18 9.53 9.51 11.51 9.47 8.93 9.89
Fe2O
a
3 10.39 10.61 10.16 10.18 11.69 11.97 11.95 11.93 12.21 12.38
MnO 0.17 0.17 0.16 0.16 0.20 0.20 0.19 0.19 0.21 0.20
MgO 15.00 15.49 14.59 14.45 12.45 12.54 10.37 12.52 12.80 12.62
CaO 7.50 7.88 7.57 8.36 11.88 11.06 10.28 11.79 12.00 12.24
Na2O 1.75 1.60 1.80 1.83 1.01 1.01 1.49 1.07 0.93 1.11
K2O 0.80 0.54 0.84 0.58 0.63 0.52 0.83 0.51 0.31 0.37
P2O5 0.16 0.14 0.16 0.14 0.17 0.17 0.20 0.16 0.17 0.17
LOI 6.00 6.12 6.08 5.24 4.69 5.52 4.88 4.52 4.79 2.66
Total 99.21 99.74 99.03 99.43 99.53 99.77 99.79 99.89 99.38 100.00
Mg# 74.10 74.30 73.98 73.75 67.85 67.47 63.23 67.51 67.47 66.87
V (ppm) 187 195 214 241 227 221
Cr 1,490 1,580 810 630 830 760
Co 60 60 54 56 63 58
Ni 520 480 180 170 200 190
Cu 50 60 60 70 60 70
Zn 70 50 70 100 70 80
Ga 12 12 12 15 13 13
Rb 13 14 17 16 10 7
Sr 214 584 285 571 381 502
Y 12.7 12.2 13.3 59.3 62.3 46
Zr 32 27 29 40 32 28
Nb 1.2 1.2 1.3 1.9 1.4 1.2
Cs 0.3 0.5 0.5 0.4 0.4 0.2
Ba 124 248 174 300 201 114
La 3.73 3.81 3.66 9.34 19.1 5.71
Ce 9.06 8.75 9.61 11.4 23 12
Pr 1.31 1.25 1.42 2.95 5.57 1.96
Nd 5.93 5.4 6.73 14.9 27.9 9.77
Sm 1.66 1.45 2.04 4.61 8.38 3.08
Eu 0.662 0.557 0.774 1.86 2.93 1.26
Gd 1.81 1.67 2.11 6.6 9.51 4.93
Tb 0.34 0.29 0.38 1.29 1.72 0.98
Dy 2.15 1.85 2.41 8.43 10.8 6.6
Ho 0.45 0.38 0.48 1.76 2.16 1.44
Er 1.36 1.17 1.42 5.15 6.13 4.28
Tm 0.204 0.178 0.207 0.716 0.875 0.615
Yb 1.33 1.17 1.35 4.24 5.37 3.87
Lu 0.203 0.182 0.209 0.609 0.795 0.569
Hf 1 0.7 0.9 1.2 1.1 1
Ta 0.08 0.07 0.07 0.08 0.03 0.06
Pb <5 <5 <5 <5 <5 <5
(Continues)
DAMPARE ET AL. 17
TABLE 3 (Continued)
D8301A D8301B D8301C D8302 D90211A D90211B D90211C D90212A D90212B D90212C
Th 0.39 0.32 0.38 0.51 0.37 0.37
U 0.14 0.11 0.17 0.26 0.47 0.33
Note. Magnesium number (Mg#) = 100 × molar Mg2+/(Mg2+ + Fe2+).
aTotal iron as Fe2O3.
FIGURE 6 SiO2 vs. FeO*/MgO diagram for the mafic–ultramafic
plutonic rocks from the southern Ashanti greenstone belt. Previous
chemical data of Loh and Hirdes (1999) have been included in the plot4.2 | Sr–Nd isotopic data
Sr–Nd isotopic data, 87Sr/86Sr initial ratios, εNd (T), and Nd model ages
(TDM1 and TDM2) are presented in Table 5. The initial
87Sr/86Sr ratios
and εNd values were calculated at an age of 2.1 Ga, which represents
the crustal formation time during the Eburnean orogeny (e.g.,
Abouchami et al., 1990; Boher et al., 1992). The initial isotopic ratios
were derived using Rb/Sr and Sm/Nd ratios from the ICP‐MS data.
The rocks display low 87Sr/86Sr initial ratios (0.693178–0.702423).
The low 87Sr/86Sr initial values of D9023A (0.693178) and D9038
(0.699468) probably suggest alteration of feldspar. The Sr initial ratios
values are comparable to those previously reported on the Birimian
basaltic and granitic rocks of West Africa (e.g., Abouchami et al., 1990;
Boher et al., 1992; Gasquet et al., 2003; Taylor et al., 1992). These data
show the primitive isotopic Sr signature of the rocks, thereby precluding
significant magmatic contributions from Rb‐enriched continental crust.
The granitoids show a range of initial εNd values from −1.01 to +2.92
at 2.1 Ga. The gabbros have initial εNd (T) of −1.23 to +5.23. The Ahama
pyroxenites show negative initial εNd (T) values from −0.58 to −5.68.
The meta‐peridotites from the Aketakyi ultramafic/ophiolitic complex
show high positive εNd (T) values from +3.69 to +4.93.
Crustal residence ages (TDM) may be calculated using diverse
models of depleted mantle evolution. Nd models give ages which vary
by several hundreds of Ma on the basis of previous studies (Albarède
& Brouxel, 1987; Ben Othman et al., 1984; DePaolo, 1981; Goldstein,
O'Nions, & Hamilton, 1984; McCulloch, 1987; Nelson & DePaolo,
1985). The model of Ben Othman et al. (1984) is considered to be rep-
resentative of the Birimian depleted mantle (Abouchami et al., 1990)
and was subsequently adopted by some workers including Abouchami
et al. (1990), Boher et al. (1992), and Gasquet et al. (2003). However,following the approach of some earlier workers in the Birimian terrane
of Ghana, one‐stage model ages (TDM1) have been calculated for the
rocks, using the model of DePaolo (1981). The Nd model age of crustal
rocks indicates their time of extraction from the mantle. However, the
assumption that a Sm–Nd model age represents an average crustal
residence time is valid only if no fractionation of Sm/Nd has taken
place since the first separation of its protolith from the mantle source.
This is, however, not always the case. In order to minimize the bias as
a result of Sm/Nd fractionation, which might lead to either underesti-
mation or overestimation of one‐stage Nd model ages, a two‐stage Nd
model age (TDM2) may be calculated for the rocks. The two‐stage
model age was computed following the procedure of Keto and
Jacobsen (1987), and using the expression below:
TDM2 ¼ TDM − ðTDM1 − tÞ ½ðƒcc − ƒsÞ=ðƒcc − ƒDMÞ;
where ƒcc, ƒs, and ƒDM are the ƒSm/Nd values of the continental crust, the
sample, and depleted mantle, respectively, and t is the formation age
of the rock; and ƒSm/Nd is given by the expression:
ƒ ¼ 
  
 147Sm=144Nd 147 144Sm=Nd Sample = Sm= NdCHUR –1:
In this calculation, ƒcc = −0.4, ƒDM = 0.086426, and t = 2.1 Ga.
It is difficult to choose between one‐stage and two‐stage model
ages as each has its own uncertainties (e.g., DePaolo, Linn, & Schubert,
1991; Wu, Zhao, Simon, Wilde, & Sun, 2005). Wu et al. (2005) indi-
cated that more accurate results could be obtained for the single‐stage
model age, if the Sm/Nd fractionation, expressed as ƒSm/Nd, is limited
to the range of −0.2 to −0.6.
In this study, the two‐stage Nd model age (TDM2) is preferred to
the single‐stage model age (TDM1) as it displays a more regular pattern.
The model ages are 2.12–2.45 Ga for the granitoids and 2.03–2.47 Ga
for the gabbros. The Group I pyroxenites yielded TDM2 values of 2.43
to 2.61 Ga, whereas no meaningful Nd model ages were obtained for
the Group II pyroxenites, as they yielded extremely high TDM values.
The analysed meta‐peridotite samples from the Aketakyi ophiolitic
complex yielded Nd model ages of 1.96 to 2.01 Ga.5 | DISCUSSION
5.1 | Alteration effects
The rocks of the southern Ashanti greenstone belt have experienced
peak amphibolite facies to retrograde greenschist‐facies metamor-
phism, and mobility of some elements may have occurred. Alteration
18 DAMPARE ET AL.
FIGURE 7 Selected major (wt.%) and trace elements (ppm) variations with MgO (wt.%) for the analysed mafic–ultramafic rocks in this study and
that of Loh and Hirdes (1999)
FIGURE 8 Al2O3–CaO–MgO plot of the mafic–ultramafic rocks
compared with ophiolitic rocks (fields after Coleman, 1977). Loh and
Hirdes' (1999) data have also been plottedof plutonic rocks is a common phenomenon, especially in older rocks.
It is usually evidenced by high loss‐on‐ignition (LOI) values and
increased scatter and mobility of major and large‐ion lithophile ele-
ments (LILE). Some studies have noted, however, that even for rocks
of ancient heritage, the concentrations of such “mobile” elements
are commonly not significantly changed from their primary abun-
dances (e.g., Whalen, Syme, & Stern, 1999). The extent of alteration
of rocks could also be investigated using the chemical index of alter-
ation (CIA) calculations of Nesbitt and Young (1982). The CIA is
defined as molar (100*Al2O3/(Al2O3 + CaO* + Na2O + K2O)), where
CaO* is the contribution of CaO from silicates only. According to
Nesbitt and Young (1982), the CIA values for unaltered granite and
basic rock are 50 and 42, respectively, while a CIA value of any rock
exceeding 60 indicates a significant alteration. Therefore, rocks with
less CIA values could be used for petrogenetic interpretations. Owing
to the relatively immobile nature of HFSE and REEs under most con-
ditions (e.g., Whalen et al., 1999), they have often been used for igne-
ous petrogenetic and tectonic studies. Major elements have mostly
been used to give background information due to their high suscepti-
bility to mobility during processes such as metamorphism and
DAMPARE ET AL. 19
TABLE 4 Representative whole‐rock major and trace element chemical analyses of the Palaeoproterozoic mafic rocks from the southern Ashanti
volcanic belt
Aketakyi gabbroic rocks
D9042a D9043a D9044a AKT 2a AKT 3a AKT 4a AKT 5a AKT 6a AKT 7a
SiO2 (wt.%) 45.73 45.80 45.30 46.35 45.60 45.16 43.19 44.99 47.94
TiO2 0.49 0.62 0.95 0.47 0.47 0.76 0.50 0.52 0.54
Al2O3 14.20 12.32 13.56 12.99 13.11 11.37 13.88 14.15 16.68
Fe O b2 3 11.01 11.27 13.53 10.51 11.87 13.54 12.79 11.49 10.31
MnO 0.16 0.18 0.21 0.17 0.22 0.19 0.19 0.18 0.15
MgO 9.18 11.03 8.60 9.32 10.95 10.66 10.03 9.41 6.59
CaO 13.12 13.41 10.50 14.14 11.73 12.71 13.37 13.52 10.87
Na2O 0.64 0.58 2.01 1.18 0.92 1.22 0.75 1.07 2.79
K2O 0.17 0.40 0.28 0.20 0.11 0.20 0.09 0.11 0.46
P2O5 0.04 0.04 0.07 0.05 0.04 0.09 0.04 0.07 0.20
LOI 4.30 3.79 3.94 3.78 4.63 3.38 4.12 3.60 3.17
Total 99.05 99.43 98.95 99.17 99.65 99.27 98.95 99.10 99.71
Mg# 62.27 65.96 55.71 63.73 64.63 60.92 60.83 61.85 55.86
V (ppm) 246 292
Cr 770 500
Co 70 49
Ni 270 250
Cu 60 40
Zn 60 60
Ga 12 11
Rb 6 12
Sr 158 173
Y 13.4 14.8
Zr 7 10
Nb 0.3 0.5
Cs 0.3 0.6
Ba 27 36
La 0.59 1.44
Ce 1.87 3.49
Pr 0.33 0.71
Nd 2.15 4.22
Sm 0.88 1.47
Eu 0.411 0.643
Gd 1.42 2.01
Tb 0.32 0.41
Dy 2.26 2.76
Ho 0.5 0.59
Er 1.57 1.77
Tm 0.247 0.26
Yb 1.66 1.64
Lu 0.257 0.244
Hf 0.3 0.5
(Continues)
20 DAMPARE ET AL.
TABLE 4 (Continued)
Aketakyi gabbroic rocks
D9042a D9043a D9044a AKT 2a AKT 3a AKT 4a AKT 5a AKT 6a AKT 7a
Ta 1.67 0.02
W 50.7 0.8
Pb <5 <5
Th <0.05 <0.05
U <0.01 0.12
Aketakyi gabbroic rocks Axim gabbroic rocks Kegyina gabbroic rocks
AKT 8a AKT 12a AKT 13a AKT 14a D9031a D9035a D9039a D90310a D90311Ba
SiO2 (wt.%) 45.36 46.29 48.33 47.48 49.83 48.57 46.41 51.02 47.45
TiO2 0.46 0.63 0.65 0.57 0.62 0.79 1.06 0.34 0.20
Al2O3 13.74 7.62 16.29 11.44 12.24 13.85 18.44 4.29 20.05
Fe O b2 3 11.35 14.08 10.69 10.52 9.59 11.58 11.32 11.65 7.17
MnO 0.18 0.21 0.16 0.18 0.16 0.19 0.15 0.21 0.12
MgO 9.20 14.98 6.90 14.17 10.12 6.78 4.71 14.96 6.85
CaO 14.20 9.73 9.58 9.37 8.72 9.66 9.39 13.02 9.06
Na2O 0.83 0.41 3.13 1.57 2.03 2.51 3.59 0.62 3.06
K2O 0.02 0.06 0.52 0.51 2.19 1.73 0.37 0.19 1.56
P2O5 0.04 0.06 0.26 0.05 0.27 0.28 0.25 0.03 0.03
LOI 3.96 5.21 3.61 4.07 3.76 3.42 3.50 2.99 3.32
Total 99.33 99.29 100.12 99.92 99.54 99.35 99.19 99.32 98.86
Mg# 61.63 67.81 56.10 72.74 67.64 53.70 45.17 71.78 65.44
V (ppm) 201 291 211 188 99
Cr 880 340 270 440 390
Co 53 59 41 61 40
Ni 250 140 160 430 200
Cu 130 70 80 290 40
Zn 70 80 90 < 30 50
Ga 14 18 20 7 20
Rb 61 50 8 6 63
Sr 532 815 520 61 538
Y 12.1 19.5 23.4 15 4
Zr 44 58 50 20 12
Nb 2.2 3.6 2.1 0.4 0.4
Cs 1.1 0.9 0.3 0.2 1.4
Ba 857 596 88 38 189
La 9.39 12.6 7.9 13.9 2.66
Ce 21.1 29.4 23.5 23.8 5.36
Pr 2.67 3.82 3.81 3.73 0.7
Nd 11 15.8 19.8 15.7 2.75
Sm 2.46 3.76 4.94 3.88 0.65
Eu 0.89 1.15 1.59 1.07 0.407
Gd 2.32 3.72 5.2 3.67 0.68
Tb 0.38 0.62 0.83 0.6 0.11
(Continues)
DAMPARE ET AL. 21
TABLE 4 (Continued)
Aketakyi gabbroic rocks Axim gabbroic rocks Kegyina gabbroic rocks
AKT 8a AKT 12a AKT 13a AKT 14a D9031a D9035a D9039a D90310a D90311Ba
Dy 2.17 3.59 4.71 3.36 0.64
Ho 0.42 0.72 0.93 0.61 0.13
Er 1.25 2.11 2.66 1.74 0.38
Tm 0.181 0.318 0.378 0.265 0.058
Yb 1.17 2.08 2.33 1.75 0.4
Lu 0.174 0.309 0.348 0.249 0.063
Hf 1.3 1.8 1.6 0.8 0.4
Ta 0.78 2.39 1.7 0.02 0.03
W 27.5 68 50.8 1.5 1.2
Pb 8 5 <5 <5 <5
Th 0.99 1.23 0.19 0.26 0.17
U 0.42 0.54 0.16 0.33 0.05
Note. Magnesium number (Mg#) = 100 × molar Mg2+/(Mg2+ + Fe2+).
aSample.
bTotal iron as Fe2O3.hydrothermal activities. However, if the mobility of major elements is
minimal, they could still reflect the primary igneous processes involved
in the formation of the rocks. Although crustal reworking can result in
inherited chemical signatures, it is less a problem here as isotopic char-
acter of the rocks is largely juvenile.
The analysed gabbroic rocks show variable alteration with LOI
values ranging from 3.0 to 5.2 wt.%. Al; Ca; Mg; high‐field‐strength
elements (HFSE) such as Th, Zr, Hf, Nb, Ta, Ti, Y; and REE may have
remained immobile (e.g., Barnes et al., 1985). Potassium and rubidium
may be easily changed during alteration; however, their nearly con-
stant K2O/Rb ratios (298–478) indicate that their original chemical
compositions were not significantly modified. The CIA values for the
gabbros range from 15 to 46 and that of Prince's Town rocks range
from 47 to 50, indicating minimum alteration. Hence, geochemical
compositions of the gabbroic rocks can approximately reflect their
source characteristics. The LOI values of the ultramafic rocks are high
due to alteration; microscopic studies also indicate that these rocks
are variably altered and serpentinized. The low CIA values of the
Ahama pyroxenite (27–36) and the Aketakyi UMC (16–41) suggest
that they have not been severely altered. In what follows, the major
oxides of the rocks in data plots are presented on a volatile‐free basis
to reduce the effect of variable dilution resulting from
serpentinization. Also, we have mainly used the relatively immobile
elements (HFSE, REE, and transition elements) in our interpretations.5.2 | Petrogenesis
5.2.1 | Prince's Town pluton
The Prince's Town granitoids form part of the Dixcove‐ (or belt‐type)
granitoids of Ghana, which are metaluminous, Na‐rich, dominantlyhornblende‐ to biotite‐bearing granodiorite to diorite, monzonite and
syenite. Generally, the Palaeoproterozoic granitoids intruding the
Birimian terrane of the Leo‐Man Shield have been classified into two
main groups according to the presence or absence of amphibole
(Lompo, 2009 and references therein): (a) the amphibole‐bearing gran-
itoids (PAG) and (b) the biotite ± muscovite‐bearing granitoids without
amphibole (PBG). The Dixcove‐type granitoids may be included in the
PAG‐type granitoids in the Birimian terrane.
Following the approach of Altherr et al. (2000) and using molar
ratios of CaO/(MgO + FeOtot) and Al2O3/(MgO + FeOtot), Dampare et al.
(2005) indicated that the granitoids were likely derived from partial
melting of metaplutonic or metavolcanic rocks, possibly with a contribu-
tion from metagreywacke. The trace element data of the Prince's Town
granitoids are broadly similar to the sodic calc‐alkaline Birimian granit-
oids (NaCG) reported from central Cote d'Ivoire (Doumbia et al., 1998;
Figure 9). Doumbia et al. (1998) suggested that NaCG may have been
generated from partial melting of the basaltic protolith metamorphosed
under the amphibolite facies. It is, therefore, likely that the Prince's
Town granitoids may have formed by similar processes. The Prince's
Town granitoids, however, show a lower degree of REE fractionation
(LaN/YbN = 11.93–15.30) and HREE fractionation (YbN = 4.88–11.53),
compared to the NACG. The REE patterns with unobtrusive Eu anoma-
lies as well as the negative Nb–Ta anomalies indicate that hornblende
was the main residual phase, but with some garnet at the site of partial
melting at least for Samples D9021 and D9025.
Also, the major and trace elements of some of the rocks within the
Prince's Town pluton, D9021 and D9025, show some geochemical
similarities to adakitic rocks, such as their high Na2O/K2O, (La/Yb)N
and Sr/Y ratios. The petrogenesis of adakitic rocks is controversial,
and different models have been proposed to account for their origin.
Notably among these models are included (a) partial melting of a
22 DAMPARE ET AL.
TABLE 5 Sr–Nd isotopic compositions of Palaeoproterozoic rocks from the southern Ashanti volcanic belt, Ghana
Rb Sr 87Rb/ 2σ (87Sr/ Sm Nd 147Sm/ 143Nd/ 2σ εNd TDM1 TDM2
Sample (ppm) (ppm) 86Sr 87Sr/86Sr (X10−6) 86Sr)2.1 (ppm) (ppm)
144Nd 144Nd (X10−6) (2.1 Ga) (Ga) (Ga) ƒSm/Nd
Axim gabbroic body
D9031 (Gabbro) 61 532 0.3319 0.711016 14 0.700971 2.46 11 0.1353 0.511726 9 −1.23 2.55 2.47 −0.312
D9035 (Gabbro) 50 815 0.1775 0.706410 12 0.701038 3.76 15.8 0.1440 0.511849 8 −1.17 2.61 2.47 −0.268
Kegyina gabbroic body
D9039 (Gabbro) 8 520 0.0445 0.703013 11 0.701666 5.04 18.2 0.1675 0.512187 5 −0.94 2.86 2.47 −0.148
D90310 (Gabbro) 6 61 0.2846 0.710010 11 0.701394 3.88 15.7 0.1495 0.512193 8 4.07 2.00 2.03 −0.240
D90311B (Gabbro) 63 538 0.3390 0.712596 11 0.702336 0.65 2.75 0.1430 0.511919 10 0.46 2.40 2.32 −0.273
Aketakyi gabbroic body
D9043 (Gabbro) 12 173 0.2007 0.708362 14 0.702287 1.47 4.22 0.2107 0.513099 7 5.23 5.16 2.19 0.071
Prince's Town granitoid
D9023A (Qtz diorite) 77 380 0.5864 0.710929 10 0.693178 3.97 20.9 0.1149 0.511656 9 2.92 2.12 2.12 −0.416
D9025 (Granodiorite) 46 655 0.2032 0.707690 13 0.701540 2.76 14.9 0.1121 0.511416 8 −1.01 2.43 2.45 −0.430
D9038 (Granodiorite) 75 366 0.5934 0.717431 15 0.699468 3.95 21.6 0.1106 0.511456 7 0.16 2.33 2.35 −0.438
Ahama mafic‐ultramafics
D8301B (Pyroxenite I) 13 214 0.1757 0.706628 14 0.701309 1.66 5.93 0.1694 0.512133 7 −2.49 3.20 2.61 −0.139
D8302 (Pyroxenite I) 14 584 0.0693 0.703508 11 0.701409 1.45 5.4 0.1625 0.512135 9 −0.58 2.71 2.43 −0.174
D90211C (Pyroxenite II) 16 571 0.0810 0.703693 12 0.701240 4.61 14.9 0.1872 0.512269 9 −4.66 5.53 3.05 −0.048
D90212C (Pyroxenite II) 7 502 0.0403 0.702932 13 0.701711 3.08 9.77 0.1907 0.512266 7 −5.68 ‐ ‐ −0.030
Aketakyi UMC
D90410 (Metaperidotite) 1 22 0.1315 0.704456 12 0.700477 0.70 2.17 0.1952 0.51284 8 4.38 1.63 2.01 −0.008
D90411B (Metaperidotite) <1 10 ‐ 0.705412 14 ‐ 0.40 1.28 0.1891 0.512784 8 4.93 1.57 1.96 −0.039
D90413 (Metaperidotite) <1 10 ‐ 0.705089 14 ‐ 0.11 0.32 0.2080 0.512982 14 3.69 0.63 2.01 0.057
Note. εNd values calculated at 2.1 Ga, relative to the present‐day chondritic values of
143Nd/144Nd = 0.512638 and 147Sm/144Nd = 0.1967 ƒ = [(147Sm/144NdSample)/(147Sm/Nd Sm/144NdCHUR)]‐1. TDM1 and TDM2
are single‐stage and two‐stage Nd model ages; TDM1 is computed according to the depleted mantle model of Ben Othman, Polvé, and Allègre (1984) and TDM2 following the approach of Keto and Jacobsen
(1987).
DAMPARE ET AL. 23
FIGURE 9 Primitive‐mantle‐normalized multi‐element patterns of Prince's Town granitoids (PTG) compared with sodic calc‐alkaline Birimian
granitoids (NaCG) reported from central Cote d'Ivoire (Doumbia et al., 1998). Normalization values are from Sun and McDonough (1989)
TABLE 6 Composition of the Prince's Town granItoids compared with the average compositions of adakites, TTGs, sanukitoids, closepet‐type
and experimental melts
TTG1a TTG2a TTG3a Sanukitoidsa Closepeta Experimental
D9021 D9023A D9025 D9038 LSAa HASa >3.5 Ga 3–3.5 Ga <3.5 Ga <62% <62% meltsa
In wt.%
SiO2 66.66 62.96 70.49 68.30 56.25 64.8 69.59 69.65 68.36 58.76 56.39 68.94
TiO2 0.34 0.36 0.27 0.38 1.49 0.56 0.39 0.36 0.38 0.74 1.2 0.78
Al2O3 16.05 16.33 15.72 14.50 15.69 16.64 15.29 15.35 15.52 15.8 15.79 17.7
Fe2O3 3.91 5.75 3.10 4.23 6.47 4.75 3.26 3.07 3.27 5.87 7.34 2.42
MnO 0.08 0.11 0.06 0.08 0.09 0.08 0.04 0.06 0.05 0.09 0.13 0.05
MgO 2.05 2.72 1.53 2.14 5.15 2.18 1 1.07 1.36 3.9 3.38 0.84
CaO 3.90 5.63 3.30 3.48 7.69 4.63 3.03 2.96 3.23 5.57 5.45 2.06
Na2O 4.74 3.74 4.75 3.95 4.11 4.19 4.6 4.64 4.7 4.42 3.94 4.92
K2O 1.98 1.64 1.87 2.95 2.37 1.97 2.04 1.74 2 2.78 3.17 2.53
P2O5 0.13 0.12 0.10 0.11 0.66 0.2 0.13 0.14 0.15 0.39 0.72
In ppm
V 73 74 49 76 184 95 39 43 52 95 129 25
Cr 200 200 160 170 157 41 34 21 50 128 50
Co 13 13 27 13
Ni 90 80 80 70 103 20 12 15 21 72 38 16
Cu 40 20 20 40
Zn 60 50 50 50
Ga 20 16 20 16
Ge 1 1.1 1 1.2
Rb 55 77 46 75 19 52 79 59 67 65 93 98
Sr 667 380 655 366 2,051 565 360 429 541 1,170 978 333
Y 9.3 23.9 9.7 24.3 13 10 12 14 11 18 37 11.9
Zr 104 118 98 130 188 108 166 155 154 184 323 196
(Continues)
24 DAMPARE ET AL.
TABLE 6 (Continued)
TTG1a TTG2a TTG3a Sanukitoidsa Closepeta Experimental
D9021 D9023A D9025 D9038 LSAa HASa >3.5 Ga 3–3.5 Ga <3.5 Ga <62% <62% meltsa
Nb 3.2 4 2.9 4 11 6 8 6 7 10 18 11.4
Sn <1 1 <1 1
Cs 1.1 1.4 1.1 1.4
Ba 712 837 681 845 1,087 721 449 523 847 1,543 1,441 651
La 15.8 32.7 17.7 32.7 41.1 19.2 35.3 31.4 30.8 59.9 90.9 28.65
Ce 33.8 42.4 33.2 41.5 89.8 37.7 61.7 55.1 58.5 126 188 53.56
Pr 4 6.18 4.18 6.16
Nd 14.9 20.9 14.9 21.6 47.1 18.2 25.8 19.6 23.2 54.8 84.9 25.05
Sm 2.76 3.97 2.76 3.95 7.8 3.4 4.2 3.3 3.5 9.8 14.5
Eu 0.812 1.09 0.775 1.08 2 0.9 1 0.8 0.9 2.3 3.2 1.23
Gd 2.16 3.85 2.16 3.82 4.8 2.8 3.2 2.4 2.3 6 9.2
Tb 0.31 0.59 0.29 0.6
Dy 1.65 3.31 1.5 3.33 2.8 1.9 1.8 1.9 1.6 3.2 5.6 2.35
Ho 0.3 0.68 0.28 0.67
Er 0.9 2.07 0.8 2.08 1.21 0.96 0.77 0.77 0.75 1.41 2.68 1.21
Tm 0.134 0.303 0.121 0.306
Yb 0.95 1.95 0.83 1.96 0.93 0.88 0.78 0.63 0.63 1.32 2.05 0.94
Lu 0.148 0.303 0.13 0.31 0.08 0.17 0.2 0.13 0.12 0.26 0.34
Hf 2.8 3.3 2.6 3.5
Ta 0.29 0.35 2.66 0.4
W <0.5 0.7 80.1 2.1
Tl 0.36 0.36 0.24 0.42
Pb 12 9 9 11
Bi 0.6 0.4 0.5 0.4
Th 2.9 4.01 3.77 4.03
U 0.95 1.15 0.89 1.1
aData from Martin et al. (2005).subducting oceanic slab (e.g., Defant & Drummond, 1990; Martin,
Smithies, Rapp, Moyen, & Champion, 2005), (b) crustal assimilation
and fractional crystallization (AFC) processes from parental basaltic
magmas (Castillo et al., 1999), (c) partial melting of a stalled (or dead)
slab in the mantle (e.g., Qu, Hou, & Li, 2004), (d) partial melting of
delaminated lower crust (e.g., Kay & Mahlburg Kay, 1993; Wang
et al., 2006), and (e) partial melting of mafic rocks in the lower part
of thickened crust (Atherton & Petford, 1993; Petford, Atherton, &
Halliday, 1996). On the basis of an extensive adakite geochemical
database, Martin et al. (2005) have grouped adakites into two groups:
high‐SiO2 (HSA) and low‐SiO2 (LSA) adakites. High‐SiO2 (HSA) have
SiO2 (>60 wt.%), MgO (0.5 to 4 wt.%), CaO + Na2O (<11 wt.%), and
Sr (<1,100 pm), compared to LSA with SiO2 (<60 wt.%), MgO (4 to
9 wt.%), CaO + Na2O (>10 wt.%) and Sr (>1,000 ppm).
The HSA are considered to be product of subducted basaltic slab
melts that have reacted with peridotite during its ascent through man-
tle wedge, whereas LSA are believed to have been generated bymelting of a peridotitic mantle wedge whose composition has been
modified by reaction with felsic slab melts (Martin et al., 2005). Sam-
ples D9021 and D9025 are similar to HSA (Table 6; Figure 10) and
may have been generated by similar processes.
Mauer (1990) indicated that the belt‐type granitoids and tholeiitic
basalts of Ghana are genetically closely related and that the granitoids
were derived from partial melting of the tholeiitic basalts. This claim is
corroborated by Hirdes et al. (1992), who also indicated from their iso-
topic data that the belt volcanic rocks and belt granitoids are coeval
and comagmatic, and were formed at different stages of the same
igneous event. The major element data also suggest that the Prince's
Town plutonics were mostly likely derived from the metavolcanics.
The analysed granitoids are characterized by negative and positive
εNd (2.1 Ga) values ranging from −1.01 to +2.92 and TDM2 of 2.12–
2.45 Ga. The relatively high positive initial εNd of +2.92 and
Protererozioc Nd model age of 2.17 Ga obtained for a quartz diorite
(D9023A) suggest that it was juvenile at its time of formation. For
DAMPARE ET AL. 25
FIGURE 10 Compositions of the Prince's Town granitoids compared
with those of the average adakites (after Martin et al., 2005)the other granodioritic samples (D9025 and D9038), the εNd (2.1 Ga)
values of −1.01 and +0.16 with Nd model ages of 2.45 and 2.35 Ga
suggest partly enriched mantle sources. The granitoids, however, dem-
onstrate primitive initial 87Sr/86Sr ratios of ~0.700. Therefore, the iso-
topic signatures of the Prince's Town granitoids are indicative of
juvenile crust which may have been contaminated by some amount
of an older crustal component. A slightly crustal contamination of
the granitoids is also demonstrated by the inverse correlation between
εNd (2.1) and SiO2 (Figure 11).5.2.2 | Mafic and ultramafic rocks
Fractional crystallization and crustal contamination
The Ahama pyroxenites, Aketakyi, Axim, and Kegyina gabbroic rocks
exhibit chemical characteristics different from the Aketakyi UMC in
MgO variation diagrams (Figure 7). Loh and Hirdes (1999) performed
XRF analysis on these rocks, except the Kegyina gabbros, and their
chemical data are included in the MgO variation plots. The Aketakyi
gabbroic rocks display a common evolutionary trend in MgO variation
diagrams (Figure 7). The Al2O3 and CaO contents are negativelyFIGURE 11 Plots of εNd (2.1 Ga) versus SiO2, TiO2, and P2O5 (wt.%)
for the Palaeoproterozoic rocks from the Ashanti volcanic beltcorrelated with MgO, which suggest fractionation of pyroxene. The
K2O and Na2O display a negative correlation with MgO, suggesting
fractionation of hornblende. The role played by plagioclase is envisaged
in the evolution of these magmas, as well as other gabbroic bodies,
based on their variable Eu anomalies (Figure 4b). In the Axim and
Kegyina gabbros as well as the Ahama pyroxenites, Al2O3 and CaO
are again correlated negatively with MgO, whereas the Cr and Ni corre-
late positively with MgO. These trends suggest fractionation of olivine
and pyroxene (Figure 7). The rocks also show similar evidence of frac-
tionation of hornblende as the Aketakyi gabbroic rocks. A well‐defined
negative correlation between TiO2 and MgO suggests fractionation of
Fe–Ti oxides. Although the Ahama pyroxenites, Aketakyi, Axim, and
Kegyina gabbroic rocks demonstrate some similarities in the MgO vari-
ation diagrams (Figure 7), some chemical variations, including REE pat-
terns (Figure 4b,c), exist among these individual bodies, which suggest
that they cannot be related in a single line of descent from a single
parental magma via fractional crystallization. The variation in the Nd iso-
topic composition of these magmatic rocks (Table 5) also precludes
them from being part of a single magmatic system. Variations in the
Sr–Nd isotopic composition of magmatic rocks are commonly explained
by either crustal contamination or heterogeneous mantle sources.
Crustal contamination in the magmatic rocks was visualized in the
plot of εNd (T) versus SiO2 (Figure 11). The lack of systematic trend of
decreasing εNd (T) values with increasing SiO2, as would be expected
in rocks affected by crustal contamination, suggests that crustal con-
tamination was minimal in the magmas of these mafic and ultramafic
bodies. The mafic and ultramafic rocks display negative Nb and Ta
anomalies relative to LILE and LREE in the primitive‐mantle‐normalized,
trace element diagrams (Figure 5b,c). Crustal contamination could also
produce such anomalies but would also result in positive Zr–Hf anoma-
lies due to enrichment of these elements in crustal materials (Zhao &
Zhou, 2007). Taylor and McLennan (1985) have also indicated that
upper crustal rocks are typically enriched in U, Zr and Hf elements,
and depleted in Nb, Ta, and Ti. Thus, the negative U, Th, Zr, and Hf
anomalies in the rocks (Figure 5b,c) cannot be explained by upper
crustal contamination. Therefore, the negative Zr–Hf anomalies in the
rocks suggest that only little or no crustal contamination occurred in
them. Crustal components are also rich inTh and Pb. Thus, low and con-
stant Th and Pb, low Th/Yb, and high Nb/Th ratios in the rocks are also
inconsistent with crustal contamination (Figure 12). However, compared
to the gabbroic rocks, the Axim gabbros have slightly higher Th and Pb
contents, and higher Th/Yb and ratios, suggesting some small degree of
crustal contamination might have occurred (Table 2; Figure 12a,c).
Nature of the mantle source region
High Nb and Ta values in magmatic rocks have been interpreted to
indicate oceanic island basalt (OIB) component in those rocks (e.g.,
Edwards et al., 1994; Zou, Zindler, Xu, & Qu, 2000). The Axim,
Kegyina, and Aketakyi rocks show depletion in Nb and Ta. The Ahama
pyroxenitic rocks also display negative Nb–Ta anomalies (Figure 5).
Negative Nb–Ta anomalies may also be caused by crustal contamina-
tion. However, the negative Zr–Hf anomalies observed in the rocks
are inconsistent with crustal contamination, as crustal materials are
26 DAMPARE ET AL.
FIGURE 12 Plots of (a) Nb/Ta vs. Th/Yb, (b) Nb/Ta vs. Zr/Hf, (c) Nb/Th vs. Th, and (d) Zr/Ndpm vs. Nb/Lapm for the gabbroic and pyroxenitic
rocks from the southern Ashanti greenstone belt. Values of N‐MORB, OIB, and primitive mantle are from Sun and McDonough (1989). Values
for the upper and lower crust are from Wedepohl (1995)enriched in elements such as Th, Zr, Hf, and so on (Figure 5b,c). Neg-
ative Zr–Hf anomalies are not common among intraplate basalts
(Zhou, Kennedy, Sun, Malpas, & Lesher, 2002). Thus, the negative
Nb and Ta and Zr–Hf anomalies indicate the lack of an OIB compo-
nent in their source region. Rather, the gabbroic and pyroxenitic rocks
may have been derived from lithospheric mantle sources.
Pearce and Peate (1995) have indicated that the budget of moder-
ately incompatible elements (HREE, HFSE, and Ti) is mainly controlled
by partial melting processes, and such elements may be used to esti-
mate the degree of the source depletion (Woodhead, Eggins, & Gam-
ble, 1993). It has been indicated that the HFSE may be used to
constrain the nature of the mantle sources which may have been
depleted by previous melt extraction in back‐arc basins (Elliott, Plank,
Zindler, White, & Bourdon, 1997; Woodhead et al., 1993) or in arc set-
tings (Grove, Parman, Bowring, Price, & Baker, 2002). Experimental
studies have shown that Nb–Ta and Zr–Hf have considerably different
partition coefficients in the system clinopyroxene/anhydrous silicatemelt (Hart & Dune, 1993; Johnson, 1998), where Zr is mostly more
incompatible than Hf by a factor of 1.5–2, and the DNb/DTa ratio is less
than 1. Consequently, Nb/Ta and Zr/Hf could be significantly fraction-
ated and would be positively correlated during partial melting of the
upper mantle. Plank and White (1995) have indicated that mantle
wedge‐derived arc basalts may inherit such ratios, reflecting variable
degrees of mantle depletion produced prior to episodes of melt gener-
ation. In view of this, very low HFSE contents and small Nb/Ta ratios
in arc volcanic rocks are believed to reflect previous melt extraction
from the mantle wedge (Woodhead et al., 1993). Woodhead et al.
(1993) and Pearce and Peate (1995) have indicated the process
involved in the melt extraction would also lead to positive correlations
of sub‐chondritic Nb/Ta ratios with La/Yb, Th/Yb, Zr/Yb, Zr/Ti, Ti/V,
and Y/Sc ratios, which would decrease with increasing depletion of
the mantle wedge. However, the pyroxenitic and gabbroic rocks
largely do not exhibit such consistency in the elemental correlations
(Figure 12b,d). It appears, therefore, that the mantle source region of
DAMPARE ET AL. 27the rocks was a normal lithospheric mantle wedge, which had not
been depleted by previous melt extraction.
Gabbros D9042 and D9043 from the Aketakyi gabbroic body
reveal LREE‐depletion as in MORB yet an overabundance of fluid‐
soluble elements that indicate a slab‐derived component (Figures 4b
and 5b). The high initial εNd value of +5.23 in the D9043 indicates
depleted upper mantle sources. Its Nd model age of 2.15 Ga suggests
a juvenile character. The initial 87Sr/86Sr ratio of 0.702287 obtained
for D9043 suggest a small degree of crustal contamination of the
mantle melt. The LREE‐depleted gabbroic samples probably originated
at a back‐arc‐basin spreading centre where LREE depleted rocks with
enrichment of fluid mobile elements are ubiquitous (e.g., Bach,
Hegner, & Erzinger, 1998; Jenner, Cawood, Rautenschlein, & White,
1987). The analysed Axim gabbros show consistent negative initial
εNd values (−1.23 and −1.17) but low initial
87Sr/86Sr ratios
(0.700971 and 0.701038). The Nb depletion relative to the LREEs,
negative Nb–Ta and Zr–Hf anomalies (Figures 4b and 5b), relatively
high LILE/HREE (or LILE/HFSE) ratios as well as their low initial εNd
values could be interpreted as the signatures of the addition of
subduction‐related fluids into a depleted mantle source (Kepezhinskas
et al., 1997). The high Nd model ages (TDM2) of 2.54 and 2.55 Ga
obtained for the Axim gabbros suggest an input of a pre‐Birimian
crustal component in their formation, probably through subduction.
The Axim gabbros most likely originated from a subduction‐related
lithospheric mantle. The Kegyina gabbroic body demonstrates a wide
range of initial εNd values (−0.94, 0.46, and 4.07) which indicates partly
enriched mantle and depleted mantle sources. Sample D90310 with
high initial εNd value (+4.07) show similar characteristics as the
Aketakyi gabbro D9043 but shows LREE‐enrichment. D90310 dem-
onstrates a Proterozoic Nd model age (2.08 Ga), suggesting a juvenile
character. The εNd (2.1) values of −0.94 and +0.46 obtained for the
other Kegyina rocks, D9039 and D90311B, with respective Nd model
ages of +2.57 and +2.39 are indicative of some contamination of their
mantle belts by older crustal materials. Like the Axim gabbros, they
could have also originated from a subduction‐related lithospheric
mantle. The analysed Ahama pyroxenitic rocks have initial 87Sr/86Sr
ratios of 0.701211–0.701711, which indicate minimum crustal con-
tamination of the mantle melt. The εNd (2.1) values of these samples
are −0.58, −2.49, −4.66, and −5.68, indicative of partly enriched to
enriched mantle products. The pyroxenitic rocks display Nb depletion
relative to the LREEs, negative Nb–Ta and Zr–Hf anomalies (Figure 5
c), relatively high LILE/HREE (or LILE/HFSE) ratios, and these features
combined with their low initial εNd values could point to subduction‐
related lithospheric mantle sources. Their high negative εNd (2.1)
values and Nd model ages of 2.52 and 2.75 Ga are possibly indicative
of a greater subduction input of an older crustal material.
Source modification by subduction components
The mantle wedge above a subduction zone may be modified by com-
ponents from the down‐going slab by processes such as (a) fluxing of
fluids derived from dehydration of altered oceanic crust
(Hawkesworth, Turner, McDermott, Peate, & van Calsteren, 1997) or
subducted sediments (e.g., Class, Miller, Goldstein, & Langmuir,2000) and (b) addition of melts of subducted sediment (e.g., Münker,
2000) or melts of the MORB portion of the slab (Peacock, Rushmer,
& Thompson, 1994). As demonstrated earlier, the trace element geo-
chemical signatures of the gabbroic and pyroxenitic rocks suggest that
they may have been derived from a metasomatized lithospheric man-
tle source above a subduction zone.
The abundance of highly incompatible elements in magma is con-
trolled by fluid influx (Stolper & Newman, 1994). Rare earth elements
and HFSE (e.g., Th, Zr, Hf, Nb, and Ta) are relatively immobile in aque-
ous fluids (Keppler, 1996; Turner et al., 1997), compared with LILE.
Therefore, the enrichment of REE and HFSE in a mantle wedge is
attributed to inputs of slab melts rather than aqueous fluids (Elliott
et al., 1997; Plank & Langmuir, 1992). Mantle sources modified by slab
melts may have lower Th/Zr, Rb/Y (Kepezhinskas et al., 1997), Ba/Nb
and Ba/Th ratios (Hawkesworth et al., 1997) than those modified by
fluids. The chemical compositions of the gabbroic and pyroxenitic
rocks suggest that their mantle sources experienced largely fluid and
some melt metasomatism above a subduction zone.
The Ahama Group I pyroxenites show Th–U and Nb–Ta troughs,
positive Sr anomalies, and minor negative Zr–Hf and Ti anomalies,
whereas the Group II pyroxenites exhibit negative Th anomalies, pro-
nounced negative Nb, Zr–Hf and Ti anomalies, and spikes in Sr
(Figure 5c). The Th/U ratios of the Ahama Group I pyroxenites vary
from 2.24 to 2.91 and are relatively higher than those of the Group
II rocks (0.79–1.96); both types have Th/U ratios less than those of
OIB (3.5–3.8) and enriched mantle (EMI) which range from 4.5 to 4.9
(Weaver, 1990). It is noted that Zr/Nb ratios in mafic rocks range from
about 40 in mid‐ocean ridge basalt (MORB) to 10 in E‐type MORB and
to <10 in OIB and rift‐related lavas (Pearce & Norry, 1979). The Zr/Nb
ratios of the pyroxenites vary from 21 and 27, falling between OIB
and MORB, but comparable with the Zr/Nb ratios of the associated
LREE‐enriched basalts and andesites from the southern Ashanti volca-
nic belt (Dampare, Shibata, Asiedu, Osae, & Banoeng‐Yakubo, 2008).
Loh and Hirdes (1999) obtained Zr/Nb ratios of 5–34 for the pyroxe-
nites, similar to this study. They also have high Ba/Nb (95–207), Ba/Zr
(4–9), Ba/Th (308–588), and U/Th ratios (0.3–0.7) relative to the prim-
itive mantle. Loh and Hirdes (1999) also obtained high Ba/Nb (39–
305), Ba/Zr (8–9), Ba/Th (194–305), and U/Th (1) ratios for the pyrox-
enites. Such high elemental ratios in the pyroxenites, which are also
higher than those of the average upper crust, suggest that they were
more likely derived from a mantle source strongly modified by hydrous
fluids. The Axim gabbroic rocks also exhibit high ratios of Ba/Nb (166–
390), Ba/Zr (10–19), Ba/Th (485–866), and U/Th ratios (0.4) relative
to the primitive mantle. These observations are consistent with the
previous work by Loh and Hirdes (1999), who obtained Ba/Nb,
Ba/Zr, Ba/Th, and U/Th ratios of 421–565, 7–9, 22–565, and 0.1–1,
respectively, for the Axim gabbroic rocks. The Kegyina gabbroic rocks
have higher ratios of Ba/Nb (42–473), Ba/Zr (2–16), Ba/Th (146–
1,112), and U/Th ratios (0.3–1.27) relative to the primitive mantle, also
indicating that their mantles were strongly modified by hydrous fluid.
The Aketakyi gabbroic rocks have higher ratios of Ba/Nb (72–90)
and Ba/Zr (4) but retain very low Th concentrations. Loh and Hirdes
(1999) obtained Ba/Nb (47–93), Ba/Zr (1–7), Ba/Th (47–93), and
28 DAMPARE ET AL.U/Th ratios of 47–93, 1–7, 47–93, and 1, respectively. Thus, the man-
tle source of the Aketakyi gabbroic rocks may not have been modified
significantly by slab‐derived fluids.
It can be inferred from the chemical and isotopic composition of
the rocks that the Ahama pyroxenites were extracted from a more
fluid‐metasomatized mantle different from that of Axim, Kegyina,
and Aketakyi gabbroic rocks, with the mantle source of the Aketakyi
gabbros being the least metasomatized. These mafic and ultramafic
rocks, which are genetically unrelated and have different emplacement
records, were later juxtaposed. The Aketakyi gabbros are associated
with the LREE‐depleted, back‐arc basin basalts, the sheet‐dyke com-
plex, and the Aketakyi ultramafic rocks in the study area, and they
could be part of an obducted piece of a back‐arc basin. The Prince's
Town pluton intrudes the western flank of the Aketakyi ultramafic
complex (Attoh et al., 2006), and the Aketakyi gabbro is inferred to
have been emplaced earlier than the Prince's Town pluton. In the field,
the Axim gabbroic body occurs to the west of the Prince's Town plu-
ton and is associated with argillite and andesitic lavas in the transition
zone between the Kumasi basin and the Ashanti belt. The Axim body
is believed to have been emplaced later than the granitoids. The
Ahama ultramafic body is located to north of the Axim volcanic belt
(Figure 2) is associated with the LREE‐enriched volcanic arc basalts
and andesites (Dampare, Shibata, Asiedu, Osae, & Banoeng‐Yakubo,
2008). The volcanic rocks are also intruded by the Prince's Town plu-
ton, but the relationship between the Ahama ultramafic body and
the Prince's Town pluton is uncertain. The Kegyina mafic body occurs
at the margin of the Prince'sTown pluton and could be coeval with the
latter. The Aketakyi gabbro is inferred to have been emplaced earlier
than the Prince's Town pluton, as the latter intrudes the western flank
of the Aketakyi ultramafic complex, of which the Aketakyi gabbro is
considered to be part.
Aketakyi ultramafic complex
Ultramafic–mafic rocks have diverse origins and could represent one
of the following: (a) fertile upper mantle, (b) residual upper mantle,
and (c) ultramafic cumulate in large magma chambers at the base of
the crust. Ultramafic–mafic rocks can occur as complete or incomplete
ophiolitic complexes as well as rootless bodies tectonically emplaced
into lower and upper crustal rocks.
The chemical compositions of the analysed rocks from the
Aketakyi UMC have provided some useful petrogenetic information
about the rocks. Petrographically, the rocks from the Aketakyi UMC
appear to be crystal cumulates, and this is confirmed by the geochem-
ical data (Table 4). Trace elements are depleted in the rocks as a result
of the cumulate effect. Crystal fractionation is suggested by MgO var-
iation diagrams (Figure 7), which include the chemical data of the pres-
ent study and that of Loh & Hirdes (1999). The compatible trace
elements Ni, Co, and Cr are positively correlated with MgO. Nickel is
mainly controlled by fractionation and accumulation of olivine, unless
sulphide is present in the magma. Cobalt is also controlled mainly by
fractionation and accumulation of olivine. Chromium is controlled by
fractionation of chromite and diopside, and V is controlled by fraction-
ation and accumulation of magnetite. The Ga and V contents of therocks show an inverse correlation with MgO, reflecting their incom-
patible behaviour. The abundances of Al2O3, Na2O, CaO, and SiO2
are inversely correlated with MgO (Figure 7). The negative correlation
of SiO2 and Al2O3 with MgO indicates fractionation of silica‐poor,
magnesia‐rich minerals such as orthopyroxene and olivine. The
observed trends in variation diagrams suggest fractional crystallization
of olivine and pyroxene. The pyroxenitic‐gabbroic varieties have
higher Sr contents than the peridotites, signifying the importance of
plagioclase in their formation.
The nature of the parental magma of the ultramafics has been
investigated following the method of Roeder and Elmslie (1970). The
olivine composition is used to estimate the MgO/FeO value of the pri-
mary magma using Mg–Fe distribution coefficient (Kd = [(FeO/MgO)
ol/(FeO/MgO)magma]molar = 0.3 ± 0.03). Although Kd is relatively insen-
sitive to temperature, pressure, and liquid composition (Roeder &
Elmslie, 1970), the cumulate olivine can be less magnesium‐rich than
that initially formed because of re‐equilibration with the residual
magma. Hence, the calculated MgO/(MgO + FeO) value might be
lower than that of the parent magma. Therefore, it is better to select
the most magnesian olivine in order to constrain the MgO and FeO
values.
Some information on olivine compositions for the Aketakyi UMC
has been provided by Attoh et al. (2006), and combining their results
with those newly obtained in this study, the most magnesian olivine
in this complex is Fo85. Using the distribution coefficient of 0.3, Fo85
corresponds to a Mg# (molar ratio of MgO/(MgO + FeO)) of 0.65 in
the liquid. Therefore, it can be concluded that the parent magma of
the Aketakyi UMC was high in magnesium, although the precise con-
centrations of SiO2, MgO, and FeO could not be obtained using this
method. Thus, the ultramafic complex probably originated from a
mafic magma in the greenstone belt of which the tholeiitic basalt is a
possible candidate. The analysed Aketakyi meta‐peridotites have Nd
model ages (1.99–2.04 Ga), which suggest that they were juvenile at
their time of formation. Their εNd (2.1 Ga) values (+3.69 to +4.93) sug-
gest they derived from depleted mantle magmas.
Brown amphibole in the Aketakyi UMC shows postcumulus tex-
tural characteristics (i.e., postcumulus phase enclosing olivine, pyrox-
ene, and serpentinized olivine grains). The occurrence of primary
magmatic amphibole suggests the presence of relatively high water
contents in the parental magma. Thus, hydrous partial melting of a
mantle source which had previously been metasomatized is envisaged.5.3 | Tectonic setting considerations
5.3.1 | Southern Ashanti volcanic belt
Several schemes exist for assigning granitoids to various tectonic envi-
ronments based on their geochemistry (e.g., Batchelor & Bowden,
1985; Maniar & Piccoli, 1989; Pearce, Harris, & Tindle, 1984). In this
study, the trace element characteristics of the Prince'sTown granitoids
have been compared with a volcanic arc granite reference sample from
Chile and a within‐plate granite reference sample from Oslo (Pearce
et al., 1984). The trace elements of the granitoids show an overall
DAMPARE ET AL. 29negative slope from the LILE to the HFSE and depletion of Nb and Ta,
similar to those of granitoids emplaced in a modern‐day active conti-
nental margin context, such as depicted by the reference pattern of
volcanic arc granite (VAG) from Chile (Figure 13). Enrichment in LILE
and depletion in Nb and Ta usually characterize active margin‐related
magmas, whereas the development in a within‐plate context would
result in enrichment in Nb and Ta such as observed in the within‐plate
granitoid from Oslo (Figure 13). The observed tectonic setting of the
Prince's Town pluton is consistent with the previous studies carried
out on the granitoids from the southern Ashanti belt (Dampare et al.,
2005; Loh & Hirdes, 1999) and other Birimian volcanic belts of Ghana.
The LILE and LREE enrichment and HFSE depletion in the Ahama
pyroxenites and, the Axim, Kegyina and Aketakyi gabbroic rocks indi-
cate involvement of subduction‐related components in their source
region (Figure 5). Their geochemical signatures resemble those of the
associated metavolcanic rocks, which formed in an intra‐oceanic island
arc–back‐arc environment (Dampare, Shibata, Asiedu, Okano, et al.,
2008), and those of island arc intrusions (e.g., Zhou et al., 2002), sug-
gesting a subduction zone setting. Thus, these mafic and ultramafic
intrusions were derived from arc‐related magmatism during the
Palaeoproterozoic.
Ultramafic rocks may be subdivided according to their geotectonic
setting into the subcontinental and suboceanic mantle, including man-
tle underneath oceanic ridges, oceanic islands, oceanic plateaus, island
arcs, fore‐ and back‐arc regions, and so on (e.g., Bonatti & Michael,
1989). Also, the upper mantle can be pervasively metasomatized, or
affected by melt–rock, melt–fluid, and rock–fluid interactions. The
Aketakyi ultramafic complex has previously been suggested to repre-
sent a fragment of a Palaeoproterozoic ophiolitic complex emplaced
in a supra‐subduction zone (SSZ) tectonic setting (Attoh et al., 2006).
The inferred tectonic setting of the complex is further supported byFIGURE 13 Ocean ridge granite (ORG)‐normalized spider diagram of
the Prince's Town granitoid. The ORG normalization values and data
for volcanic arc and within‐plate granites are from Pearce et al. (1984)the following lines of evidence: (a) some of the rocks display LREE‐
depleted MORB‐like patterns; such patterns are not only restricted
to MORB but could also be observed in magmas at back‐arc basin;
(b) some of the rocks show slight depletion in Nb, Ta, Zr–Hf, and Ti
(i.e., mild subduction‐zone signature; Figure 5c). Almost all the rocks
demonstrate strong enrichment in Cs; however, Cs enrichment could
also result from serpentinization. Although these arc‐/subduction‐
related features are not observed in all the rocks analysed, it is difficult
to envisage how they could have formed in different tectonic settings;
(c) the basaltic rocks associated with the ultramafic complex show
geochemical signatures characteristic of back‐arc basin basalts; and
(d) the Prince's Town granitoids, which intruded into the western flank
of the ultramafic complex, are island arc‐related granitoids. Thus, the
Aketakyi UMC formed in a supra‐subduction zone environment.
The geotectonic environment in which the plutonic rocks were
formed or emplaced has also been investigated in the plot of ƒSm/Nd
versus εNd (2.1 Ga). Most of the rocks fall in and close to the field of
the Proterozoic arc crust (Figure 14; Roddaz et al., 2007 and the refer-
ences therein). Indeed, their εNd (2.1 Ga) values overlap the fields of
Palaeoproterozoic arc crust and MORB. The Aketakyi UMC meta‐
peridotites plot in the arc field, thereby demonstrating their affinity
for a supra‐subduction environment.
The southern Ashanti volcanic belt consists of serpentinized peri-
dotites, pyroxenites, gabbros, granitoids, and metavolcanic rocks.
However, the real relationship between the rocks is obscured because
of lack of definite isochron ages and poor outcrop exposure. The
supra‐subduction zone setting proposed for these rocks is mainly
based on geochemical, isotopic, and field observations outlined here:
(a) The association of pillow basalt and gabbros, pronounced develop-
ment of andesitic flows and pyroclastic rocks as well as occurrence of
cherty rocks in the belt suggest an intra‐oceanic depositional environ-
ment. (b) The high initial εNd (2.1) isotopic values (e.g., +4.38 to +5.23)
of some of the plutonic rocks indicate a long‐term LREE‐depleted
mantle source(s), similar to the source of modern N‐MORB (see Hof-
mann, 2003). High initial εNd (2.1) isotopic values of +3.89 to +7.21
were also obtained for tholeiitic basalts in the volcanic belt (Dampare,
Shibata, Asiedu, Osae, & Banoeng‐Yakubo, 2008). However, some of
the rocks have Nb‐depleted, relative to Th and La, trace element pat-
terns (Dampare, Shibata, Asiedu, Osae, & Banoeng‐Yakubo, 2008; this
study), consistent with a subduction zone geodynamic setting. (c)
Large variations in the initial εNd values may reflect either mantle
source heterogeneity or crustal contamination. As discussed in the
previous sections, contamination by continental crust during magma
ascent, rather than contamination of their source regions by
subducted crustal material, may be minimal. (d) Adakitic (this study)
and boninitic (Dampare, Shibata, Asiedu, Osae, & Banoeng‐Yakubo,
2008) signatures observed in some of the rocks also support
subduction‐zone environment. In this context, the generated supra‐
subduction zone (SSZ) oceanic crust in the back‐arc basin consists of
the Aketakyi ultramafic‐mafic complex, Aketakyi gabbro, and BABB‐
type basalts (massive and pillowed lavas). This probably represents
an upper mantle transition and crustal section of a back‐arc basin
ophiolitic complex. The arc magmatism in the study area is
30 DAMPARE ET AL.
FIGURE 14 Plot of εNd (2.1 Ga) versus
fractionation parameter (ƒSm/Nd). Fields of old
(Archaean) continental crust, MORB, and the
arc rocks are from Roddaz et al. (2007) and
the references thereincharacterized by island arc calc‐alkaline series volcanics (Dampare
et al., 2008), mafic (Axim and Kegyina gabbroic rocks), and ultramafic
(Ahama pyroxenitic rocks) rocks as well as the Prince's Town granitoid.
These supra‐subduction rocks were accreted to the continental mar-
gin. Thus, the volcanic and plutonic rocks of the southern Ashanti
greenstone belt consist of an association of island arc and back‐arc
sequences that have been amalgamated during subduction–accretion
and collisional obduction.
5.3.2 | Ashanti belt and the WAC
As was mentioned in the earlier section, the geotectonic setting of the
Palaeoproterozoic rocks of the West African Craton has been debated
and various interpretations have emerged, even from the same area. In
Ghana, for example, Leube et al. (1990) have proposed an
intracontinental rift model for the Palaeoproterozoic rocks, whereas
arc or subduction related models are preferred by some other workers
(Asiedu et al., 2004; Attoh et al., 2006; Dampare et al., 2008; Dampare
et al., 2005; Loh & Hirdes, 1999; Pohl & Carlson, 1993; Sylvester &
Attoh, 1992). Feybesse et al. (2006) developed a metallogenesis model
for the occurrence of gold deposits in Ghana which invoked a combi-
nation of continental margin, juvenile magmatism and convergence
and collision between an old continent and a juvenile crust for the tec-
tonic environments in which the Birimian greenstone belts were gen-
erated. In this context, the Ashanti volcanic belt is considered to mark
the boundary between an Archaean continental domain and a Birimian
oceanic domain. For Harcouët et al. (2007), the basement in the
Ashanti area is likely to be of continental type rather than oceanic,
as the results of their numerical modelling using thermal parameters
such as thermal conductivity values and heat‐production rates are
more consistent with the continental basement scenario. Leube et al.
(1990) have indicated that most Birimian tholeiites have N‐MORB
chemistry with few showing slight to moderate LREE enrichment.
The authors interpret the coexistence of both flat and LREE‐enriched
patterns of tholeiites and also the εNd value of +2 (Taylor, Moorbath,
Leube, & Hirdes, 1988) as indications of slight crustal contamination
in the petrogenesis of the tholeiites. For Sylvester and Attoh (1992),the trace element chemistry of Birimian volcanic belts in Ghana is
comparable to that of Archaean greenstone belts, and that intermedi-
ate calc‐alkaline units show the high Ba/La and Ta depletion which is
characteristic of similar rocks in modern subduction environments and
distinct from those formed in intraplate settings. The authors there-
fore suggest that Birimian volcanic rocks probably originated as imma-
ture island arcs built on oceanic crust. Dampare, Shibata, Asiedu,
Okano, et al. (2008) have, on the basis of geochemistry, inferred an
intra‐oceanic island arc–fore‐arc–back‐arc setting for the
Palaeoproterozoic metavolcanic rocks from the southern Ashanti vol-
canic belt. They recognize basalts with both flat to slightly LREE‐
depleted (N‐MORB‐like) and LREE‐enriched patterns, and LREE‐
enriched calc‐alkaline andesites in the volcanic suite, with the rocks
displaying various degrees of subduction‐related trace element char-
acteristics. The formation of most of the Palaeoproterozoic volcanic
belts in the oceanic context has also been proposed by other workers
(e.g., Abouchami et al., 1990; Ama Salah, Liégeois, & Pouclet, 1996;
Béziat et al., 2000; Boher et al., 1992; Davis et al., 1994; Egal et al.,
2002; Liégeois et al., 1991; Vidal & Alric, 1994). Moreover, geochro-
nological and isotopic data (e.g., Abouchami et al., 1990; Ama Salah
et al., 1996; Boher et al., 1992; Doumbia et al., 1998; Gasquet et al.,
2003; Hirdes et al., 1996; Liégeois et al., 1991; Taylor et al., 1992)
indicate a juvenile source for the Palaeoproterozoic volcanic and gra-
nitic rocks, with the exceptions of plutons from the Winneba area in
the southeast of Ghana (Taylor et al., 1992) and the vicinity of the
Man nucleus (e.g., Kouamelan, Peucat, & Delor, 1997), where there
is evidence of a stronger influence of recycled Archaean basement.
In terms of crustal growth event in the West African Craton,
Abouchami et al. (1990) and Boher et al. (1992) proposed a mantle
plume model, in which extensive plateaus were formed by a mantle
plume event, followed by formation of island arcs on the top of the oce-
anic plateaus that then collided with the Man Archaean craton. In the
case of Ghana, Taylor et al. (1992) have indicated, from geochronologi-
cal and isotopic data, that the Birimian of Ghana represents a major
Early Proterozoic magmatic crust‐forming event around 2.3–2.0 Ga by
differentiation from a slightly depleted mantle source. Thus, the
Birimian of Ghana forms part of the major Proterozoic (Eburnean)
DAMPARE ET AL. 31
FIGURE 15 Plot of εNd (2.1 Ga) versus formation age (time) for the
Palaeoproterozoic rocks from the Ashanti volcanic belt. Data of
Archaean continental crust is from Kouamelan, Peucat, and Delor
(1997) CHUR = Chondritic uniform reservoir. Symbols as in Figure 14episode of juvenile crustal accretionwhich is recognized inWestAfrican
Craton and dated at 2.2–2.1 by Abouchami et al. (1990). Davis et al.
(1994), also following the accretionary model, have proposed that the
Birimian sedimentary basins resulted from accretion of arcs and oceanic
plateaus now represented by allochthonous Birimian volcanic belts.
Our present geochemical and isotopic data as well as field observa-
tions are inconsistent with the mantle plume model proposed for
Palaeoproterozoic rocks in Ghana (e.g., Leube et al., 1990) but rather
plead in favour of a supra‐subduction setting in the study area.
Magmatism associated with mantle plumes is usually voluminous, con-
sists of basaltic lavas, high‐Mg komatiites and picrites, and may be
associated with radiating mafic dyke swarms (e.g., Park et al., 1995).
Although some komatiites have been identified among the volcanic
rocks in Ghana (Leube et al., 1990), they are generally insignificant in
volume and are scarce in the southern Ashanti volcanic belt.
Palaeoproterozoic igneous assemblages in the southern Ashanti belt
include granitoids with volcanic‐arc‐like geochemical features (Loh &
Hirdes, 1999; Dampare et al., 2005; this study). Thus, evidence for a
Palaeoproterozoic mantle plume is lacking. The Palaeoproterozoic
mafic and ultramafic rocks in the belt do not have OIB signatures,
ruling out a plume‐related origin. Instead, the source region was
strongly modified by fluids and melts derived from a subducted slab.
The question arises as to whether this modified mantle was a relic pro-
duced by older subduction or if subduction was coeval with produc-
tion of the mafic magmas. An active subduction environment is
consistent with the geology of the area, a back‐arc volcanic sequence
associated with the ophiolitic plutonics (Dampare, Shibata, Asiedu,
Osae, & Banoeng‐Yakubo, 2008). It is also known that some of the
granitoid plutons in the region have adakitic affinities (this study),
and such adakitic melts could have produced the enrichment of the
mantle source region from which the some of the gabbros were
derived. These observations support the interpretation that the intru-
sions were related to active subduction during the Palaeoproterozoic.
Many subducted slabs experience multistage dehydration or melting
(Rollinson & Tarney, 2005), which can occur separately in space and
time (Elliott et al., 1997; Turner et al., 1997). Slab dehydration nor-
mally occurs at shallower depths than slab melting (Rollinson &
Tarney, 2005). Slab melting (including sediment and altered basalt) is
usually thought to be limited to young and relatively hot oceanic crust
and typically produces adakitic (Stern & Kilian, 1996) and high‐Mg
andesitic magmas (e.g., Tsuchiya, Suzuki, Kimura, & Kagami, 2005).
Thus, removal of mobile LILE from the slab by fluids is typically asso-
ciated with the early stages of subduction, whereas later melting pro-
duces adakites with high K/Rb ratios and low Th, K, and Rb contents
(Rollinson & Tarney, 2005; Stern & Kilian, 1996). Consequently, rocks
associated with slab melt‐modification are generally distributed
behind those derived from fluid enriched mantle sources (Rollinson
& Tarney, 2005). In the southern Ashanti belt, the close association
of the gabbros and the granitoids, as well as the pyroxenites, suggests
that both dehydration and slab melting occurred approximately simul-
taneous at the same crustal level. Therefore, we suggest that the sub-
duction zone was steep, and the mantle was relatively warm due to
convection at an active continental margin.5.4 | Pre‐Birimian (or Archaean?) crust in Ghana
As indicated earlier, the εNd (2.1 Ga) values of the plutonic rocks range
from −5.68 to +4.93, suggesting that the mantle source of the rocks
have been contaminated to various degrees by crustal components.
The Nd model ages (TDM2), 1.99–2.75 Ga, provide further evidence
for a pre‐Birimian crust in the study area. The juvenile character or
otherwise of the analysed rocks was visualized in the plot of εNd
(2.1 Ga) versus formation age (t; Figure 15). The Nd isotopic evolution-
ary trend of Archaean crust is also shown for comparison. The isotopic
values used are from Kouamelan, Peucat, and Delor (1997), who esti-
mated the composition of the Archaean rocks form Cote d'Ivoire
which forms part of the Archaean domain of the WAC, since no
Archaean crust is known in Ghana, at least for now. These Archaean
rocks have εNd (2.1 Ga) values ranging from −10 to −13. The
metaperidotites from the Aketakyi ultramafic complex, Aketakyi gab-
bro, Prince's Town quartz diorite (D9023A), and the Kegyina gabbro
(D90310) either fall in or close to the depleted mantle curve, which
suggest that they were juvenile at their time of formation
(Figure 15).
The Palaeoproterozoic Nd model ages (1.83–2.17 Ga; Table 5) of
these rocks support their juvenile character. The Prince's Town granit-
oids (D9025 and D9038), Axim gabbro and Kegyina gabbro and
Ahama pyroxenitic rocks are dominated by pre‐Birimian (and
Neoarchaean) Nd model ages of 2.41–2.51, 2.54–2.55, 2.39–2.57,
and 2.52–2.75 Ga, respectively (Table 5). Their respective εNd
(2.1 Ga) values of −1.01 to +0.16, −1.23 to −1.17, −0.94 to +0.46,
and −0.58 to −5.68 indicate variable contributions from pre‐Birimian
crustal materials, with the most significant contribution recorded in
the Ahama pyroxenitic rocks. Thus, the Ahama pyroxenites show a
clear evidence of older crustal/Archaean inputs, which is probably
32 DAMPARE ET AL.supported by their εNd (2.1 Ga) values being relatively closer to the
Archaean crust, compared to the other rocks. The occurrence of some
pre‐Birimian ages in detrital zircons (2,245–2,266 Ma) from the
Tarkwaian sediments (Davis et al., 1994; Loh & Hirdes, 1999) lends
support to the pre‐Birimian crustal growth episode in Ghana. These
data are not exceptional to Ghana, as the occurrence of some pre‐
Birimian ages have also been reported in two zircon grains (2,208
and 2,220 Ma) from the Tafolo tonalite in central Cote d'Ivoire
(Doumbia et al., 1998), in zircon cores from the Issia granite (2,212–
2,305 Ma) in south‐western Cote d'Ivoire (Kouamelan, 1996;
Kouamelan, Peucat, & Delor, 1997), and in zircon cores from the
Dabakala tonalite (2,312 ± 17 Ma) in central Cote d'Ivoire (Gasquet
et al., 2003).
The Nd isotopic data provide evidence for a possible contamina-
tion of the juvenile Birimian crust of the southern Ashanti belt by a
significant amount of a pre‐Birimian crustal material. Does it indicate
an Archaean crust beneath the southern Ashanti greenstone belt?
Contrary to other isotopic studies on the Birimian rocks of Ghana
(Taylor et al., 1992), which tend to preclude the involvement of nota-
ble contributions from an older crust in their genesis with the excep-
tion of the Winneba granitoids, the present data may implicate a
recycled Archaean crustal material component in the genesis of some
of the igneous rocks in the study area. The low initial 87Sr/86Sr ratios
tend to argue against an Archaean terrane at lower crustal levels.
Alternatively, the Archaean component might have resulted from sub-
duction of sediments from Archaean cratons or small Archaean ter-
ranes. In this scenario, the retained strontium isotopic signature of
the older crustal material could probably be similar to that of the plu-
tonic rocks. This latter interpretation, where the Palaeoproterozoic
mantle‐derived rocks were contaminated by crustal materials derived
from subduction, is favoured since the geochemistry of the rocks
(including HFSE depletion) does not support major crustal contamina-
tion during magmatic emplacement and crystallization.6 | CONCLUSIONS
The trace element signatures as well as the field relations of the stud-
ied granitoids and mafic–ultramafic rocks associated with
metavolcanic rocks in the southern Ashanti greenstone belt have led
to the conclusions presented below. The I‐type Prince's Town granit-
oids show geochemical features characteristic of subduction‐related
granitoids as found worldwide in intra‐oceanic arcs. The Ahama pyrox-
enites, Axim gabbro, and Kegyina gabbroic rocks are subduction‐
related ultramafic and mafic intrusions which occur in the Axim volca-
nic branch. These gabbroic and Ahama pyroxenitic rocks may have
been derived from lithospheric mantle sources, with the Ahama pyrox-
enites being extracted from a more fluid‐metasomatized mantle differ-
ent from that of Axim and Kegyina and Aketakyi gabbroic. These rocks
share similar characteristics with the LREE‐enriched intra‐oceanic‐
island arc‐fore‐arc basalts/basaltic andesites, which also occur in the
Axim volcanic lobe (see Dampare, Shibata, Asiedu, Osae, & Banoeng‐
Yakubo, 2008).The Aketakyi ultramafic complex demonstrates a supra‐subduction
affinity and is associated with the LREE‐depleted, arc‐like Aketakyi
gabbro and the LREE‐depleted, back‐arc‐basin basalts in the Cape
Three Points volcanic branch. The ultramafic complex probably origi-
nated from a mafic magma in the greenstone belt of which the tholei-
itic basalt is a possible candidate. Brown amphibole in the Aketakyi
ultramafic complex shows postcumulus textural characteristics (i.e.,
postcumulus phase enclosing olivine, pyroxene, and serpentinized oliv-
ine grains). The occurrence of primary magmatic amphibole suggests
the presence of relatively high water contents in the parental magma.
Thus, hydrous partial melting of a mantle source which had previously
been metasomatized is envisaged.
The analysed plutonic rocks commonly have low initial 87Sr/86Sr
ratios consistent with previous studies on Palaeoproterozoic rocks
from the West African Craton. The granitoids have εNd (2.1 Ga) values
ranging from −1.01 to +2.92 and TDM2 of 2.17–2.51 Ga, whereas the
gabbros have εNd (2.1 Ga) ranging from −1.23 to +5.23 and TDM2
values from 2.08 to 2.57 Ga. The Group I pyroxenites from the Ahama
ultramafic body show negative initial εNd (2.1 Ga) values from −0.58 to
−2.48 and ΤDM2 values of 2.52 to 2.75 Ga, whereas the Group II
pyroxenites show relatively higher negative εNd (2.1 Ga) values ranging
from −4.66 to −5.68 but no meaningful TDM2 ages. The
metaperidotites from the Aketakyi ultramafic/ophiolitic complex show
high positive εNd (2.1 Ga) values from +3.69 to +4.93, and ΤDM2 values
of 1.99 to 2.04 Ga, which suggest that they were juvenile at their time
of formation. Their high positive εNd (2.1 Ga) values suggest they
derived from depleted mantle magmas.
The Nd isotopic data provide evidence for a possible contamina-
tion of the juvenile Birimian crust of the southern Ashanti belt by
some amount of a pre‐Birimian crustal material. The low initial 87Sr/
86Sr ratios tend to argue against an Archaean terrane at lower crustal
levels. Alternatively, the Archaean component might have resulted
from subduction of sediments from Archaean cratons. This latter inter-
pretation, where the Palaeoproterozoic mantle‐derived rocks were
contaminated by crustal materials derived from subduction, is
favoured since the geochemistry of the rocks does not support major
crustal contamination during magmatic emplacement and
crystallization.
In conclusion, the ultramafic to mafic bodies and granitoid intru-
sives which are associated with metavolcanic rocks in the southern
Ashanti greenstone belts of Ghana were not plume‐generated but
were derived through supra‐subduction‐related magmatism during
the Palaeoproterozoic. This is consistent with the island arc complex
model in which the Palaeoproterozoic terranes of West Africa evolved
through subduction–accretion processes.ACKNOWLEDGEMENTS
Field work was sponsored by the National Nuclear Research Institute
(NNRI) of the Ghana Atomic Energy Commission (GAEC). The first
author gratefully acknowledges the financial support provided by the
Japanese Government (Monbukagakusho: MEXT) and Okayama Uni-
versity for laboratory work. The XRF analyses were performed at the
DAMPARE ET AL. 33Department of Earth Sciences, Okayama University, Japan, with the
assistance of Ms. Mayumi Usui. Critical comments by anonymous
reviewers are gratefully appreciated.
ORCID
Samuel B. Dampare https://orcid.org/0000-0002-7314-9581
Patrick A. Sakyi https://orcid.org/0000-0002-7536-6264
REFERENCES
Abouchami, W., Boher, M., Michard, A., & Albarede, F. (1990). A major 2.1
Ga event of mantle magmatism in West Africa: An early stage of crustal
accretion. Journal of Geophysical Research, 95, 17605–17629. https://
doi.org/10.1029/JB095iB11p17605
Albarède, F., & Brouxel, M. (1987). The Sm–Nd secular evolution of the
continental crust and the depleted mantle. Earth and Planetary Science
Letters, 82, 25–36. https://doi.org/10.1016/0012‐821X(87)90104‐X
Allibone, A., McCuaig, C. T., Harris, D., Etheridge, M., Munroe, S., Byrne, D.,
… Gyapong, W. (2002). Structural controls on gold mineralization at the
Ashanti Gold Deposit, Obuasi, Ghana. In R. J. Goldfarb, & R. L. Neilson
(Eds.), Integrated methods for discovery: Global exploration in the 21st
century (Vol. 9). Society of Economic Geologists Special Publication. (pp.
65–93). Littleton: Society of Economic Geologists, Inc. (SEG).
Altherr, R., Holl, A., Hegner, E., Langer, C., & Kreuzer, H. (2000). High
potassium, calc‐alkaline I‐type plutonism in the European Variscides:
northern Vosges (France) and northern Schwarzwald (Germany). Lithos
50, 51–73.
Ama Salah, I., Liégeois, J.‐P., & Pouclet, A. (1996). Evolution d'un arc
insulaire océanique birimian précoce au Liptako nigérien (Sirba):
géologie, géochronologie et géochimie. Journal of African Earth Sci-
ences, 22(3), 235–254. https://doi.org/10.1016/0899‐
5362(96)00016‐4
Asiedu, D. K., Dampare, S. B., Asamoah, S. P., Banoeng‐Yakubo, B., Osae, S.,
Nyarko, B. J. B., & Manu, J. (2004). Geochemistry of Paleoproterozoic
metasedimentary rocks from the Birim diamondiferous field, southern
Ghana: Implications for provenance and crustal evolution at the
Archean–Proterozoic boundary. Geochemical Journal, 38, 215–228.
https://doi.org/10.2343/geochemj.38.215
Atherton, M. P., & Petford, N. (1993). Generation of sodium‐rich magmas
from newly underplated basaltic crust. Nature, 362, 144–146. https://
doi.org/10.1038/362144a0
Attoh, K., & Ekwueme, B. (1997). The West African shield. In M. J. De Wit,
& L. D. Ashwall (Eds.), Greenstone belts (pp. 517–528). New York:
Oxford University Press.
Attoh, K., Evans, M. J., & Bickford, M. E. (2006). Geochemistry of an
ultramafic‐rodingite rock association in the Paleoproterozoic Dixcove
greenstone belt, southwestern Ghana. Journal of African Earth Sciences,
45, 333–346. https://doi.org/10.1016/j.jafrearsci.2006.03.010
Bach, W., Hegner, E., & Erzinger, J. (1998). Chemical fluxes in the Tonga
subduction zone: Evidence from the southern Lau Basin. Geophysical
Research Letters, 25, 1467–1470. https://doi.org/10.1029/98GL00840
Barnes, S.‐J., Naldrett, A. J., & Gorton, M. P. (1985). The origin of the frac-
tionation of platinum‐group elements in terrestrial magmas. Chemical
Geology, 53, 303–323.
Bartholomew, R. W. (1961). The geology of 1/4° field sheets 3 and 5 Axim
NE,1. Ghana Geological Survey Arch Rep, 18(1991), 78.
Bassot, J. P. (1987). Le complexe volcano‐plutonique calcoalcalin de la
rive're Daléma (Est Sénegal): Discussion de sa signification
géodynamique dans le cadre de l'orogénie eburnéenne (Protérozoicinférieur). Journal of African Earth Sciences, 6(4), 505–519. https://doi.
org/10.1016/0899‐5362(87)90091‐1
Batchelor, B., & Bowden, P. (1985). Petrogenetic interpretation of granitoid
rock series using multicationic parameters. Chemical Geology, 48,
43–55. https://doi.org/10.1016/0009‐2541(85)90034‐8
Ben Othman, D., Polvé, M., & Allègre, C. J. (1984). Nd–Sr isotopic compo-
sition of granulites and constraints on the evolution of the lower
continental crust. Nature, 307, 510–515. https://doi.org/10.1038/
307510a0
Bessoles, B. (1977). Geologie de l'Afrique: le craton ouest‐African. Memoire
BRGM, 88, 403.
Béziat, D., Bourges, F., Debat, P., Lompo, M., Martin, F., & Tollon, F. (2000).
A paleoproterozoic ultramafic‐mafic assemblage and associated volca-
nic rocks of the Boromo greenstone belt: Fractionates originating
from island‐arc volcanic activity in the West African craton. Precam-
brian Research, 101, 25–47. https://doi.org/10.1016/S0301‐
9268(99)00085‐6
Béziat, D., Dubois, M., Debat, P., Nikiema, S., Salvi, S., & Tollon, F. (2008).
Gold metallogeny in the Birimian craton of Burkina Faso (West Africa).
Journal of Earth Science, 50, 215–233.
Blenkinsop, T. G., Schmidt‐Mumm, A., Kumi, R., & Sangmor, S. (1994).
Structural geology of the Ashanti gold mine, Obuasi, Ghana. Geological
Jahrbuch, D100, 131–153.
Boher, M., Abouchami, W., Michard, A., Albarède, F., & Arndt, N. T. (1992).
Crustal growth in West Africa at 2.1 Ga. Journal of Geophysical
Research, 97, 345–369. https://doi.org/10.1029/91JB01640
Bonatti, E., & Michael, P. J. (1989). Mantle peridotites from continental rifts
to ocean basins to subduction zones. Earth and Planetary Science Let-
ters, 91, 297–311. https://doi.org/10.1016/0012‐821X(89)90005‐8
Castillo, R. P., Janney, P. E., & Solidum, R. S. (1999). Petrology and geo-
chemistry of Camiguia Island, Southern Phillippines: Insight to the
source of adakites and other lavas in a complex arc setting. Contribu-
tions Mineralogy and Petrology, 134, 33–51.
Class, C., Miller, D. M., Goldstein, S. L., & Langmuir, C. H. (2000).
Distinguishing melt and fluid subduction components in Umnak Volca-
nics, Aleutian Arc. Geochemistry, Geophysics, Geosystems, 1, 1–28.
Coleman, R. G. (1977). Ophiolites: Ancient oceanic lithosphere? (p. 229). New
York: Springer‐Verlag. https://doi.org/10.1007/978‐3‐642‐66673‐5
Dampare, S., Shibata, T., Asiedu, D., & Osae, H. (2005). Major‐element geo-
chemistry of Proterozoic Prince's Town granitoid from the southern
Ashanti volcanic belt, Ghana. Okayama University Earth Science Reports,
12, 15–30.
Dampare, S. B., Shibata, T., Asiedu, D. K., Okano, O., Manu, J., & Sakyi, P.
(2008). Sr–Nd isotopic compositions of Paleoproterozoic metavolcanic
rocks from the southern Ashanti volcanic belt, Ghana. Okayama Univer-
sity Earth Science Reports, 16, 9–28.
Dampare, S. B., Shibata, T., Asiedu, D. K., Osae, S., & Banoeng‐Yakubo, B.
(2008). Geochemistry of Paleoproterozoic metavolcanic rocks from
the southern Ashanti volcanic belt, Ghana: Petrogenetic and tectonic
setting implications. Precambrian Research, 162, 403–423.
Davis, D. W., Hirdes, W., Schaltegger, U., & Nunoo, E. A. (1994). U–Pb age
constraints on deposition and provenance of Birimian and gold‐bearing
Tarkwaian sediments in Ghana, West Africa. Precambrian Research, 67,
89–107. https://doi.org/10.1016/0301‐9268(94)90006‐X
Debat, P., Nikiema, S., Mercier, A., Lompo, M., Beziat, D., Bourges, F., …
Wenmenga, U. (2003). A new metamorphic constraint for the
Eburnean orogeny from Paleoproterozoic formations of the Man shield
(Aribinda and Tampelga countries, Burkina Faso). Precambrian Research,
123, 47–65. https://doi.org/10.1016/S0301‐9268(03)00046‐9
34 DAMPARE ET AL.Defant, M. J., & Drummond, M. S. (1990). Derivation of some modern arc
magmas by melting of young subducted lithosphere. Nature, 347,
662–665. https://doi.org/10.1038/347662a0
DePaolo, D. J. (1981). Neodymium isotopes in the Colorado Front Range
and crust‐mantle evolution in the Proterozoic. Nature (London), 291,
193–196. https://doi.org/10.1038/291193a0
DePaolo, D. J., Linn, A. M., & Schubert, G. (1991). The continental crust age
distribution: Methods of determining mantle separation ages from Sm–
Nd isotopic data and implication to the southwestern United States.
Journal of Geophysical Research, 96(B2), 2071–2088. https://doi.org/
10.1029/90JB02219
Doumbia, S., Pouclet, A., Kouamelan, A., Peucat, J. J., Vidal, M., & Delor, C.
(1998). Petrogenesis of juvenile‐type Birimian (Palaeoproterozoic)
granitoids in Central Côte d'Ivoire, West Africa: Geochemistry and
Geochronology. Precambrian Research, 87, 33–63.
Edwards, C. M. H., Menzies, M. A., Thirlwall, M. F., Morris, J. D., Leeman,
W. P., & Harmon, R. S. (1994). The transition to potassic alkaline volca-
nism in island arcs: The Ringgit‐Beser complex, east Java, Indonesia.
Journal of Petrology, 35, 1557–1595. https://doi.org/10.1093/petrol-
ogy/35.6.1557
Egal, E., Thiéblemont, D., Lahondère, D., Guerrot, C., Costea, C. A., Iliescu,
D., … Kolié, P. (2002). Late Eburnean granitization and tectonics along
the western and northwestern margin of the Archean Kénéma–Man
domain (Guinea, West African Craton). Precambrian Research, 117,
57–84. https://doi.org/10.1016/S0301‐9268(02)00060‐8
Eisenlohr, B. N. (1989). Conflicting evidence on the timing of mesothermal
and paleoplacer gold mineralisation in early Proterozoic rocks from
southwest Ghana, West Africa. Mineral Deposita, 27, 23–29.
Eisenlohr, B. N., & Hirdes, W. (1992). The structural development of the
early Proterozoic Birimian and Tarkwaian rocks of southwest Ghana,
West Africa. Journal of African Earth Sciences, 14, 313–325. https://
doi.org/10.1016/0899‐5362(92)90035‐B
Elliott, T., Plank, T., Zindler, A., White, W., & Bourdon, B. (1997). Element
transport from slab to volcanic front at the Mariana arc. Journal of Geo-
physical Research, 102, 14991–15019. https://doi.org/10.1029/
97JB00788
Feybesse, J. L., Billa, M., Guerrot, C., Duguey, E., Lescuyer, J. L., Jean‐Pierre
Milési, J. P., & Bouchot, V. (2006). The paleoproterozoic Ghanaian
province: Geodynamic model and ore controls, including regional stress
modeling. Precambrian Research, 149, 149–196. https://doi.org/
10.1016/j.precamres.2006.06.003
Feybesse, J. L., & Milési, J. P. (1994). The Archean/proterozoic contact
zone in West Africa: A mountain belt of décollement thrusting and
folding on a continental margin related to 2.1 Ga convergence of
Archean cratons? Precambrian Research, 69, 199–227. https://doi.org/
10.1016/0301‐9268(94)90087‐6
Gasquet, D., Barbey, P., Adou, M., & Paquette, J. L. (2003). Structure, Sr–
Nd isotope geochemistry and zircon U–Pb geochronology of the gran-
itoids of the Dabakala area (Côte d'Ivoire): Evidence for a 2.3 Ga crustal
growth event in the Paleoproterozoic of West Africa? Precambrian
Research, 127, 329–354. https://doi.org/10.1016/S0301‐
9268(03)00209‐2
Goldstein, S. L., O'Nions, R. K., & Hamilton, P. J. (1984). A Sm–Nd isotopic
study of atmospheric dusts and particulate from major river systems.
Earth and Planetary Science Letters, 70, 221–236. https://doi.org/
10.1016/0012‐821X(84)90007‐4
Grove, T. L., Parman, S. W., Bowring, S. A., Price, R. C., & Baker, M. B.
(2002). The role of an H2O‐rich fluid component in the generation of
primitive basaltic andesites and andesites form the Mt. Shasta region,
N. California. Contributions to Mineralogy and Petrology, 142,
375–396. https://doi.org/10.1007/s004100100299Harcouët, V., Guillou‐Frottier, L., Bonneville, A., Bouchot, V., & Milesi, J.‐P.
(2007). Geological and thermal conditions before the major
Palaeoproterozoic gold‐mineralization event at Ashanti, Ghana, as
inferred from improved thermal modeling. Precambrian Research, 154,
71–87. https://doi.org/10.1016/j.precamres.2006.11.014
Hart, S. R., & Dune, T. (1993). Experimental cpx/melt partitioning of 24
trace elements. Contributions to Mineralogy and Petrology, 113, 1–8.
https://doi.org/10.1007/BF00320827
Hawkesworth, C. J., Turner, S. P., McDermott, F., Peate, D. W., & van
Calsteren, P. (1997). U–Th isotopes in arc magmas: Implication for ele-
ment transfer from the subducted crust. Science, 276, 551–555.
https://doi.org/10.1126/science.276.5312.551
Hirdes, W., & Davis, D. W. (2002). U–Pb geochronology of
Paleoproterozoic rocks in the southern part of the Keougou‐Keneba
inlier, Senegal, West Africa: Evidence for diachronous accretionary
development of the Birimian Province. Precambrian Research, 118,
83–99.
Hirdes, W., Davis, D. W., & Eisenlohr, B. N. (1992). Reassessment of Prote-
rozoic of Proterozoic granitoid ages in Ghana on the basis of U/Pb
zircon and monzanite dating. Precambrian Research, 56, 89–96.
https://doi.org/10.1016/0301‐9268(92)90085‐3
Hirdes, W., Davis, D. W., Lüdtke, G., & Konan, G. (1996). Two generations
of Birimian (Paleoproterozoic) volcanic belts in northeastern Côte
d'Ivoire (West Africa): Consequences for the ‘Birimian controversy’.
Precambrian Research, 80, 173–191. https://doi.org/10.1016/S0301‐
9268(96)00011‐3
Hofmann, A. W. (2003). Sampling mantle heterogeneity through oceanic
basalts: Isotopes and trace elements. In R. W. Carlson (Ed.) The Mantle
and Core (Vol. 2). (pp. 61–101). Elsevier: New York.
Jenner, G. A., Cawood, P. A., Rautenschlein, M., & White, W. M. (1987).
Composition of backarc basin volcanics, Valu Fa Ridge, Lau Basin: Evi-
dence for a slab‐derived component in their mantle source. Journal of
Volcanology and Hydrothermal Research, 32, 209–222. https://doi.org/
10.1016/0377‐0273(87)90045‐X
John, T., Klemd, R., Hirdes, W., & Loh, G. (1999). The metamorphic evolu-
tion of the paleoproterozoic (Birimian) volcanic Ashanti belt (Ghana,
West Africa). Precambrian Research, 98, 11–30. https://doi.org/
10.1016/S0301‐9268(99)00024‐8
Johnson, K. T. M. (1998). Experimental determination of partition coeffi-
cients for rare earth and high‐field‐strength elements between
clinopyroxene, garnet, and basaltic melt at high pressure. Contributions
to Mineralogy and Petrology, 133, 60–68. https://doi.org/10.1007/
s004100050437
Junner, N. R. (1940). Geology of the Gold Coast and Western Togoland.
Gold Coast Geological Survey Bulletin, 11, 40.
Kay, R. W., & Mahlburg Kay, S. (1993). Delamination and delamination
magmatism. Tectonophysics, 219, 177–189. https://doi.org/10.1016/
0040‐1951(93)90295‐U
Kepezhinskas, P., McDermott, F., Defant, M., Hochstaedter, A., Drummond,
M. S., Hawdesworth, C. J., … Bellon, H. (1997). Trace element and Sr–
Nd–Pb isotopic constraints on a three‐component model of Kamchatka
Arc petrogenesis. Geochimica et Cosmochimica Acta, 61, 577–600.
https://doi.org/10.1016/S0016‐7037(96)00349‐3
Keppler, H. (1996). Constraints from partitioning experiments on the com-
position of subduction‐zone fluids. Nature, 380, 237–240. https://doi.
org/10.1038/380237a0
Keto, L. S., & Jacobsen, S. B. (1987). Nd and Sr isotopic variations of early
Paleozoic oceans. Earth and Planetary Science Letters, 84, 27–41.
https://doi.org/10.1016/0012‐821X(87)90173‐7
Kouamelan, A. N. (1996). Géochronologie et géochime des formations
archéennes et protérozoїques de la dorsale de Man en Côte d'Ivoire.
DAMPARE ET AL. 35Implications Pour la Transition Archéen‐Protérozoїque. Mém. Géosciences
Rennes, 73, 290.
Kouamelan, A. N., Delor, C., & Peucat, J.‐J. (1997). Geochronological evi-
dence for reworking of Archean terrains during the Early Proterozoic
(2.1 Ga) in the western Cote d'Ivoire (Man Rise‐West African Craton).
Precambrian Research, 86, 177–199. https://doi.org/10.1016/S0301‐
9268(97)00043‐0
Kouamelan, A. N., Peucat, J. J., & Delor, C. (1997). Reliques archéennes
(3,15 Ga) au sein du magmatisme birimien (2,1 Ga) de Côte d'Ivoire,
craton Ouest‐African. Comptes rendus de l'Académie des sciences. Série
2. Sciences de la terre et des planètes, 324(9), 719–727.
Ledru, P., Johan, V., Milési, J. P., & Tegyey, M. (1994). Markers of the last
stages of the Palaeoproterozoic collision: evidence for a 2 Ga continent
involving circum‐South Atlantic provinces. Precambrian Research, 69,
169–191. https://doi.org/10.1016/0301‐9268(94)90085‐X
Lemoine, S., 1988. Evolution géologique de la région de Dabakala (NE Côte
d'Ivoire) au Protérozoïque inférieur. Unpublished thesis, Univ.
Clermont‐Ferrand, 338 p.
Leube, A., Hirdes, W., Mauer, R., & Kesse, G. O. (1990). The early Protero-
zoic Birimian Supergroup of Ghana and some aspects of its associated
gold mineralization. Precambrian Research, 46, 139–165. https://doi.
org/10.1016/0301‐9268(90)90070‐7
Liégeois, J. P., Claessens, W., Camara, D., & Klerkx, J. (1991). Short‐lived
Eburnian orogeny in southern Mali. Geology, tectonics, U–Pb and
Rb–Sr geochronology. Precambrian Research, 50, 111–136. https://
doi.org/10.1016/0301‐9268(91)90050‐K
Loh, G., & Hirdes, W. (1999). Explanatory notes for the geological map of
southwest Ghana 1:100,000: sheets Sekondi (0402A) and Axim
(0403B): Ghana. Geological Survey Bulletin, 49, 149.
Lompo, M. (2009). Geodynamic evolution of the 2.25–2.0 Ga
Palaeoproterozoic magmatic rocks in the Man‐Leo Shield of the West
African Craton. A model of subsidence of an oceanic plateau. In S. M.
Reddy, R. Mazumder, D. A. D. Evans, & A. S. Collins (Eds.),
Palaeoproterozoic supercontinents and global evolution (Vol. 323). Geo-
logical Society of London Special Publication. (pp. 231–254). London:
Geological Society.
Maniar, P. D., & Piccoli, P. M. (1989). Tectonic discrimination of granitoids.
Geological Society of America Bulletin, 101, 635–643. https://doi.org/
10.1130/0016‐7606(1989)101<0635:TDOG>2.3.CO;2
Martin, H., Smithies, R. H., Rapp, R., Moyen, J.‐F., & Champion, D. (2005).
An overview of adakite, tonalite–trondhjemite–granodiorite (TTG), and
sanukitoid: relationships and some implications for crustal evolution.
Lithos, 79, 1–24. https://doi.org/10.1016/j.lithos.2004.04.048
Mauer, R. (1990). Petrographische und geochemische Unttersuchungen der
präkambrischen (Birimian) Granitoide Ghanas. Diss. Techn. Univ. Berlin,
202 pp.
McCulloch, M. T. (1987). Sm–Nd isotopic constraints on the evolution of
Precambrian crust in the Australian continent. In A. Kröner (Ed.), Prote-
rozoic lithospheric evolution (Vol. 17). Geodyn. Series. (pp. 115–130).
Washington, D.C.: AGU.
McDonough, W.F., Frey, F.A., (1989). Rare earth elements in upper mantle
rocks. In: Lipin, B., McKay, G. (Eds.), Geochemistry and mineralogy of rare
earth elements. Reviews in Mineralogy vol. 21, pp. (99–145). Washing-
ton: The Mineralogical Society of America.
Milési, J. P., Ledru, P., Feybesse, J. L., Dommanget, A., & Marcoux, E. (1992).
Early Proterozoic ore deposits and tectonics of the Birimian orogenic
belt, West Africa. Precambrian Research, 58, 305–344. https://doi.
org/10.1016/0301‐9268(92)90123‐6
Miyashiro, A. (1974). Volcanic rock series in island arcs and active conti-
nental margins. American Journal of Science, 274, 321–355. https://
doi.org/10.2475/ajs.274.4.321Mortimer, J. (1992). Lithostratigraphy of the early Proterozoic Toumodi
Volcanic Group in Central Côte d'Ivoire: Implications for Birimian stra-
tigraphy models. Journal of African Earth Sciences, 14, 81–91. https://
doi.org/10.1016/0899‐5362(92)90057‐J
Münker, C. (2000). The isotope and trace element budget of the Cambrian
Devil River Arc System, New Zealand: Identification of four source
components. Journal of Petrology, 41, 759–788. https://doi.org/
10.1093/petrology/41.6.759
Nelson, B. K., & DePaolo, D. J. (1985). Rapid production of continental
crust 1.7 to 1.9 b.y. ago: Nd isotopic evidence from the basement of
the North America mid continent. Geological Society of America Bulletin,
96, 746–754. https://doi.org/10.1130/0016‐7606(1985)96<746:
RPOCCT>2.0.CO;2
Nesbitt, H. W., & Young, G. M. (1982). Early Proterozoic climates and plate
motions inferred from major element chemistry of lutites. Nature, 299,
715–717.
Oberthür, T., Vetter, U., Davis, D. W., & Amanor, J. A. (1998). Age con-
straints on gold mineralization and Palaeoproterozoic crystal
evolution in the Ashanti belt of southern Ghana. Precambrian Research,
89, 129–143.
Opare‐Addo, E., John, B. E., Mukasa, S. B., & Browning, P. (1993). Field and
geochronologic (U–Pb) constraints on the age and generation of gran-
itoids and migmatites in southern Ghana. EOS Transactions of American
Geophysical Union, 74, S301.
Park, J. K., Buchan, K. L., & Harlan, S. S. (1995). A proposed giant radiating
dyke swarm fragmented by the separation of Laurentia and Australia
based on paleomagnetism of ca. 780 Ma mafic intrusions in western
North America. Earth and Planetary Science Letters, 132, 129–139.
Pawlig, S., Gueye, M., Klischies, R., Schwarz, S., Wemmer, K., &
Siegesmund, S. (2006). Geochemical and Sr–Nd isotopic data on the
Birimian of the Kedougou‐Kenieba Inlier (Eastern Senegal): Implications
on the Palaeoproterozoic evolution of the West African Craton. South
African Journal of Geology, 109, 411–427. https://doi.org/10.2113/
gssajg.109.3.411
Peacock, S. M., Rushmer, T., & Thompson, A. B. (1994). Partial melting of
subducting oceanic crust. Earth and Planetary Science Letters, 121,
227–244. https://doi.org/10.1016/0012‐821X(94)90042‐6
Pearce, J. A., Harris, N. B. W., & Tindle, A. G. (1984). Trace element dis-
crimination diagrams for the interpretation of granitic rocks. Journal
of Petrology, 25, 956–983. https://doi.org/10.1093/petrology/
25.4.956
Pearce, J. A., & Norry, M. J. (1979). Petrogenetic implications of Ti, Zr, Y,
and Nb, variation in volcanic rocks. Contributions to Mineralogy and
Petrology, 69, 33–47. https://doi.org/10.1007/BF00375192
Pearce, J. A., & Peate, D. W. (1995). Tectonic implications of the composi-
tion of volcanic arc magmas. Annual Review of Earth and Planetary
Sciences, 23, 251–285. https://doi.org/10.1146/annurev.
ea.23.050195.001343
Petford, N., Atherton, M. P., & Halliday, A. N. (1996). Rapid magma produc-
tion rates, underplating and remelting in the Andes: Isotopic evidence
from northern‐central Peru (90–110 s). Journal of South American Earth
Sciences, 9, 69–78. https://doi.org/10.1016/0895‐9811(96)00028‐4
Pigois, J.‐P., Groves, D. I., Fletcher, I. R., McNaughton, N. J., & Snee, L. W.
(2003). Age constraints on Tarkwaian palaeoplacer and lode‐gold for-
mation in the Tarkwa‐Damang district, SW Ghana. Mineralium
Deposita, 38, 695–714. https://doi.org/10.1007/s00126‐003‐0360‐5
Plank, T., & Langmuir, C. H. (1992). Sediments melt and basaltic crust dehy-
drates at subduction zones. Eos, Transactions, American Geophysical
Union, 73, 637.
Plank, T., & White, W. M. (1995). Nb and Ta in arc and mid‐ocean basalts.
Eos, 76, 655.
36 DAMPARE ET AL.Pohl, D., & Carlson, C. (1993). A plate tectonic re‐interpretation of the 2.2–
2.0 Ga Birimian province, Tarkwaian System and metallogenesis in
West Africa. In J. W. Peters, G. O. Kesse, & Acquah (Eds.), Regional
trends in African geology (pp. 378–381). Accra: Geological Society of
Africa.
Qu, X. M., Hou, Z. Q., & Li, Y. G. (2004). Melt components derived from a
subducted slab in late orogenic ore‐bearing porphyries in the Gangdese
copper belt, southernTibetau plateau. Lithos, 74, 131–148. https://doi.
org/10.1016/j.lithos.2004.01.003
Roeder, P. L., & Elmslie, R. F. (1970). Olivine‐liquid equilibrium. Contribu-
tions to mineralogy and petrology, 29, 275–289. https://doi.org/
10.1007/BF00371276
Roddaz, M., Debat, P., & Nikiéma, S. (2007). Geochemistry of Upper
Birimian sediments (major and trace elements and Nd—Sr isotopes)
and implications for weathering and tectonic setting of the Late
Paleoproterozoic crust. Precambrian Research, 159, 197–211.
Rollinson, H. R., & Tarney, J. (2005). Adakites—The key to understanding
LILE depletion in granulites. Lithos, 79, 61–81. https://doi.org/
10.1016/j.lithos.2004.04.050
Stern, C. R., & Kilian, R. (1996). Role of the subducted slab, mantle wedge
and continental crust in the generation of adakites from the Austral
Volcanic Zone. Contributions to Mineralogy and Petrology, 123,
263–281. https://doi.org/10.1007/s004100050155
Stolper, E., & Newman, S. (1994). The role of water in the petrogenesis of
Mariana trough magmas. Earth and Planetary Science Letters, 121,
293–325. https://doi.org/10.1016/0012‐821X(94)90074‐4
Sun, S. S., & McDonough, W. F. (1989). Chemical and isotopic systematics
of oceanic basalts: Implication for mantle composition and processes.
In A. D. Saunders, & M. J. Norry (Eds.), Magmatism in ocean basins
(Vol. 42). Geological Society of London Special Publication. (pp.
313–345). London: Geological Society.
Sylvester, P. J., & Attoh, K. (1992). Lithostratigraphy and composition of
2.1 Ga greenstone belts of the West African craton and their bearing
on crustal evolution and Archean‐Proterozoic boundary. Journal of
Geology, 100, 377–393. https://doi.org/10.1086/629593
Tagini, B. (1971). Esquisse structurale de la Cote d'Ivoire. Essai de
géotectonique régionale. These Univ. Lausanne. Soc. Dev. Min. Cote
d'Ivoire (SODEMI), 302 pp.
Taylor, N. P., Moorbath, S., Leube, A., & Hirdes, W. (1992). Early Protero-
zoic crustal evolution in the Birimian of Ghana: Constraints from
geochronology and isotope geochemistry. Precambrian Research, 56,
97–111. https://doi.org/10.1016/0301‐9268(92)90086‐4
Taylor, P.N., Moorbath, S., Leube, A., & Hirdes, W. (1988). Geochronology
and crustal evolution of Early Proterozoic granite‐greenstone terrains
in Ghana/West Africa. Abstract, International Conference on the geol-
ogy of Ghana with special emphasis on Gold. 75th Anniversary of
Ghana Geol. Surv. Dept., Accra, pp. 43–45.
Taylor, S. R., & McLennan, S. M. (1985). The continental crust: Its composi-
tion and evolution (p. 312). Oxford: Blackwell Scientific.
Tsuchiya, N., Suzuki, S., Kimura, J. I., & Kagami, H. (2005). Evidence for slab
melt/mantle reaction: petrogenesis of Early Cretaceous and Eocene
high‐Mg andesites from the Kitakami Mountains, Japan. Lithos, 79,
179–206. https://doi.org/10.1016/j.lithos.2004.04.053
Tunks, A. J., Selley, D., Rogers, J. R., & Brabham, G. (2004). Vein minerali-
zation at the Damang Gold Mine, Ghana: Controls on mineralization.
Journal of Structural Geology, 26, 1257–1273. https://doi.org/
10.1016/j.jsg.2003.11.005Turner, S., Hawkesworth, C., Rogers, N., Barlett, J., Worthington, T., Hergt,
J., … Smith, I. (1997). 238U–230Th disequilibria magma petrogenesis, and
flux rates beneath the depleted Tonga–Kermadec island arc.
Geochimica et Cosmochimica Acta, 61, 4855–4884. https://doi.org/
10.1016/S0016‐7037(97)00281‐0
Vidal, M., & Alric, G. (1994). The Palaeoproterozoic (Birimian) of Haute‐
Comoé in the West African craton, Ivory Coast: A transtensional
back‐arc basin. Precambrian Research, 65, 207–229.
Wang, Q., Xu, J. F., Jian, P., Bao, Z. W., Zhao, Z. H., Li, C. F., … Ma, J. L.
(2006). Petrogenesis of adakitic porphyries in an extensional tectonic
setting, Dexing, south China: Implications for the genesis of porphyry
copper mineralization. Journal of Petrology, 47, 119–144. https://doi.
org/10.1093/petrology/egi070
Weaver, B. L. (1990). Geochemistry of highly‐undersaturated ocean island
basalt suites from the South Atlantic Ocean: Fernando de Noronha and
Trinidade islands. Contributions to Mineralogy and Petrology, 105,
502–515. https://doi.org/10.1007/BF00302491
Wedepohl, K. H. (1995). The composition of the continental crust.
Geochimica et Cosmochimica Acta, 59, 1217–1232.
Whalen, J. B., Syme, E. C., & Stern, R. A. (1999). Geochemical and Nd iso-
topic evolution of Paleoproterozoic arc‐type granitoid magmatism in
the Flin Flon Belt, Trans‐Hudson orogen, Canada. Canadian Journal of
Earth Sciences, 36, 227–250. https://doi.org/10.1139/e98‐026
Woodhead, J. D., Eggins, S., & Gamble, J. (1993). High field strength and
transition element systematics in island arc and back‐arc basin basalts:
Evidence for a multiphase melts extraction and a depleted mantle
wedge. Earth and Planetary Science Letters, 144, 491–504.
Wu, F., Zhao, G., Simon, A., Wilde, S. A., & Sun, D. (2005). Nd isotopic con-
straints on crustal formation in the North China Craton. Journal of Asian
Earth Sciences, 24, 523–545. https://doi.org/10.1016/j.
jseaes.2003.10.011
Yao, Y., Murphy, P. J., & Robb, L. J. (2001). Fluid characteristics of granit-
oid‐hosted gold deposits in the Birimian terrane of Ghana: A fluid
inclusion microthermometric and Raman spectroscopic study. Economic
Geology, 96, 1611–1643.
Zhao, J.‐H., & Zhou, M.‐F. (2007). Geochemistry of Neoproterozoic mafic
intrusions in the Panzhihua district (Sichuan Province, SW China):
Implications for subduction‐related metasomatism in the upper mantle.
Precambrian Research, 152, 27–47. https://doi.org/10.1016/j.
precamres.2006.09.002
Zhou, M.‐F., Kennedy, A. K., Sun, M., Malpas, J., & Lesher, C. M. (2002).
Neo‐proterozoic arc‐related mafic intrusions in the northern margin
of South China: implications for accretion of Rodinia. Journal of Geol-
ogy, 110, 611–618. https://doi.org/10.1086/341762
Zou, H. B., Zindler, A., Xu, X. S., & Qu, Q. (2000). Major, trace element, and
Nd, Sr and Pb isotope studies of Cenozoic basalts in SE China: Mantle
sources, regional variations, and tectonic significance. Chemical Geol-
ogy, 171, 33–47. https://doi.org/10.1016/S0009‐2541(00)00243‐6
How to cite this article: Dampare SB, Shibata T, Asiedu DK,
et al. Ultramafic–mafic and granitoids supra‐subduction
magmatism in the southern Ashanti volcanic belt, Ghana: Evi-
dence from geochemistry and Nd isotopes. Geological Journal.
2019;1–37. https://doi.org/10.1002/gj.3512
DAMPARE ET AL. 37APPENDIX A
ANALYTICAL TECHNIQUESMajor and trace element analyses
Samples were collected from mafic, ultramafic, and granitoid out-
crops associated with metavolcanic rocks in the Axim and Cape
Three Points volcanic branches of the southern Ashanti greenstone
belt. Analysis of major elements was carried out on fused discs by
the automated X‐ray fluorescence spectrometer, Philips model PW
1480, at the Department of Earth Sciences, Okayama University.
All samples were crushed and milled in an agate mortar as tungsten
carbide crushing/milling vessels are a well‐known source of Ta con-
tamination. The procedure for preparation of glass beads for X‐ray
fluorescence (XRF) analysis, analytical techniques, and quality control
on the XRF analyses are described elsewhere (Dampare, Shibata,
Asiedu, Osae, & Banoeng‐Yakubo, 2008). The precision was gener-
ally better than 1%. Loss on ignition (LOI) was computed from
weight loss by heating an oven‐dried (110°C) whole‐rock powder
to 1000°C for 2 hr.
The trace elements analyses selected samples were performed at
the Activation Laboratories Ltd. (Actlabs), Ontario, Canada, using
fusion inductively coupled plasma and inductively coupled plasma‐
mass spectrometry (ICP‐MS). In all, 26 samples collected from the
granitoid intrusion and various mafic and/or ultramafic bodies in
the study area were selected, taking into consideration the level of
alteration, as indicated by their LOI values. Details of the analyticalprocedure (i.e., accuracy, precision, and standards) can be obtained
from the Activation Laboratories Ltd. The detection limits (in ppm)
for the measured trace elements are 5 (V, As, Pb), 20 (Cr, Ni), 1
(Co, Ga, Rb, Zr, Sn), 10 (Cu), 30 (Zn), 0.5 (Ge, Y, W), 2 (Sr, Mo),
0.2 (Nb, Sb), 0.1 (In, Cs, Yb, Hf, Bi), 3 (Ba), 0.05 (La, Ce, Nd, Tl,
Th), 0.01 (Pr, Sm, Gd, Tb, Dy, Ho, Er, Ta, U), 0.005 (Eu, Tm), and
0.002 (Lu).Sr–Nd isotopic analyses
Sr–Nd isotope analyses were conducted at the Center of Instrumental
Analysis, Okayama University, Japan. Approximately 70 mg of whole‐
rock powdered samples were dissolved in a mixture of purified
HNO3–HF–HClO4 in Teflon vials on a hot plate (100–140°C). Separa-
tion of Sr and Nd was carried out using a standard two‐column ion‐
exchange technique. Isotopic analyses were carried out at the Center
of Instrumental Analysis using a five‐collector Finnigan MAT‐262 mass
spectrometer. Sr and Nd were loaded in 2% H3PO4 (repeatedly puri-
fied using a cation exchange column) on double Ta–Re filaments and
analysed in the static mode. For mass fractionation corrections, the
87Sr/86Sr and 143Nd/144Nd isotopic data were normalized against
the values of 86Sr/88Sr = 0.1194 and 146Nd/144Nd = 0.7219, respec-
tively. During the period of data acquisition, the mean of the mea-
sured 87Sr/86Sr ratios of NBS 987 standard gave 0.710295 ± 11
(2σm; n = 13 analyses). The mean of the measured 143Nd/144Nd ratios
of JNdi‐1 standard gave 0.5121428 ± 9 (2σm; n = 5 analyses), which
corresponds to a La Jolla Nd standard value of 0.511885.