Gondwana Research 109 (2022) 183–204
Contents lists available at ScienceDirect

Gondwana Research

journal homepage: www.elsevier .com/locate /gr
Provenance and depositional setting of the Buem structural unit
(Ghana): Implications for the paleogeographic reconstruction of the
West African and Amazonian cratons in Rodinia
https://doi.org/10.1016/j.gr.2022.04.020
1342-937X/� 2022 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.

⇑ Corresponding author.
E-mail address: dkwayisi@ug.edu.gh (D. Kwayisi).
Daniel Kwayisi a,b,⇑, Jeremie Lehmann b, Marlina Elburg b

aDepartment of Earth Science, University of Ghana, Main Campus, Legon, Ghana
bDepartment of Geology, University of Johannesburg, Auckland Park Kingsway Campus, Johannesburg, South Africa

a r t i c l e i n f o a b s t r a c t
Article history:
Received 9 August 2021
Revised 11 March 2022
Accepted 28 April 2022
Available online 4 May 2022
Handling Editor: J.G. Meert

Keywords:
Detrital zircon geochronology
Gondwana
Rodinia
West African Craton
Amazonian Craton
Passive margin
We present new field, petrological and geochemical data, combined with U-Pb zircon ages and Lu-Hf
isotope compositions for the sandstones of the Neoproterozoic Buem structural unit (BSU) of the
Dahomeyide belt, and whole-rock geochemical data of BSU shale to investigate their provenance and
depositional setting. The BSU contains siliciclastic sequences and fragments of oceanic lithosphere
(pillow lavas, gabbro, and peridotite) that have archived the entire evolution of the Ediacaran West
Gondwana Orogen (WGO), from early accretion to final amalgamation of the West African Craton
(WAC) with the Benino-Nigerian Shield. Three broad groups of samples exist within the BSU in Ghana:
those with dominantly older age fractions of 2300 – 1800 Ma, represented by samples from the base
of the BSU; those with prominent 1700 – 1100 Ma zircons, occupying the middle part of the BSU; and
those with significant 1000 – 970 Ma age fractions, forming the uppermost part of the BSU. Results from
this study, together with published data on the BSU, and adjacent Togo structural unit and Voltaian
Supergroup, reveal two main sedimentary sequences in the Dahomeyide belt, i.e., passive margin and
foreland basin sequences with three potential provenances: Amazonian Craton, Benino-Nigerian
Shield, ± WAC. This interpretation resembles that for the sedimentary rocks of the Borborema
Province, NE Brazil, which implies similar evolution along the WGO. Thus, a long, >2500 km passive mar-
gin basin developed at 1000 – 700 Ma, which was subsequently inverted during the Braziliano/Pan-
African plate convergence and collision, and the concomitant formation of the foreland basin during
the assembly of the supercontinent Gondwana. Because a greater portion of the detritus in the BSU is
probably from the Amazonian Craton, we propose that the WAC and the Amazonian Craton adjoined each
other from the Paleoproterozoic onward within the supercontinents Rodinia and Gondwana, until the
opening of the Atlantic Ocean.
� 2022 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.
1. Introduction

The Ediacaran West Gondwana Orogen (WGO) resulted in the
amalgamation of four cratonic nuclei, i.e., the West African and
Amazonian cratons to the west, and São Francisco Craton, and
Sahara Metacraton to the east, during the Brasiliano/Pan-African
orogeny (Fig. 1 inset; Caby, 2003; Cordani et al., 2013; Ganade de
Araujo et al., 2014a). The WGO developed fromWest Africa to Cen-
tral Brazil over >4000 km and records a long-lived (780 to 640 Ma,
U-Pb ages of arc-related granitoids) subduction system since the
early Neoproterozoic (Caby, 2003; Cordani et al., 2013; Ganade
de Araujo et al., 2014a, 2016). The WGO accretionary orogenic evo-
lution was terminated at � 620 – 610 Ma when several newly
formed sections of oceanic crust, intra-oceanic and continental arcs
were caught up in the collision between the Amazonian and West
African cratons on the one side, against the São Francisco Craton
and Saharan Metacraton on the other during the Brasiliano/
Pan-African orogeny (Caby, 2003; Cordani et al., 2013; Ganade de
Araujo et al., 2014a, Guillot et al., 2019). Four segments (Dahome-
yide, Pharusian and Brazilia belts and Borborema Province) form
the WGO (Fig. 1; Ganade de Araujo et al., 2016). The Buem struc-
tural unit (BSU) of the Dahomeyide belt is a key crustal component
for unravelling the evolution of the WGO system from early pas-
sive margin formation to final amalgamation (Fig. 1a). This is
because the BSU is situated in the lower plate, and contains silici-
clastic sequences and fragments of oceanic lithosphere (Fig. 2 and
Fig. 3) that archive the entire evolution of the WGO, from early

http://crossmark.crossref.org/dialog/?doi=10.1016/j.gr.2022.04.020&domain=pdf
https://doi.org/10.1016/j.gr.2022.04.020
mailto:dkwayisi@ug.edu.gh
https://doi.org/10.1016/j.gr.2022.04.020
http://www.sciencedirect.com/science/journal/1342937X
http://www.elsevier.com/locate/gr


Fig. 1. (a) Geological map of West Africa (after Guillot et al., 2019) and (b) Geological map of South America (after Cordani and Teixeira, 2007) arranged in Gondwana position
(before the opening of the Atlantic Ocean; Gray et al., 2008). The inset is the map of West Gondwana Orogen (). adapted from Ganade de Araujo et al., 2016

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
accretion to final amalgamation of the West African Craton (WAC)
with the Benino-Nigerian Shield (i.e., the leading western edge of
the Saharan Metacraton; Fig. 1a; e.g. Affaton et al., 1991; Attoh,
1998; Trompette, 2000).

Based on petrographic and geochemical studies of the sand-
stones from the BSU (upper and lower parts of the stratigraphy),
Osae et al. (2006) proposed that the sandstones were deposited
in a passive margin setting and received detritus probably from
the WAC. In contrast, Kalsbeek et al. (2008) suggested that a signif-
icant amount of the detritus for the sandstones of the BSU was
sourced from the Amazonian Craton, based on the abundance of
U-Pb detrital zircon age fractions between 1700 and 1000 Ma from
the upper and lower parts of the BSU, which are absent on the
WAC, but common in the Amazonian Craton. Ganade de Araujo
et al. (2016) inferred that the uppermost part of the BSU formed
as foreland deposits with detritus originating from the BNS based
on the significant proportion of detrital zircon U-Pb ages of 900 –
600 Ma, an age group that is absent as a proto-source on the
WAC. These studies suggest that the tectonostratigraphic evolution
of the BSU possibly records temporal variations in the provenance
of the sediments (WAC, Amazonian Craton, Benini-Nigerian
Shield), and competition between different sources at the same
time (WAC vs Amazonian Craton). They also highlight the need
for a detailed geochemical and provenance investigation of the
entire BSU stratigraphy for constraining (i) the crustal evolution
of the proto-source(s), (ii) the possible connection with the Amazo-
nian Craton, (iii) timing of tectonic inversion from rifting to con-
vergence, and (iv) the overall geodynamic evolution of the
external zone of the Dahomeyide belt.
184
To this end, this study presents, in addition to detrital U-Pb ages
for the entire BSU, the first Lu-Hf isotope zircon compositions of
the sandstone, and whole-rock geochemical data of shale from
the BSU. The results of this study show that the BSU sedimentary
rocks were deposited in a passive margin basin, with detritus accu-
mulated from both the WAC and Amazonian Craton, with a greater
proportion from the latter. The results provide constraints for the
discussion on the assembly of West Gondwana during the
Brasiliano/Pan-African orogeny and reconstruction of the pre-
Gondwana position of the WAC with respect to the Amazonian
Craton in the supercontinent Rodinia.
2. Geological setting

2.1. Regional Geological background

The West African Craton (WAC) is composed of granitoids and
greenstone belts with associated sedimentary basins (Fig. 1a;
Abouchami et al., 1990; Kouamelan et al., 1997; Baratoux et al.,
2011). The WAC formed during three main tectonomagmatic and
metamorphic events at � 3200 – 3000 Ma (Leonian events), 2900
– 2700 Ma (Liberian events), and� 2250 – 2060Ma (Eburnean oro-
geny, e.g. Baratoux et al., 2011; Tshibubudze et al., 2013; Sakyi
et al., 2014; Anum et al., 2015; Kouamelan et al., 2015; Block
et al., 2016). The WAC was stabilized during the Eburnean orogeny
and was not affected by any other major orogenic events until its
incorporation into West Gondwana during the Neoproterozoic
Pan-African orogeny (Villeneuve and Cornée, 1994; Deynoux



Fig. 2. Geological map of Ghana, Togo and Benin illustrating the context of the Volta Basin, and the Dahomeyide belt (modified after Ganade de Araujo et al., 2016).

Fig. 3. West-east cross-section of the Dahomeyide belt showing the relationship between the Buem and Togo structural units, and Voltaian Supergroup (modified after
Guillot et al., 2019).

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
et al., 2006; Ennih and Liégeois, 2008). The WAC is covered by the
late Meso- and Neoproterozoic sedimentary rocks of the Taoudeni
basin (Rooney et al., 2010; Kah et al., 2012) and transected by sev-
eral generations of dolerite dyke swarms and associated sills with
ages ranging from � 1700 to 200 Ma (Jessel et al., 2015; Baratoux
et al., 2019).

The Amazonian Craton (Fig. 1b) consist of four main Archean
blocks (Imataca, Amapá, Carajás, and Xingu-Iricoumé blocks) dated
185
at 3300 – 2600 Ma, which were stabilized around 2100 Ma during
the Transamazonian orogeny (2180 – 1950 Ma) (Ledru et al., 1994;
Tassinari et al., 2001; Lofan et al., 2003; Cordani et al., 2009; Neto
and Lofan, 2019). The Transamazonian orogeny resulted in the for-
mation of greenstones and TTGs (2190 – 2130 Ma), high-grade
metamorphic (2070 – 2030 Ma) and felsic volcanic
(2000 – 1800 Ma) belts of the Maroni-Itacaiúnas and
Ventuari-Tapajós provinces (Fig. 1b; Delor et al., 2003;



D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
Rosa-Costa et al., 2006; Cordani and Teixeira, 2007). Other Protero-
zoic orogenic belts found within the Amazonian Craton include the
1780–1550MaRioNegro-Juruena, 1500–1300Mabelt Rondonian-
San Ignacio, and 1250 – 900Ma Sunsan-Aguapeí orogenic belts (e.g.,
Santos, 2003; Tassinari et al., 2004; Boger et al., 2005; Cordani and
Teixeira, 2007). Theeasternmarginof theAmazonianCraton is occu-
pied by the 700 – 500 Ma Brasiliano orogenic belt (Borborema Pro-
vince), which forms the southern section of the WGO, marking the
collision between the Amazonian and São Francisco cratons
(Cordani and Teixeira, 2007; Ganade de Araujo et al., 2014a, b).

The WAC and Amazonian Craton have been proposed to be con-
nected from 2.0 Ga until their breakup in the Cretaceous (Rogers,
1996; Trompette, 1994; Rogers and Santosh, 2002). Nonetheless,
the connection of the WAC with the Amazonian Craton in the
Mesoproterozoic and early Neoproterozoic is poorly constrained
(Tohver et al., 2006). From paleomagnetic data, the Amazonian
Craton is positioned close to the WAC in almost all the reconstruc-
tions of the supercontinent Rodinia (1200 – 1000 Ma) (e.g., Evans,
2009, 2013). An Amazonian-WAC connection during the Mesopro-
terozoic and early Neoproterozoic is also supported by Baratoux
et al. (2019) and Antonio et al. (2021) using U-Pb baddeleyite ages
and paleomagnetic data from dolerite dyke swarms respectively.
The WAC-Amazonian connection is only supported by paleomag-
netic data and correlated dyke swarms. Thus, a detailed sedimen-
tary/stratigraphic study to test this proposed connection is
necessary.

The Dahomeyide belt, which forms the central section of the
WGO, occupies the southeastern margin of the WAC (Fig. 1a and
2; e.g., Attoh et al., 1997; Attoh and Nude, 2008; Ganade de
Araujo et al., 2014a). This belt extends from the southeastern coast
of Ghana to Nigeria through Togo and Benin and is about 1000 km
long (Affaton et al., 1991; Attoh et al., 1997; Agbossoumondé et al.,
2001). The Dahomeyide belt formed as a result of the closure of the
Pharusian Ocean during convergence, leading to continent–conti-
nent collision between the WAC and Benino-Nigerian Shield
(Affaton et al., 1991; Attoh et al., 1997; Agbossoumondé et al.,
2001; Cordani et al., 2003; Duclaux et al., 2006). The Dahomeyide
belt comprises three main zones. The external zone to the west is
in a lower plate position and consists of inverted passive margin
sequences of the Buem and Togo structural units, with fragments
of oceanic lithosphere in the former (Fig. 2 and Fig. 3; Ghana
National Geological Mapping Project, 2009; Kwayisi et al., 2020).
The basement of the external zone is interpreted to be underthrust
WAC Eburnean crust, based on geophysical and structural interpre-
tations (Kwayisi et al., 2020), and in agreement with Paleoprotero-
zoic gneisses occurring in tectonic windows within the passive
margin sequences (the Ho-gneisses e.g., Agyei et al., 1987; Attoh
and Nude, 2008; Aidoo et al., 2014). The internal zone to the east
is composed of Paleoproterozoic granitoids and gneisses (2190 –
2140 Ma) of the Benino-Nigerian Shield that were intruded by
magmatic arc plutons at 670 – 610 Ma and post-collisional plutons
at 580 – 540 Ma (Kalsbeek et al., 2012; Attoh et al., 2013; Ganade
de Araujo et al., 2016). Arc magmatism may be as old as 780 Ma,
which is the age of detrital zircon grains from syn-orogenic migma-
tite (Ganade de Araujo et al., 2016). The external and internal zones
are separated by a well-defined suture zone of high-pressure (HP:
eclogite and granulite) metamorphic rocks, with protolith ages
at � 800 Ma, and peak metamorphism at 610 ± 5 Ma (Attoh
et al., 1991; Affaton et al., 2000; Berger et al., 2011). Exhumation
of the HP rocks at 600 – 570 Ma marks the end of collision
(Attoh et al., 1997, 2007).

2.2. Geology of the study area

Correlation between the Buem and Togo structural units and
the Voltaian Supergroup is a matter of ongoing debate. Although
186
the contacts between these three units are tectonic (Figs. 2 and
3), there is a general increase in metamorphic conditions from
west to east, from unmetamorphosed in the Voltaian Supergroup,
prehnite-pumpellyite to greenschist facies in the BSU (Affaton
et al., 1997; Nude et al., 2015; Kwayisi et al., 2020) and greenschist
to lower amphibolite facies in the Togo structural unit (Adjei and
Tetteh, 1997; Attoh et al., 1997). This is somewhat correlated with
an eastward increase in deformation intensity, as the generally
flat-lying Voltaian Supergroup shows steep dips at its eastern mar-
gin (Grant, 1969).

Fig. 4 summarises the various correlations and classification
schemes proposed in the literature. Pioneering work considered
the Buem and Togo structural units to be older than the Voltaian
Supergroup, because of their higher degree of deformation and
metamorphism (Fig. 4a; Junner and Hirst, 1946). However, the
recognition of the characteristic ‘‘Triad” of the Kodjari-Buipe Sub-
group of the Oti Group in both the Buem and Togo structural units
led Affaton (1990), Kalsbeek et al. (2008), and Anani et al. (2019) to
infer that the Buem and the Togo structural units are the lateral
deformed and metamorphosed equivalents of the Voltaian Super-
group (Table 1; Fig. 4g). Based on the similarity in detrital zircon
age distributions between the Voltaian Supergroup, the Togo and
Buem structural units, Kalsbeek et al. (2008) and Ganade de
Araujo et al. (2016) proposed that the three units were deposited
in similar passive margin and foreland basins (Table 1). The maxi-
mum depositional age for the BSU rocks with passive margin char-
acteristics is � 950 Ma, and these rocks correlate with the passive
margin rocks of the Bombouaka Group of the Voltaian Supergroup
and quartzite of the TSU (Kalsbeek et al., 2008; Ganade de Araujo
et al., 2016). The maximum depositional age for the uppermost
sedimentary rocks of BSU is � 600 Ma, suggesting derivation from
the BNS, and correlation with the Oti and Obosum groups of the
Voltaian Supergroup and the Kanti schist of the TSU (Ganade de
Araujo et al., 2016).

The 5 – 7 km thick Voltaian Supergroup is considered analogous
to the sedimentary rocks of the Taoudeni basin (Fig. 1a; Affaton
et al., 1980; Villeneuve and Cornée, 1994; Anani, 1999). Weak
deformation, lithofacies, extensive borehole coverage, airborne
radiometric and geochemical data allowed for the stratigraphic
division of the Voltaian Supergroup in three unconformable units,
from bottom to top: Bombouaka, Oti, and Obosum groups (Table 1
and Fig. 2; Deynoux et al., 2006; Carney et al., 2010; Kalsbeek and
Frei, 2010).

The Bombouaka Group comprises two sandstone-dominated
units, which are separated by a middle mudstone unit. The rocks
of the Bombouaka Group are interpreted as passive margin units
of the Pharusian Ocean from lithofacies and geochemical studies
(Kalsbeek et al., 2008; Anani et al., 2017). The WAC was inferred
to be the source of the Bombouaka Group sediments (Anani,
1999; Anani et al., 2013, based on sandstone petrology and geo-
chemical data), but Kalsbeek et al. (2008) and Anani et al. (2017)
have inferred a significant contribution of sediments from the
Amazonian Craton, based on geochemical data and U-Pb detrital
zircon age fractions between 1700 and 1000 Ma, which are missing
on the WAC but are abundant on the Amazonian Craton.

The Oti Group lies, with a glacial unconformity, on the Bom-
bouaka Group, or locally, directly on the WAC basement rocks.
The Oti Group comprises the Kodjari-Buipe and the Afram-
Bimbila subgroups (Carney et al., 2010). The Kodjari-Buipe Sub-
group corresponds to a ‘‘Triad” of glaciogenic deposits, baryte-
bearing dolomitic limestone, and thinly bedded siliceous shale
(Affaton, 1990;). The Afram-Bimbila Subgroup is composed of
a rhythmic alternation of shale and siltstone, with various sand-
stone or greywacke interlayers, and lenses of clayey limestone.
The depositional setting of the Kodjari-Buipe Subgroup is con-
sidered to be a passive margin of the Pharusian Ocean, based



Fig. 4. The subdivisions and correlations of the Voltaian Supergroup, the Buem structural unit (BSU) and the Togo structural unit (TSU) proposed by (a) Junner and Hirst
(1946), (b) Black (1967), (c) Grant (1969), (d) Bozhko (1969), (e) Saunders (1970), (f) Blay (1983), and (g) Affaton (1990).

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
on lithofacies and geochemical composition (Affaton, 1990;
Carney et al., 2010; Amedjoe et al., 2018). The depositional set-
ting of the Afram-Bimbila Subgroup is interpreted to be a
peripheral (lies on the lower plate) foreland basin, based on
the asymmetrical, wedge-like geometry that are typical of fore-
land basins (Jordan, 1981) and a significant proportion of detri-
tal zircons in the age range between 900 and 600 Ma (Carney
et al., 2010; Ganade de Araujo et al., 2016). The Afram-
Bimbila Subgroup has been interpreted as flysch-derived
(argillaceous sediments with rhythmic repetition of shale and
siltstone) deposits from the Dahomeyide belt (Carney et al.,
2010).

The Obosum Group is interpreted as a foreland basin molasse-
type deposit, deposited at the late stage of the Pan-African conti-
nental collision (Carney et al., 2010; Kalsbeek and Frei, 2010).
The Obosum Group is subdivided into the Yendi and Kebia sub-
groups (Table 1). The Yendi Subgroup includes a lower polymictic,
micaceous and mainly sandy interval, overlain by a thick sequence
of shale, impure limestone, and siltstone. The Kebia Subgroup is
composed of various polymictic sandstone and conglomerate units,
with lenses of shale and siltstone (Carney et al., 2010).

The Voltaian Supergroup was therefore deposited in two dis-
tinct depositional settings at different times. The Bombouaka
Group and the Kodjari-Buipe Subgroup were deposited in passive
margins at 959 ± 62 Ma (Rb-Sr isochrons on clay minerals) and
ca. 635 Ma respectively (Table 1: Clauer, 1976; Carney et al.,
2010). The depositional age of the Afram-Bimbila Subgroup is
576 ± 13 Ma (Lu-Hf dating on phosphorite; Barfod et al., 2004);
however, the age of the Obosum Group is unknown, both deposited
in a peripheral foreland basin.
187
The Neoproterozoic Togo structural unit (Fig. 2) consists of
monocyclic arenaceous and argillaceous sedimentary rocks, which
have been metamorphosed into quartzite, phyllite and schist
(Junner and Hirst, 1946; Grant, 1969; Adjei and Tetteh, 1997).
Anani et al. (2019) indicated that the geochemistry of the phyllite
of the Togo structural unit signifies derivation from a passive con-
tinental margin. Detrital zircon age data suggest that the quartzite
and schist (referred to as the Kanti schist) of the Togo structural
unit were deposited in a passive margin and foreland basin respec-
tively, with sediments sourced probably from the WAC-Amazonian
Craton, and BNS respectively (Table 1; Kalsbeek et al., 2008;
Ganade de Araujo et al., 2016).

The BSU, a fold and thrust belt, comprises dominantly shale and
sandstone, and subordinate volcanic and mafic–ultramafic plu-
tonic rocks, chert, carbonate, and ironstone (Fig. 5a; Ghana
National Geological Mapping project, 2009; Kwayisi et al., 2020).
Different workers have proposed different stratigraphic divisions
for the BSU, because of intense folding and duplication by thrusting
that obscure the original lithostratigraphic architecture (e.g., Sup-
plementary Table S1 in Kwayisi et al., 2020). The BSU is deformed
with generally steeply east-dipping S2 foliation that is axial planar
to near-isoclinal F2 folds and local top-to-the west D2 shear zones
(Kwayisi et al., 2020), similar to the Togo structural unit (Agyei and
Tetteh, 1997). The D2 fabrics have locally been overprinted by D3

deformation expressed as F3 kink bands and open folds, with NE
plunging fold axes (Kwayisi et al., 2020). From detailed fieldwork
and petrographic studies, interpretation of airborne geophysical
data, and qualitative restoration of orogenic structures, Kwayisi
et al. (2020) produced a new map of the BSU in Ghana (Fig. 5a).
This map reveals the occurrence of the ultramafic (mainly serpen-



Ta
bl
e
1

Co
m
pi
la
ti
on

of
av

ai
la
bl
e
lit
ho

lo
gi
ca
l,
ag

e,
an

d
de

po
si
ti
on

al
se
tt
in
g
da

ta
on

th
e
V
ol
ta
ia
n
Su

pe
rg
ro
up

,t
he

Bu
em

an
d
To

go
st
ru

ct
ur

al
un

it
s.

U
n
it

D
ep

o
si
t
ty
p
e

Y
o
u
n
ge

st
zi
rc
o
n

U
-P

b
ag

e

D
ep

o
si
ti
o
n
al

ag
e

P
ro

ve
n
an

ce

V
o
lt
ai
an

Su
p
er
gr

o
u
p

O
bo

su
m

G
ro
u
p

K
eb

ia
Su

bg
ro
u
p

Fo
re
la
n
d

(M
ol
as
se
-t
yp

e)
1
,2

59
1
M
a1

?
B
N
S1

,2

Y
en

di
Su

bg
ro
u
p

O
ti
G
ro
u
p

A
fr
am

-B
im

bi
la

Su
bg

ro
u
p

Fo
re
la
n
d
(F
ly
sc
h
-t
yp

e)
2
,3

60
0
M
a2

,3
57

6
±
13

M
a4

(L
u
/H

f,
ph

os
ph

or
it
e)

B
N
S3

K
oj
ar
i-
B
u
ip
e

Su
bg

ro
u
p

Pa
ss
iv
e
m
ar
gi
n
2
,5

ca
.6

35
M
a2

(c
on

st
ra
in
ed

by
U
-P
b
zi
rc
on

ag
e
of

M
ar
in
oa

n
gl
ac
ia
ti
on

ev
en

t)

W
A
C
5

A
m
az

on
ia
n
C
ra
to
n
an

d
W

A
C
1

B
om

bo
u
ak

a
G
ro
u
p

Pa
ss
iv
e
m
ar
gi
n

1
,2
,3
,6
,7
,8

11
00

M
a1

95
9
±
62

M
a9

(R
b/
Sr
,c

la
y

fr
ac
ti
on

s)
W

A
C
6
,7

A
m
az

on
ia
n
C
ra
to
n
an

d
W

A
C
1
,8

B
u
em

st
ru

ct
u
ra
l
u
n
it

U
pp

er
m
os

t
Fo

re
la
n
d3

60
0
M
a3

ca
.6

50
M
a1

1
(R
b/
Sr
,

gl
au

co
n
it
e)

B
N
S3

U
pp

er
an

d
lo
w
er

Pa
ss
iv
e
m
ar
gi
n

1
,3
,1
0

95
0
M
a1

W
A
C
1
0

A
m
az

on
ia
n
C
ra
to
n
an

d
W

A
C
1

To
go

st
ru

ct
u
ra
l
u
n
it

K
an

ti
sc
h
is
t

Fo
re
la
n
d3

60
0
M
a3

70
3
±
8
M
a1

3
(U

-P
b
zi
rc
on

,
m
et
ab

as
al
t)

B
N
S3

Q
u
ar
tz
it
e
an

d
ph

yl
li
te

Pa
ss
iv
e
m
ar
gi
n
1
,3
,1
2

95
0
M
a1

,3
A
m
az

on
ia
n
C
ra
to
n
an

d
W

A
C
2
,1
2

1
K
al
sb

ee
k
et

al
.(
20

08
),

2
C
ar
n
ey

et
al
.(
20

10
),3

G
an

ad
e
de

A
ra
u
jo

et
al
.(
20

16
),

4
B
ar
fo
d
et

al
.(
20

04
),

5
A
m
ed

jo
e
et

al
.(
20

18
),

6
A
n
an

i(
19

99
),

7
A
n
an

ie
t
al
.(
20

13
),

8
A
n
an

ie
t
al
.(
20

17
),

9
C
la
u
er

(1
97

6)
,1

0
O
sa
e
et

al
.(
20

06
),

1
1
C
la
u
er

et
al
.(
19

82
),

1
2
A
n
an

i
et

al
.(
20

19
),

1
3
G
an

ad
e
de

A
ra
u
jo

et
al
.(
20

14
c)
.

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204

188
tinised peridotite) rocks in top-to-the west thrust zones, which
mark major tectonic contacts: to the west between the BSU and
Voltaian Supergroup and the east between the BSU and Togo struc-
tural unit and within the sedimentary rocks of the BSU. It also led
to an updated tectonostratigraphic division of the BSU into lower
clastic units, followed by an ultramafic complex that is capped
by chemical sedimentary units (chert and carbonate). These are
overlain by gabbro and intermediate to mafic volcanic rocks, which
are in turn overlain by the chemical units. The top part of the
sequence consists of the upper clastic unit (Fig. 5b). All contacts
between units are tectonic, expect for thin (3 m-thick) chemical
units capping the ultramafic rocks, and gabbro locally in intrusive
contact with volcanic rocks. Two main groups of sedimentary rocks
have been identified in the BSU based on petrographic and geo-
chemical data, and detrital zircon age fractions of the sediments;
passive margin and foreland deposits (Table 1; Kalsbeek et al.,
2008; Ganade de Araujo et al., 2016).
3. Field relations and petrography

Sedimentary rocks dominate the study area, occupying nearly
70% of the BSU (Fig. 5a). The sedimentary rocks of the BSU can
be grouped into clastic and chemical units (Fig. 5b). The clastic
units are sandstone, shale, turbiditic shale and diamictite, occupy-
ing the bottom and upper parts of the BSU, with the mafic–ultra-
mafic and chemical units in the middle. The lower clastic units
are overlain in thrust contact by the mafic–ultramafic rocks,
whereas the upper clastic units make up the top of the tectonos-
tratigraphic column. Figs. 6 and 7 illustrate the field and petro-
graphic characteristics of the BSU clastic units respectively. The
lower clastic units have undergone weak metamorphism defined
by albite, quartz and muscovite and related to the ubiquitous D2

deformation event described by Kwayisi et al. (2020). The upper
clastic units are generally unmetamorphosed. The lower clastic
units are largely fine- to medium-grained, consisting of interca-
lated quartzite and slate (Fig. 6a), usually occupying low topo-
graphical levels (forming lowlands). The lower clastic units are
abundant in the eastern part of the study area, towards the contact
with the Togo structural unit. The upper clastic units, on the other
hand, consist generally of massive units (Fig. 6b) that occupy
higher topographic levels (forming high ridges) than the lower
clastic units, and consist of thick beds of well- to poorly-sorted
units including diamictite, sandstone and shale (Fig. 6b and 6c).
They are dominant in the western part of the study area, towards
the contact with the Voltaian Supergroup.

The features of the framework grains of the BSU clastic units,
characterised by abundant quartz, appreciable amounts of feldspar
(plagioclase and K-feldspar) together with lithic fragments (vol-
canic rocks, slate, sandstone, greywacke, schist), classify them as
quartz arenite, sublitharenite, subarkose, lithic subarkose and
litharenite (diamictite) (Fig. 8). The main criterion for distinguish-
ing lower and upper clastic units’ quartz arenites is that the lower
clastic units are bedded whereas the upper clastic units are gener-
ally massive. The shales vary in colour, ranging from dark grey, red,
purple, green, to creamy white. The shales are grouped into upper
shale and lower shale based on their position in the tectonostratig-
raphy (Fig. 5b). The upper shales occur as interbeds within the
upper clastic units, while the lower shales occur within the lower
clastic units.
3.1. Lower clastic units

The lower clastic units of the BSU are quartz-arenite (quartzite;
NTDK3) and shales (slate; lower shale, NTDK124B and NTDK242B).
The quartz arenite is thickly to thinly bedded, fine- to medium-



Fig. 6. Representative field photographs of the BSU clastic units shown in Fig. 3b. (a) Intercalation of quartz arenite and shale at the base of the BSU, (b) thickly bedded
sandstone of the upper clastic units, (c) diamictite, and (d) thick beds of sandstone with diamictite interbed.

Fig. 5. (a) Geological map of the Buem structural unit, and (b) tectono-stratigraphic sequence of the Buem structural unit, showing structural relationships (modified from
Kwayisi et al., 2020). Locations of new samples for geochemistry and U-Pb dating (bold) are shown in the map and stratigraphy.

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204

189



Fig. 8. QFL diagram (Dickinson et al., 1983) for the Buem sandstones and diamictite
Q = quartz, F = feldspar, and L = lithic fragment (excluding polycrystalline quartz).

Fig. 7. Photomicrographs of (a) deformed quartz arenite at the base of the BSU showing elongated grains, (b) quartz arenite from the UCU, (c) diamictite, (d) deformed
diamictite showing elongated clasts of schist, defining the S2 slaty cleavage, (e) sublitharenite with volcanic clast and moderately- to well-sorted grains, (f) moderately sorted
sublitharenite, (g) poorly-sorted sublitharenite, (h) subarkose with angular and poorly-sorted grains, (i) lithic subarkose showing significantly altered feldspars into sericite.
Qz = Quartz, Ksp = K-feldspar, Pl = Plagioclase, Ser = Sericite (from Whitney and Evans, 2010) and Lt = Total lithic fragment (including polycrystalline quartz; after Dickinson
et al., 1983).

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
grained with sub-rounded to rounded clasts (Fig. 7a). It is com-
posed of about 94 – 96% quartz grains, 1 – 3% lithic fragments
(quartzite), 1 – 2% K-feldspar, and 1% muscovite (Fig. 7a). The
framework grains of the quartz arenite are cemented mainly with
silica. The quartz grains show undulose extinction and deformation
190
lamellae. The K-feldspar is significantly altered to sericite. The
lower shales are fine-grained, laminated and composed domi-
nantly of clay minerals. They are intercalated with the quartz aren-
ite (Fig. 6a). Lower shales seldom are in tectonic contact with the
mafic–ultramafic plutonic rocks.
3.2. Upper clastic units

The upper clastic units include quartz arenite (NTDK237),
diamictite or litharenite (NTDK165C), sublitharenite (NTDK60,
NTDK165B, NTDK209B), subarkose (NTDK328B), lithic subarkose
(NTDK269), and U-shale (NTDK59, NTDK270). The quartz arenite
is generally massive, medium-grained, moderately- to well-
sorted with sub-rounded to rounded grains (Fig. 6b and Fig. 7b).
Framework grains in the quartz arenite are dominantly quartz
(96 – 98%), and minor K-feldspar (2 – 4%), which are mainly
cemented with silica. The diamictite occurs as massive beds that
have coarse-grained (>2 mm) clasts embedded in fine-grained sand
or clay matrix (Fig. 6c and d, and Fig. 7c). It is poorly sorted with
sub-angular to rounded clasts. In a few samples, the grains are
elongated and show a slightly preferred orientation (Fig. 7d). The
diamictite is composed of quartz (45 – 55%), rock fragments (35
– 50%), plagioclase (1 – 2%), and muscovite (<1%). Rock fragments
are clasts of schist (garnet-staurolite, biotite, quartz-sericite, and
chlorite-quartz schist), slate, granite, greywacke, and volcanic
material (Fig. 7c). The diamictite contains a sericite matrix and sil-
ica cement.

Three varieties of sublitharenites can be distinguished, based on
texture and composition. Sample NTDK209 is medium to coarse-
grained, moderate- to well-sorted with sub-rounded clasts
(Fig. 7e). It is composed of quartz (75%), lithic fragments (15%



Table 2
Major and trace elements composition of the Buem sedimentary rocks. Sandstone.

Sample Id NTDK3 NTDK237 NTDK60 NTDK209B NTDK165B NTDK165C NTDK260 NTDK328B

Longitude 0.489 0.3864 0.4771 0.2790 0.4292 0.4160 0.4342 0.3840

Latitude 8.092 7.6553 7.6811 6.9880 7.3751 7.3800 8.0422 7.0023

LOWER CLASTIC UNIT UPPER CLASTIC UNIT

Quartz arenite Sublitharenite Diamictite Lithic subarkose Subarkose

SiO2 92.88 96.3 92.38 73.98 79.44 95.55 89.5 74.14
TiO2 0.19 0.12 0.16 0.56 0.33 0.12 0.17 0.68
Al2O3 3.44 1.81 3.7 11.51 9.02 2.7 4.59 10.66
Fe2O3 1.76 0.55 1.27 4.37 3.74 0.5 2.72 4.69
MnO 0.04 0.05 0.04 0.08 0.18 0.4 0.4 0.07
MgO 0.26 0.12 0.26 1.51 1.15 0.11 0.21 1.6
CaO 0.04 0.05 0.04 0.68 0.24 0.04 0.04 0.96
Na2O 0.12 0.09 0.04 3.63 1.73 0.4 0.04 1.7
K2O 0.49 0.4 1.13 1.35 2.11 0.74 1.04 2.05
P2O5 0.04 0.05 0.4 0.14 0.08 0.4 0.4 0.16
LOI 0.74 0.72 0.96 1.85 1.71 0.74 1.3 2.73
Sum 99.88 100.26 99.85 99.71 99.74 100.46 99.52 99.43
CIA 78 69 66 59 60 60 72 61
Sc 3.3 0.3 3.2 12 8.5 1.9 4.7 14
V 33 28 29 87 55 22 30 83
Co 1.5 0.6 3.4 9 14.5 0.5 4.7 12.2
Ni 12 7 11 22 28 4 10 38
Rb 22 4 47 39 72 27 45 92
Sr 30 1 0 136 91 5 16 63
Y 6.5 2.6 8.2 24.9 13.2 5.8 12.2 39.8
Zr 42 24 77 103 99 57 95 202
Nb 2.24 0.61 2.79 6.34 4.45 1.92 2.7 12.26
Cs 0.65 0.05 1.82 0.94 3.11 0.43 0.98 2.91
Ba 153 10 190 561 428 181 178 341
La 15.7 3.7 9.4 17.8 13.6 7.2 14.2 39.6
Ce 28.5 6.9 22.3 35.0 29.6 13.6 27.7 67.7
Pr 3.39 0.82 2.24 4.63 3.52 1.59 3.46 9.3
Nd 12.0 2.8 8.1 17.7 13.4 5.7 12.7 34.7
Sm 2.15 0.52 1.41 3.8 2.56 1.1 2.36 7.08
Eu 0.48 0.09 0.29 0.86 0.56 0.23 0.5 1.41
Gd 1.76 0.4 1.28 3.66 2.26 1.02 2.14 6.5
Tb 0.23 0.06 0.19 0.58 0.32 0.15 0.33 0.94
Dy 1.16 0.36 1.25 3.63 1.97 0.88 2 5.51
Ho 0.23 0.08 0.26 0.76 0.41 0.18 0.39 1.09
Er 0.60 0.20 0.79 2.25 1.23 0.49 1.04 2.93
Tm 0.09 0.03 0.13 0.33 0.18 0.07 0.17 0.42
Yb 0.57 0.23 0.87 2.26 1.21 0.49 1.13 2.82
Lu 0.08 0.03 0.15 0.34 0.19 0.08 0.17 0.44
Hf 1.2 0.69 2.2 2.83 2.74 1.63 2.64 5.73
Ta 0.18 0.04 0.2 0.47 0.34 0.15 0.2 0.86
Pb 3.64 0.96 3.79 9.72 11.21 5.17 11.13 16.47
Th 1.92 0.74 3.04 4.59 4.18 1.88 3.34 10.88
U 0.51 0.22 0.52 1.13 0.82 0.42 0.59 2.35
Eu/Eu* 0.75 0.59 0.67 0.70 0.71 0.66 0.68 0.64

Shale

Sample Id NTDK59 NTDK270 NTDK124B NTDK242B

Longitude 0.4754 0.3240 0.5112 0.3991

Latitude 7.6782 7.2860 7.3574 7.6945

U-Shale L-Shale

SiO2 69.02 68.98 62.98 69.52
TiO2 0.63 0.63 0.78 0.66
Al2O3 15 14.99 16.46 11.84
Fe2O3 6.23 6.22 7.11 8.51
MnO 0.04 0.4 0.08 0.05
MgO 0.64 0.65 2.54 1.77
CaO 0.04 0.04 0.31 0.27
Na2O 0.04 0.04 1.59 2.09
K2O 3.72 3.72 4.15 2.71
P2O5 0.11 0.11 0.16 0.17
LOI 4.09 4.01 3.57 2.06
Sum 99.54 99.42 99.8 99.66
CIA 72 72 64 61
Sc 11 15 20 16
V 91 96 90 82
Co 14 14.2 17.9 8.9
Ni 38 39 189 26

(continued on next page)

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204

191



Table 2 (continued)

Shale

Sample Id NTDK59 NTDK270 NTDK124B NTDK242B

Longitude 0.4754 0.3240 0.5112 0.3991

Latitude 7.6782 7.2860 7.3574 7.6945

U-Shale L-Shale

Rb 79 111 176 124
Sr 18 30 14 19
Y 9.6 16.8 36.6 33.8
Zr 139 138 159 245
Nb 8.72 8.84 13.88 23.33
Cs 6.25 6.06 6.36 4.09
Ba 884 919 590 175
La 6.8 20.6 14.5 30.3
Ce 36.4 47.1 59.9 63.2
Pr 1.92 5.36 3.86 7.6
Nd 7.4 19.5 14.7 28.1
Sm 1.64 3.7 3.61 5.61
Eu 0.44 0.91 0.82 1.09
Gd 1.82 3.38 3.9 5.11
Tb 0.31 0.5 0.67 0.83
Dy 1.89 2.92 4.56 5.25
Ho 0.4 0.61 1.02 1.08
Er 1.25 1.74 3.14 3.18
Tm 0.19 0.27 0.48 0.49
Yb 1.35 1.8 3.12 3.33
Lu 0.21 0.29 0.51 0.52
Hf 4 3.88 4.52 6.49
Ta 0.65 0.63 0.97 1.4
Pb 2.14 2.66 18.54 14.65
Th 4.23 5.34 12.17 11.79
U 1.42 1.5 2.04 1.89
Eu/Eu* 0.78 0.79 0.67 0.62

LOI = loss on ignition, CIA = chemical index of alteration, and ICV = index of compositional variability. The diamictite show very high SiO2 content probably because most of
the rock fragments and matrix are of silica-rich materials (Fig. 6c and d).

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
mainly volcanic clast), and plagioclase (10%). Sample NTDK60 is
fine to medium-grained, moderately sorted with subangular to
subrounded clasts (Fig. 7f). It is composed of quartz (85%), lithic
fragments (8%) and K-feldspar (7%). Sample NTDK165B contains
clasts ranging from fine to coarse-grained (Fig. 7g). It is poorly
sorted and contains sub-angular to subrounded clasts. The frame-
work grains of NTDK165B are quartz (84%), lithic fragments
(14%), and K-feldspar (2%). The lithic fragments are sandstone
and slate. Matrix content in the sublitharenites varies from 2 to
4% and is mainly sericite. The subarkose (NTDK328B) is generally
poorly to moderately sorted and contains fine-grained angular
clasts with abundant matrix (Fig. 7h).

The subarkose is quartz- and K-feldspar-dominated, making up
84% and 10% respectively of the total framework grains. Matrix
content is about 6%. The K-feldspar is significantly altered to seri-
cite. The lithic subarkose is medium-grained (NTDK260, Fig. 7i).
Texturally, the lithic subarkose contains sub-angular to sub-
rounded grains that are moderately sorted and slightly elongated.
It is composed of quartz (70%), K-feldspar (12%) and lithic frag-
ments (16%). K-feldspar is significantly altered to sericite. A signif-
icant amount of muscovite (2%) and trace amounts of zircon are
observed. The matrix content of the lithic subarkose is about 2 –
5% and generally composed of sericite that is deformed between
framework grains. Generally, the upper shales are fine-grained,
laminated and composed dominantly of clay minerals. Thin layers
of upper shale commonly occur as interbeds within the diamictite
and the sublitharenite (Fig. 5b).
4. Sampling strategy and analytical techniques

To determine the depositional setting and potential prove-
nance, 12 fresh and representative samples (each sample repre-
sent one sample block) of the BSU sedimentary rocks were
192
analysed for their whole-rock major and trace element concen-
trations at the Spectrum Analytical Facility, University of Johan-
nesburg, South Africa. Major and trace elements analyses were
done by XRF spectrometer (Philips Panalytical MagiX Pro) and
ICP-MS (Perkin Elmer NexION 300D) respectively. Their position
in the stratigraphy is reported in Fig. 5b. Three samples from the
lower clastic units include a quartz arenite (NTDK3) and two
intercalated lower shales (NTDK124B, and NTDK242B). Nine
samples were selected from the upper clastic units. They include
one basal quartz arenite (NTDK237), three sublitharenites
(NTDK60, NTDK165B, and NTDK209B), one diamictite
(NTDK165C), one subarkose (NTDK328B), one lithic subarkose
(NTDK260) and two upper shales interlayered with the sub-
litharenite (NTDK59 and NTDK270). Six samples of siliciclastic
rocks, in stratigraphic order: NTDK3, and NTDK237, NTDK60,
NTDK260, NTDK165C, and NTDK165B were selected for zircon
U-Pb and Lu-Hf- analyses. The multi-collector inductively cou-
pled plasma mass spectrometer (MC-ICP-MS; Nu Plasma II
instrument) was used for the U-Pb and Hf analyses at the
University of Johannesburg, South Africa. A detailed description
of sample preparation and analytical procedure and protocol
can be found in Supplementary Data A. The following parame-
ters were used in calculating the epsilon-Hf (eHf(t)) and TDM
model age; a decay constant of 1.867 � 10-11 (Söderlund et al.,
2004), depleted mantle values of 176Hf/177Hf = 0.28325 and
176Lu/177Hf = 0.0388 (Griffin et al., 2004), and CHUR parameters
176Lu/177Hf = 0.0336, 176Hf/177Hf = 0.282785 of Bouvier et al.
(2008). The one-stage Hf model ages (TDM1), calculated from
the 176Lu/177Hf and 176Hf/177Hf ratios measured for the zircons
of the BSU samples, can only give a minimum age for the
proto-source material of the magma from which the zircon crys-
tallised. Therefore, for each zircon, a two-stage Hf model age
(TDM2), or a ‘‘crustal” model age (TDMC ), which assumes that its



Fig. 9. Chondrite-normalized REE plot for (a) Buem sandstones and diamictite, (b) Buem shales. Multi-elements plot normalized to UCC (c) Buem sandstones and diamictite,
and (d) Buem shales. Normalising values for UCC from Rudnick and Gao (2003) and chondrite from Taylor and McLennan (1985). PAAS is from McLennan (1989).

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
parental magma was produced from average continental crust
(176Lu/177Hf = 0.015; Griffin et al., 2004) that originally was
derived from the depleted mantle, was calculated.
5. Major element and trace element characteristics

Table 2 presents the whole-rock major and trace element con-
centrations of the sedimentary rocks of the BSU. Post-diagenetic
alteration and metamorphism can affect the concentration of the
major elements, especially MgO, CaO, Na2O and K2O, and increase
the volatile content, as measured by the loss on ignition (LOI) of
sedimentary rocks (Jiang et al., 2017). Therefore, it is important
to evaluate the influence of post-diagenetic alteration and meta-
morphism on the mobility of the elements on the BSU sedimentary
rocks. All the samples of the BSU have low LOI content (0.7 – 4.1 wt
%), which makes it less likely that significant remobilisation of
these major elements has occurred during post-diagenetic alter-
ation and metamorphism.

The chemical index of alteration (CIA) permits the assessment
of the intensity of chemical weathering at the source of the sedi-
mentary rocks, particularly the conversion of feldspars to clay min-
erals (Nesbitt and Young, 1982). The CIA is defined by the molar
ratio [Al2O3/(Al2O3 + CaO* + Na2O + K2O)]/100, with CaO* signifying
Ca in the silicate fraction i.e. exclusive of calcite, dolomite and apa-
tite (CaO* = CaO – CO2 (cc) – (0.5 � CO2 (dol)) � 10/3 (P2O5) (ap))
(Nesbitt and Young, 1982). In this expression, cc is calcite, dol is
dolomite and ap is apatite. McLennan (1993) indicated that in
cases where CO2 is not measured, the expression for CaO* becomes
CaO* = CaO � 10/3 (P2O5) (ap); we use this approach here, as CO2

was not measured, and carbonates is not noted within thin sec-
tions. Low CIA (50) values show a weak intensity of weathering
while high (�100) values signify high chemical weathering
(Nesbitt and Young, 1982). The CIA values calculated for the BSU
sedimentary rocks vary between 69 and 78 for the quartz arenites
and 59 to 66 for the sublitharenites. The diamictite, subarkose, and
lithic subarkose respectively have CIA values of 60, 61 and 72
193
(Table 2). The shales have CIA values of 72 (upper shale), and
61–64 (lower shale). The CIA values for the BSU sedimentary rocks
indicate low to moderate degrees of chemical weathering.

Because the Rare Earth Elements (REEs), High Field Strength
Elements (HFSEs) and transition metals are least affected by
post-depositional alteration and metamorphism (Taylor and
McLennan, 1985; McLennan et al., 1993), potential provenance
and depositional setting interpretations will be based on these ele-
ments only. The BSU sedimentary rocks exhibit similar REE pat-
terns, comparable to Upper Continental Crust (UCC) and Post-
Archean Australian Shale (PAAS), with concentrations of the Light
REE (LREE) up to 100 times chondrite, nearly flat Heavy REE
(HREE), and negative Eu anomalies on the chondrite-normalised
REE diagram (Fig. 9a and 9b). Two of the shales show pronounced
positive Ce anomalies. Fig. 9c and 9d are the diagrams of the
incompatible trace element concentrations normalized to UCC for
the BSU sedimentary rocks. On these diagrams, the BSU sedimen-
tary rocks are similar to PAAS but display stronger Sr depletion
(Fig. 9c). The diamictite, sublitharenites and lithic subarkose have
positive K peaks although absolute values are lower than for PAAS
and UCC. Upper clastic units’ quartz arenite, sample NTDK237, has
the lowest trace element concentration, showing a Nb-Ta trough,
depletion in Ba, and enrichment in K, P and Ti. A very weak Nb-
Ta trough is also evident for the lithic subarkose, diamictite, and
two sublitharenite samples (NTDK60 and NTDK165B). The lithic
subarkose, diamictite, and a sublitharenite (NTDK209B) exhibit a
positive P anomaly. The overall trace element patterns for the
shales resemble that of PAAS and the concentrations are almost
the same as UCC (Fig. 9d).
6. Zircon morphology, U-Pb age distribution, and Lu-Hf isotope
composition

Cathodoluminescence (CL) images of detrital zircon grains
together with spot ages and eHf(t) values are given in Supplemen-
tary Data B and the U-Pb zircon analysis results for the samples of



Fig. 10. Cumulative age distribution plots for near-concordant (10% discordance limit) zircons from the BSU samples. The letters A to K represent major magmatic and
metamorphic events on the WAC (purple), BNS (green), and the Amazonian Craton (AC, grey). A = Leonian, B = Liberian, C = Eburnean, D = Archean events, E = Ventuary-
Tapajós, F = Rio Negro-Juruena, G = Rondonian-San Ignacio, H = Sunsas-Aguapeí, I = Brasiliano, J = Paleoproterozoic basement, K = magmatism in the upper plate (i.e., BNS)
during Pan-African plate convergence of the Dahomeyide belt. The dashed lines are the confidence intervals.

Table 3
Zircon age distributions of the Buem samples.

Sample Classification Position Age fraction (Ma) Percentage

NTDK165B Sublitharenite Top part of upper clastic unit 1650 – 900 61%
2400 – 1740 34%
2980 – 2560 5%

NTDK165C Diamictite Upper clastic unit 1590 – 890 54%
2190 – 1620 33%
2990 – 2450 13%

NTDK60 Sublitharenite Upper clastic unit 1550 – 970 46%
2150 – 1620 49%
3110 – 2500 5%

NTDK260 Lithic subarkose Upper clastic unit 1600 – 1130 48%
2150 – 1700 52%

NTDK237 Quartz arenite Lower part of the upper clastic unit 1550 – 1140 7%
2240 – 1640 71%
2960 – 2470 22%

NTDK3 Quartz arenite Lower clastic unit 1530 – 1370 4%
2350 – 1741 95%
2600 1%

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204

194



D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
the BSU are available in Supplementary Data C. In this study, a total
of 581 zircons were analysed in six BSU samples for their U-Pb
ages, of which 432 analyses provided ages within ± 10% concor-
dance and only these near-concordant data will be discussed. The
reported U-Pb ages are 207Pb/206Pb ages, although 235U/206Pb ages
are shown in Supplementary Data C. The Lu-Hf isotope composi-
tions for the near-concordant zircons from the BSU samples are
given in Supplementary Data D. The results are illustrated in the
cumulative age distribution (Fig. 10), probability density and ker-
nel density estimate plots (Supplementary Data E). The results
are presented in stratigraphic order, starting with the lowermost
sample as shown in Fig. 5b.

The zircons separated from sample NTDK3 (quartz arenite) col-
lected within the lower clastic units, generally have regular oscilla-
tory or sector zoning, and mostly exhibit rounded grain
morphologies. The zircons have 1:1 to 2:1 length to width ratios
and range between 55 and 160 lm in size. They dominantly have
Th/U ratios between 0.3 and 1.0 with very few grains having Th/U
ratios of 0.08 – 0.27. The dominant 207Pb/206Pb age fraction is 2350
– 1740 Ma (95%), with two zircon grains at 1530 and 1370 Ma (4%)
and one single 2600 Ma (1%) zircon grain for the 69 near-
concordant zircons analysed for this sample (Table 3). Most of
the zircon grains in this sample are between 2200 and 2040 Ma
(Fig. 10a, and Supplementary Data E), with a peak at 2100 Ma.

The initial 176Hf/177Hf isotopic ratios of this sample vary from
0.28100 to 0.28191. The Paleoproterozoic zircon grains gave nega-
tive and positive eHf(t) values from �14.4 to + 8.6, with the major-
ity in the range �5 to + 5 (Fig. 11a) corresponding to Hf crustal
model ages (TDMC ) from 3.6 to 2.2 Ga (Supplementary Data D).
Fig. 11. Epsilon-Hf vs. age diagram for the Buem samples, showing the detrital zircon
published data for the WAC, BNS, AC and the Santa Quitéria Arc (SQA), (a) samples NDT
NTDK165C, (d) comparison of the BSU samples with the Neoproterozoic Pan-African Anti-
et al. (2018) for the WAC, Pepper et al. (2016), Neto and Lofan (2019) for the AC, Ganade d
(2014b) for SQA, and Abati et al. (2012) for Anti-Atlas belt. The letters A to K represent ma

195
The two zircons of Mesoproterozoic age have positive eHf(t) values
close to + 3 and Hf crustal model ages of ca. 2.1 Ga.

The quartz arenite at the base of the upper clastic unit (Fig. 5b),
sample NTDK237, predominantly contains rounded zircon grains
that are mostly zoned (oscillatory zoning) and have a length to
width ratios of 1:1 to 2:1 and size of 80 to 160 lm. Th/U ratios
of these zircons are mainly between 0.3 and 2.0, with few grains
between 0.2 and 0.3. The 73 near-concordant zircons analysed
from this sample yielded a prominent zircon 207Pb/206Pb age frac-
tion of 2240 – 1640 Ma (71%), with a dominant fraction of 2240 –
1900 and a peak at 2150 Ma (Table 3, Fig. 10b, and Supplementary
Data E). Minor amounts of zircon with ages of 2960 – 2470 Ma
(22%) and 1550 – 1170 Ma (7%) are also recorded.

This quartz arenite sample records a wider range of initial
176Hf/177Hf isotopic ratios of 0.28047 – 0.28195 than sample
NTDK3. The zircon grains with Archean ages have eHf(t) values
of + 1.1 to �17.2 and Hf TDMC of 4.3 – 3.0 Ga (Supplementary Data
D). Paleoproterozoic and Mesoproterozoic zircon grains yielded
negative and positive eHf(t) values, �20.0 to + 4.3 and �4.3
to + 3.0 respectively (Fig. 11a), corresponding to Hf TDMC ages = 3.9
– 2.4 Ga and 2.4 – 2.0 Ga, respectively.

Sample NTDK260 (lithic subarkose), which conformably over-
lies quartz arenite sample NTDK237, contains dominantly euhedral
and a few rounded zircons, which display internal zoning. The zir-
cons range in size from 60 to 100 lm and have between 1:1 and
2:1 length to width ratios. These zircon grains have Th/U ratios
of mostly 0.3 – 2.0 and minor 0.07 – 0.30. Sample NTDK260 mainly
contains zircons of 2150 – 1700 Ma (52%) (with two peaks, at
1950 Ma and 1800 Ma), and of 1600 – 1130 Ma (48%) (with peaks
data relative to CHUR and Depleted Mantle evolution curves and comparison to
K3 and NTDK237, (b) samples NTDK60 and NTDK260, (c) samples NTDK165B and
Atlas belt in Morocco. Published data are from Parra-Avila et al. (2017) and Peterson
e Araujo et al. (2016) and Kalsbeek et al. (2020) for the BNS, Ganade de Araujo et al.
jor magmatic and metamorphic events on theWAC, BNS, and the Amazonian Craton.



D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
at 1550 Ma and 1200 Ma) (Table 3, Fig. 10c, and Supplementary
Data E).

In terms of Hf isotopic composition, this sample shows a wide
range of initial 176Hf/177Hf isotopic ratios (0.28112 – 0.28212), with
positive and negative eHf(t) values of �9.7 to + 5.3 (Paleoprotero-
zoic grains) and �15.3 to + 6.0 (Mesoproterozoic grains) (Fig. 11b).
Hf crustal model ages for the Paleoproterozoic and Mesoprotero-
zoic zircon grains are 3.5 – 2.3 Ga, and 3.1 – 1.9 Ga, respectively
(Supplementary Data D).

Sample NTDK60 (sublitharenite) conformably overlies sample
NTDK260 and is separated from it by a shale layer. Zircons sepa-
rated from sample NTDK60 generally have regular internal zoning,
exhibiting both euhedral and rounded grain morphologies. The zir-
cons have 1:1 to 2:1 length to width ratios and range between 70
and 150 lm in size. They have Th/U ratios dominantly between 0.3
and 1.7 with very few grains having Th/U ratios of 0.1 – 0.3. Sample
NTDK60 contains zircon 207Pb/206Pb age fractions in the ranges
3110 – 2500 Ma (5%), 2150 – 1620 Ma (49%), and 1550 – 970 Ma
(46%) from the 72 near-concordant zircons analysed (Table 3,
and Fig. 10d).

Initial 176Hf/177Hf of this sample range between 0.28075 and
0.28218 with the following eHf(t) values and Hf crustal model ages
for the various zircon age fractions: �0.8 to �2.4 and TDMC = 3.6 –
3.2 Ga for Archean zircons, �13.0 to + 3.6 and TDMC = 3.4 – 2.3 Ga
for Paleoproterozoic zircons, �17.6 to + 5.1 and TDMC = 3.1 –
1.7 Ga for Mesoproterozoic zircons, and –22.6 and TDMC = 3.2 Ga
for the one Neoproterozoic zircon grain analysed (Fig. 11b, Supple-
mentary Data D).

The diamictite sample (NTDK165C) that conformably overlies
the sublitharenite (NTDK60) has the largest zircon grains (100–
230 lm) compared to the other samples. The zircon grains are
mostly internally zoned and predominantly of rounded morphol-
ogy. The zircon grains have Th/U ratios predominantly in the
range 0.3 – 2.0 and subordinately 0.01 – 0.30. Zircon 207Pb/206Pb
Fig. 12. (a) Pairwise comparison of the confidence intervals of zircon U-Pb age distributi
and with published data, using the 1-O parameter of Andersen et al. (2016) as a measur
lower part is for U-Pb age, (b) Pairwise comparison of the upper clastic units of the BSU s
and (c) Pairwise comparison of the lower clastic unit and the lower part of the upper c
quartzite of the TSU. Green = 0 (statistically indistinguishable at the 95% confidence level)
clastic units. Published data are from Kalsbeek et al. (2008) and Ganade de Araujo et al

196
age fractions recorded from the 90 near-concordant analysed zir-
cons are, from least to most prominent: 2990 – 2450 Ma (13%),
2190 – 1620 (33%), and 1590 – 890 Ma (53%), (Table 3, and
Fig. 10e).

The initial 176Hf/177Hf isotopic ratios of the zircons vary from
0.28158 to 0.28095. The eHf(t) values and depleted mantle extrac-
tion ages are �5.7 to + 2.3 and TDMC = 3.5 – 3.1 Ga for Archean zir-
cons; �9.8 to + 3.9 and TDMC = 3.2 – 2.2 Ga for Paleoproterozoic
zircons; �19.4 to + 4.7 and TDMC = 3.0 – 1.7 Ga for Mesoproterozoic
zircons; and –22.5 to + 0.4 and TDMC = 3.0 – 1.8 Ga for the Neopro-
terozoic zircons (Fig. 11c, and Supplementary Data D).

The sublitharenite, sample NTDK165B, conformably overlying
the diamictite (NTDK165C) with a stratigraphic contact, contains
both rounded and euhedral zircon grains that are mostly zoned
(oscillatory zoning) and have a length to width ratios of 1:1 to
2:1 and sizes of 70 to 230 lm. Th/U ratios of these zircons are
mainly between 0.3 and 2.0, with few grains between 0.04 and
0.30. The 77 near-concordant zircons analysed from this sample
yielded zircons 207Pb/206Pb age fractions at 2980 – 2560 Ma (5%),
2400 – 1740 Ma (34%), and 1650 – 900 Ma (61%) (Table 3,
Fig. 10f, and Supplementary Data E).

In terms of Hf isotopic composition, this sample yielded initial
176Hf/177Hf isotopic ratios of 0.28067 – 0.25211. Hf model ages
for Archean grains are TDMC = 4.0 – 3.7 Ga with mainly negative
eHf(t) values from �8.0 to �6.7 (Supplementary Data D,
Fig. 11c). Hf model ages for Paleoproterozoic and Mesoproterozoic
zircon grains respectively vary from TDMC = 3.4 – 2.5 Ga, and 2.7 –
1.8 Ga and both positive and negative eHf(t) values of �11.9
to + 3.8 (Paleoproterozoic grains) and �3.7 to + 5.3 (Mesoprotero-
zoic grains). Neoproterozoic zircon grains have mainly negative
eHf(t) values (-21.3 to �4.7) and Hf model ages of TDMC = 3.1 –
2.1 Ga. Summary of the petrographic, geochemical, and geochrono-
logical characteristics of the BSU sedimentary rocks are shown in
Supplementary Data F.
ons and Hf-in zircon isotopic composition among the individual samples of the BSU
e of difference. Note the upper part of the plot is Hf-in zircon comparison while the
howing extensive similarity to the Bombouaka Group, and the quartzite of the TSU,
lastic units of the BSU showing extensive similarity to Bombouaka Group and the
, white = 0–0.05, red = >0.05 (statistically different), and grey = no data. LCU = Lower
. (2016).



Fig. 13. Upper quartile vs. lower quartile plots, after Andersen et al. (2018), of zircon U-Pb age data for the BSU samples and published data. Published data are from Kalsbeek
et al. (2008) and Ganade de Araujo et al. (2016). The orthogonal lines on each sample point are confidence limit (2 sigma).

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
7. Discussion

7.1. Tectonostratigraphic record of the BSU

A first-order examination of the cumulative age distribution
curves and the U-Pb detrital zircon age vs. eHf(t) patterns in Figs. 10
and 11 allows us to distinguish three main provenance signatures
within the BSU: 1. Samples NDTK3 and NTDK237, from the base of
the lower and upper clastic units respectively, contain prominent
2300 – 1800 Ma (60 – 89%) zircons.; 2. Samples NTDK60 and
NTDK260, in the upper clastic units, contain predominantly 1800
– 1100 Ma (64 – 67%) zircons and a minor fraction at 2300 –
1800 (35%), but lack Neoproterozoic zircons grains; 3. Samples
NTDK165B and NTDK165C, occupying the top part of the upper
clastic units, contain a significant fraction of Neoproterozoic grains
(1000–970 Ma, 7%). Some of the 1200 – 900 Ma zircons from the
upper clastic units with pronounced negative eHf(t) values
between �10 and –23 and TDMC model ages between 3.1 and
2.5 Ga are arranged in a linear array with a slope of 176Lu/177Hf
of �0.001 (Fig. 11b and c). This suggests the main phase of crustal
growth between 3.1 and 2.5 Ga and subsequent resetting of the U-
Pb system during the early Neoproterozoic. Some of these late
Mesoproterozoic and Neoproterozoic zircons, however, show juve-
nile characteristics, having eHf(t) values of + 5.3 to 0 (Fig. 11b and
c).

As a measure of similarity between samples, the 1-O method as
proposed by Andersen et al. (2016) was employed because it takes
the uncertainty on the distribution into account. Where O repre-
sents the part of the cumulative age distribution curve over which
the confidence intervals of the two samples overlap (0 � O � 1).
O = 1.0 signifies that the samples have indistinguishable age distri-
bution patterns (Andersen et al., 2016). The pairwise O values for
the samples from the BSU are shown in Supplementary Data G,
and Fig. 12 represents a graph of 1-O illustrating the similarity pat-
tern of the BSU samples. On the graph, the two lowermost samples
NTDK3 and NTDK237 are consistently distinct from the other sam-
ples of the BSU. The Lower Quartile (LQ) vs Upper Quartile (UQ)
197
plot (Fig. 13) reinforces that samples NTDK3 and NTDK237 are
similar by having a narrow age (2100 – 1800 Ma) distribution of
their zircons, which is significantly different from the other sam-
ples (from 900 to 1800 Ma).

Although from Figs. 12 and 13 samples NTDK165B and
NTDK165C, and samples NTDK260 and NTDK60 are undistinguish-
able, only samples, NTDK165B and NTDK165C, contain Neopro-
terozoic zircons with only moderately negative eHf(t) values of
�1 to �5. It is thus clear that three broad groups of samples exist
within the BSU, those with older age fractions dominantly 2300 –
1800 Ma, represented by samples from the lower clastic units and
base of the upper clastic units, those with prominent 1800 –
1100 Ma zircons occupying the middle part of the upper clastic
units, and those with significant 1000 – 970 Ma age fractions,
forming the uppermost part of the upper clastic units of the BSU.
7.2. Depositional setting

7.2.1. Maturity and recycling of the sediments
There is a large variation in texture and mineralogical content

observed for the siliciclastic rocks of the BSU. This calls for an
investigation on the maturity of the sediments, using joint textural
observations and chemical maturity indexes, such as Th/Sc and
Zr/Sc. A plot of Th/Sc versus Zr/Sc is an important discriminator
to differentiate between the contrasting effects of proto-source
composition and sedimentary sorting or recycling on the composi-
tion of siliciclastic sedimentary rocks (McLennan et al., 1993). The
BSU sedimentary rocks display wide variations in their Zr/Sc and
Th/Sc ratios (Fig. 14a). Except for the quartz arenite at the base
of the upper clastic units which plot far above UCC and PAAS, the
other sandstone samples plot close to UCC and PAAS. This may
reflect unrecycled upper continental crust material to intensely
recycled detritus, suggestive of immature to mature sediments.
All the BSU shale samples have Zr/Sc and Th/Sc ratios lower than
UCC and PAAS, indicating that sedimentary recycling process was
not pronounced in the shales and thus the shales are composed
of immature sediments. The large variation in the Zr/Sc and



Fig. 14. (a) Th/Sc versus Zr/Sc after (McLennan et al., 1993). Note data points from
TTG, granite, felsic volcanic rocks, andesite, and basalt are from Condie (1993), and
(b) plot of La/Th against Hf for the BSU sedimentary rocks (composition fields after
Floyd and Leveridge, 1987).

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
Th/Sc ratios exhibited by the BSU sedimentary rocks (Fig. 14a) may
indicate poor mixing of the sediments in these rocks, a reflection of
the intercalation of shale and sandstone within the lower clastic
unit, and sandstones with shale interbeds within the upper clastic
units (Fig. 5b).
7.2.2. Composition of the source area
The detrital zircons of the BSU sandstones and diamictite show

a large range of U-Pb ages and Hf isotope model ages of 3110 –
970 Ma and 4.3 – 1.7 Ga respectively, a reflection of the accumula-
tion of sediments from a mixture of young and old continental
crust made up of juvenile mantle and reworked crustal materials
or pre-existing sedimentary rocks. This, coupled with the fact that
weathering and recycling rates do not follow stratigraphic succes-
sion, indicates the contribution of detritus to the sedimentary
basin from different proto-sources (proximal and distal) of differ-
ent emplacement ages or sediments thereof. Mafic igneous rocks
usually have minimal negative Eu anomalies (Eu/Eu* = 0.8 – 1),
whereas negative Eu anomalies (Eu/Eu* = 0.5 – 0.8) are often
observed in felsic igneous rocks (Taylor and McLennan, 1985;
Cullers, 1994; Cullers and Podkovyrov, 2000). Thus, the negative
Eu anomalies shown by all the BSU sedimentary rocks may suggest
the accumulation of detritus from felsic igneous rocks or sedi-
ments. The Th/Sc versus Zr/Sc ratios suggest that the BSU sedimen-
tary rocks were sourced from mixed felsic and intermediate proto-
198
source rocks of dominantly TTG and felsic volcanic rocks, with
minor contributions from granite and/or andesite or sediments
thereof (Fig. 14a). To further assess the composition of the BSU sed-
imentary rocks, the ratio of La/Th versus Hf content was employed.
The low concentration of Hf with correspondingly low La/Th ratios
in the BSU sedimentary rocks implies sediment derivation from an
intermediate source rock plotting close to the mixed felsic/mafic
source regions (Fig. 14b). Overall, the BSU sedimentary rocks were
derived from mixed sources of felsic and intermediate source
rocks.

7.2.3. Tectonic setting of the BSU sedimentary rocks
Two tectonic setting models have been proposed for the BSU

sedimentary rocks. Osae et al. (2006) proposed a passive margin
setting for the sandstones (upper and lower) of the BSU, based
on petrography and major element composition. Kalsbeek et al.
(2008) indicated that the sandstone for the upper and lower part
of the BSU with passive margin geochemical characteristics have
detrital zircon ages of 950 Ma and older. Foreland basin sedimen-
tary rocks of the Dahomeyide belt are characterised by detrital zir-
con ages of 900 – 600 Ma (Kalsbeek et al., 2008; Ganade de Araujo
et al., 2016). Thus, Ganade de Araujo et al. (2016) considered a fore-
land basin for the uppermost part of the BSU sandstones because of
the significant amounts of detrital zircon grains with ages 900 –
600 Ma. Detrital zircon grains obtained from the BSU sandstones
and diamictite yielded a wide range of ages from 3110 to
970 Ma. Youngest detrital zircon ages of passive and foreland sed-
imentary rocks of the Dahomeyide belt are at ca. 950 and 600 Ma
respectively (Kalsbeek et al., 2008; Ganade de Araujo et al., 2016).
Detrital zircons with ages between 1000 and 950 Ma constitute
about 7% on average of the total zircons analysed from the BSU
samples (this study). Thus, the absence of detrital zircons of ages
between 900 and 600 Ma of the BSU sandstones (this study) may
suggest their deposition in a passive margin depositional setting.

7.3. Correlations of the Buem structural unit with Voltaian Supergroup
and Togo structural unit

The detrital zircon and geochemical record of the BSU is com-
pared with the Voltaian Supergroup (Kalsbeek et al., 2008;
Ganade de Araujo et al., 2016; Anani et al., 2017; Abu and Zongo,
2017; Amedjoe et al., 2018) and Togo structural unit (Kalsbeek
et al., 2008; Ganade de Araujo et al., 2016; Anani et al., 2019).
The quartzite of the Togo structural unit and Bombouaka Group
of the Voltaian Supergroup have zircon age fractions of 2200 –
930 Ma and 2200 – 1130 Ma respectively (Figs. 12, 13, and 15a
and b; Kalsbeek et al., 2008; Ganade de Araujo et al., 2016). Neo-
proterozoic zircon grains of 900 – 600 Ma have been recorded in
the sandstones of the Oti and Obosum groups of the Voltaian
Supergroup and the Kanti schist of the Togo structural unit (Figs.
12, 13, and 15c; Kalsbeek et al., 2008; Ganade de Araujo et al.,
2016). The BSU has been divided into upper and lower sections
by Affaton (1990), using a diamictite unit as a marker. Affaton
(1990) suggested that the lower (below the diamictite) and upper
(above the diamictite) sections of the BSU correlate with the Bom-
bouaka and Oti groups, respectively.

Kalsbeek et al. (2008) inferred a possible correlation between
the sandstones of the BSU with the Bombouaka Group and the
quartzite of the TSU on the basis of detrital zircon ages
of > 950 Ma obtained for the BSU sandstones. In this study, samples
of the BSU contain detrital zircon with ages of ca. 970 Ma (upper
clastic units) and 1130 Ma (lower clastic units) and older, similar
to the quartzite of the Togo structural unit and Bombouaka Group
respectively (Figs. 12, 13 and 15). The zircon age population from
the BSU (this study) is however, distinct from the Oti and Obosum
groups, and the Kanti schist (Figs. 12, 13 and 15). The absence of



Fig. 15. Cumulative age distribution plots for the BSU samples compared to published data for the Voltaian Supergroup, TSU and BSU and correspond schematic diagram
illustrating the geodynamic evolution of the external zone of the Dahomeyide belt. (a) Lower passive margin, (b) upper passive margin, and (c) foreland basin. Published data
are from Kalsbeek et al. (2008) and Ganade de Araujo et al. (2012, 2016). The letters A to K represent major magmatic and metamorphic events on the WAC, BNS, and the AC,
see Fig. 10 for details.

D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
900 – 600 Ma zircon grains from the upper clastic units, even for
the units above the diamictite (Fig. 5b; this study) may suggests
that the upper section of the BSU in Ghana may not correlate with
the Oti Group as suggested by Affaton (1990). However, zircons of
such ages have been recorded for one sandstone sample (DKE-
361B) of the upper part of the BSU in Benin (Fig. 2, Ganade de
Araujo et al., 2016). The Benin sandstone (DKE-361B) is distinct
from the sandstone samples analysed in this present study, as
shown in Figs. 12 and 13, and may thus correspond to a part of
the BSU that is missing from the stratigraphy of the Ghana section
of the BSU, or it is yet to be sampled.

Based on geochemical composition and detrital zircon ages
of > 950 Ma, Kalsbeek et al. (2008), Ganade de Araujo et al.
(2016), Abu and Zongo (2017) and Anani et al. (2017; 2019) pro-
posed that the Bombouaka Group and the Togo structural unit
formed in the same passive margin basin of the Dahomeyide
belt. Based on the youngest detrital zircon at 600 Ma, Ganade
de Araujo et al. (2016) inferred a foreland (peripheral) basin set-
ting for the Oti Group and Kanti schist. However, Amedjoe et al.
(2018) considered the Oti Group to have been deposited in a
passive margin setting, based on geochemical characteristics.
Although some of the Oti samples have active margin signature,
Amedjoe et al. (2018) interpreted that as inherited from an old
island-arc source area, i.e. the WAC basement. Considering all
the evidence, it is therefore proposed that the BSU, Togo struc-
tural unit and Bombouaka Group possibly formed in the same
depositional basin, as reflected by similar geochemical features
and detrital zircon age distributions.
199
7.4. Potential provenance of the BSU sedimentary rocks

Detrital zircons recognized to be of igneous origin can be used
to determine major magmatic events in the proto-source areas
(Condie et al., 2009), and obtain information regarding continental
crust formation and evolution from their Lu-Hf isotopic composi-
tion (Andersen et al., 2016). A majority (>75%) of the zircons of
the BSU sedimentary rocks have regular internal zoning (Supple-
mentary Data B) with Th/U ratios between 0.3 and 2.0 (Supplemen-
tary Data C), in agreement with an igneous origin (e.g. Kinny et al.,
1990; Hoskin and Schaltegger, 2003; Corfu et al., 2003). Thus, the
U-Pb ages and Lu-Hf isotopic composition obtained from the
igneous zircons of the BSU sedimentary rocks reflect the timing
of their crystallisation, which in turn provides important con-
straints on the crustal evolutionary history of their proto-source.

A minor amount of 3110 – 2500 Ma (7%) zircons is recorded in
most of the analysed samples of the BSU. All the samples from the
BSU contain zircons in the age range from 2200 to 1800 Ma (33% –
93%). The potential sources for these zircons are the basement
rocks of the WAC, Amazonian Craton, and/or the Benino-Nigerian
Shield. The WAC is divided into a western Archean basement and
eastern Paleoproterozoic domain; these two areas are charac-
terised by greenstones, granitoids, and granitic gneisses of ages
between 3500 and 2500 Ma, and between 2200 and 2000 Ma
respectively (e.g. Abouchami et al., 1990; Taylor et al., 1992;
Potrel et al., 1998; Egal et al., 2002; Sakyi et al., 2014; Anum
et al., 2015; Kouamelan et al., 2015; Rollison et al., 2016;
Grenholm et al., 2019) (Fig. 11). Metasedimentary rocks of Paleo-



D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
proterozoic age (2200 – 1900 Ma) are associated with the Paleo-
proterozoic greenstones, granitoids and granitic gneisses. Also,
Archean and Paleoproterozoic greenstones, granitoids and granitic
gneisses (3500 – 1800 Ma) are exposed on the Amazonian Craton
(Cordani et al., 2009; Neto and Lofan, 2019). Therefore, potential
sources for the Archean and the Paleoproterozoic zircons of the
BSU sedimentary rocks are the granitoids and granitic gneisses of
the Archean and Paleoproterozoic basement and/or sedimentary
rocks associated with the Paleoproterozoic granitoids and granitic
gneisses of the WAC and the Amazonian Craton. Major and trace
element concentrations of the BSU sedimentary rocks and the Hf
isotopic ratios of their zircons indicate that they originated from
a felsic/intermediate continental basement with minimal to
intense sedimentary recycling.

The BSU has been reported to be underthrust by the Paleopro-
terozoic Birimian rocks based on field data (Attoh, 1998; Attoh
and Nude, 2008) and processing of aeromagnetic data (Kwayisi
et al., 2020). Additionally, the low positive and low to high negative
eHf(t) values for the Archean and Paleoproterozoic zircons of the
BSU sedimentary rocks are within the range for the Archean and
Paleoproterozoic granitoids and sedimentary rocks of the WAC
and more especially the Amazonian Craton (Fig. 11a to c; Parra-
Avila et al., 2017; Pepper et al., 2016; Petersson et al., 2018; Neto
and Lofan, 2019). The Benino-Nigerian Shield is a source area that
could be considered as it consists of 2200 – 2000 Ma granitic
gneisses and metasedimentary rocks (Attoh et al., 2013; Ganade
de Araujo et al., 2016; Kalsbeek et al., 2020), which were reworked
during the Pan-African orogeny (Attoh et al., 2013; Ganade de
Araujo et al., 2016). However, this is considered unlikely because
a signature of Benino-Nigerian Shield provenance is Pan-African
zircons (900 – 600 Ma, Kalsbeek et al., 2008; Ganade de Araujo
et al., 2016), which are absent in our samples (Figs. 11 and 15).
Moreover, all paleogeographic reconstructions suggest that the
WAC and Benino-Nigerian Shield were not attached until the
assembly of Gondwana at ca. 500 Ma (Rogers and Santosh, 2002;
Abdelsalam et al., 2002; Meert et al., 2012).

The WAC provenance does not explain our age fraction at 1700
– 1000 Ma, which is present in most of the BSU samples, because
the WAC did not record any major tectono-metamorphic events
between 1700 and 1000 Ma (Ennih and Liégeois, 2008). The pro-
portion of zircons with ages of 1700 – 1000 Ma in the BSU samples
varies with stratigraphic position, from bottom to top 9% (NTDK3),
18% (NTDK237), 64% (NTDK260), 55% (NTDK60), 54% (NTDK165C)
and 56% (NTDK165B) (Table 3). Other peripheral basins to the
WAC, such as the Neoproterozoic sedimentary rocks from the
Pan-African Anti-Atlas belt, southern Morocco, lack detrital zircons
of age fractions between 1700 and 800 Ma (Fig. 11d; Abati et al
2012), suggesting that this age fraction could be sourced from
areas external to the WAC such as the Amazonian Craton.

Late Paleoproterozoic and Mesoproterozoic (1700 – 1000 Ma)
granitoids and granitic gneisses are widespread on the Amazonian
Craton (Fig. 11a to c; e.g. Santos et al., 2000, 2008; Tassinari et al.,
2000; Cordani et al., 2009; Pepper et al., 2016). In the Rondonian–
San Ignacio and Sunsas-Aguapeí belts of the Amazonian Craton
(Fig. 1b), rocks of these ages are dominant (e.g., Sadowski and
Bettencourt, 1996; Cordani and Teixeira, 2007). Besides recent
paleogeography reconstruction of Rodinia from U-Pb baddeleyite
ages and paleomagnetic data on dolerite dyke swarms put the
Amazonian and West African cratons together (Baratoux et al.,
2019; Antonio et al., 2021). The Amazonian Craton is positioned
to the north of the WAC. This is consistent with paleocurrent data
of Bombouaka sandstones which, indicates NW- and NE-directed
sediment flow (Carney et al., 2010). Thus, the accumulation of
greater proportions of the detritus into the BSU from the Amazo-
nian Craton during Rodinia break-up is most likely. Moreover,
the positive and negative eHf(t) values for the zircons with ages
200
ranging from 1700 to 1000 Ma of the BSU samples are within the
range for the granitoids and sedimentary rocks of similar ages
within the Amazonian Craton (Fig. 11a to c; Pepper et al., 2016).

Kalsbeek et al. (2008) and Ganade de Araujo et al. (2016) have
proposed an Amazonian Craton contribution to the sediment sup-
ply for the sandstones of the Bombouaka Group of the Voltaian
Supergroup and the quartzite of the Togo structural unit based
on detrital zircon age fraction of 1700 – 1000 Ma (Fig. 15). The
Neoproterozoic rocks of the Taoudeni basin (Fig. 1a), which are
correlative to the Voltaian Supergroup, also record detrital zircon
with ages between 1700 and 1000 Ma, which have been inter-
preted to have been sourced probably from the Amazonian Craton
(Straathof, 2011).

More recently, Kalsbeek et al. (2020) obtained a 1146 ± 4 Ma
U-Pb zircon age from an augen-gneiss in contact (unclear the type
of contact) with the quartzite of Togo structural unit to the east.
Kalsbeek et al. (2020) inferred that this age could indicate a Meso-
proterozoic magmatic event that has not yet been identified on the
WAC, and thus this augen-gneiss could have contributed to the
sediment supply to the sandstones of the Bombouaka Group,
quartzite of the Togo structural unit and sandstones of the BSU.
However, there is still the need to account for the detrital zircon
age fractions between 1700 and 1300 Ma, 1000 and 900 Ma, and
the pronounced negative eHf(t) values, between � 10 and � 20
of the 2200 – 1800 age fraction (Fig. 11), which are missing from
the geological records of the WAC. Hence, the record of 1700 –
1000 Ma zircon age fractions in the BSU strengthens the
Amazonian-WAC connection. Thus, during Rodinia time, the WAC
and the Amazonian Craton were probably not separated by any
major seas and these cratons may have been connected since the
Paleoproterozoic until the breakup of Pangea as have been pro-
posed by Rogers (1996), Trompette (1994), Rogers and Santosh
(2002), and Evans (2009, 2013).

Neoproterozoic zircons are only recorded in the uppermost part
of the UCU (NTDK165B and NTDK165C) of the BSU in low propor-
tions (average 7%; Table 3). As discussed earlier, about 50% of these
zircons (980 – 890 Ma), having eHf(t) values between � 10
and � 23 and TDMC model ages between 2.5 and 3.1 Ga may have
lost Pb whereas only a few (990 – 970 Ma) show more juvenile
characteristics, having eHf values of + 5.3 to 0. These few juvenile
Neoproterozoic zircon grains could have been sourced from geo-
logical domains on the Amazonian Craton. Widespread over the
Amazonian Craton is the 1000 – 970 Ma anorogenic granitoids
which formed at the end of the Sunsas orogeny with comparable
eHf(t) values (+7 to 0; Pepper et al., 2016). In addition, the Goais
magmatic arc that provided 900 – 750 Ma detrital zircons to the
passive margin sedimentary rocks in eastern Amazonia could be
a possible source for these Neoproterozoic zircons (Fig. 11;
Cordani and Teixeira, 2007; Pepper et al., 2016).

In a summary, the detrital U-Pb zircon ages and Lu-Hf isotope
compositions of the BSU reveal the Amazonian Craton as the
potential provenance for its sedimentary rocks with or without
any contribution from the WAC.

7.5. Evolution of the external zone of the Dahomeyide belt

Results from this study, together with published data on the
BSU, Togo structural unit and the Voltaian Supergroup, reveal
two main sedimentary sequences in the external zone of the
Dahomeyide belt, i.e., passive margin and foreland basin sequences
with three potential provenances: Amazonian Craton, Benino-
Nigerian Shield, ± WAC (Fig. 15; Kalsbeek et al, 2008; Ganade de
Araujo et al., 2016; this study). The 3110 to 930 Ma zircon fraction
correspond to an Amazonian Craton provenance, with a minor con-
tribution from the WAC, whereas the 900 – 600 Ma is the signature
of the Benino-Nigerian Shield provenance (Fig. 11 and Fig. 15). The



D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
evolution of the passive margin sequence of the Dahomeyide belt
started with the deposition of the Bombouaka Group (with deposi-
tional age at 959 ± 65 Ma from a Rb–Sr isochron on clay, Clauer,
1976). This was followed by the deposition of the Togo and Buem
structural units (Fig. 15a), with depositional ages respectively at
703 ± 8 Ma (zircon U-Pb on a metavolcanic rock in the Togo struc-
tural unit; Ganade de Araujo et al., 2014a) and at � 650 Ma (Rb/Sr
in glauconite on a shale unit in the BSU; Clauer, 1982 in Guillot
et al., 2019). This large variation in depositional ages reflects a long
period of passive margin basin existence (from � 1000
to � 700 Ma) (Fig. 15b; Ganade de Araujo et al., 2016, this study).
Passive margin sedimentation was halted by the onset of the Neo-
proterozoic Pan-African orogeny, which resulted in subduction and
collision between 750 and 570 Ma (Affaton et al., 2000; Hirdes and
Davis, 2002; Guillot et al., 2019). Deposition of foreland basin
sequences (Oti and Obosum groups and the Kanti schist) occurred
during and after Pan-African continent–continent collision
(Fig. 15c). First, there was the deposition of the Kodjari-Buipe Sub-
group of the Oti Group with a � 635 Ma (Marinoan glaciation)
depositional age. Second, the Afram-Bimbila Subgroup of the Oti
Group and the Kanti schist were deposited, based on the 600 Ma
youngest detrital zircon age. The final one to be deposited was
the Obosum Group, which overlies the Afram-Bimbila Subgroup,
and which has a maximum depositional age of � 591 Ma, based
on the youngest detrital zircon (Ganade de Araujo et al., 2016).
7.6. Implications for the paleogeographic reconstruction of the West
African and Amazonian cratons in Rodinia

Passive margin units consisting of pre- and syn- to post-rift
metasedimentary rocks occur in the Pharusian belt, the northern
segment of the WGO (Caby, 2014). Although no radiogenic isotope
nor geochronological data are available on these passive margin
units, their deposition may have occurred before 790 Ma, which
is the age of an intra-oceanic arc gabbro (793 – 790 U-Pb zircon
ages) that intruded them (Caby, 2014). In the Borborema Province,
which is the southeastern segment of the WGO, the sedimentary
rocks of the Martinópole Group show similar detrital zircon age
distribution, depositional settings, and provenance patterns
(Fig. 1b; Ganade de Araujo et al., 2012) as the external zone of
the Dahomeyide belt (Fig. 15). The São Joaquim Formation of the
Martinópole Group was deposited in a passive margin (Fig. 15a),
with a depositional age of 777 Ma (U-Pb zircon age of interleaved
felsic volcanic rocks) and sediments sourced from the Amazonian
Craton and WAC (Ganade de Araujo et al., 2012; 2016). Foreland
basin sediments of the Goiabeira schist of the Martinópole Group
and Jaibaras Basin in the Borborema Province share similar detrital
zircon age fractions to that of the Oti and Obosum groups and Kanti
schist (Fig. 15c; Ganade de Araujo et al., 2012; 2016). The similarity
in provenance pattern and depositional setting of the external zone
of the Dahomeyide belt and sedimentary rocks of the Borborema
Province provides important constraints on the assembly of West
Gondwana during the Brasiliano/Pan-African orogeny and the
reconstruction of pre-Gondwana paleogeography of West Africa
and South America. A > 2500 km-long passive margin basin devel-
oped between ca. 1000 – 700 Ma before the onset of the assembly
of West Gondwana, which was inverted during the Brasiliano/Pan-
African plate convergence and collision, and the formation of the
peripheral foreland basin.
8. Conclusions

Detailed field, petrographic, geochemical, and geochronological
data have been used to constrain the provenance and depositional
setting of the Buem structural unit sedimentary rocks. The Buem
201
structural unit sedimentary rocks were deposited in a passive mar-
gin setting with sediments derived dominantly from the Amazo-
nian Craton and minor contributions from the West African
Craton. The geochemical and geochronological data reveal a strong
correlation between the Buem structural unit, Togo structural unit
and Voltaian Supergroup, an indication of similar depositional set-
ting and provenance. Results from this study combined with previ-
ous data suggest two main sedimentary rocks in the external zone
of the Dahomeyide belt, passive margin, and peripheral foreland
basin which received detritus from probably the Amazonian Cra-
ton, Benino-Nigeran Shield, ± West African Craton. Geochronolog-
ical data suggest the formation of a passive margin between
1000 and 700 Ma, followed by a foreland basin formation during
subduction-collision at 750 – 570 Ma. Comparable data are avail-
able for the sedimentary rocks of the Borborema Province, NE Bra-
zil, which imply similar evolution along the West Gondwana
Orogen during the break-up of Rodinia and subsequent assembly
of the supercontinent Gondwana. Because a greater portion of
the detritus is from the Amazonian Craton, the coexistence of the
West African and the Amazonian cratons from the Paleoprotero-
zoic with no major oceans between them until the opening of
the Atlantic Ocean is proposed.

CRediT authorship contribution statement

Daniel Kwayisi: Conceptualization, Methodology, Investigation,
Writing – original draft. Jeremie Lehmann: Supervision, Writing –
review & editing. Marlina Elburg: Supervision, Writing – review &
editing.

Declaration of Competing Interest

The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.

Acknowledgments

The authors acknowledge the financial support provided by the
National Research Foundation (NRF) of South Africa, grant
(105451), the African-German Network of Excellence in Science
(AGNES) Mobility Grant 2018, Paleoproterozoic Mineralization
(PPM) group of the University of Johannesburg, Department of
Geology, DSI-NRF Centre of Excellence for Integrated Mineral and
Energy Resource Analysis (CIMERA), and the College of Basic and
Applied Sciences (CBAS), University of Ghana. Our heartfelt appre-
ciation goes to Dr Jacob M. Kutu, the Nyavor family, Mr Emmanuel
Nyavor, Mr Emmanuel Kwaku Awunyo, the Chief and people of
Bontibor, Nkwanta, Chiaso, Nkonya-Mangoase, Aburuburuwa, Asu-
kawkaw, Kanease, and Akyem for their support during the field-
work. We are grateful to the Geological Survey Department of
Ghana for the release of vehicle for the fieldwork. The authors
would like to show appreciation to the Editor and two reviewers
Udo Zimmermann and Daniel T. Brennan for taking their time to
review the manuscript.

Appendix A. Supplementary material

Supplementary data to this article can be found online at
https://doi.org/10.1016/j.gr.2022.04.020.

References

Abati, J., Aghzer, A.M., Gerdes, A., Ennih, N., 2012. Insights on the crustal evolution of
the West African Craton from Hf isotopes in detrital zircons from the Anti-Atlas
belt. Precambr. Res. 212-213, 263–274.

https://doi.org/10.1016/j.gr.2022.04.020
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0005
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0005
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0005


D. Kwayisi, J. Lehmann and M. Elburg Gondwana Research 109 (2022) 183–204
Abdelsalam, M.G., Liégeois, J.-P., Stern, R.J., 2002. The Saharan metacraton. J. Afr.
Earth Sci. 34 (3-4), 119–136.

Abouchami, W., Boher, M., Michard, A., Albarede, F., 1990. A major 2.1 Ga event of
mafic magmatism in West Africa: an early stage of crustal accretion. J. Geophys.
Res. Solid Earth 95 (B11), 17605–17629.

Abu, M., Zango, M.S., 2017. Geochemical characteristics of the Neoproterozoic
Anyaboni sandstone of the southeastern Voltaian Basin Ghana. Scientific Res. J.
(SCIRJ) 5 (11), 54–65.

Adjei, A.O., Tetteh, G.M., 1997. Deformational phases of the Togo series, Ho-Nyive-
Honuta area Ghana. Ghana Mining J. 3 (1), 1–9.

Affaton, P., 1990. Le bassin des Volta (Afrique de l’Ouest): une marge passive, d’âge
protérozoïque supérieur, tectonisée au Panafricain (600 plus ou moins 50 Ma).
In: Thèse Doctorat d’Etat, collections Etude et Thèse, p. 499.

Affaton, P., Kröner, A., Seddoh, K.F., 2000. Pan-African granulite formation in the
Kabye Massif of northern Togo (West Africa): Pb� Pb zircon ages. Int. J. Earth
Sci. 88 (4), 778–790.

Affaton, P., Aguirre, L., Ménot, R.P., 1997. Thermal and geodynamic setting of the
Buem volcanic rocks near Tiéle, Northwest Benin West Africa. Precambr. Res.
82, 191–209.

Affaton, P., Rahaman, M.A., Trompette, R., Sougy, J., 1991. The Dahomeyide Orogen:
tectonothermal evolution and relationships with the Volta Basin. In: Dallmeyer,
R.D., Lécorché, J.P. (Eds.), The West African Orogens and Circum-Atlantic
Correlatives. Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 107–122.
https://doi.org/10.1007/978-3-642-84153-8_6.

Affaton, P., Sougy, J., Trompette, R., 1980. The tectonostratigraphic relationships
between the upper Precambrian and lower Paleozoic Volta Basin and the pan-
African Dahomeyide orogenic belt (West Africa). Am. J. Sci. 280 (3), 224–248.

Agbossoumonde, Y., Menot, R.-P., Guillot, S., 2001. Metamorphic evolution of
Neoproterozoic eclogites from south Togo (West Africa). J. Afr. Earth Sci 33 (2),
227–244.

Agyei, E.K., Van Landewijk, J.E.J.M., Armstrong, R.L., Harakal, J.E., Scott, K.L., 1987.
Rb/Sr and K/Ar geochronometry of southeastern Ghana. J. Afr. Earth Sci. (1983),
6(2), 153–161.

Aidoo, F., Nude, P.M., Dampare, S.B., Agbossoumondè, Y., Salifu, M., Appenteng, M.K.,
Tulasi, D., 2014. Geochemical Characteristics of Granitoids (Ho Gneiss) from the
Pan-African Dahomeyide Belt, Southeastern, Ghana: Implications for
Petrogenesis and Tectonic Setting. J. Environ. Earth Sci. 4, 46–65.

Amedjoe, C.G., Gawu, S.K.Y., Ali, B., Aseidu, D.K., Nude, P.M., 2018. Geochemical
compositions of Neoproterozoic to Lower Palaeozoic (?) shales and siltstones in
the Volta Basin (Ghana): Constraints on provenance and tectonic setting.
Sediment. Geol. 368, 114–131.

Anani, C., 1999. Sandstone petrology and provenance of the Neoproterozoic
Voltaian Group in the southeastern Voltaian Basin Ghana. Sediment. Geol. 128
(1-2), 83–98.

Anani, C.Y., Bonsu, S., Kwayisi, D., Asiedu, D.K., 2019. Geochemistry and provenance
of Neoproterozoic metasedimentary rocks from the Togo structural unit
Southeastern Ghana. J. Afr. Earth Sci. 153, 208–218.

Anani, C.Y., Mahamuda, A., Kwayisi, D., Asiedu, D.K., 2017. Provenance of sandstones
from the Neoproterozoic Bombouaka Group of the volta Basin, northeastern
Ghana. Arabian J. Geosci. 10 (21), 465.

Anani, C., Moradeyo, M., Atta-Peters, D., Kutu, J., Asiedu, D., Boamah, D., 2013.
Geochemistry and provenance of sandstones from Anyaboni and surrounding
areas in the voltaian basin, Ghana. Int. Res. J. Geol. and Min. 6, 206–212.

Andersen, T., Kristoffersen, M., Elburg, M.A, 2016. How far can we trust provenance
and crustal evolution information from detrital zircons? A South African case
study. Gondwana Res. 34, 129–148.

Andersen, T., Kristoffersen, M., Elburg, M.A., 2018. Visualizing, interpreting and
comparing detrital zircon age and Hf isotope data in basin analysis–a graphical
approach. Basin Res. 30 (1), 132–147.

Anum, S., Sakyi, P.A., Su, B.-X., Nude, P.M., Nyame, F., Asiedu, D., Kwayisi, D., 2015.
Geochemistry and geochronology of granitoids in the Kibi-Asamankese area of
the Kibi-Winneba volcanic belt, southern Ghana. J. Afr. Earth Sci. 102, 166–179.

Antonio, P.Y., Baratoux, L., Trindade, R.I., Rousse, S., Ayite, A., Lana, C., Macouin, M.,
Adu, E.W., Sanchez, C., Silva, M.A., Firmin, A.S., 2021. West Africa in Rodinia:
High quality paleomagnetic pole from the� 860 Ma Manso dyke swarm
(Ghana). Gondwana Res., 28–43

Attoh, K., 1998. High-pressure granulite facies metamorphism in the Pan-African
Dahomeyide orogen West Africa. The J. Geol. 106 (2), 236–246.

Attoh, K., Nude, P.M., 2008. Tectonic significance of carbonatite and ultrahigh-
pressure rocks in the Pan-African Dahomeyide suture zone, southeastern
Ghana. Geol. Soc. Lond., Spec. Publ. 297 (1), 217–231.

Attoh, K., Dallmeyer, R.D., Affaton, P., 1997. Chronology of nappe assembly in the
Pan-African Dahomeyide orogen, West Africa: evidence from 40Ar-39Ar mineral
ages. Precambrian Res. 82 (1–2), 153–171.

Attoh, K., Hawkings, G., Bowring, S.A., 1991. U-Pb zircon ages from the Pan-African
Dahomeyide orogen, West Africa. EOS Trans. Am. Geophys. Union Spring Meeting
72, abstract.

Attoh, K., Corfu, F., Nude, P., 2007. U-Pb zircon age of deformed carbonatite and
alkaline rocks in the Pan-African Dahomeyide suture zone West Africa.
Precambr. Res. 155 (3-4), 251–260.

Attoh, K., Samson, S., Agbossoumondé, Y., Nude, P.M., Morgan, J., 2013. Geochemical
characteristics and U-Pb zircon LA-ICPMS ages of granitoids from the Pan-
African Dahomeyide orogen West Africa. J. Afr. Earth Sci. 79, 1–9.

Baratoux, L., Metelka, V., Naba, S., Jessell, M.W., Grégoire, M., Ganne, J., 2011.
Juvenile Paleoproterozoic crust evolution during the Eburnean orogeny (� 2.2-
2.0 Ga), western Burkina Faso. Precambr. Res. 191 (1-2), 18–45.
202
Baratoux, L., Söderlund, U., Ernst, R.E., De Roever, E., Jessell, M.W., Kamo, S., Naba, S.,
Perrouty, S., Metelka, V., Yatte, D., Grenholm, M., 2019. New U–PB baddeleyite
ages of mafic dyke swarms of the West African and amazonian cratons:
Implication for their configuration in supercontinents through time. In Dyke
Swarms of the World: A Modern Perspective, Springer, Singapore, 263–314.

Barfod, G.H., Vervoort, J.D., Montanez, I.P., Riebold, S., 2004. Lu-Hf geochronology of
phosphates in ancient sediments. 13th Goldschmith conference, Copenhagen,
Geochim. Cosmochim. Acta, abstract, 68(11), A336.

Berger, J., Caby, R., Liégeois, J.-P., Mercier, J.-C., Demaiffe, D., 2011. Deep inside a
neoproterozoic intra-oceanic arc: growth, differentiation and exhumation of
the Amalaoulaou complex (Gourma, Mali). Contrib. Miner. Petrol. 162 (4), 773–
796.

Black, R., 1967. Sur l’ordonnance des chaînes métamorphiques en Afrique
occidentale. Chronicle of Mines Research, Min. Paris 364, 225–238.

Blay, P.K., 1983. The stratigraphic correlation of the Afram Shales of Ghana West
Africa. J. Afr. Earth Sci. 1 (1), 9–16.

Block, S., Baratoux, L., Zeh, A., Laurent, O., Bruguier, O., Jessell, M., Ailleres, L., Sagna,
R., Parra-Avila, L.A., Bosch, D., 2016. Paleoproterozoic juvenile crust formation
and stabilisation in the south-eastern West African Craton (Ghana); New
insights from U-Pb-Hf zircon data and geochemistry. Precambr. Res. 287, 1–30.

Boger, S., Raetz, M., Giles, D., Etchart, E., Fanning, C., 2005. U-Pb age data from the
Sunsas region of Eastern Bolivia, evidence for the allochthonous origin of the
Paragua Block. Precambr. Res. 139 (3-4), 121–146.

Bouvier, A., Vervoort, J.D., Patchett, P.J., 2008. The Lu-Hf and Sm-Nd isotopic
composition of CHUR: constraints from unequilibrated chondrites and
implications for the bulk composition of terrestrial planets. Earth Planet. Sci.
Lett. 273 (1-2), 48–57.

Bozhko, N.A., 1969. Division et corrélation des dépôts du Précambrien supérieur de
la plateforme africaine. Traduit de ‘’Vestnik Moskoskogo Universiteta. Geologiia
i Rasviedka 5, 21–26.

Caby, R., 2003. Terrane assembly and geodynamic evolution of central-western
Hoggar: a synthesis. J. Afr. Earth Sci. 37 (3-4), 133–159.

Caby, R., 2014. Nature and evolution of Neoproterozoic ocean-continent transition:
Evidence from the passive margin of the West African craton in NE Mali. J. Afr.
Earth Sci. 91, 1–11.

Carney, J.N., Jordan, C.J., Thomas, C.W., Condon, D.J., Kemp, S.J., Duodo, J.A., 2010.
Lithostratigraphy, sedimentation and evolution of the Volta Basin in Ghana.
Precambr. Res. 183 (4), 701–724.

Clauer, N., 1976. Isotopic geochemistry of strontium in sedimentary environments.
Application to the geochronology of the cover of the West African craton.
Persée-Portal of scientific j. SHS 45, (1).

Clauer, N., Caby, R., Jeannette, D., Trompette, R., 1982. Geochronology of
sedimentary and metasedimentary Precambrian rocks of the West African
craton. Precambr. Res. 18 (1-2), 53–71.

Condie, K.C., 1993. Chemical composition and evolution of the upper continental
crust: contrasting results from surface samples and shales. Chem. Geol. 104 (1-
4), 1–37.

Condie, K.C., Belousova, E., Griffin, W.L., Sircombe, K.N., 2009. Granitoid events in
space and time: constraints from igneous and detrital zircon age spectra.
Gondwana Res. 15 (3-4), 228–242.

Cordani, U.G., Teixeira, W., 2007. Proterozoic accretionary belts in the Amazonian
Craton. Geol. Soc. Am. Mem. 200, 297–320.

Cordani, U.G., Pimentel, M.M., Ganade De Araujo, C.E., Basei, M.A.S., Fuck, R.A.,
Girardi, V.A.V., 2013. Was there an Ediacaran Clymene ocean in central South
America? Am. J. Sci. 313 (6), 517–539.

Cordani, U.G., Teixeira, W., D’Agrella-Filho, M.S., Trindade, R.I., 2009. The position of
the Amazonian Craton in supercontinents. Gondwana Res. 15 (3-4), 396–407.

Cordani, U.G., D’Agrella-Filho, M.S., Brito-Neves, B.B., Trindade, R.I.F., 2003. Tearing
up Rodinia: the Neoproterozoic palaeogeography of South American cratonic
fragments. Terra Nova 15 (5), 350–359.

Corfu, F., Hanchar, J.M., Hoskin, P.W., Kinny, P., 2003. Atlas of zircon textures. Rev.
Miner. Geochem. 53 (1), 469–500.

Cullers, R.L., 1994. The controls on the major and trace element variation of shales,
siltstones, and sandstones of Pennsylvanian-Permian age from uplifted
continental blocks in Colorado to platform sediment in Kansas, USA. Geochim.
Cosmochim. Acta 58 (22), 4955–4972.

Cullers, R.L., Podkovyrov, V.N., 2000. Geochemistry of the Mesoproterozoic
Lakhanda shales in southeastern Yakutia, Russia: implications for
mineralogical and provenance control, and recycling. Precambr. Res. 104 (1–
2), 77–93.

Delor, C., de Roever, E.W.F., Lafon, J.M., 2003. The Bakhuis ultrahigh-temperature
granulite belt (Suriname): Transamazonian crustal stretching in the revised
Guiana Shield framework. Geol. Fr. 2, 207–230.

Deynoux, M., Affaton, P., Trompette, R., Villeneuve, M., 2006. Pan-African tectonic
evolution and glacial events registered in Neoproterozoic to Cambrian cratonic
and foreland basins of West Africa. J. Afr. Earth Sci. 46 (5), 397–426.

Dickinson, W.R., Beard, L.S., Brakenridge, G.R., Erjavec, J.L., Ferguson, R.C., Inman, K.
F., Knepp, R.A., Lindberg, F.A., Ryberg, P.T., 1983. Provenance of North American
Phanerozoic sandstones in relation to tectonic setting. Geol. Soc. Am. Bull. 94
(2), 222. https://doi.org/10.1130/0016-7606(1983)94<222:PONAPS>2.0.
CO;210.1130/8302.

Duclaux, G., Ménot, R.P., Guillot, S., Agbossoumondé, Y., Hilairet, N., 2006. The mafic
layered complex of the Kabyé massif (north Togo and north Bénin): Evidence of
a Pan-African granulitic continental arc root. Precambr. res. 151 (1–2), 101–118.

Egal, E., Thiéblemont, D., Lahondère, D., Guerrot, C., Costea, C.A., Iliescu, D., Delor, C.,
Goujou, J.-C., Lafon, J.M., Tegyey, M., Diaby, S., Kolié, P., 2002. Late Eburnean

http://refhub.elsevier.com/S1342-937X(22)00131-9/h0010
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0010
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0015
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0015
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0015
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0020
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0020
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0020
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0025
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0025
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0030
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0030
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0030
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0035
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0035
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0035
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0035
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0040
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0040
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0040
https://doi.org/10.1007/978-3-642-84153-8_6
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0050
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0050
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0050
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0060
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0060
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0060
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0070
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0070
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0070
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0070
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0075
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0075
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0075
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0075
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0080
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0080
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0080
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0085
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0085
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0085
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0090
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0090
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0090
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0095
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0095
http://refhub.elsevier.com/S1342-937X(22)00131-9/h0095
http://refhub.elsevier.com/S1342-937X(22)