The Late Palaeozoic Ice Age unconformity in southern Namibia viewed as a patchwork mosaic

Abstract

The expansion of ice masses across southern Africa during the Late Palaeozoic Ice Age has been known for 150 years, including the distribution of upland areas in controlling the configuration of glaciation. In Namibia, increasing attention has focussed on long and deep palaeovalley networks in the Kaokoland region in the north, but comparatively little work has been attempted in the topographically subdued plains of the south, in the Aranos and Karasburg basins. The desert terrain of the Aranos area exposes diamictites of the Dwyka Group discontinuously over about 300 km, extending further south to the Karasburg area at the Namibian-South African border along the Orange River. Whilst examined at a stratigraphic level, the nature of the contact between the Dwyka glacial rocks and underlying lithologies has not been systematically investigated. This paper presents the results from fieldwork in austral winter 2019, in which a highly varying basal contact is described that records the processes of growth, flow and expansion of ice masses across this part of Gondwana. At the basin margins, subglacially-produced unconformities exhibit classic glacially striated pavements on indurated bedrock. In comparison, the basal subglacial unconformity in the more basinward regions is characterised by soft-sediment striated surfaces and deformation. In the Aranos Basin, soft-sediment shear zones originated in the subglacial environment. This type of subglacial unconformity developed over well differentiated, unconsolidated, siliciclastic materials. Where ice advanced over more poorly sorted material or cannibalised pre-existing diamictites, “boulder-pavements” recognized as single clast-thick boulder-dominated intervals formed. Importantly, these boulder-pavements are enriched in clasts, which were facetted and striated in-situ by overriding ice. By integrating measurements of striation orientations, fold vergence and palaeocurrent information, former ice flow pathways can potentially be reconstructed over a wide area.

Full text

Depositional Rec. 2022;8:419–435.
|
419wileyonlinelibrary.com/journal/dep2
Received: 15 January 2021
|
Revised: 13 June 2021
|
Accepted: 16 June 2021
DOI: 10.1002/dep2.163
ORIGINAL ARTICLE
The Late Palaeozoic Ice Age unconformity in southern Namibia
viewed as a patchwork mosaic
Daniel P.Le Heron1
|
ChristophKettler1
|
Neil P.Griffis2
|
PierreDietrich3
|
Isabel P.Montañez2
|
David A.Osleger2
|
AxelHofmann4
|
GuilhemDouillet5
|
RolandMundil6
This is an open access article under the terms of the Creat ive Commo ns Attri bution License, which permits use, distribution and reproduction in any medium, provided
the original work is properly cited.
© 2021 The Authors. The Depositional Record published by John Wiley & Sons Ltd on behalf of International Association of Sedimentologists.
1Institüt für Geologie, Universität Wien,
Vienna, Austria
2Earth and Physical Sciences, University
of California Davis, Davis, California,
USA
3Géosciences Rennes, UMR6118,
Université de Rennes 1, Rennes Cedex,
France
4Department of Geology, University of
Johannesburg, Johannesburg, South
Africa
5Institut für Geologie, Bern, Switzerland
6Berkeley Geochronology Center,
Berkeley, California, USA
Correspondence
Daniel P. Le Heron, Institüt für Geologie,
Althanstraße 14, Universität Wien, 1190
Vienna, Austria.
Email: daniel.leheron@univie.ac.at
Abstract
The expansion of ice masses across southern Africa during the Late Palaeozoic Ice
Age has been known for 150years, including the distribution of upland areas in
controlling the configuration of glaciation. In Namibia, increasing attention has
focussed on long and deep palaeovalley networks in the Kaokoland region in the
north, but comparatively little work has been attempted in the topographically sub-
dued plains of the south, in the Aranos and Karasburg basins. The desert terrain
of the Aranos area exposes diamictites of the Dwyka Group discontinuously over
about 300km, extending further south to the Karasburg area at the Namibian- South
African border along the Orange River. Whilst examined at a stratigraphic level,
the nature of the contact between the Dwyka glacial rocks and underlying litholo-
gies has not been systematically investigated. This paper presents the results from
fieldwork in austral winter 2019, in which a highly varying basal contact is described
that records the processes of growth, flow and expansion of ice masses across this
part of Gondwana. At the basin margins, subglacially produced unconformities ex-
hibit classic glacially striated pavements on indurated bedrock. In comparison, the
basal subglacial unconformity in the more basinward regions is characterised by
soft- sediment striated surfaces and deformation. In the Aranos Basin, soft- sediment
shear zones originated in the subglacial environment. This type of subglacial uncon-
formity developed over well- differentiated, unconsolidated, siliciclastic materials.
Where ice advanced over more poorly sorted material or cannibalised pre- existing
diamictites, ‘boulder- pavements’ recognised as single clast- thick boulder- dominated
intervals formed. Importantly, these boulder- pavements are enriched in clasts,
which were facetted and striated in- situ by overriding ice. By integrating measure-
ments of striation orientations, fold vergence and palaeocurrent information, former
ice flow pathways can potentially be reconstructed over a wide area.
KEYWORDS
glacial, ice sheets, Late Palaeozoic

420
|
LE HERON et al.

|
INTRODUCTION
The Dwyka Group of southern Africa preserves a world-
class archive of the Late Palaeozoic Ice Age (LPIA), whose
glacial record has enjoyed a rich history of research for
more than a century (Lomas et al., 1905; Sutherland,
1870). Recently, important work has focussed on the geo-
chronology of the LPIA succession, in particular regional
linkages with conjugate basins in South America (Griffis
et al., in press). The emerging picture of isochronous de-
glacial processes across the region, from South Africa to
Brazil (Griffis et al., 2019a), is providing much sharper
focus than that afforded by older studies that emphasised
continental- scale diachroneity as the LPIA record shifted
across Gondwana from Bolivia via southern Africa to
Australia as a result of the migration of Gondwana away
from polar regions (Eyles et al., 1993). The LPIA was an
interval of complex faunal turnover and extinction, and
hence interdisciplinary earth systems approaches have
been taken to shape the evolving paradigm (Montañez &
Poulsen, 2013). In this, there remain key aspects where
specific sedimentological, stratigraphic and geomor-
phological investigations play a lead role. These include
accurate palaeogeographic reconstructions, including pa-
laeotopography, whereby ice mass type, thickness and re-
treat patterns can be characterised. These endeavours are
important, particularly as it remains the case that ‘the lo-
cation, longevity and geographic extent of late Palaeozoic
ice centres in west- central Gondwana remain ambiguous’
(Fedorchuk et al., 2019).
In spite of the rich tradition of investigation, the size
of the main Karoo Basin of South Africa and neighbour-
ing basins to the north and across southern and central
Africa (Catuneanu et al., 2005), means that large regions
still remain under- investigated, and hence many new in-
sights remain possible through ongoing field investiga-
tions. In the western part of South Africa, for example,
subglacially striated surfaces record a stepwise evolution
at individual localities, with the potential to reveal phases
of flow, decoupling and re- incision at the base of LPIA
marine- terminating ice sheets (Le Heron et al., 2019). In
northern Namibia (Kaokoland), deep glacial palaeovalleys
first identified by Martin and Schalk (1953) have been re-
appraised and reinterpreted as fjords, with the palaeo- ice
thickness during the melt phase estimated.
The Aranos Basin of central Namibia and the
Karasburg Basin at the South Africa- Namibia border are
major Dwyka diamictite depocentres located between the
Kaokoland and main Karoo Basin of South Africa and
have been the focus of numerous studies over the last
century (Bangert et al., 1999; Du Toit, 1921; Visser, 1983,
1987, 1997; Stollhofen et al., 2000, 2008; Stratten, 1977;
Werner, 2006; Zieger et al., 2019). Since the turn of the
century, the palaeo- ice stream concept has revolutionised
the approach to ancient glacial sequences, where data are
integrated at all scales to characterise the former glacier
bed, and thereby understand its flow character (Stokes &
Clark, 2001). For the LPIA record, this has included satel-
lite image interpretation of mega- scale glacial lineations
(Andrews et al., 2019; Assine et al., 2018; Le Heron, 2018),
the aerial photograph approach (Le Heron et al., 2019),
outcrop description and micromorphological analysis of
thin sections (Henry, 2013). The present paper integrates
new data from central and southern Namibia and the data
of previous workers to (a) characterise the basal uncon-
formity and (b) develop a model for ice flow behaviour for
these regions during the earliest phase of the LPIA.
1.1
|
Study area and
geological background
Late Palaeozoic glaciogenic rocks are well- documented
across southern and central Africa (Du Toit, 1921; Linol
et al., 2016; Rogers & Du Toit, 1904). One of the largest,
laterally continuous and most studied late Palaeozoic ba-
sins in Africa is the main Karoo Basin of South Africa, at
the base of which the glaciogenic rocks referred to as the
Dwyka Group occur (Dietrich & Hofmann, 2019; Du Toit,
1921; Isbell et al., 2008; Rogers & Du Toit, 1904; Visser,
1997). Karoo Basin- like sedimentation occurs in intracra-
tonic basins, across central and southern Africa, with gla-
cial deposits largely occurring within the Dwyka Group
although some regional derivatives exist (see Catuneanu
et al., 2005). In Namibia, Dwyka Group rocks outcrop
along the entire western and southern region and in sub-
crop across the Kalahari Basin of Namibia and Botswana
(Miller, 2008; Wilson, 1964). This study focusses on the
Aranos and Karasburg subbasins of the greater Kalahari
region.
The Aranos Basin and the Karasburg Basin (some-
times referred to as the Warmbad, Noerdoewer or Orange
River Basin) (Figure1A– C) can be considered co- eval to
the main Karoo Basin in South Africa, which contains
a record of four deglacial sequences in the centre of the
basin (Grill, 1997; Isbell et al., 2008; Stolhoffen et al.,
2008; Visser, 1987, 1997). In contrast to the Kaokoland
in northern Namibia, which comprises an upland re-
gion deeply dissected by 300Myr old fjords, the Aranos
Basin (Figure1B) comprises a subdued topography
with the Dwyka Group resting unconformably on red
Cambrian sandstones of the upper Nama Group, which
are hereafter referred to as ‘basement’. In the Aranos
Basin, the Gibeon Formation, representing the basal
Dwyka Group and the lowermost deglacial sequence,
dips gently westward. In this region, several localities

|
421
LE HERON et al.
FIGURE  A series of maps placing the study in context. (A– C) Study areas of the Dwyka Group in southern Namibia, compiled
from Martin (1981), Visser (1983), Bangert et al. (1999) and Geiger (1999). (D) Simplified regional- scale geological map emphasising the
connections between Namibia and South America during the LPIA. Slightly modified from Griffis et al. (in press)
WINDHOEK
JOHANNESBURG
CAPE TOWN
NAMIBIA
SOUTH AFRICA
LESOTHO
Main Karoo Basin
BOTSWANA
km
0500
Aranos
Basin
Dwyka Group
Post Dw yka Ka ro o Su pergroup
Keetmanshoop
Mariental
17°21° 25°29°
33°
30°
34°
26°
21°25° 29°
22°S
study areas
Karoo
Karasburg
Wa rmbad
Noordoewer
Karasburg Basin
Namibia -
South Africa border
Schlip
Mariental
Tses
Keetmanshoop
Karas Mts.
18° 20°
24°
26°
“Inselberg”
Gibeon
Fish River tributary
“Airport canyon”
Noordoewer
Map B
Map C
B
A
C
Aranos Basin
N m abi
A
ngola
Huab Basin
Owambo Basin
Waterbe rg Basin
Aranos
Basin
Karasburg
Basin
n
n
Owambo Basin
n
W
W
aterbe
aterb
W
W
W
W
r
g Basin
g B
Ba
Namibia
Angola
South Africa
Botswana
Damara Belt
Rio Grande do Sol
state
Uruguay
Neoproterozoic to Camb. (<900 Ma)
L. Neoproterozoic - L. Mesoprot.
(550-1300 Ma)
Paleozoic Sed. Basins
Karoo Basin
Mesoproterozoic and older
Phanerozoic cover
Legend
Studied basins
Aranos Basin
Karasburg Basin
Palaeovalleys (Martin, 1981)
2500 km
*
*
*
Santa Catarina
state
Paraná state
D
0km 100
0km 50

422
|
LE HERON et al.
expose the basal contact with the Dwyka, including
the so- called Airport Canyon south of Mariental, a set
of buttes, sections at Gibeon, outcrops along the Fish
River and north of Keetmanshoop (Figure1B). In these
studied sections, individual outcrops of Dwyka glacio-
genic deposits do not exceed a few tens of metres in
thickness, although elsewhere within the basin, gla-
ciogenic deposits are tens to hundreds of metres thick
(Martin, 1981). Recent U– Pb detrital zircon investiga-
tions (Griffis et al., in press) explore the regional- scale
connections between Namibia and South America
during the LPIA (Figure1D), a connection that has
been posited for a long time given the occurrence of
westward- directed bedrock palaeovalleys in north-
ern Namibia (Martin, 1981). In the Karasburg Basin,
along the South Africa- Namibia border at the Orange
River in the Noordoewer region the basal contact of the
Dwyka Group is well- exposed. There, recent study has
revealed that the Dwyka Group is over a 100m thick,
and sampling of multiple ash horizons has yielded
good chronostratigraphic constraints (Griffis et al., in
press). The focus throughout this paper is on the basal
contact between the Dwyka Group and the basement.
The motivation of this study is to reveal the processes
at work at the ice- bed interface to resolve the style and
dynamics of LPIA in this region of Gondwana.
FIGURE  The Mariental “Airport Canyon” sections at approximately 24°39′42.90″S 17°55′51.30″E. Profile (A) exhibits a spectacular
suite of soft- sediment deformation. These include convolute bedding, deformation bands in wave rippled sandstone, complex thrust
geometries and recumbent folds. Profile (B) lacks all these facies, but a pebbly diamictite is present. The profiles are bracketed by presumed
Lower Palaeozoic sandstone with nodules beneath and young (presumed Pleistocene- Recent) river terrace gravels on top. The assemblage
of deformation structures compares closely to subglacial facies described elsewhere (Melvin, 2019). Profile (B) lacks these characteristics,
and preserves a massive diamictite in place of the sandstone facies at (A). Note that owing to scree- covered sections the nature of the lateral
relationship between (A) and (B) is uncertain at that stratigraphic level

|
423
LE HERON et al.
2
|
DESCRIPTIONS
In this section, descriptions are given on a locality- by-
locality basis as it is important to assess the spatial vari-
ations across the region in terms of the morphology of
the sub- Dwyka unconformity and immediately overlying
deposits. At the several localities where it was possible to
measure clast striation orientations, the measurements
refer to data from multiple clasts.
2.1
|
Airport Canyon
This section comprises a ca 300m long intermittent expo-
sure along the banks of the Fish River, immediately south
of Mariental. Two logged sections from this exposure,
produced 250m apart, serve to highlight the heterogene-
ity of the 2– 4m thick basal Dwyka succession (Figure2).
Resting on basement, the basal Dwyka comprises
lonestone- free shale overlain by deformed sandstone
FIGURE  Lithofacies from the
Airport Canyon section, profile (A).
Photographs as follows: (A) Overall
profile view. (B) Soft- sediment injection
structure. (C) Recumbently folded
wave rippled sandstones, traversed by
deformation bands. (D) Ladderback
ripples. (E) Deformation bands.
Collectively, the deformation structures
are interpreted to record strain within a
subglacial substrate, that is within the
deforming bed (Evans et al., 2006)

424
|
LE HERON et al.
facies to the south (section a, Figure2) and poorly ex-
posed diamictite to the north (section b, Figure2). In the
southern section, the deformed sandstones are capped by
undeformed sandstones: at outcrop, this manifests as a
pseudo- coarsening upward profile (Figure3A). A number
of deformation features are noted within the sandstones.
Flame structures occur at the contact between beds, and
are developed at the decimetre- scale (Figure3B). At a
larger, metre- scale, recumbent folds occur (Figure3C).
These folded sandstones contain both the flame struc-
tures together with interference ripples on the bedding
planes (Figure3D). In addition, the folded bedding is
FIGURE  (A) Aerial image
revealing the relationship between pebbly
sandstone inselbergs of the basal Dwyka
and underlying Cambrian sandstones.
(B) Outcrop level view of the Cambrian-
Dwyka contact, with inset photograph
showing large boulder of Cambrian
sandstone sitting at the top of the Butte.
(C) Stacked sets of pebbly, trough cross-
bedded sandstone (note lens cap for scale,
circled) Interpreted as a fluvial channel
system cut into basement, probably
recording initial downcutting and incision
into the surface prior to glacial advance.
Outcrop at 25°4′49.20″S 17°44′13.40″E

|
425
LE HERON et al.
dissected by deformation bands (Figure3E). The lateral
relationship between these sandstone facies at this sec-
tion and the diamictites is unclear owing to intermittent
exposure (Figure2).
2.2
|
Butte
A series of basal Dwyka sandstone outcrops are exposed
south of Mariental where the contact with basement
is well- defined. From aerial imagery captured using a
DJI Mavic Pro drone, the relationship between the red-
weathering Cambrian strata and the yellow- weathering
Dwyka deposits is apparent (Figure4A). The Dwyka
weathers out as a series of disconnected buttes that to-
gether belong to a sinuous outcrop belt. The basal Dwyka
is characterised by considerable (at least metre- scale)
downcutting into the Cambrian basement (Figure4B).
The Dwyka sandstones comprise metre- scale trough
cross- bedded, locally pebbly sandstones (Figure4C). A
dominant SW trend is suggested from the trough cross-
beds; multistorey cut- and- fill cycles are observed, the top
of which also provide evidence of climbing ripple cross-
lamination at some levels (Figure5). The degree of down-
ward incision within the multistorey? sandstones locally
exceeds 2m (Figure5), and the entire succession is or-
ganised into a ca 10m thick fining upward motif that is
interrupted by a 30cm interval of deformed lonestone- free
mudstone. The succession is capped by boulders of facet-
ted, and locally striated, red Cambrian sandstone (inset,
Figures4B and 5).
2.3
|
Gibeon
The Dwyka succession at Gibeon, the locality- type of the
basal Dwyka Group for the Aranos Basin (Miller, 2008),
rests directly on basement, like that at Airport Canyon
and Butte outcrops, forming an 8 m thick heterogene-
ous diamictite- dominated succession (sedimentary log,
Figure6A). The succession comprises yellow, grey and
red- weathering deposits that transition upwards from
massive via stratified diamictites and are capped by pebble
conglomerates (Figure6 and accompanying sedimentary
log). The stratified diamictites contain excellent examples
of striated clasts (Figure6C). The orientations of striations
on the upper (sub- horizontal) surfaces of the clasts reveal
a prominent bimodal orientation, with striations oriented
NNE- SSW and SNE- WSW (Figure6B). Striations are also
well- developed on boulder- sized clasts that cluster at the
base of the succession (Figure6E). In between stratified
diamictite horizons, delicately laminated siltstones occur
(Figure6D). The uppermost pebble conglomerates are
clast- supported and dominated by equant, sub- angular
clasts (Figure6E).
2.4
|
Fish River
The so- called Fish River successions comprise outcrops,
about 7.5km apart, between Tses to the east and Brukkaros
volcano to the west. At the studied outcrops, diamictites
measure a few metres thick (Figure7A). Away from the
creek beds, the 10– 20 m thick Ganigobis shales are pre-
served. A U- Pb zircon CA- ID- TIMS age of 299.31±0.35Ma
is reported from the Ganigobis shales, sampled ca 5 m
above the diamictite (Griffis et al., in press). Both stratified
(Figure7B) and massive diamictite subfacies are observed,
FIGURE  Logged section through the basal Dwyka
succession at the inselberg, showing characteristic stacked pebbly
sandstones arranged into fining up intervals. Rose diagram shows a
well- developed south- westward palaeocurrent dispersal

426
|
LE HERON et al.
with excellent striated clasts throughout (Figure7C).
Striation orientations from the upper (sub- horizontal) sur-
faces of clasts reveal a very weak N– S trend for the eastern
section (Figure7A), and a more clearly developed NW- SE
orientation for the western section (Figure7B).
2.5
|
Keetmanshoop
A roadside section in the vicinity of Keetmanshoop
(Figure8) exposes the contact between clinoform- bearing
sandstone basement and basal diamictites of the Dwyka
above. The morphology of the unconformity can be dem-
onstrated in panoramic photographs, particularly with
vertical exaggeration (Figure8A) and are shown to con-
sist of irregular undulations of approximately 2– 3m with
a wavelength of about 20m. At a finer scale, decimetre-
scale undulations are also recorded (Figure8A). The con-
tact between the basement and the Dwyka is extremely
sharp (Figure8B). In this section, the basal deposits of the
Dwyka are a maximum of 1.5m thick and comprise peb-
bly, clast- poor silty diamictites (Figure8C). In a nearby
outcrop, these basal diamictites encompass numerous
pebbles either derived from the underlying basement or
exotic lithologies. Some pebbles are conspicuously fac-
eted and striated, and have been found deformed, form-
ing an elongated ridge with a striated top (NNE- SSW
trend).
FIGURE  The basal Dwyka at Gibeon. The logged section shows the dominance of massive diamictites at the base, stratified
diamictites in the middle, and pebbly conglomerates at the top. The numbered intervals to the left of the log correspond to each of the
beds shown in photograph (A), that is an interval showing the full transition from massive via stratified diamictite, capped by pebbly
conglomerate. (B) Detailed view of the massive to stratified diamictite, with photograph taken approximately 8m along strike from
photograph A. Note the excellent 3D exposure of clasts. Striations on their upper surface were measured (N=30): on the rose diagram,
a strong NW- SE orientation is observed, with a secondary W– E orientation. (C) Detail of striations on the upper surface of an embedded
sandstone clast. (D) Detail of virtually clast- free, delicately stratified interval. (E) Clast- supported conglomerate at the top of the succession

|
427
LE HERON et al.
FIGURE  The basal Dwyka at the Fish River sections. (A) Simple sedimentary log showing a predominantly stratified diamictite and
a weak N– S trend of striation orientations on the upper surface of clasts. (B) Typical view of stratified muddy diamictite in the Fish River
Tributary region. Note lens cap for scale. (C) Example of a striated clast: throughout this study, orientation measurements were only taken
from similar sub- horizontal surfaces, and not from the side of the clast. Note lens cap for scale. (D) Aerial view of the relationship between
Cambrian basement and Dwyka. Clast orientations from the basal diamictite record a more prominent NW- SE trend in this locality

428
|
LE HERON et al.
2.6
|
Noordoewer
This section lies along the Orange River, with out-
crops straddling the Namibian- South African border.
A 100m thick section of the Dwyka was measured in
which four distinct facies can be recognised: (a) a boul-
der conglomerate, (b) diamictites, with both massive
and stratified variants observed, (c) a lonestone- bearing
shale facies and (d) a lonestone- free shale facies. The
boulder conglomerate (Figure9A,B), interpreted as
the lowermost ‘tillite’ unit of the Gibeon Formation
(Miller, 2008), dominates the lowest 1m of the Dwyka;
the contact between this and sandstone basement is
sharp and irregular at the outcrop scale (Figure9A).
The orientation of boulder A- axes were measured and
found to be of variable orientations with weak NW- SE
and NE- SW trends tentatively identified (Figure9). The
first of two lonestone- bearing shale intervals, probably
corresponding to the Zwartbas Formation, drapes the
boulder bed, and exposes excellent examples of pebble-
sized clasts below which shale laminae are warped and
deformed, and above which layers are undisturbed
(Figure9C,D). Above a stratified diamictite, a sec-
ond lonestone- bearing shale, a bedding plane exposes
chisel- shaped incisions (Figure9E– G). These features
consist of ca 1cm wide grooves with a sharp termina-
tion at one end and a tapering, gradational termination
at the other. The incisions are predominantly NE- SW
oriented (Figure9). These lonestone- bearing shales
pass upward into the lonestone- free shales in which
two volcanic ash layers yield ages of 300.45±0.37Ma
and 299.41±0.24Ma, 5 and 7m above the base of the
black shales (Griffis et al., in press). These ages were
determined using high- resolution U– Pb zircon CA- ID-
TIMS analyses (Griffis et al., in press).
3
|
INTERPRETATIONS
The consensus of previous workers is that a large, initial ice
mass expanded over central and southern Namibia, with a
subsequent re- advance overprinting specific areas (Griffis
FIGURE  Nature of the basal Dwyka unconformity and basal diamictites in a road cut near Keetmanshoop. (A) Photomosaic of
the road cut together with a vertically exaggerated version to emphasise topography at the base of the Dwyka. This topography consists of
long wavelength (ca 20– 30) undulations, with smaller metre- scale to decimetre- scale undulations superimposed. (B) Close up photograph
(position shown in A) of the sharp nature of the contact, also exhibiting well- developed clinoforms in the Cambrian strata. (C) Detail of
massive, basal diamictite at the contact (see (A) for position of photograph, lens cap for scale)

|
429
LE HERON et al.
et al., 2019b, in press; Stollhofen et al., 2000; Visser, 1983).
In the following, the data are integrated from each of the
sections described above to develop a process- based model
for the earliest phases of the LPIA in central- southern
Namibia. Before doing so, it is emphasised that although
the data quality at the described outcrops are high, there is
some uncertainty regarding lateral facies transitions both
at the small scale (e.g. Airport Canyon) and at the large
scale (e.g. between the sandstone facies at Butte and the
other sections). These uncertainties are incorporated into
the following depositional models. First, the nature of the
contacts is interpreted, followed by an analysis of the fa-
cies constituting the basal Dwyka succession.
3.1
|
The basal unconformity
The basal Dwyka unconformity is interpreted as a com-
plex, composite surface which includes elements of (a)
incision and fill of bedrock channels by proglacial river
deposits and (b) shearing and deformation of subglacial
and proglacial sediments. Basal facies exposed at the
FIGURE  Nature of the basal Dwyka contact and its relationship with underlying Cambrian sandstones at Noordoewer, along the
Orange River on the border with South Africa. The Dwyka crops out on both sides of the border. Sedimentary log shows the position of each
of the photographs and features alongside. (A) Lens cap for scale sits on Cambrian sandstone, with the Dwyka sitting irregularly upon it.
The basal Dwyka occupies fractures and fissures defining a highly irregular contact. (B) Series of elongated boulders immediately above the
basal contact, with lens cap circles for scale. The orientation of these is shown on the rose diagram to the right of the log. (C) Dropstones in
shales approximately 1.5m above the basal contact. (D) Dropstone in stratified muddy diamictite. (E), (F) and (G) correspond to three views
of chisel- shaped incisions found along bedding planes in lonestone- bearing shales. Each of the photographs shows the structures on the
upper surface of the beds. The orientation of these features is shown (‘bedding plane scour marks’) to the right of the log
A
B
C
D
E
F
G
A, B
C
D
E, F, G
1 m
N = 22
Bedding plane
scour marks
N = 56
Boulder A-axis
orientations

430
|
LE HERON et al.
buttes locality is interpreted to record an unconformity
cut through fluvial incision. This interpretation is sup-
ported by (a) trough cross- bedded sandstones (Figure4C)
which form the bulk of the Dwyka Group interpreted as
fluvial deposits, (b) multi- metre scale downcutting and a
geometric relationship between the morphology of the in-
cision and overlying trough cross- bedded strata, together
with (c) the organisation of individual sandstone buttes
into a sinuous morphology at the landscape scale. The
latter is interpreted to record a palaeovalley incision. At
Gibeon, Fish River and Keetmanshoop, however, subgla-
cial shearing and deformation are observed between the
basement and the Dwyka diamictites. At Keetmanshoop,
however, the contact between basement and Dwyka is
considered to record warm- based, subglacial erosion or
groove- ploughing to explain long wavelength basin struc-
tures seen at outcrop (Figure8), which are tentatively
interpreted to record cross- sections through subglacial
bedforms. In particular, the elongated, striated ridge is
interpreted as a drumlin formed subglacially (Ely et al.,
2016). The smaller metre- scale undulations are posited
to represent subglacial quarrying. The deformed (folded)
boulder pavement found along the bank of the Fish River
is also interpreted as subglacial striation, deformation
and shearing. Among the other localities, the basal un-
conformity exposed at Noordoewer (Figure9A) is also
proposed to represent a combination of subglacial quar-
rying and subglacial sediment injection into bedrock via
hydrofracturing. The sharp, locally irregular contact be-
tween the Dwyka and basement is explained as a series
of either quarried or fractured bedrock weaknesses in-
jected with subglacial sediment. It is notable that in con-
trast to abundant striated pavements found on the rim
of the Karasburg Basin (Miller, 2008) these appear to be
extremely rare at the margins of the Aranos Basin, with
none recorded in this study. It is believed that only a sin-
gle striated pavement has been found in the entire Aranos
Basin (Heath, 1972), in spite of ideal outcrop conditions
(very low angle dipping surfaces and abundant basement
outcrops). At Airport Canyon, the contact is disconform-
able and consists of Dwyka shale resting on basement, but
the poorly exposed nature of the contact precludes de-
tailed interpretation.
3.2
|
The basal Dwyka succession
The complexity of the basal Dwyka unconformity pales
in comparison to the lithofacies which are found imme-
diately above in the Gibeon Formation. The diamictites of
the Gibeon outcrop are interpreted as an assemblage of
subglacial tillites (Evans et al., 2006), with multiple sub-
glacial depositional episodes dominated by lodgement
interrupted by ice- bed separation events in which stratified
diamictites were deposited. The same section also records
evidence for more energetic meltwater events, explaining
the presence of clast- supported pebble conglomerates.
The evidence for subglacial origin for the diamictites is
provided by convincing preferred striation orientations on
the upper surface of clasts. The bimodal trend of the stria-
tions invites the proposal that shear was applied from two
separate directions. The Fish River sections also reveal
complex striation orientations on clast surfaces, perhaps
also suggestive of cross- cutting flow orientations. In the
diamictite facies, the occurrence of bedding- parallel fab-
rics exemplified at Gibeon is best interpreted as an assem-
blage of subglacial shear surfaces. Alternatively, it could
be proposed that striated clasts were rotated during trac-
tion in response to changing strain rates or rheology (e.g.
wetting and drying) of the deforming layer. Underscoring
the interpretation of ice- bed separation events in the lam-
inites is (a) the deflection and piercing of underlying lami-
nae, diagnosing the lonestones as dropstones, and (b) the
excellent preservation of laminae, requiring a waterlain
origin. Less insight is possible for similar diamictites such
as those at Keetmanshoop, yet given the context a subgla-
cial origin for the massive, clast- poor diamictites is also
possible. Further, the boulder- bearing conglomerate at
Noordoewer is interpreted as a subglacial deposit. The two
weak trends in terms of the clast long axes at this locality
potentially indicate directions of subglacial shearing, with
the long axes expected to align parallel to palaeo- ice flow.
If this is the case, then the data at Noordoewer can be in-
terpreted to record two cross- cutting flow orientations.
Both the Airport Canyon and Butte localities yield fa-
cies whose origins require special attention. The Airport
Canyon assemblage is interpreted to record the build out
of a small wave- dominated delta that was cannibalised
by subglacial deformation during its development. In
this context, basal shales are interpreted as bottomsets,
the deformed sandstones belong to foresets, and the un-
deformed, stacked wave rippled, well- sorted sandstones
belong to the topsets. This interpreted succession corre-
sponds to profile a (Figure2); profile b (Figure2) bears a
massive diamictite in the same stratigraphic position as the
deformed sandstone in profile a. In a non- glacial context,
it might be possible to appeal to downslope mass wasting
and slumping to explain the relationships. A mass wasting
hypothesis is nevertheless stymied by the following issues.
First, this does not explain the diamictites occurring at a
slightly higher elevation than the laterally equivalent, de-
formed sandstones (Figure2). Second, the assemblage of
structures is well- known from other glacial successions
including the so- called Zarqa facies of the Late Ordovician
in Saudi Arabia (Melvin, 2019; Tofaif et al., 2019). In that
example, deformation bands cross- cutting soft- sediment

|
431
LE HERON et al.
folds, also developed in sandstone, together with laterally
equivalent diamictites, pass into laterally equivalent stri-
ated pavements. A similar interpretation for the Airport
Canyon succession is proposed whereby, collectively, the
deformation structures are interpreted to record strain
within a subglacial substrate. The metre- scale folds were
probably produced during a more ductile phase of sub-
glacial deformation, overprinted by more brittle products,
that is the deformation bands, once deformation ‘locked
up’.
At Butte, the occurrence of stacked, multistorey trough
cross- bedded sandstones characterised by erosive, pebble
lined bases testify to processes of repeated cut and fill in
a fluvial environment. Given the large- scale context dis-
cussed above (the infill of a palaeovalley) these deposits
probably record the migration of braid bars in a generally
high energy environment. The overall fining upward pro-
file of the Butte succession implies progressive shallow-
ing of the channel system. The presence of boulders of
Cambrian red sandstones on the uppermost level replete
with clast facets may imply that the Butte section was
overridden during a subsequent glacial advance.
3.3
|
Constraints on water depth at
Noordoewer?
The intercalation of three facies above the basal deposits
(i.e. both massive and stratified diamictites, lonestone-
bearing shale and lonestone- free shale) provide excel-
lent insight into both process and palaeoenvironment
following the emplacement of the subglacial boulder
conglomerates. In contrast to the massive basal diam-
ictites in localities such as Gibeon or Keetmanshoop,
massive diamictites at Noordoewer that are intercalated
with shale are interpreted as waterlain rather than sub-
glacial (in contrast to the basal conglomerates with in-
jectites). This interpretation was reached because of the
associated well- stratified facies (both lonestone- bearing
and lonestone- free shale). The deflection and warping
of laminae beneath the lonestone diagnoses these as
dropstones. In ancient rocks, most dropstones are in-
terpreted to have been deposited beneath ice shelves
(Lechte & Wallace, 2016) or from floating icebergs
(Condon et al., 2002; Le Heron et al., 2020; Rodríguez-
López et al., 2016). Numerical modelling also shows
promise in terms of estimating water depth through ex-
amination of dropstone impact structures (Bronikowska
et al., in press).
On account of the chisel- shaped incisions (Figure9E–
G) below one of the dropstone- bearing horizons, it is
suggested that at least some of the dropstones were de-
posited from shorefast ice rather than by icebergs. The
chisel- shaped incisions are tentatively interpreted as fur-
rows cut by an expanding mass of shorefast ice which ex-
panded southward during the winter. The predominant
NE- SW orientation is thus proposed to reveal the direction
of expansion of the ice: the sharp end of the structures is
interpreted to record the point of downcutting of ice into
the sea floor, with incision becoming less pronounced with
distance (Figure9). These interpretations imply that, ini-
tially at least, shallow water conditions prevailed during
early Dwyka deposition. Thereafter, however, given the
thickness of the succession above in the Owl Gorge region
1km to the north, it is likely that the succession records
a transition to deeper water (Griffis et al., in press) during
an initial deglaciation cycle.
4
|
DISCUSSION
4.1
|
Development of the basal Dwyka
surface: a complex, composite origin
In southern Namibia, the development of the basal un-
conformity beneath the Dwyka was complex and is in-
terpreted as a combination of direct subglacial erosion
of basement, deformation of subglacial sediment, the
incision of bedrock through pressurised subglacial melt-
water systems, and to proglacial fluvial deposition. In the
following, specific reference is made to the earliest phase
of glaciation to affect the region, that is the lowermost
glacial cycle of Grill (1997) and Stollhofen et al. (2000).
Since Du Toit (1921), phases of data collection on striation
orientations, roches moutonnees, cross- bedding orienta-
tions and tillite fabrics (cf. Stollhofen et al., 2000, 2008;
Stratten, 1977; Visser, 1983; Werner, 2006) have painted
a complex picture. Several authors (Griffis et al., 2019a,
2019b; Stolhoffen et al., 2000, 2008; Stratten, 1967, 1977;
Visser, 1983; Zieger et al., 2019) have pointed to two con-
flicting directions of ice movements, namely the so- called
Namaland ice sheet flowing from the Windhoek Highlands
in the north toward the south, and the Transvaal Ice sheet
that flowed from east to west. Further, some workers have
proposed both southward (Namaland) and northward
(Transvaal) flowing ice lobes emerging from glacial val-
leys over the Aranos Basin, with the relative contributing
of a northward and southward flow shifting over space
and time (Visser, 1987).
The data presented here suggest that support for the
‘Namaland ice sheet’ as well as the ‘Transvaal ice sheet’
can be found within a single deposit implying that these
fabric orientations must have developed in a single phase.
This applies to multiple localities where two ice flow ori-
entations can be interpreted from the same basal Dwyka
succession, namely at Gibeon, the Fish River sections,

432
|
LE HERON et al.
and at Noordoewer. These seemingly conflicting flow
orientations are best interpreted as evolving in one, time-
transgressive phase of deformation in the subglacial envi-
ronment. Through this process, the direction of shear in
the subglacial environment switched over time from N– S
to E– W, or vice versa, recording the local flow conditions
such as the development of bedrock obstacles or stick- slip
behaviour in the deforming bed (Boulton & Hindmarsh,
1987). Translation of part of the subglacial bed must ac-
count for this, in a manner similar to the mosaic concepts
of deforming spots proposed by Piotrowski et al. (2004).
Note that a complex, evolving subglacial environment
appears to conflict with the model of Zieger et al. (2019)
developed from detrital zircon work, who envisaged two
separate flow orientations: a southward ice flow in an
early phase (lower Gibeon Formation) and a westward
flow in a later phase (upper Gibeon Formation).
4.2
|
Reappraising regional ice flow and
ice sheet reconstructions
At the scale of Gondwana, if emphasis is placed on detri-
tal zircon provenance, a very simple picture has emerged:
Craddock et al. (2019) interpret two westward flowing
ice caps. The northern ice cap- named the Dwyka ice
cap- connected Namibia to south- eastern Brazil, whereas
the southern Ellsworth ice cap was proposed to nourish
Uruguay and eastern Argentina with glacial sediments
(Craddock et al., 2019; Fedorchuk et al., 2020; Griffis et al.,
2019b; Zieger et al., 2019).
At a slightly smaller scale, based on new data together
with a swathe of data from previous authors, Dietrich et al.
(2019) presented a regional reconstruction (their figure
10) which showed both south- west (Aranos Basin) and
south and west (Karasburg Basin) flows inferred from stri-
ated pavements. Paradoxically, the Kalahari- Karoo Basin
(to the east) was depicted as an ‘ice- influenced sea’ where
iceberg and ice shelf sedimentation prevailed. Therefore,
the palaeoglaciological connection between the upland
areas and both the Aranos and Karasburg basins requires
further investigation.
Consideration of the palaeoglaciology of central and
southern Namibia during the LPIA cannot be under-
taken without giving detailed attention to South America,
where a number of important papers have been published
in the past 6years. Traditionally, attention has focussed
on the connection between an ice mass considered to
be centred on the Windhoek Highlands of Namibia (i.e.
north of the Aranos Basin) and the conjugate Paraná
Basin of Brazil. There, predominantly NNW- SSE oriented
striations mostly cut into Devonian sandstone in Paraná
State or in bedrock to the states in the south, together
with p- forms carved into Neoproterozoic granites, have
been interpreted to indicate a NNW directed ice flow
(Rocha- Campos et al., 2008). This regional flow pattern
is approximately 180° to that interpreted in the Aranos
Basin regional NNW- SSE trend, yet is supported by recent
fieldwork and satellite image interpretation that convinc-
ingly demonstrates a suite of polished, streamlined and
asymmetrical structures developed on granite in Uruguay
(Assine et al., 2018). In Brazil, the record of striation is not
only preserved on hard bedrock but also on soft sediment
where some structures are interpreted as ice keel turbates
(Vesely & Assine, 2014). At a regional scale, including in
Namibia, it is vital to separate those structures and fea-
tures that are produced subglacially from those produced
from floating ice.
At the south- eastern margin of the Paraná Basin on the
Rio Grande do Sul Shield, Fedorchuk et al. (2019) reinter-
preted palaeovalleys with supposed fjord fills as non- glacial
lacustrine to estuarine deposits. In that study, however, core
did not penetrate the base of the palaeovalley precluding its
context with basement to be evaluated. It also proposed that
rather than the eastern part of the Paraná Basin being ‘fed’
by ice lobes centred on Namibia, ice lobes from Uruguay
delivered sediment to the north. Nevertheless, based on a
provenance study the following year (Fedorchuk et al., 2020)
it was acknowledged that although Uruguayan sediment
sources could be recognised in the Paraná Basin, glaciers
derived from the highlands of Namibia continued to be a
valid interpretation. In contrast, investigation of the fill to
the east– west oriented Mariana Pimentel palaeovalley at the
southern margin of the Paraná Basin reveals a record dom-
inated by rhythmites, interpreted as varves and associated
with dropstones in core (Tedesco et al., 2020). Thus, appar-
ent distance from the supposed ice centres in Namibia does
not have an obvious bearing on whether palaeovalley fill
contains glacial facies or not. At the larger scale, this empha-
sises the complexity of the basal LPIA phases. Indeed, ba-
sins with thick glaciomarine sequences (e.g. the Rio Bianco
Basin, Argentina: 1.4km thick) favour rather nuanced in-
terpretations for earlier phases of the LPIA, with climate
phases hierarchically organised with glacial advance, retreat
and non- glacial intervals (Ezpeleta et al., 2020).
New investigations in the Kaokoland in northern
Namibia have reinforced earlier studies, where com-
pelling evidence for deep palaeovalley incision has long
been known (Martin, 1981). There, in conjunction with
the glacial geomorphology (roches moutonnees, striated
bedrock, etc.), the upward facies transitions from subgla-
cial through nearshore to deeper marine diagnose a wide-
network of palaeo- fjords: the first pre- Pleistocene fjord
landsystems positively identified in the pre- Cenozoic
record. The palaeogeographic and palaeoglaciological
context is vital because modern fjord systems are major

|
433
LE HERON et al.
carbon sinks, the burial of which has a positive feedback
effect on ongoing cold climate cycles, and this is also ex-
pected to have had a major impact on the LPIA climate
dynamic. These processes compound the complexity of
LPIA glacial cycles that are envisaged, including the role
of equilibrium line altitude as Gondwana (Isbell et al.,
2012).
5
|
CONCLUSIONS
Reappraisal of the basal Dwyka surface at several loca-
tions at the margins of the Aranos and Karasberg basins
in Namibia reveal that a complex suite of processes was
responsible for the generation of the basal unconformity.
Analysis of a number of outcrops at the western margin
of the Aranos Basin allows us to recognise the products of
subglacial erosion, subglacial shearing of unconsolidated
sediments, together with channelisation and palaeovalley
development compatible with a fluvial genesis. The basal
LPIA unconformity in southern Namibia must therefore be
viewed as a patchwork mosaic that evolved from these pro-
cesses. Re- examination of diamictites and conglomerates
immediately above this unconformity paints an increas-
ingly complex picture. Evidence for both N– S and E– W ice
flow can be found within these deposits: these orientations
have traditionally been associated with the products of ice
sheets coming from these directions. Instead, however, it is
proposed that these orientations simply reflect shearing of
sediment beneath the LPIA ice masses.
ACKNOWLEDGEMENTS
Le Heron and Dietrich are grateful to the South Africa–
Austria joint project of the National Research Foundation
(NRF) and the Österreichischer Austauschdienst (OEAD
Project no. ZA 08/2019) for funding. No additional data
are available for consultation beyond the data presented
in this paper. We are very grateful to Jonathan Lee and
Anonymous for constructive reviews, to Benjamin Walter
for handling the paper and for further suggestions, and to
Greta Mackenzie for copyediting.
DATA AVAILABILITY STATEMENT
No additional data are available for consultation beyond
the data presented in this paper.
ORCID
Daniel P. Le Heron https://orcid.org/0000-0002-7213-5874
REFERENCES
Andrews, G.D., McGrady, A.T., Brown, S.R. & Maynard, S.M. (2019)
First description of subglacial megalineations from the late
Paleozoic ice age in southern Africa. PLoS One, 14, e0210673.
Assine, M.L., de Santa Ana, H., Veroslavsky, G. & Vesely, F.F.
(2018) Exhumed subglacial landscape in Uruguay: erosional
landforms, depositional environments, and paleo– ice flow
in the context of the late Paleozoic Gondwanan glaciation.
Sedimentary Geology, 369, 1– 12.
Bangert, B., Stollhofen, H., Lorenz, V. & Armstrong, R.L. (1999) The
geochronology and significance of ash- fall tuffs in the glaci-
genic, Carboniferous- Permian Dwyka Group of Namibia and
South Africa. Journal of African Earth Sciences, 29, 33– 49.
Boulton, G.S. & Hindmarsh, R.C.A. (1987) Sediment deformation be-
neath glaciers: rheology and geological consequences. Journal
of Geophysical Research, 92, 9059– 9082.
Bronikowski, M., Pisarska- Jamroży, M. & van Loon, A.J. (in press).
Dropstone deposition: results of numerical process modeling of
deformation structures, and implications for the reconstruction
of the water depth in shallow lacustrine and marine succes-
sions. Journal of Sedimentary Research, 91(5), 507– 519. https://
doi.org/https://doi.org/10.2110/jsr.2020.111
Catuneanu, O., Wopfner, H., Eriksson, P.G., Cairncross, B., Rubidge, B.S.,
Smith, R.M.H. & Hancox, P.J. (2005) The Karoo basins of south-
central Africa. Journal of African Earth Sciences, 43, 211– 253.
Condon, D.J., Prave, A.R. & Benn, D.I. (2002) Neoproterozoic glacial-
rainout intervals: observations and implications. Geology, 30,
35– 38.
Craddock, J.P., Ojakangas, R.W., Malone, D.H., Konstantinou,
A., Mory, A., Bauer, W., Thomas, R.J., Craddock Affinati, S.,
Pauls, K., Zimmerman, U., Botha, G., Rochas- Campos, A., dos
Santos, P.R., Tohver, E., Riccomini, C., Martin, J., Redfern, J.,
Horstwood, M. & Gehrels, G. (2019) Detrital zircon provenance
of Permo- Carboniferous glacial diamictites across Gondwana.
Earth- Science Reviews, 192, 285– 316.
Dietrich, P., Franchi, F., Setlhabi, L., Prevec, R. & Bamford, M. (2019)
The nonglacial diamictite of Toutswemogala Hill (lower Karoo
Supergroup, central Botswana): implications on the extent of
the late Paleozoic ice age in the Kalahari- Karoo Basin. Journal
of Sedimentary Research, 89, 875– 889.
Dietrich, P. & Hofmann, A. (2019) Ice- margin fluctuation sequences
and grounding zone wedges: the record of the late Paleozoic ice
age in the eastern Karoo Basin (Dwyka Group, South Africa).
The Depositional Record, 5, 247– 271. https://doi.org/10.1002/
dep2.74
Du Toit, A.L. (1921) The Carboniferous glaciation of South Africa.
South African Journal of Geology, 24, 188– 227.
Ely, J.C., Clark, C.D., Spagnolo, M., Stokes, C.R., Greenwood, S.L.,
Hughes, A.L.H., Dunlop, P. & Hess, D. (2016) Do subglacial bed-
forms comprise a size and shape continuum? Geomorphology,
257, 108– 119.
Evans, D.J.A., Phillips, E.R., Hiemstra, J.F. & Auton, C.A. (2006)
Subglacial till: formation, sedimentary characteristics and clas-
sification. Earth Science Reviews, 78, 115– 176.
Eyles, C.H., Eyles, N. & Franca, A.B. (1993) Glaciation and tecton-
ics in an intracratonic basin: the late Palaeozoic Itararé Group,
Paraná Basin, Brazil. Sedimentology, 40, 1– 25.
Ezpeleta, M., Rustán, J.J., Balseiro, D., Dávila, F.M., Dahlquist, J.A.,
Vaccari, N.E., Sterren, A.F., Prestianni, C., Cisterna, G.A. &
Basei, M. (2020) Glaciomarine sequence stratigraphy in the
Mississippian Río Blanco Basin, Argentina, southwestern
Gondwana Basin analysis and palaeoclimatic implications for
the Late Paleozoic ice age during the Tournaisian. Journal of the
Geological Society, 177, 1107– 1128.

434
|
LE HERON et al.
Fedorchuk, N.D., Griffis, N.P., Isbell, J.L., Goso, C., Rosa, E.L.M.,
Montañez, I.P., Yin, Q.- Z., Huyskens, M.H., Sanborn, M.E.,
Mundil, R., Vesely, F.F. & Iannuzzi, R. (2020) Provenance
of late Paleozoic glacial/post- glacial deposits in the eastern
Chaco- Paraná Basin, Uruguay and southernmost Paraná Basin,
Brazil. Journal of South American Earth Sciences, 106, 102989.
Fedorchuk, N.D., Isbell, J.L., Griffis, N.P., Montañez, I.P., Vesely, F.F.,
Iannuzzi, R., Mundil, R., Yin, Q.- Z., Pauls, K.N. & Rosa, E.L.M.
(2019) Origin of paleovalleys on the Rio Grande do Sul Shield
(Brazil): implications for the extent of late Paleozoic glaciation
in west- central Gondwana. Palaeogeography, Palaeoclimatology,
Palaeoecology, 531, 108738.
Geiger, M. (1999) An Explanation of the Geological Map 1:10 000 of
the Namibian borderland along the Orange River at Zwartbas –
Warmbad District – Karas Region – Namibia. Diplomkartierung:
Bayerische Julius- Maximilians- Universität Würzburg.
Griffis, N.P., Montañez, I.P., Fedorchuk, N., Isbell, J., Mundil,
R., Vesely, F., Weinshultz, L., Iannuzzi, R., Gulbranson,
E., Taboada, A., Pagani, A., Sanborn, M.E., Huyskens, M.,
Wimpenny, J., Linol, B. & Yin, Q.Z. (2019b) Isotopes to ice:
constraining provenance of glacial deposits and ice centers in
west- central Gondwana. Palaeogeography, Palaeoclimatology,
Palaeoecology, 531, 108745.
Griffis, N.P., Montañez, I.P., Mundil, R., Le Heron, D.P., Dietrich, P.,
Kettler, C., Linol, B., Mottin, T., Vesely, F., Iannuzzi, R., Huyskens,
M. & Yin, Q.- Z. (in press). High- latitude ice and climate control on
sediment supply across the late Carboniferous and early Permian
of SW Gondwana. GSA Bulletin. https://doi.org/10.1130/B35852.1
Griffis, N.P., Montañez, I.P., Mundil, R., Richey, R., Isbell, J., Fedorchuk,
N., Linol, B., Iannuzzi, R., Vesely, F., Mottin, T., da Rosa, E., Keller,
B. & Yin, Q.- Z. (2019a) Coupled stratigraphic and U- Pb zircon age
constraints on the late Paleozoic icehouse- to- greenhouse turnover
in south- central Gondwana. Geology, 47, 1146– 1150.
Grill, H. (1997) The Permo- Carboniferous glacial to marine Karoo
record in southern Namibia: sedimentary facies and sequence
stratigraphy. Beringeria, 19, 1– 98.
Heath, D.C. (1972) Die geologie van die Sisteem Karoo in die gebied
Mariental- Asab. Geological Survey of South Africa Memoir, 61,
44.
Henry, L.C. (2013) Late Paleozoic glaciation and ice sheet collapse
over Western and Eastern Gondwana: sedimentology and stra-
tigraphy of glacial to post- glacial strata in Western Argentina
and Tasmania, Australia. Theses and Dissertations 112. https://
dc.uwm.edu/etd/112
Isbell, J.L., Cole, D.I. & Catuneanu, O. (2008). Carboniferous–
Permian glaciation in the main Karoo Basin, South Africa:
stratigraphy, depositional controls, and glacial dynamics. In:
Fielding, C.R., Frank, T.D. & Isbell, J.L. (Eds.) Resolving the late
Paleozoic ice age in time and space. Geological Society of America
Special Paper, 441, 71– 82.
Isbell, J.L., Henry, L.C., Gulbranson, E.L., Limarino, C.O., Fraiser,
M.L., Koch, Z.J., Ciccioli, P.L. & Dineen, A.A. (2012) Glacial
paradoxes during the late Paleozoic ice age: evaluating the
equilibrium line altitude as a control on glaciation. Gondwana
Research, 22, 1– 19.
Le Heron, D.P. (2018) An exhumed Paleozoic glacial landscape in
Chad. Geology, 46, 91– 94.
Le Heron, D.P., Busfield, M.E. & Kettler, C. (2020) Ice- rafted drop-
stones in “postglacial” Cryogenian cap carbonates. Geology,
49(3), 263– 267. https://doi.org/10.1130/G48208.1
Le Heron, D.P., Dietrich, P., Busfield, M.E., Kettler, C.,
Bermanschläger, S. & Grasemann, B. (2019) Scratching the
surface: footprint of a late carboniferous ice sheet. Geology, 47,
1034– 1038.
Lechte, M. & Wallace, M. (2016) Sub- ice shelf ironstone deposition
during the Neoproterozoic Sturtian glaciation. Geology, 44,
891– 894.
Linol, B., de Wit, M.J., Barton, E., de Wit, M.J.C. & Guillocheau, F.
(2016) U- Pb detrital zircon dates and source provenance anal-
ysis of Phanerozoic sequences of the Congo Basin, central
Gondwana. Gondwana Research, 29, 208– 219.
Lomas, J., Molengraaff, G.N., Penck, A., Hobson, B., Beck, P.R.,
Anderson, W., Coleman, A.P., Katzenstein, F.G. & Davis, W.M.
(1905) Permian glaciation in South Africa. Science, 22(563),
475.
Martin, H. (1981) The Late Paleozoic Dwyka Group of the South
Kalahari Basin in Namibia and Botswana and the subglacial
valleys of the Kaokoveld in Namibia. In: Hambrey, M.J. &
Harland, W.B. (Eds.) Earth’s pre- Pleistocene glacial record. New
York, NY. Cambridge University Press, pp. 61– 66.
Martin, H. & Schalk, K. (1953) Gletscherschliffe an der Wand eines
U- Tales im Nördlichen Kaokofeld, Südwestafrika. Geologische
Rundschau, 46, 571– 575.
Melvin, J. (2019) Zarqa megafacies: widespread subglacial deforma-
tion in the Sarah formation of Saudi Arabia and implications
for the sequence stratigraphy of the Hirnantian glaciation. In:
Le Heron, D.P., Hogan, K.A., Phillips, E.R., Huuse, M., Busfield,
M.E. & Graham, A.G.C. (Eds.) Glaciated margins: the sedimen-
tary and geophysical archive. Geological Society, London, Special
Publications, 475, 53– 80.
Miller, R.M.G. (2008). The geology of Namibia, volume III: upper
Palaeozoic to Cenozoic. Windhoek, Namibia: Ministry of Mines
and Energy, Geological Survey. ISBN 0869767348.
Montañez, I.P. & Poulson, C.J. (2013) The late Paleozoic ice age:
an evolving paradigm. Annual Review of Earth and Planetary
Sciences, 41, 629– 656.
Piotrowski, J.A., Larsen, N.K. & Junge, F.W. (2004) Reflections on
soft subglacial beds as a mosaic of deforming and stable spots.
Quaternary Science Reviews, 23, 993– 1000.
Rocha- Campos, A.C., dos Santos, P.R. & Canuto, J.R. (2008). Late
Paleozoic glacial deposits of Brazil: Paraná Basin. In: Fielding,
C.R., Frank, T.D. & Isbell, J.L. (Eds.), Resolving the Late Paleozoic
ice age in time and space. Geological Society of America Special
Paper, 441, 97– 114.
Rodríguez- López, J.P., Liesa, C.L., Pardo, G., Meléndez, N., Soria,
A.R. & Skilling, I. (2016) Glacial dropstones in the western
Tethys during the late Aptian- early Albian cold snap: pa-
laeoclimate and palaeogeographic implications for the mid-
Cretaceous. Palaeogeography, Palaeoclimatology, Palaeoecology,
452, 11– 27.
Rogers, A.W. & Du Toit, A.L. (1904). Geological survey of parts of the
divisions of Ceres, Sutherland and Calvinia. Annual Report of
the Geological Commission. Cape Times Limited, Government
Printers, Cape Town, 9- 71.
Stokes, C.R. & Clark, C.D. (2001) Palaeo- ice streams. Quaternary
Science Reviews, 20, 1437– 1457.
Stollhofen, H., Stanistreet, I.G., Bangert, B. & Grill, H. (2000)
Tuffs, tectonism and glacially related sea- level changes,
Carboniferous- Permian, southern Namibia. Palaeogeography,
Palaeoclimatology, Palaeoecology, 161, 127– 150.

|
435
LE HERON et al.
Stollhofen, H., Werner, M., Stanistreet, I.G. & Armstrong, R.A.
(2008). Single- zircon U– Pb dating of Carboniferous– Permian
tuffs, Namibia, and the intercontinental deglaciation cycle
framework. In: Fielding, C.R., Frank, T.D. & Isbell, J.L. (Eds.),
Resolving the late Paleozoic ice age in time and space. Geological
Society of America Special Paper, 441, 83– 96.
Stratten, T. (1967) A preliminary report on a directional study of the
Dwyka Tillites in the Karroo Basin of South Africa. Gondwana
Stratigraphy, lUGS Symp. Buenos Aires, UNESCO, pp. 741– 761.
Stratten, T. (1977) Conflicting directions of Dwyka ice flow in the
Western Cape Province and southern south west Africa.
Geological Society of South Africa Transactions, 80, 79– 86.
Sutherland, P.C. (1870) Notes on an ancient boulder- clay of Natal.
Quarterly Journal of the Geological Society of London, 26, 514– 517.
Tedesco, J., Cagliari, J. & Aquino, C.D. (2020) Late Paleozoic ice- age
rhythmites in the southernmost Paraná Basin: a sedimentolog-
ical and paleoenvironmental analysis. Journal of Sedimentary
Research, 90, 969– 979.
Tofaif, S., Le Heron, D.P. & Melvin, J. (2019). Development of a palae-
ovalley complex on a Late Ordovician glaciated margin in NW
Saudi Arabia. In: Le Heron, D.P., Hogan, K.A., Phillips, E.R.,
Huuse, M., Busfield, M.E. & Graham, A.G.C. (Eds.), Glaciated
margins: the sedimentary and geophysical archive. Geological
Society, London, Special Publications, 475, 81– 107.
Vesely, F.F. & Assine, M.L. (2014) Ice- keel scour marks in the geo-
logical record: evidence from Carboniferous soft- sediment stri-
ated surfaces in the Paraná Basin, Southern Brazil. Journal of
Sedimentary Research, 84, 26– 39.
Visser, J.N.J. (1983) An analysis of the Permo- Carboniferous
glaciation in the marine Kalahari Basin, southern Africa.
Palaeogeography, Palaeoclimatology, Palaeoecology, 44,
295– 315.
Visser, J.N.J. (1987) The palaeogeography of part of southwest-
ern Gondwana during the Permo- Carboniferous glaciation.
Palaeogeography, Palaeoclimatology, Palaeoecology, 61, 205– 219.
Visser, J.N.J. (1997) Deglaciation sequences in the Permo-
Carboniferous Karoo and Kalahari basins of southern Africa:
a tool in the analysis of cyclic glaciomarine basin fills.
Sedimentology, 44, 507– 521.
Werner, M. (2006) The stratigraphy, sedimentology, and age of the
Late Palaeozoic Mesosaurus Inland Sea, SW Gondwana – New
implications from studies on sediments and altered pyroclastic
layers of the Dwyka and Ecca Group (Lower Karoo Supergroup)
in Southern Namibia. PhD thesis. University of Würzburg,
Würzburg, 425 p.
Wilson, E.J. (1964) Core description, Vreda 281, No. 1: Santa Maria,
California. Artnell Exploration Company Report, 18 p.
Zieger, J., Rothe, J., Hofmann, M., Gärtner, A. & Linnemann, U.
(2019) The Permo- Carboniferous Dwyka Group of the Aranos
Basin (Namibia) – how detrital zircons help understanding
sedimentary recycling during a major glaciation. Journal of
African Earth Sciences, 158, 103555.
How to cite this article: Le Heron DP, Kettler C,
Griffis NP, et al. The Late Palaeozoic Ice Age
unconformity in southern Namibia viewed as a
patchwork mosaic. Depositional Rec. 2022;8:419–
435. https://doi.org/10.1002/dep2.163