See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/231828368

Stratigraphy and morphotectonics of Karoo deposits of the northern Selous

Basin, Tanzania

Article in Geological Magazine · July 1991

DOI: 10.1017/S0016756800017593

CITATIONS READS

35 1,598

2 authors, including:

Helmut . (Heli) Wopfner

University of Cologne
60 PUBLICATIONS 2,089 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

kAROO BASINS OF eAST aFRICA View project

All content following this page was uploaded by Helmut . (Heli) Wopfner on 11 August 2017.

The user has requested enhancement of the downloaded file.

 Geol. Mag. 128 (4), 1991. pp. 319-334. Printed in Great Britain 319

Stratigraphy and morphotectonics of Karoo deposits of the

northern Selous Basin, Tanzania
H. WOPFNER* & C. Z. KAAYAt
* Geologisches Institut, Universitat zu Koln, Zulpicherstr. 49. D 5000 Koln 1, Germany
f Geology Department, University of Dar es Salaam, P.O. Box 35052, Dar es Salaam, Tanzania

(Received 11 September 1990; accepted 5 February 1991)

Abstract - Late Permian Karoo deposits of the northern Selous Basin in south-central Tanzania
comprise conglomerates and diamictitic boulder beds of alluvial and scarp-foot fan origin. These
merge with grey to greenish sandstones, siltstones and black shales of deltaic and lacustrine
environments. Microflora assemblages indicate a late Permian age. Lateral changes and interfingering
of various lithofacies units are common. Depositional development was controlled by syndepositional
faulting and variations of gradients resulting from fault movements. The position of the fault scarp
separating the basin area from the elevated basement horst to the west roughly corresponded with the
present boundary between the Selous Basin and the Precambrian metamorphics of the Uluguru
Mountains.
The late Permian Karoo succession of this part of the Selous Basin apparently overlaps older
Karoo deposits contained in the north-northeast trending narrow graben structures. It is therefore
regarded as a new depositional event which was initiated by renewed tensional tectonism in late
Permian time. During this tectonic episode the narrow early Karoo graben structures were expanded
into much broader rift basins. Material eroded from the rift shoulders and associated highlands was
literally dumped across the fault scarps, forming debris aprons and scarp-foot fans. Rivers emanating
from the highlands formed large alluvial fans and, further afield, deltas issued into freshwater lakes.
Some of these late Permian faults were rejuvenated by late Cretaceous to early Tertiary tectonism.
Thermal waters circulating along fractures converted feldspars, biotites and hornblendes to prehnite.
Further tectonic adjustments in mid Tertiary time led to the present-day morphology.

normal faults, but occasionally unconformable con-

1. Introduction
tacts between Precambrian metamorphics and Karoo
About 150 km west of Dar es Salaam, the monotonous sediments are exposed.
undulating coastal region is interrupted by the The area of the Selous Basin was first studied by
Uluguru Mountains, rising to a maximum height of Bornhardt (1900), who described the succession
2400 m above sea level. The mountains are composed transected by the Rufiji River and exposed around
of meta-igneous and meta-sedimentary rocks con- Hatambulo Hill (also spelled Nantumbula). More
sisting of granulites, graphitic gneisses, meta-anortho- recent workers include Spence (1957), who produced
sites, charnockites, marbles and pegmatites of the the first geological map of the area; Kreuser (1983),
middle Precambrian Usagaran System (Sampson, who investigated the area north of Rufiji River and
1958). A belt of marble is prominently exposed along Hankel (1987), who reappraised the succession south
the eastern margin of the mountain range, in some of Rufiji River. Certain discrepancies between strati-
areas weathering to bizarre karst topography. The graphic schemes proposed by these authors have
ranges are abutted by late Palaeozoic to early prompted us to re-investigate key sections of the
Mesozoic Karoo sediments and younger sedimentary northwestern Selous Basin. In addition borehole
rocks (Kent, Hunt & Johnstone, 1971; Fig. la, b). profiles from the proposed dam site at Stieglers
The juvenile morphology of the Uluguru Mountains Gorge, held in the internal files of RUBADA, and
combined with high level remnants and knick-points core material stored at Stieglers Gorge Camp were
of old surfaces attest to young tectonic uplift along examined. Observations made by the senior author
north-northeast trending faults. Occasionally these and T. Kreuser during several days offieldworkin
faults are transected by northwest striking dislocation 1981 are also incorporated into this paper.
planes (Fig. 1). The Uluguru Mountains thus form a
horst block which separates the Karoo deposits of the
Selous Basin on the east from those of the Mikumi
2. Basin definition
Basin to the west. The contacts between the Karoo
deposits of these basins and the Precambrian rocks of The nomenclature of Tanzanian sedimentary basins
the Uluguru Mountains are generally formed by still lacks proper definition, particularly since little
22-2

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 320 H. WOPFNER & C. Z. KAAYA

L E G EN D

Cainozoic

Jurassic &
Cretaceous

Karoo
X X Proterozoic
X
X
X (Usagaran&
Ubandian)
Precambrian
granites and
migmatites
50 100 150km

i36°

•7°

5>VNy akatitu

Gr est Ru ah a

Usagaran
Basement

Figure 1. For legend see facing page.

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 Karoo deposits of Selous Basin, Tanzania 321

attention has been given to the temporal changes of exclusion of post-Karoo strata (Fig. 1 a). The western
basin dynamics caused by variations of regional stress limit of the Selous Basin is defined by the fault contact
patterns. This has led to a rather ambiguous ter- with - or by the disconformable overlap of the
minology, the area under discussion being no ex- sediments onto - the Precambrian rocks of the
ception. Usagaran System. We include the Karoo deposits
The term 'Selous Basin' is currently used by the between Mgeta and Ruvu rivers previously referred to
Tanzanian Petroleum Development Corporation as Mvuha Basin by Spence (1957), within the Selous
(TPDC). Spence (1957) introduced the name Mvuha Basin and refer to this part as Mvuha Lobe, using the
Basin for the small lobe of Karoo exposure between term purely in a geometrical sense of the present
Mgeta and Ruvu rivers but left the Selous Basin outcrop area. Towards the east the Karoo deposits of
unnamed. Hankel (1987), following the practice of the Selous Basin lie beneath the younger sedimentary
Uranerzbergbau (Kogler, Bianconi & Biittner, 1983), blanket of the post-Karoo 'coastal basins' (Fig. la,
used the term Luwegu Basin. He extended the basin b).
boundary south into the headwaters of the Luwegu
and Mbarangandu rivers, but without defining the
southern basin limits. He included both Karoo and 3. Stratigraphy
post-Karoo deposits within the one basin name.
Kreuser (1983) and Kreuser et al. (1990) referred to The stratigraphy of the area was first described by
the northern part of the Selous Basin as Rufiji Basin. Bornhardt (1900), who roughly mapped the dis-
The same term was also used by Haldemann (1962) tribution of the Karoo rocks and measured a 680 m
and is used by the Rufiji Basin Development Authority section between the Rufiji River and Hatambulo Hill.
(RUBADA) to identify the present drainage network Subsequent accounts were published by Janensch
of the Rufiji River. To make matters worse, all these (1927), Stockley (1936), Spence (1957), Kreuser (1983)
and other Karoo basins not mentioned here have been and Hankel (1987). Most of these workers accepted
included within the overall term of 'coastal basins of the original rock units established by Bornhardt with
Tanzania' (Kent, Hunt & Johnstone, 1971), but only minor modifications (Fig. 2). Hankel (1987)
without giving any specific definition how far inland attempted to formalize these units according to
the coastal region extended. modern nomenclature procedures but was handi-
According to Wopfner (1972), a sedimentary basin capped by the lack of a base to the classical section of
is defined by its geographic limits and the duration Bornhardt. He placed the total Karoo succession of
over which basin dynamics were controlled by a the Rufiji area within two formations, the Hatambulo
uniform, genetically related set of tectonic forces. and the Rufiji Formation. He subdivided the former
Obviously, the stresses controlling an intracratonic in ascending order into four members: Kidahi,
rift basin are quite different from those acting in a Pangani, Sumbadzi and Ruaha. The relationship of
pericratonic basin. Basins resulting from such differing his units to those of Bornhardt (1900) and Spence
controlling mechanisms should not, in our opinion, be (1957) is shown in Figure 2.
combined under a common name. The importance of As mentioned above, the classical section described
differentiation between the various stages during the by Bornhardt (1900) has no clearly defined lower
transformation of an intracratonic rift to the peri- boundary as the base of the Kidahi Member is not
cratonic situation of a passive margin has been exposed. However, maroon basal conglomerates
demonstrated by Falvey (1974). Applying these resting uncomformably on gneiss were recorded by
principles to Tanzanian sedimentary basins, we restrict Janensch (1927) from the upper reaches of Sumbadzi
the term 'Karoo' to the depositional event which River. Beds of similar lithology exposed along
commenced with the late Palaeozoic glaciation and Behobeho River and abutting the basement near the
which was terminated by the middle-late Triassic to headwaters of that river were termed Viransi Beds by
pre-middle Jurassic unconformity. A similar division Spence (1957). Hankel rejected the term Viransi Beds
has been practised by Spence (1957). The importance as a legitimate unit. Our own investigations and
of the middle to late Triassic event, apparently having reappraisal of published literature indicate that the
a global significance, has recently been pointed out by Viransi Beds, in the sense of Spence (1957), form the
Veevers (1989). base of the Karoo succession of the northwestern
In this paper we use the term Selous Basin pertaining Selous Basin and that they are probably equivalents of
strictly to the receptacle of Karoo deposits and to the the Kidahi Member.

Figure 1. (a) Generalized geological map of southern Tanzania showing present outcrop limits of Karoo basins related to
Precambrian basement provinces and post-Karoo sediment cover. The area of part (b) is outlined in the upper centre of the
map. (b) Generalized geological map showing relationship between Uluguru Mountains (Uluguru Horst) and adjoining Karoo
basins.

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 322 H. WOPFNER & C. Z. KAAYA

BORNHARDT, 1900 SPENCE 19SS; 19S7 HANKEL, 1987 THIS PAPER

SCHIC HTEN itH iiltill

Stufe a Bambawe Beds z
Ruaha Member
z
Ruaha Member

o o

FO R M A T
FO R M A T
Stufe b Mkulinzo Beds Sumbadzi Member Sumbadzi

Pangani Member <-jS~

Pangani
HATA MBULO

Stufe c <C7 Pangani

H A T A M E ULO
Member

H A T A M B ULO
Sandstone
^ ^ Member

Stufe d Hame Beds

v. f^~> Kidahi
Viransi -C____^
Kidahi
Stufe e Kidahi Beds
Member Member ^ Member

Viransi Beds
No base
Basement No base Basement \ No base

Figure 2. Stratigraphic terminology of northern Selous Basin. The age of the succession is discussed in the text.

olive green, highly micaceous, usually biotitic sand-

3.a. Hatambulo-Rufiji area
stones. Plant debris and indistinct bioturbations are
The Karoo deposits of this area are generally flat- common. The siltstones are generally ripple laminated,
lying or broadly warped, with structural dips rarely scalloped ripples being the most common form.
exceeding 15°. The main structural feature is the Frequently, the siltstones have undergone convolute
Hatambulo syncline, a shallow, southerly plunging deformations of various intensity, particularly near
structure, the axial core of which is occupied in the the contact with the overlying sandstone unit (Fig. 4).
north by Hatambulo Hill (Spence, 1957). From here The third and uppermost lithofacies type of a cycle
the axis runs south-southwesterly to a position just consists of light grey sandstones. Near the base, they
west of the confluence of Sumbadzi and Rufiji rivers. are generally medium grained and ill-sorted with high
The area has been affected by a pronounced northwest matrix content, but become coarser and better sorted
trending joint pattern of post-Karoo origin, but this higher up. They exhibit either sigmoidal current
has little influence on the overall distribution of the bedding or large scale trough-type structures. The
Karoo strata. latter are usually associated with lenses of granule
conglomerates.
The framework of the sandstones consists of 85 %
3.a. 1. Kidahi Member
quartz and 7-15% feldspars. The grains are angular
The oldest rocks (units e and d) of the classical to subrounded, those of feldspars are partly sericitized
Bornhardt section (see Fig. 2) are exposed at Kidahi and corroded. The rest is made up of reworked rock
Crossing, some 9 km downstream (east) of Stieglers fragments and mica. Chlorite, which imparts the
Gorge (Fig. 1 b). The section is located on the eastern green colour of thefinegrained sediments, is essentially
limb of Hatambulo Syncline and the beds dip westerly of early diagenetic origin, forming after biotite and
at about 18°. This unit, which is now referred to as the from the remobilization of the detrital clay. The
Kidahi Member, is at least 300 m thick (Spence, 1957; upward coarsening cycles are commonly 8-15 m thick
Hankel, 1987) but total thickness is not known as the with sandstone thicknesses varying between 3 and
lower boundary of this unit is not exposed. 8 m. Foreset azimuths from 15 measurements range
The Kidahi Member consists generally of between 260° and 340° with a maximum towards the
coarsening-upward cycles of olive green silty shales west.
and siltstones to coarse and gritty sandstones. A We interprete the succession of the Kidahi Member
typical cycle is depicted in Figure 3. It commences at as deposits of delta front and distributary channel
the base with dark grey to olive green parallel bedded environments of a fluvial dominated homopycnal
silty shale with wedge shaped intercalations of fine to delta. This interpretation is supported by the presence
very fine grained calcareous sandstones. The sand- of convolute bedding and by the specific grain size
stones are up to 15 cm thick, well sorted and without variations within the individual cycles (see Fig. 3).
any internal structure. Locally this interbedded unit Convolute structures as depicted in Figure 4 are
changes laterally into chaotic, convolute masses of typical for point bar deposits, but they are also
shale and sandstone. The succeeding unit consists of prevalent in channel fill environments (Coleman &

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 Karoo deposits of Selous Basin, Tanzania 323

K I D A H I CROSSING
Bed L it h oIogy
siltstone,green
10 micaceous and
shaley sandstone
sandstone, medium-fine

10- sandstone, light grey,

coarse to gritty, feld -
spathic, current
bedded

sandstone, light grey,

medium grained
sandstone,grey, fine
grained,parallel or
festoon bedding
siltsfone,olive green,
micaceous, laminated
with lenses of brown,
fine grained sandstone
sandstone.dirty, f ine
v grained, micaceous
siltstone, olive green,
micaceous & biqtitic,
convolute bedding
near top
shale, silty.micaceous
greenish grey.inter-
bedded with brown,
fine, calcareous
sandstone, parallel
bedding, laterally
thixotropically
deformed
sandstone, medium
feldspathic & kaolin

Figure 3. Examples of coarsening-upward cycle of lower Kidahi Member at Kidahi crossing. Fining near top of cycle is a
characteristic feature of distributary channel sandstones.

Gagliano, 1970; Coleman & Prior, 1982). The current observing the 90 m sheer drop of Stieglers Gorge is
bedded sandstones (beds 7 and 8 in Fig. 3) are that of a massive sandstone unit, closer scrutiny of the
diagnostic features of distributary channels (Allen, exposures reveals a number of intercalations of green,
1970) and the upward coarsening fabric of the micaceous, very fine grained sandstones to siltstones
sandstones, followed by a diminution of grain size in 7-35 m thick beds. This alternation is clearly
near the top, is typical for bar-finger sands (Fisk, displayed in Figure 5, which shows a cross-section
1961). based on core logs at one of the proposed RUBADA
dam sites. In outcrop, the sandstones are coarse
grained to gritty, forming individual banks with
3.a.2. Pangani Member thicknesses of 80 cm to 2 m. Generally they are planar
The Pangani Member, which corresponds to unit c of in shape. Internally the sandstones exhibit planar or
Bornhardt (1900), rests conformably on the Kidahi sigmoidal current-foresets and troughs of up to 1 m in
Member. Its base is defined by a thick sandstone width. Some of the erosive bases of troughs are filled
exposed at the foot of the escarpment leading up to with thin layers of granules of up to 2 cm in diameter.
Mtundusi Hill. According to Hankel (1987) the West of Stieglers Gorge Camp at the confluence of
Pangani Member is at least 180 m thick. The best Zema and Rufiji rivers, where the sandstone beds of
exposures of the Pangani Member are at Stieglers the lower Pangani Member are well exposed and
Gorge, where the Rufiji River has cut a 90 m chasm accessible, foreset azimuths from 32 measurements
into the flat-lying rocks. Whilst the first impression on are concentrated around 160° and around 50°.

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 324 H. WOPFNER & C. Z. KAAYA

Figure 4. Convolute bedding in olive green, micaceous siltstone of Kidahi Member. Lens cap on right of the picture measures
58 mm; the vertical diameter of the structure at the right of the picture measures 175 mm. Note erosive truncation of siltstone
by overlying sandstone.

NW SE
cot STIEGLERS GORGE

— 100

— 50

50 100
siltstone & mudstone Pangani Formation
scale in metres

Figure 5. Section of Pangani Member across Stieglers Gorge, showing inter-bedding of coarse grained sandstones with dark
mudstones and siltstones. The latter are lithologically identical to siltstones of Sumbadzi Member. This section has been
interpreted from diamond core profiles of RUBADA. The logs of CD4, CD2, and DH5 are projected into the plane of section.

Variations are quite rapid with changes from 200° in grain supported with up to 50% feldspars and 10%
a lower bed to 42° in the erosively incised set rock fragments mostly derived from the basement.
immediately above. Quartz and feldspar overgrowth constitutes the bulk
The sandstone is composed of subangular to of the cement. In some horizons there is an abundance
rounded quartz, subangular feldspar and rock frag- of coalified tree trunks with diameters of up to 20 cm
ments. It contains little matrix. Sorting is fair but the and length of up to 2 m. The stems show a regular
general grain size spectrum is broad, ranging from 0.5 cross-segmentation which is typical for Equisetaceae.
to 6 mm, the largest grains being feldspar. Under the The interbedded greenish grey to olive siltstones
microscope the sandstones are texturally immature, and mudstones vary in thickness from thin slivers to

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 Karoo deposits of Selous Basin, Tanzania 325

about 100°-120° are diagnostic of distributary mouth

bar and beach ridge barrier deposits, interfingering
with proximal pro-delta sediments (Allen, 1970). Some
of the beds may also be attributed to Gilbert-type
delta mechanisms (Fouch & Dean, 1982). •

3.a.3. Sumbadzi Member

This member corresponds in part to unit b of
Bornhardt (1900), comprising the succession of fine
grained micaceous sediments between the uppermost
sandstone bank of Pangani Member and the
sandstone-mudstone section of unit a, exposed at
Hatambulo Hill. However, to the west of Rufiji River,
this interval is dominated by black silty shale and
micaceous siltstones which according to Hankel (1987)
reach a thickness of at least 200 m. A section composed
of several fining upwards cycles, forming a transition
between the black siltstone facies and the basal
Hatambulo profile has been described by Kreuser
(1983) from the upper reaches of Zema River.
The exposures along the middle and upper reaches
of Sumbadzi River are of particular significance for
the understanding of the lithofacies relationships.
About 13-15 km upstream from its confluence with
the Rufiji River, the beds dip 10° to the south. At the
crossing of the old motor track, about 12 m of black,
micromicaceous fissile shale and dark grey, ripple
laminated silty mudstones are exposed. Spence (1957)
recorded thin coal seams from this section. Further to
the west these fine grained sediments become increas-
ingly interspersed with green calcareous sandstones,
forming 2-4 m thick banks (Fig. 6a). The sandstones
are fine grained to conglomeratic, some of them
containing cobble size components. The intervening
intervals consist of flaser bedded, grey, micaceous
siltstones with abundant plant debris, interbedded
with ripple laminated, fine to very fine grained
sandstones (Fig. 6b). The current directions of the
conglomeratic sandstones are generally towards the
south and the azimuths of the current ripple marks
Figure 6. (a) The planar upper surface of a thick coarse to
medium grained sandstone of the lower Sumbadzi Member vary from 120° to 350°.
is overlain on the left-hand side of the picture by thinly The mineral composition of the sandstones is almost
parallel bedded and ripple laminated black siltstones. The identical with those of the Pangani Member, except
dry bed of the Sumbadzi River is seen in the background, (b) for an increase in matrix content to about 10% and
Ripple marks and black ripple laminated silty shale of lower the presence of diagenetic chlorite, some kaolinite and
Sumbadzi Member in Sumbadzi River. calcite cement. The sediments in this section are
interpreted as lacustrine deposits interfingering with
alluvial fan deposits (cf. Fouch & Dean, 1982).
10 m beds. They are micaceous and sometimes Towards the base of this succession coarse grained
calcareous, may show parallel laminations, ripple sediments become dominant, and near the headwaters
marks and convolute and diapiric deformations. They of Sumbadzi River green and maroon coloured
contain some trace fossils and small pieces of fossil conglomerates rest unconformably on basement
plants. (Janensch, 1927). These basal rudaceous beds, in part
Hankel (1987) suggested a delta plain environment described as 'reddish conglomerates', were correlated
for the Pangani Member. In our opinion, the general by Spence (1957) with his Viransi Beds exposed at the
geometry of individual sandstone beds and the rapidly headwaters of Behobeho River, northwest of
changing current directions with a divergence of Hatambulo Hill (Fig. 1 b).

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 326 H. WOPFNER & C. Z. KAAYA

HATAMBULO HILL L ith o1 og y

sandstone,
350- yellowish brown,
• : • . • • . • • • : • . • • . • • • • . • . • . • • • ^ /
fine to medium
— — .—__—- [
•'• • '.•'•'•• • • ' • : • ' • ' • ' • . ' • • ' • ; • ' . - • ' • • ' : J j ) sandstone: v.fine- medium
pinkish-grey, cur rent
lineations,parallel &
300-
_1 777_1 771_1
wavy bedding,
convoluted, f eldspathic.
micaceous,
r7Z-7-ZJ—r-
250- siltstones & mudstones:
_^_7T__ 77" _^_ 7 massive,grey
-•.'•*' • ' ' ' . ' •'• ' 1

siltstone/mudstone
200-
Z~rZ"— -Z1 grey-green
micaceous with
sandy streaks
Figure 7. Hatambulo hill viewed from south, exposing — .. — .. /sandstone, medium
about 360 m of Ruaha Member. 150 : • ' . • : • : • . • : • : • . - . -j grey,ripple marks and
\wavy bedding
3.a.4. Ruaha Member
grey silty
The Ruaha Member is equated by Hankel (1987) to mudstone
unit a of Bornhardt (1900). This is the youngest 100-
lithostratigraphic unit exposed in the area north of
•Jy.^'-.-M-^-A calcareous
Rufiji River. It rests conformably on green siltstones sandstones,
of the Sumbadzi Member and builds up Hatambulo z-z—z(
::X:-::-S;-:-:Zi.-::./.-.--.:J-:l
grey, fine
Hill (Figs, 7, 8). Due to the southerly plunge of the massive to current
50 bedded and
Hatambulo Syncline, the same beds are exposed again siltstones, grey-green
south of the Ruaha River (Spence, 1957) where they 77~__~7 ^\ -volcanic
are overlain by the Rufiji Formation (Hankel, 1987). breccias
..' .. ..
At Hatambulo Hill, the Ruaha Member commences calc. nodules
with some 110 m of olive green, parallel laminated,
Q ^- '• ~-- .'©.' •'
3
0m- carbonaceous siltstone
micaceous siltstones and silty, occasionally carbon-
aceous mudstones, alternating with greenish to Figure 8. Columnar section of Ruaha Member exposed at
brownish grey micaceous sandstones with plant Hatambulo Hill.
remains. The next 150 m consist predominantly of
greenish grey mudstones and siltstones and few beds partially sericitized and kaolinized, biotite is partially
of brownish red to yellowish grey, very fine to fine altered to chlorite and hematite which together with
grained sandstones. The overlying uppermost 110 m quartz form the bulk of the cement.
of Hatambulo Hill are composed of a fairly regularly The immature composition of the sandstones
alternating sequence of boldly weathering, greenish indicates the proximity of an active source area and
grey to tan sandstones and grey micaceous siltstones. the internal structures like current lineations and
The sandstones which have bed thicknesses between 8 parallel bedding suggest deposition within the upper
and 15 m are fine to medium grained, grey and level of plane bed transport energy. The succession is
generally exhibit a pinkish mottling caused by cal- interpreted as deposits of a prograding upper delta
careous nodules of 2-4 mm diameter. They are thinly plain to alluvial floodplain. The green micaceous silt-
parallel bedded with current lineations and partings and mudstones appear to be overbank deposits and
on the bedding planes. Rarely, low angle foreset beds the intervening calcareous sandstones represent la-
are observed. Near the base, occasional convolute custrine delta fill as described by Coleman & Prior
structures are present. (1982).
Under the microscope, the grains are angular to
subangular, moderately to well sorted with 10-15%
3.a.5. Viransi Member
matrix. The framework consists of 50-80% quartz,
20-50% feldspar, up to 5 % rock fragments, 1-4% A section along Behobeho River, between the base-
micas and 1-5% heavy minerals. Feldspars are ment contact and the road from Kisaki to Stieglers

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 Karoo deposits of Selous Basin, Tanzania 327

Gorge (Fig. 1 b), exposes about 200 m of green and and pink, coarse grained to conglomeratic sandstones
reddish brown, current bedded, pebble and cobble near Kisaki road. It makes up 5-7% of the rock
conglomerates with intercalations of greenish, very volume, occuring as a fibrous to platy mineral in pore
coarse to medium grained current bedded calcareous spaces and replacing feldspar (plagioclase), horn-
sandstones and green-grey to pink coarse grained blende and quartz. It is in turn corroded by calcite and
feldspathic sandstones. This succession roughly cor- in some horizons calcite appears to replace it almost
responds with the type section of the Viransi Beds as completely or inhibit its formation. Prehnite belongs
proposed by Spence (1957). We suggest that this to the prehnite-pumpellyite suite of low grade
stratigraphic unit be given member status and be metamorphism (Liou, 1971). Its occurrence in the
incorporated into the Hatambulo Formation as shown generally unmetamorphosed surrounding indicates
in Figure 2. localized heating and metasomatism related to the
The section at Behobeho River commences with small volcanic bodies in the area which have been
maroon coloured conglomerates embedded in a described by Spence (1957), Kreuser (1983) and
micaceous and silty matrix and fine grained micaceous Hankel (1987). The latter author reported radiometric
sandstones. Grains are angular, poorly sorted with ages obtained from these volcanic pipes, dykes and
occasional high matrix content. The framework is pipe breccias of about 40 Ma. The hot springs
made up of 35-50% feldspar, 35% quartz, 10% emanating along faults near Viransi Hills and else-
basement-derived rock fragments, including mica, where are probably late effects of this tectono-volcanic
and up to 15% heavy minerals which comprise in phase. The significance of this late Cretaceous-early
decreasing order garnet, hornblende, opaque minerals, Tertiary event will be discussed below.
epidote, zircon, kyanite, pyroxenes and rutile.
The abundance of micaceous matrix, the diamictitic 3.b. Mvuha Lobe
texture and the subrounded nature of the clasts (Fig. 3.b.l. Mvuha Beds
9 a) is indicative of alluvial fan deposits, emplaced by
high suspension flash floods or debris flow (Nilsen, The Mvuha Beds, first named by Spence (1957), form
1982). The contact with the underlying Usagaran the westernmost Karoo outcrops of the Mvuha Lobe
basement is faulted, but the similarity of the lithology (see Fig. 1). They are exposed north and south of
with the basal deposits near the headwaters of Mvuha Village, where they are in fault contact with
Sumbadzi River (Janensch, 1927) suggests that the marbles and schists of the Precambrian complex of
actual depositional contact was not too far from the the Uluguru Mountains. Kreuser (1983) estimated the
present position of the fault. thickness of the total, generally west-dipping, Karoo
About 48 m above the fault contact, the maroon succession of the Mvuha Lobe at about 1370 m. His
coloured conglomerates are succeeded by pale green uppermost 'unit V corresponds roughly with the
pebble conglomerates and light grey, green and pink, Mvuha Beds of Spence (1957).
coarse grained feldspathic sandstones with calcareous A fairly comprehensive section of the Mvuha Beds
sandy nodules and horizons of cobble conglomerate is exposed between a point on the road 6.3 km south
(Fig. 9 b, c). The feldspathic sandstones show cut-and- of Mvuha Village and Bambalawe trigonometric
fill structures, trough and sigmoidal current bedding point, where the Karoo rocks dip with 20°-30° west
with graded asymptotic foresets which are usually towards the fault contact with the basement. The
truncated on top by a layer of cobbles (Fig. 9d). The succession commences with reddish arkoses and
foreset azimuths vary between 70° and 115°. This part purple, fine grained calcareous sandstones and lamin-
of the section, extending to Kisaki road, is composed ated siltstones followed by about 40 m of grey boulder
of lithofacies which are suggestive of outer fan to conglomerate. The clasts consist mainly of marble
braided stream deposits (Nilsen, 1982). Higher up in derived from the basement immediately underlying
the section, fine grained intercalations become more the Mvuha Beds. The marble clasts are also the largest
frequent although occasional cobble layers are still components with diameters of up to 30 cm. Basement
present. The sandstones are very calcareous and cobbles reach up to 10 cm in diameter and quartz
contain large calcareous concretions (Kreuser, 1983). pebbles range between 1 and 3 cm. All components
These beds gradually lead up to sandstone-siltstone are well rounded. This is followed by a reddish,
interbeds which may be facies equivalents of the arkosic pebble conglomerate which in turn is overlain
Pangani Member. by another thick cobble to boulder conglomerate with
Diagenetic minerals include hematite forming after well rounded clasts of up to 30 cm diameter.
biotite and, filling the pore spaces, chlorite, sericite, The upper portion is exposed further south and
micro-crystalline quartz and calcite. Calcite has consists of several cycles, each starting with a cobble
invariably corroded feldspar, quartz and hornblende conglomerate grading upward to sandstones and green
grains. Hematite and matrix content decreases up- biotitic siltstones with calcareous concretions. The
wards and feldspar content increases to about 65%. conglomerates have a sandy matrix consisting of
Of interest is the occurrence of prehnite in the green subangular grains of marble, feldspar, quartz, amphi-

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 328 H. WOPFNER & C. Z. KAAYA

Figure 9. Examples of lithofacies types of Viransi Member exposed along Behobeho River west of the motor track to Stieglers
Gorge, (a) Unsorted diamictitic boulder conglomerate in coarse sandstone matrix, (b) Pale greenish grey medium grained
calcareous sandstone interdigitating with conglomeratic beds, (c) Conglomerate with planar interbeds of coarse grained
sandstone. Note the generally subangular shape of clasts. (d) Upper surface of well sorted single pebble-layer in gritty
feldspathic sandstone.

boles and gneiss fragments. The framework, which others contain nodules possibly of coprolite origin.
becomes dominant in the higher cycles, consists of Frequently the beds are contorted due to loading or
well rounded basement cobbles and subangular marble dewatering. The laminated siltstones also contain
fragments. In both sandstones and conglomerates, the some interbeds of hard silty dolomite whose thickness
rock fragments constitute 25-50 % of the framework, rarely exceeds 20 cm. The proportion of marble clasts
quartz provides 35-45%, feldspar 10-20% and heavy in the total rock composition increases from the lower
minerals, comprising in decreasing order hornblende, to the upper cycle. This progressive dominance of
opaque minerals, garnet, pyroxenes and zircon, make Precambrian marbles reflects syndepositional uplift of
up 2-10%. This assemblage is clearly derived from the Uluguru Horst, gradually exposing the marbles as
the Precambrian metamorphics of the Uluguru Horst. the block moved up.
In the red sediments the iron-rich minerals, in- Spence (1957) interpreted the Mvuha Beds as glacial
cluding biotite, are partially altered to hematite, deposits and placed them at the base of the Karoo
which may occur also as cement though most of the succession. He based his interpretation on the coarse
sediments are cemented by calcite. In some layers the matrix-rich composition of the boulder beds, the
green biotitic siltstones contain up to 20% biotite; freshness of the mineral constituents and the presence

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 Karoo deposits of Selous Basin, Tanzania 329

of laminated siltstones. Kreuser (1983) placed the Luekisporites is the most common, whereas the genera
Mvuha boulder beds at the top of the succession, Striatopodocarpites and Protohaploxipinus are less
which is in keeping with the overall westerly dip of the abundant. The genera Platysaccus and Alisporites are
strata. Our investigations have shown that the clasts the most common representatives of the disacciatrilete
of the boulder beds were shaped by scree movements, forms. Moderately present are also disaccitrilete and
mudflow and water transport. The generally sandy costate forms, with the genera Limitisporites and
matrix also attests to such transport mechanisms, Weylandites respectively. Very rare are apiculate
whereas indications for glacially-shaped components genera, represented by Apiculatisporites and mono-
are lacking. In our view the Mvuha Beds are products pseudosaccate forms like Densipollenites. The micro-
of scarp foot fans laterally interfingering with alluvial flora is quite diverse although not particularly well
fans. Depositional environments recognizable are preserved. Of interest is the moderate abundance of
debris flow, fan head, braided stream and distal fan Lueckisporites, indicating a certain European influ-
deposits as well as sediments laid down in intra-fan ence.
embayments. Lithofacies types identical to the ones The assemblage indicates a late Permian age for the
observed by us in the Mvuha succession have been Hatambulo Formation, probably around Kazanian-
described by Nilsen (1982) from both recent and Tatarian (early Thuringian). Supporting evidence for
ancient alluvial fans. In addition, many sub-Recent or a late Permian age has been provided by the discovery
Quaternary alluvial fans of East African rifts, as for of an early Triassic palynoflora in lateral equivalents
instance in the Rukwa Graben, exhibit the same of the Rufiji Formation (Hankel, 1990). In the Selous
sediment associations and lithofacies types as those Basin this formation overlies the Hatambulo Form-
composing the Mvuha Beds. We regard the Mvuha ation. Accordingly, the Hatambulo Formation would
Beds as equivalents of the Viransi Beds. North of be roughly a time equivalent of the 'Lower Bone
Mvuha Village, small remnants of maroon siltstones Beds' (K6) of the Karoo type section in the Ruhuhu
are exposed on the western, upthrown side of the Basin. In that basin the Karoo succession is subdivided
fault. They are of identical lithology as the basal red into eight informal units, designated in ascending
beds of the Viransi Member, which is additional order as K1-K8. The basal glacigenic unit Kl is of
confirmation for our interpretation. late Carboniferous to early Permian age, whereas the
Permian-Triassic boundary lies somewhere within the
unconformity between K6 and K7 (Kreuser et al.
4. Age and correlation
1990). Spence (1957) correlated the Viransi Beds with
The fossil content of the Hatambulo Formation the lowermost red conglomerates and red mudstones
consists almost entirely of plant material. Janensch in the Mikumi area, and with the 'Red Beds' (K3) of
(1927) recorded Estheriella bornhardti and remnants middle Permian age in the Ruhuhu Basin. However,
of Equisetales, Glossopteris and Sphenopteris from the in the Mikumi area, green conglomerates and current
black siltstones along Sumbadzi River. Glossopteris bedded sandstones, which are lithologically identical
indica and fish scales were recorded by Spence (1957). with the Viransi Member, rest on a succession of grey
None of those fossils are sufficiently diagnostic to sandy dolomites and fossiliferous siltstones. As the
permit an accurate dating. latter may be correlated with the ' Ruhuhu Beds' (K5)
Kreuser (1983) obtained small assemblages of of the Ruhuhu Basin, a correlation of the Hatambulo
Formation with the 'Lower Bone Beds; (K6) of the
palynomorphs from shale interbeds of the Pangani
Ruhuhu Basin appears more logical.
Member and Sumbadzi Member. He recorded
Lueckisporites species which he thought indicate a The stratigraphic position of the Mvuha Beds in
late Permian age. Hankel (1987) obtained a similar relation to other Karoo sediments in Tanzania is still
palynoflora characterized by the presence of Guttul- unclear. Kreuser (1983) reported on the presence of
apollenites hannonicus, which he regarded as di- varved shales with dropstones in the lower part of the
agnostic for a late Permian age. He suggested a succession which he and also Spence (1957) suspected
correlation with the lower Sakamena deposits of although without confirmation, to be of glacial origin
Madagascar. and hence correlatable with the glacial sediments of
Core samples selected from the diamond holes the Idusi Formation (Kl) in the Ruhuhu Basin. The
uppermost Mvuha conglomerates with red beds,
drilled for the dam site investigations have been
however, show no evidence of being of glacial origin.
studied by Dr R. Weiss, who provided the following
report (written comm., 1990). The most abundant
component of the assemblage is the polysaccate form 5. Facies development and palaeogeography
Guttulapollenites hannonicus, the appearance of which
is considered to be of biostratigraphic significance. On Significant lateral changes of the various lithostrati-
the other hand, Guttulapollenites gondwanensis is graphic units as described in the preceeding part of
present only occasionally. The remaining part of the this paper cannot be understood from simple super-
assemblage is dominated by striate forms; the genus position of individual units. In Figure 10 we have

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 330 H. WOPFNER & C. Z. KAAYA

no top
w HATAMBULO
FORMATION Ruaha
Member
m

Sumbadzi Mb.
Sumbadzi
. ' , • • ; ' . . . . ' . ' • • •

Member ^ Pangani
~*\ Member

•0
approximate
vertical •50
beds
scale
Usagaran LiOOm no base
Basement horizontal distance not to scale

conglomerate sandstone siltstone & mudstone

Figure 10. Schematic facies distribution with Hatambulo Formation between the headwaters of Sumbadzi River (west) and
Kidahi crossing (east). For further explanation see text.

compiled a rock relationship diagram, showing the Janensch (1927) recorded from the upper reaches of
simplified facies distribution between the basement Sumbadzi River resting on basement and which
contact in the west and Kidahi crossing in the east. Spence (1957) correlated with the Viransi Beds (now
The eastern column of the rock relationship diagram Viransi Member). This correlation is confirmed by
represents the classical Bornhardt section where the our own observations along Sumbadzi and Behobeho
individual stratigraphic members are fairly clearly rivers. North of Sumbadzi River, the succession of the
defined. Hankel (1987) placed the top of the Kidahi Sumbadzi Member becomes more sandy, consisting
Member at the foot of the breakaway of Kipalapala of fining upward cycles, each cycle generally termin-
Hill where it is sharply overlain by thick sandstones of ating in black or green micaceous siltstones. Kreuser
the Pangani Member. This relationship is confirmed (1983) ascribed these deposits to Bornhardt's (1900)
by the exposures along the old motor track to Kidahi stage b. Their position is shown on the left hand side
crossing published by Kreuser (1983) and also by our of Figure 10.
own observations. The morphotectonic situation which was respon-
To the west of Stieglers Gorge the situation is more sible for the various depositional environments and
complex. At the confluence of Sumbadzi and Rufiji which controlled the facies distribution in the northern
rivers the dark siltstones of the lower part of the Selous Basin during Hatambulo time is shown in a
Sumbadzi Member intertongue with the sandstones of generalized form in the block diagram of Figure 11.
the Pangani Member. This interfingering between The diagram depicts the situation shortly before
sandstones and dark siltstones was also observed by commencement of deposition of the Ruaha Member.
Hankel (1987). It persists with a diminishing siltstone The front panel shows the pronounced asymmetry
content into Stieglers Gorge as demonstrated by the of facies distribution within the Hatambulo Form-
cross-section shown in Figure 5. Bornhardt (1900) ation. The rudaceous deposits of the Viransi Member
correlated the dark siltstones along Sumbadzi River were laid down on alluvial fans emanating from an
with his stage d. About 13 km upstream of the elevated source in the west. This is supported by the
confluence, along Sumbadzi River, the dark ripple easterly and the southeasterly palaeocurrent directions
laminated siltstones are interbedded with coarse and observed within that unit and by the heavy mineral
conglomeratic sandstones. These represent the upper spectrum. These deposits, resulting from flash floods,
parts of the sandstones and conglomerates which mudflows, and high energy braided streams, grade

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 Karoo deposits of Selous Basin, Tanzania 331

U s a g a r a n Sumbadzi Member
Basement

Figure 11. Palaeogeography and interpreted environment of deposition of northern Selous Basin. The palaeocurrent directions
below the north arrow refer to the Viransi Member (VM), the Pangani Member (PM) and to ripple azimuths in the Sumbadzi
Member (a, b, c, d). For further explanation see text.

eastward into low sinuousity stream deposits and west of Zema River. The depositional cycle of the
lacustrine bar deposits nearer the Sumbadzi palaeo- Hatambulo Formation was concluded by the de-
lake. The ripple laminated siltstones and the fine position of the prograding delta facies of the Ruaha
grained sandstones were deposited near the shores of Member. This final event is not shown in Figure 11.
that lake. The current azimuths of the ripple marks in
the Sumbadzi Member (arrows, a, b, c and d in Fig. 6. Tectonics
11) are interpreted to be caused by wind-driven
6.a. Syndepositional rift movements
longshore and landward currents. The black siltstones
and silty shales of the Sumbadzi Member represent Aeromagnetic data (Butterham, Bullock & Hopgood,
the more distal lacustrine sediments. 1983) clearly indicate a substantial deepening of
The environment of deposition of the boulder beds magnetic basement to the east of Stieglers Gorge. The
and the coarse Mvuha conglomerates is shown in the increase in depth far exceeds the thickness of the
upper left of the diagram. These coarse rudaceous sedimentary column observed in the outcrop area.
deposits represent fault scarp debris aprons, laterally The north-northeast trend of the magnetic anomalies
interfingering with prograding, proximal fan head and the mode in which the magnetic intensity decreases
deposits. The westerly source of the clastic material is over the basin area suggests the existence of older
clearly evidenced by the boulder- and heavy mineral graben or half-graben structures at depth. Presumably
composition as described earlier. The eastern part of these are filled by older Karoo rocks equivalent to the
the succession was largely dominated by deltaic early Permian glacigene Idusi Formation (K.1) and the
deposits. The sediment intake from the east which 'Lower Coal Measures' (K2) of the Ruhuhu Basin.
prevailed during the deposition of the Kidahi Member Succeeding red beds as well as carbonaceous and
changed to southerly transport directions as the lacustrine beds apparently overlap the initial graben
Pangani delta came into existence. The situation is structures thus indicating a substantial expansion of
shown in the upper right corner of Figure 11. Existing the original depositional area. Data from seismic
evidence indicates that the delta issued into a reflection surveys and surface observations from the
freshwater body. During high rainfall periods the neighbouring Mikumi Basin confirm this interpret-
river carried a high suspension load creating slightly ation (Fig. 12). These latter deposits are correlated
hyperpycnal flow conditions, leading to rapid down with the middle Permian (about late Artinskian and
slope movements. During dry seasons the flow Early Kungurian) succession of the Ruhuhu Basin.
conditions were homopycnal and this would have The succession comprises the 'Red Beds' (K3), the
caused staged suspension settling due to the absence 'Upper Coal Measure' (K4) and the lacustrine
of electrolytic effects. Thus the Pangani Member inter- 'Ruhuhu Beds' (K5) (Kreuser et al. 1990). The
fingers laterally with the lacustrine black mudstone generally argillaceous or finely arenaceous nature of
facies of the Sumbadzi Member as well as with the the carbonaceous and lacustrine depositional cycles
overbank delta plain facies of the Sumbadzi Member
(K.4 and K5) demonstrate deposition under fairly low

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 332 H. WOPFNER & C. Z. KAAYA

energy conditions. It requires either a low erosional According to Spence (1957) the post-Karoo move-
gradient or a distal position from the sediment source ments were associated with the emplacement of small
or both. dykes, pipe breccias and carbonatite/kimberlite plugs
At the onset of deposition of the Hatambulo in the late Cretaceous. Hankel (1987) obtained
Formation the conditions of sedimentation experi- radiometric ages of around 40 Ma from some of these
enced a drastic change. The unconformable contact of igneous rocks and placed them into the late Eocene.
the Viransi Member with the Usagaran basement The presence of a pronounced bevel on the fault scarp,
demonstrates renewed overstepping and expansion of which may be correlated with laterite-covered mesas
the depositional area. The coarse, generally rudaceous and intramontane surfaces roughly 120 m above the
rocks of that unit were laid down in the immediate present plain level, indicates younger movements
proximity of an elevated crystalline source. The along the 'boundary fault'. As these movements
composition of the clasts and of the heavy mineral postdate the laterite formation they probably occurred
spectrum leaves no doubt that this elevated source in late Miocene time. The course of Behobeho River
consisted of the Usagaran basement of the Uluguru is largely controlled by the young southeasterly
Horst (Figs 11, 12). The palaeocurrent directions of trending fracture system. Between its confluence with
the Viransi Beds at Behobeho River (Fig. 11) also Rufiji River and the Kisaki Village several hot springs
demonstrate that the sediment source was in the west. occur (Spence, 1957), bearing witness to relatively
The horst which separated the Mikumi Basin from the high heat flow in the area.
Selous Basin (Figs 1,12) shed its material directly into Thermal waters circulating along the Tertiary
these adjoining basins. Repeated input of rudaceous fracture system are thought to have been responsible
material, represented by the fining upward or coars- for the formation of prehnite within some of the rocks
ening upward successions, was controlled by repeated of the Hatambulo Formation (see above). These high
movements along active faults. The existence of such heat conditions are well above the oil window.
active fault scarps is further demonstrated by the However, they should not be taken as representative
inverse stratigraphy of boulder composition of the for all Karoo basins. Rocks away from the reactivated
Mvuha Beds described above. The sandy, diamictitic faults may still have potential for the preservation of
fabric and interbedded streaky sand lenses indicate liquid hydrocarbons.
that most of the material was dumped as scree
deposits at the foot of the fault scarp.
7. Conclusions
In the case of the Mvuha Beds, the position of the
late Permian fault scarp was almost identical with the From the observations presented in this paper we
position of the present 'boundary fault'. The overall conclude that the sedimentation of the Hatambulo
reduction of average clast size from north to south Formation was governed essentially by syndeposit-
may have been caused by a slight southerly plunge of ional faulting. This tectonic event was initiated during
the horst. A technically active source for the sediments Kazanian (early Thuringian) time and created the
is also supported by the freshness of the clastic Uluguru Horst as a major source of detrital material,
minerals. The change from coarse to very coarse at the same time causing an extension of a pre-existing
rudaceous deposits in the west to fine grained rift system and the overlapping of the Hatambulo
siliciclastics in the east, as shown in Figure 11, is Formation onto basement. This situation is reflected
symptomatic for a sediment intake from an actively by the pronounced asymmetry of the lithological
rising source in the west. The southerly palaeocurrent distribution as shown in Figures 11 and 12.
directions within the deltaic Pangani Sandstone on the This late Permian tectonic event was but one of
other hand demonstrate the existence of a trunk- several tectonic pulses (Wopfner, in press, a, b) which
drainage in the direction of the graben axis. This is controlled the development of the Selous Basin and
indicated on the right-hand side of Figure 11. other Karoo basins in East Africa. Some of the stages
of this development are depicted schematically in
Figure 12. The first cross-section represents the time
6.b. Post-Karoo faulting
span from the end of early Permian to the termination
The present fault boundaries between Karoo deposits of the deposition of K4 equivalents, roughly in
and the Uluguru Horst apparently resulted from Kungurian (middle Saxonian) times. Strata of that
reactivations of dislocation planes along the same cycle are exposed in the centre of the Mvuha Lobe and
trends as the original rift movements. Concomittant also along the eastern margin of the Mikumi Basin,
with the renewal of the old fault structures a new where they rest directly on metamorphic basement
fracture pattern was established roughly at right (Kreuser, 1983).
angles to the old direction (Spence, 1957). Although The K3-K4 equivalents overlap older graben
some displacements were experienced along these structures as indicated by geophysical data. However,
fractures, their interpretation as major troughs, as the presence of older Karoo sediments in the deep
claimed by Kreuser (1983), seems to be unjustified. graben structures is conjectural as this section was

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 Karoo deposits of Selous Basin, Tanzania 333

The section showing the present situation is largely

W self explanatory. It corresponds to a cross-section
from Mikumi to a point about 10 km south of
U l u g u r u Horst Mvuha. As mentioned above, the present situation
Mikumi Basin Selous Basin
was created by rejuvenation of fault systems originally
established during K a r o o deposition in the late
Permian. These rejuvenations which took place in
older Karoo several stages in late Cretaceous and Tertiary times
about late Artinskian (K3-K4) were related to stress fields of the modern East
African Rift systems.

Acknowledgements. The project was financially supported

by Deutsche Forschungsgemeinschaft (DFG), Gesellschaft
fur Technische Zusammenarbeit (GTZ) and Deutscher
Akademischer Austauschdienst (DAAD). To these organ-
near Kazanian / Tatarian izations we record our sincere appreciations. We also thank
boundary the Tanzanian Research Council (UTAFITI) and the
Tanzanian Wildlife Department for assistance during the
field investigations in the Selous Wildlife Reserve. We
are especially grateful to the Rufiji Basin Development
Authority (RUBADA) for their permission to inspect and
extract selected core samples stored at Stieglers Gorge
Camp and for the use of unpublished information on
drilling results. Last but not least we thank Dr Roseline
present Weiss for the palynological investigation of the samples
collected during our field investigations. This paper benefited
I basement H1PHK3/K4 Late Permian
from critical remarks by Professor C. A. Kogbe, Rockview
International, Paris and Dr G. J. Nichols, Royal Holloway
Figure 12. Morphotectonic development of Uluguru Horst and Bedford New College, Egham, Surrey, U.K.
and adjacent Karoo basins from late early Permian to
Recent time.
References
never reached by drilling. The oldest beds identified by ALLEN, J. R. L. 1970. Sediments of modern Niger Delta: a
drilling so far are the (?) middle Triassic sandstones Summary and Review. Society of Economic Paleont-
encountered near total depth of 1763 m in the ologists and Mineralogists, Special Publications no. 15,
exploration well Liwale-1, drilled by the Shell oil 138-51.
BATTERHAM, P. M., BULLOCK, S. J. & HOPGOOD, D. N.
company some 185 km south-southwest of Stieglers
1983. Tanzania: integrated interpretation of aero-
Gorge (Petracca, 1986).
magnetic and radiometric maps for mineral exploration.
The second profile on Figure 12, labelled 'near Transactions, Institute of Mining and Metallurgy,
Kazanian/Tatarian boundary' (i.e. middle Thuring- London, Section B 92, 83-92.
ian), depicts the situation towards the end of the BORNHARDT, W. 1900. Zur Oberfla'chengeslallung und
deposition of the Hatambulo Formation. It shows Geologie Deutsch-Ostafrikas, 7, Berlin: Dietrich
renewed extension of the Karoo rift system associated Reimer, 595 pp.
COLEMAN, J. M. & GAGLIANO, S. M. 1970. Sedimentary
with a slight westward tilt of the pre-Hatambulo
structures: Mississippi river deltaic plain. In: Primary
strata. It also demonstrates the asymmetry of litho-
sedimentary structures and their hydrodynamic interpre-
facies distribution of the Hatambulo Formation. This tation (ed. G. V. Middleton), pp. 133-48. Society of
tectonic phase was initiated in early Kazanian (early Economic Paleontologists & Mineralogists, Special
Thuringian) time and probably remained active Publication no. 12.
towards the end of this period. These movements COLEMAN, J. M. & PRIOR, D. B. 1982. Deltaic environments
appear to have been the last ones with which rift of deposition. American Association of Petroleum
expansion was associated. The next tectonic stage, Geologists, Memoir 31, 139-78.
FALVEY, D. A. 1974. The development of continental
near the Permo-Triassic boundary, is not depicted in
margins in plate tectonic theory. Australian Petroleum
Figure 12. It bears signs of detachment mechanisms Exploration Association Journal 14 (1), 95-106.
and foundering of the eastern shoulders of the rift FISK, H. N. 1961. Bar-finger sands of the Mississippi Delta.
system. It initiated the transformation of the intra- In: Geometry of sandstone bodies (eds J. A. Peterson
cratonic rift to the pericratonic basin stage (cf. and J.C.Osmond), pp. 25-52. Tulsa: American As-
Wopfner, 1972; Falvey, 1974). The hitherto bilateral sociation of Petroleum Geologists.
rift systems changed and acquired an easterly palaeo- FOUCH, T. D. & DEAN, W. E. 1982. Lacustrine and
associated clastic depositional environments. American
gradient draining to the Malagasy trough, which had
Association of Petroleum Geologists, Memoir 31, 87-
been well established at that time (Wopfner, in press,
114.
a, b).
HALDEMANN, E. G. 1962. The geology of the Rufiji Basin
23 GEO 128

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63

 334 Karoo deposits of Selous Basin, Tanzania

with reference to proposed dam sites. Geological Survey NILSEN, T. H. 1982. Alluvial fan deposits. American As-
of Tanganyika Bulletin no. 33, 77 pp. sociation of Petroleum Geologists, Memoir 31, 49-86.
HANKEL, O. 1987. Lithostratigraphic subdivision of the PETRACCA, A. N. 1986. Oil and gas developments in central
Karoo rocks of the Luwegu Basin, Tanzania and their and southern Africa in 1985. American Association of
biostratigraphic classification based on microflora, Petroleum Geologists Bulletin 70 (10), 1412-57.
fossil wood and vertebrates. Geologische Rundschau 76 SAMPSON, D. N. 1958. The Uluguru Mountains (A summary
(2), 539-66. of the progress to 1955). Records of the Geological
HANKEL, O. 1990. Lower Triassic spores and pollen from a Survey of Tanganyika 5, 21-35.
borehole in the lower Mariakani Formation (Karoo) of SPENCE, J. 1955. Geological Map, 1:125000. Quarter Degree
Kenya. Berliner geowissenschaftliche Abhandlungen (A) Sheet 64 SE, Mkalinzo. Geological Survey of Tangany-
120, (2), 639-78. ika, G.S. 911.
JANENSCH, W. 1927. Beitrag zur Kenntnis der Karru- SPENCE, J. 1957. The geology of the eastern province of
Schichten im ostlichen Deutsch-Ostafrika. In: Wissen- Tanganyika. Geological Survey of Tanganyika Bulletin
schaftliche Ergebnisse der Tendaguru-Expedition, 28, 62 pp.
1909-1912. Palaeontographica, N. F., Supplementum STOCKLEY, G. M. 1936. A further contribution on the
7, 107-42. Karoo rocks of Tanganyika Territory. Quarterly
KENT, P. E., HUNT, J. A. & JOHNSTONE, D. W. 1971. The Journal of the Geological Society of London 92, 1-31.
geology and geophysics of coastal Tanzania. Geophysical VEEVERS, J. J. 1989. Middle/Late Triassic (230±5Ma)
paper No. 6. London: H.M.S.O., 101 pp. singularity in the stratigraphie and magmatic history of
KOGLER, K., BIANCONI, F. & BUTTNER, W. 1983. Geo- Pangean heat anomaly. Geology 17, 784-7.
chemical behaviour of uranium in lateritic profiles in WOPFNER, H. 1972. Depositional history and tectonics of
southern Tanzania. In: Uranium exploration in wet South Australian sedimentary basins. In: Proceedings
tropical environments, pp. 119-35. Vienna: Inter- 4th Symposium, Development of Petroleum Resources
national Atomic Energy Agency. of Asia and the Far East 1 (2), 251-67. ECAFE Mineral
KREUSER, T. 1983. Stratigraphie der Karroo-Becken in Ost- Resources Development Series 41, United Nations,
Tanzania. Geologisches Institut, Universitat Koln, New York.
Sonderveroffentlichung 45, 217 pp. WOPFNER, H. in press, a. Permo-Triassic sedimentary basins
KREUSER, T., WOPFNER, H., KAAYA, C. Z., MARKWORT, S., in Australia and East Africa and their relationship to
SEMKIWA, P. M., & ASLANIDIS, P. 1990. Depositional Gondwanic stress pattern. Proceedings 7th Gondwana
evolution of Permo-Triassic Karoo Basins in Tanzania Congress, Sao Paulo, Brazil, 1988.
with reference to their economic potential. Journal of WOPFNER, H. in press, b. Rifting in Tanzanian Karoo basins
African Earth Sciences 10 (1/2), 151-67. and its economic implications. 15th Colloquium on
Liou, J. G. 1971. Synthesis and stability relations of African Geology, Nancy, 1990. Extended Abstracts,
Prehnite. American Mineralogist 56, 507-31. CIFEG, Service Documentations & Editions, Orleans.

http://journals.cambridge.org Downloaded: 16 Jul 2014 IP address: 132.203.227.63