Palynofacies as a tool for high-resolution palaeoenvironmental and

palaeoclimatic reconstruction of Gondwanan post-glacial coal deposits: No.

2 Coal Seam, Witbank Coalfield (South Africa)

Alexander Wheeler1 and Annette E. Götz2,3

1The University of Queensland, School of Earth Sciences, St. Lucia, QLD 4072, Australia; e-mail:

alex.thomas.wheeler@gmail.com 2Keele University, School of Physical and Geographical Sciences, Keele, Staffordshire ST5 5BG, United Kingdom 3Kazan Federal University, 18 Kremlyovskaya St., Kazan 420008, Republic of Tatarstan, Russian Federation

Abstract The early Permian movement of Gondwana away from the South Pole caused major climatic

change across the continent. The shift from a post-glacial Carboniferous flora to a temperate Permian

flora is represented in the palynological record. Using palynofacies analysis this climate transition can

be studied at a high resolution and the palaeoenvironment can be interpreted on a local scale. Core

samples were studied from four localities of the Artinskian/Kungurian-aged No. 2 Coal Seam of the

Witbank Coalfield. At some localities, the No. 2 Coal Seam is split into an Upper Coal Seam and a

Lower Coal Seam by an intraseam parting, and samples were collected from both horizons as well as

the parting. All samples were studied with respect to palynomorph composition and phytoclast content.

The results suggest a swamp-dominated environment in the Lower Coal Seam, a river-dominated

environment in the parting and an environment which fluctuated locally from a lake-dominated to

swamp-dominated in the Upper Coal Seam due to increased input of glacial meltwater from the

hinterlands. The vegetation switched from a fern- and conifer-dominated flora in the Lower Coal Seam

to a more diverse Glossopteris-Gangamopteris flora in the Upper Coal Seam. Cordaites appears to be

limited to valleys on the northern edge of the swamplands in the Lower Coal Seam. A decrease in

monosaccate pollen grains and an increase in bisaccate pollen grains are apparent in all sample sets and

interpreted to indicate a transition from a cold to a fluctuating cool-temperate climate. This climate

signal is well documented in palynofacies of the coal seam and thus a powerful correlation tool for

high-resolution basin-wide and Gondwanan correlation.

Keywords palynofacies, palaeoenvironment, palaeoclimate, Permian, Witbank Coalfield, South

Africa

Introduction

The near continuous deposition of sediments in the Main Karoo Basin in South Africa from the late

Carboniferous (Pennsylvanian) to the Middle Jurassic captures an extensive period of climate change

from a glacial through to a dry arid climate (Falcon et al. 1984). The coal deposits of the Witbank

Coalfield and more specifically the No. 2 Coal Seam captures a crucial climatic shift in the Permian.

As glaciers receded, the climate and environment supported the development of peat-forming wetlands

in the north-eastern part of the Main Karoo Basin. In areas with such well-developed floral assemblages

one of the best methods of reconstructing the palaeoclimate and palaeoenvironment is palynology. A

palynological assemblage reflects its parent plant community which in turn is controlled by the climatic

and environmental conditions (Gastaldo 1994). This makes palynofacies a powerful tool to document

changes in climate and environment on high time resolution at multiple localities. It also allows for the

development of new methods of correlating the South African coalfields for exploration of potential

remaining coal resources in the northern part of the country which is especially important in light of the

current energy crisis affecting South Africa. The aim of this study is to use palynological samples from

multiple localities in the Witbank Coalfield to reconstruct the palaeoclimate history and to observe

changes in the local palaeoenvironment at each locality. This will allow one to identify both local

conditions, which are specific to different parts of the basin, as well as regional signatures. Previous

work on palynomorph assemblages of the Witbank coal seams (Falcon 1986) documented the overall

trend in climate amelioration during the Permian (Fig. 1). Here, we present a high-resolution

palynofacies study on the No. 2 Coal Seam to highlight the short-term changes in vegetation patterns

related to a post-glacial climate shift. Another aim of the study is to use the palynofacies and climate

signal to see if one can differentiate between the Lower and Upper Coal seams which make up the No.

2 Coal Seam even when an intraseam parting is not present. This will further develop the utilization of

palynofacies as a tool to high-resolution stratigraphic correlation even within a single coal seam.

Because of the economic significance of the Witbank Coalfield (Hancox and Götz 2014), a number of

palynological studies have already been done in the area. Hart (1967) developed the first biozonation

scheme based on palynology in the Main Karoo Basin. Four zones were defined for the Permian and

Carboniferous strata but their relatively wide time range limited their application to basin-wide

correlation (MacRae and Aitken 1997). Anderson (1977) created a more detailed biozonation scheme

consisting of eight zones and twenty one subzones. Subsequent studies which attempt interbasinal

correlation of the Main Karoo Basin with coeval basins in other parts of Gondwana use this scheme for

biostratigraphic comparison (Backhouse 1991; Stephenson and McClean 1999; Stephenson 2008).

Falcon et al. (1984) studied the palynomorph assemblages of the Witbank and Highveld coalfields,

erecting four zones and five subzones. More recently, Götz and Ruckwied (2014) and Ruckwied et al.

(2014) used palynofacies analysis to correlate the marine black shales of the Whitehill Formation with

the coals of the Vryheid Formation and investigated the palaeoclimate and palaeoenvironment

signatures of these deposits. Wheeler and Götz (2016) studied the sedimentary organic matter of the

No. 2 Coal Seam of the Highveld Coalfield south of the Witbank Coalfield to further constrain the

potential of palynofacies analysis for palaeoclimate and palaeoenvironment reconstruction.

Geological Setting

The Main Karoo Basin is a retroarc foreland basin which formed north of the Cape Fold Belt and is part

of a series of other Gondwanan basins which formed between the Late Palaeozoic to Early Mesozoic

(Catuneanu et al. 2005). The basin was bound to the north by the Cargonian Highlands and to the south

by the Cape Fold Belt (Isbell et al. 2008). The sedimentary sequence which fills the Main Karoo Basin

is known as the Karoo Supergroup and is split into five main groups (Dwyka, Ecca, Beaufort,

Stormberg, Drakensberg; Fig. 2) which were deposited over the course of approximately 120 million

years from the late Carboniferous (Pennsylvanian) to the Middle Jurassic and terminated with the

eruption of the Drakensberg basaltic lavas.

The Witbank Coalfield is regarded as the one of the most economically important coal resources in

South Africa currently being exploited (Snyman 1998; Hancox and Götz 2014). The sedimentary fill of

the coalfield is mainly made up of the fluvio-deltaic Vryheid Formation which sits on either the pre-

Karoo basement or Dwyka tillites (Cairncross 1989). The coal deposits of the Witbank Coalfield are

separated by interseam partings of sandstone or siltstone as well as intraseam partings. These partings

do not continue laterally throughout the whole coalfield and pinch out towards the south and east

(Cairncross and Cadle 1988). The coals of the Witbank Coalfield are ranked as high volatile bituminous

coals (Falcon 1986). The No. 2 Coal Seam, the main target seam for mining, is the thickest coal seam

in the coalfield averaging around 4–5 m, although varying intraseam parting thickness can push

thickness as high as 40 m and as low as 2 m (Cairncross and Cadle 1988). The No. 2 Coal Seam is

divided into an Upper Coal Seam (2U) and a Lower Coal Seam (2L) with both seams yielding a high

inertinite content and variable fusinite and vitrinite contents (Cadle et al. 1993; Götz et al. 2012). The

age of the No. 2 Coal Seam was dated as Artinskian-Kungurian (Ruckwied et al. 2014).

Materials and Methods

Core samples of the No. 2 Coal Seam were collected from four drill sites in the Witbank Coalfield (Fig.

3). Samples represent organic-rich siltstones from the Lower and Upper Seam as well as from intraseam

sand- and siltstones. A total of 42 samples were prepared using standard palynological processing

techniques as described in Vidal (1988) though with higher concentrations of acids to account for the

high mineral content of the silt- and sandstones. Organic particles were classified based on the scheme

used in Götz and Ruckwied (2014). Cuticles are considered as translucent phytoclasts (c.f. Tyson 1995).

Countings of each sample are listed in Table 1. Relative abundances of counted particles are discussed

as low (< 5%), moderate (5-20%), and high (> 20%) and together with the ratios of the organic particles

determined they are used to interpret changes in palaeoenvironment and palaeoclimate.

The Palesa locality represents the northern-most sample set taken in the Witbank Coalfield (Fig. 3). No

intraseam parting is present at this locality. Instead, siltstones of varying thicknesses are intercalated

with the coals. The total thickness of the coal seam is 8 m and 10 samples were taken from this locality.

Between 501 and 509 particles were counted for each of the palynological slides studied. The

HWH1418 locality represents the central eastern part of the Witbank Coalfield (Fig. 3). A 1 m thick

sandstone intraseam parting is present at this locality. The Lower Coal Seam is 1.5 m thick and the

Upper Coal Seam is 3.5 m thick. 14 samples were taken covering a total seam thickness of 6 m. Between

501 and 508 particles were counted for each of the palynological slides studied. The BHS14 locality

represents the central western sample set taken in the Witbank Coalfield (Fig. 3). No intraseam parting

is present at this locality. The total thickness of the coal seam is 3.6 m and 10 samples were taken from

siltstones of varying thicknesses, which are intercalated with the coals. Between 501 and 511 particles

were counted for each of the palynological slides studied. The ALBN11 locality represents the southern

part of the Witbank Coalfield close to the Smithfield Ridge (Fig. 3). A 0.4 m thick sandstone intraseam

parting is present at this locality. The Lower Coal Seam is 2 m thick and the Upper Coal Seam is 3.6 m

thick. 8 samples were taken covering a total seam thickness of 6 m. Between 501 and 591 particles were

counted for each of the palynological slides studied.

Results

Palynofacies analysis reveals similarities and differences in the composition of sedimentary organic

matter in the different parts of the Witbank Coalfield and is presented for each studied locality below.

In the northern-most part of the coalfield (Palesa), the relative abundance of palynomorphs (Fig. 4)

shows a number of prominent trends. The monosaccate pollen grains decrease in abundance replaced

by a dominant bisaccate pollen assemblage as one moves up the stratigraphy. At 4 m, even the taeniate

bisaccates have a higher abundance than the monosaccate pollen grains, although they never become

more abundant than the non-taeniate bisaccate pollen grains. Monolete spores are seen in the first 3 m

of the section and the trilete spore abundance decreases dramatically above this interval. Moderate

trilete spore abundance appears to correlate with moderate abundance of amorphous organic matter

(AOM) while higher abundances of degraded organic matter (DOM) and algae can be associated with

lower abundances of trilete spores and AOM. The moderate abundances of trilete spores and AOM also

correlate with high abundances of opaque phytoclasts (Fig. 5). At 4 m and again between 6 and 8 m,

there is a decrease in both the spore/pollen ratios as well as the opaque/translucent ratios (Fig. 5). Across

the whole coal seam, equidimensional phytoclasts are dominant especially between 6 and 8 m. The

equidimensional/blade-shaped ratio is relatively the same in samples 1 to 7 and slightly higher in

samples 9 and 10, but shows a large increase in sample 8 (Fig. 5).

In the central eastern part of the coalfield (HWH1418), monosaccate and non-taeniate bisaccate pollen

grains have similar abundances in the Lower Coal Seam (Fig. 4). No taeniate bisaccate pollen grains

are present in the Lower Coal Seam. This seam has the highest abundances of spores and AOM and low

abundances of DOM. No algae was observed. In terms of phytoclasts, this seam has high abundances

of opaque equidimensional material and slightly higher abundances of cuticles and larger plant debris

than the Upper Coal Seam. The Lower Coal Seam has spore/pollen ratios > 2 (Fig. 5). In the intraseam

parting, taeniate bisaccate pollen grains appear in low abundances for the first time. There is a relatively

high abundance of DOM and algae does appear in low abundances. The parting has the highest ratio of

opaque/translucent and equidimensional/blade-shaped phytoclasts (Fig. 5). In the Upper Coal Seam the

monosaccate pollen grains decrease and eventually disappear (Fig. 4). The bisaccate pollen grains are

most abundant in this seam and the taeniates have a high abundance. The spores decrease to low

abundances and monolete spores are not observed in some of the samples. AOM is present in low

abundances. DOM is present in high to moderate abundances and algae are moderately abundant.

Opaque equidimensional phytoclasts have their lowest abundances in this seam, while the translucent

equidimensional phytoclasts have their highest abundances. The spore/pollen, opaque/translucent and

equidimensional/blade-shaped ratios are all lower than in the Lower Coal Seam or the intraseam parting.

(Fig. 5).

In the central western part of the coalfield (BHS14), again one sees the trend of decreasing abundances

of monosaccate pollen grains and increasing abundances of bisaccate pollen grains (Fig. 4). Taeniate

bisaccate pollen grains appear in the upper 1.6 m. Spores are present in the highest abundances in the

first two meters. Monolete spores decrease and disappear above this interval, while trilete spores

decrease from high to moderate and low abundances. The AOM and DOM have relatively unchanging

abundances throughout the sequence except in sample 7 taken at 2.5 m. At this interval the AOM is

relatively low while the DOM shows a sharp increase in abundance. This is also the only sample where

algae are observed. In terms of phytoclasts, the abundances of the opaque equidimensional and the

translucent blade-shaped phytoclasts seem to change very little but the opaque blade-shaped show an

increase while the translucent equidimensional particles show a decrease. All the phytoclast categories

show large changes in sample 7 (2.5 m) and this is also reflected in the ratios (Fig. 5). The spore/pollen

ratio decreases to less than 0.08 in sample 7 (2.5 m) and remains under 0.2 in the upper three samples

(Fig. 5).

In the southern part of the coalfield (ALBN11), the monosaccate pollen grains have the highest

abundance in the Lower Coal Seam and non-taeniate bisaccates are also present (Fig. 4), as observed in

the other sample sets. No taeniate bisaccate pollen grains were observed in the Lower Coal Seam.

Monolete and trilete spores are both present in the Lower Coal Seam, however no monoletes were

observed in the second sample while triletes show the highest relative abundance. DOM is present in

relatively moderate to high abundances in the Lower Coal Seam and no AOM or algae were observed.

The equidimensional/blade-shaped ratio shows little variation throughout the sequence (Fig. 5). The

first two samples have opaque/translucent ratios quite close to 1 but the third sample has a ratio of 0.54.

The sandstone parting features monosaccate, non-taeniate bisaccate and taeniate bisaccate pollen grains.

Spore abundances remain relatively moderate. AOM is present and DOM shows a huge increase in

abundance in the parting. In the Upper Coal Seam, the monosaccate pollen grains decrease and

disappear while the bisaccate pollen grains increase (Fig. 4). Taeniate bisaccate pollen grains are only

present in the upper two samples. Monolete spores are present in moderate abundance in the Upper

Coal Seam and are not observed in sample 7 while trilete spores are highly abundant. AOM is present

with low abundances in samples 5 and 6 and increases slightly in sample 7, but is not observed in sample

8. DOM occurs in moderate to high abundances in the Upper Coal Seam. Algae are observed in samples

5, 6, and 8. The opaque/translucent ratio is very high in sample 7, and the equidimensional/blade-shaped

ratio shows little variance in the Upper Coal Seam (Fig. 5).

Discussion

The palynofacies patterns recorded in the No. 2 Coal Seam of different Witbank Coalfield localities are

related to spatio-temporal changes in vegetation, reflecting different palaeoenvironmental and

palaeoclimatic conditions.

The vegetation present in the Lower Coal Seam is controlled by the palaeoclimate as well as the local

environment. The swamp-dominated environment is characteristic of the Lower Coal Seam, which was

identified by the palynomorph assemblage as no intraseam partings are present. In the lowland areas,

the abundance of both monolete and trilete spores (e.g. Microbaculispora, Calamospora,

Punctatisporites, and Verrucosisporites) at all localities suggests a dominance of ferns, horsetail ferns

and lycopods (Anderson 1977). The highest spore abundances are found at localities HWH1418,

BHS14 and sample 2 of ALBN11. The dominant presence of monosaccate pollen grains, especially in

the lowest parts of the coal seam, is typical of post-glacial floras (Goldberg 2004; Stephenson et al.

2007). Large monosaccate pollen grains are interpreted as belonging to the genus Potonieisporites,

which has an affinity to the conifer genus Walchia in the Northern Hemisphere (Balme 1995; Traverse

2007). As noted by Falcon (1986), the lack of macrofossil evidence as well as conifers preference for

drier areas suggest that they would occupy the uplands, away from the wetlands in the south (Fig. 6).

They may have also been present on the Smithfield Ridge. Monosaccate pollen grains featuring a trilete

mark were identified in the Lower Coal Seam only at localities BHS14 and HWH1418. This pollen

grain is likely the Early Permian genus Cannanoropollis, a common species in other Gondwanan

localities (e.g., Balme 1980; Lindström 1995; Cazzulo-Klepzig et al. 2009). This genus bears a botanical

affinity to Cordaites (Balme 1995). Cordaites macrofossils have only been seen rarely in the Vryheid

Formation near the base of the formation (Seward and Leslie 1908). Cordaites is not a dominant plant

in swamps but mangrove-like forms have been observed (Traverse 2007), which suggests it is able to

occupy wetter areas than conifers are able to, allowing it to find a specific ecological niche. Thus,

Cordaites likely occupied the broad floodplains to the north of larger swamp and lake areas as it prefers

wetter conditions to the conifers but is unable to grow well in the deeper swamps located to the south

(Fig. 6). The lowland areas of the Upper Coal Seam undergo shifts in environment which would in turn

reflect the vegetation. The spore/pollen ratios in general are low in this seam (Fig. 5) with slight

increases in the Palesa and ALBN11 localities when there is a switch to a swamp-dominated

environment. This is documented in a switch from the fern-dominated lowland one sees in the Lower

Coal Seam to a more varied but tree-dominated vegetation which can include ferns, horsetail ferns,

cycads and larger glossopterid trees (Fig. 6). Glossopteris, which are associated with taeniate bisaccate

pollen grains, and Gangamopteris, which are associated with some of the non-taeniate bisaccate pollen

grains (Falcon 1986), would be mainly abundant on the broad swamps and floodplains. The low

abundance of taeniate bisaccate pollen grains at the Palesa locality and very low abundance at the

ALBN11 locality suggests a wet swampland which doesn’t support the growth of large trees. In the

uplands, glacial retreat based on the changing climate created conditions which support a greater

diversity of gymnosperms. The non-taeniate bisaccate pollen genus Pityosporites and the taeniate pollen

genus Lueckisporites are associated with different conifer species (Balme 1995), and would be present

in the drier uplands.

The palaeoenvironment of the Lower Coal Seam is generally swamp-dominated at all localities. In the

Upper Coal Seam, the palaeoenvironment varies at each locality, shifting between phases of swamp-

dominated and lake-dominated environments (Fig. 7). Given that the palaeoenvironment is controlled

by factors such as climate, topography and sedimentary input, two possible causes for the fluctuations

are either due to the climate shift or to tectonic activity. The Kaapvaal Craton, which forms the basement

of the Witbank and Highveld coalfields, is relatively stable at that point in the Early Permian (Johnson

et al. 1997). A shift in the climate from a cold to a cool-temperate climate would cause glacial retreat

and increase an influx of fresh meltwater into the basin as well as altering regional precipitation patterns.

Only at locality BHS14 do we see a lake-dominated environment throughout the Upper Coal Seam.

Palaeocurrent analysis indicates the major source of sediments is the Cargonian Highlands to the north

(Ryan 1968). The proximal depositional environment and sandstone partings are consistent with a

Gilbert-type lacustrine delta. Streams running down from the highlands merge in the valleys to form

either braided streams or meandering rivers. Large meandering rivers control sedimentation across the

broad plains to the south. The rivers are blocked from moving further south by the Smithfield Ridge

which causes them to pond into wide, shallow swamps or lakes (Fig. 6). The importance of topographic

controls on the palaeoenvironment should also be noted. The palaeotopography is likely the factor

which controlled distribution and thickness of the intraseam parting as well as thin internal partings

during flooding events. The distribution of monosaccate pollen associated with cordaitalean trees is

limited to localities associated with palaeovalleys opening to broad swamplands (HWH1418, BHS14).

These trees likely exploit the wetter conditions in the lowlands and are able to flourish on an uneven

valley floor produced by postglacial moraine deposition (Falcon 1989) but are unable to spread further

across the broad swampland due to the established stable swamp vegetation and large meandering rivers

and lakes. More studies would need to be done to determine the full extent of the distribution of these

trees in relation to the valley axes, valley flanks and the drier upland areas where coniferous vegetation

is dominant.

Distinct intervals of high abundance of opaque phytoclasts linked to coals particularly rich in inertinite

(Götz et al. 2012) may record palaeo-wildfires as recently described from other Gondwanan localities

in South America (Jasper et al. 2011; Degani-Schmidt et al. 2015; Manfroi et al. 2015; Kauffmann et

al. 2016) and India (Jasper et al. 2016). Wildfires may have been a key component of Gondwanan peat-

forming swamps (Diessel 1992), however in the case of the Witbank coals comparison with macrofossil

material would need to be done to decipher the occurrence of fossil charcoal indicative of wildfires.

All 4 localities studied in the Witbank Coalfield capture the climate signal indicated by the switch in

pollen-producing flora as previously reported by Götz and Ruckwied (2014) from the No. 2 Coal Seam

of the Inyanda Mine (northern Witbank Coalfield) and recently also detected in the No. 2 Coal Seam of

the Highveld Coalfield (Wheeler and Götz 2016). Monosaccate-producing floral elements are dominant

in the Lower Coal Seam while bisaccate (striate)-producing floral elements are the dominant group of

the Upper Coal Seam (Fig. 7). This climate signal was proposed as tool for cross-basin correlation

(Ruckwied et al. 2014) and the recent studies confirm the potential of using climate signatures recorded

in palynomorph assemblages for high-resolution correlation.

The switch from a monosaccate-producing to a bisaccate- and striate-producing flora is also seen in

other parts of Gondwana. Post-glacial floral shifts in the early-mid Permian are reported from Australia,

Brazil and Oman (Backhouse 1991; Iannuzzi et al. 2007, 2010; Stephenson et al. 2005). This shift in

floral composition in locations across Gondwana suggests a continent-wide change in climate (López-

Gamundí 1997; Wopfner 1999; Goldberg 2004; Stephenson et al. 2007). The likeliest reason for this is

the northward movement of the continent across the equator with the margin of the Karoo Basin moving

from 80°S in the late Carboniferous (Pennsylvanian)/early Permian (Cisuralian) to 40°S in the Late

Triassic/Early Jurassic (Caputo and Crowell 1985; Visser 1986).

Conclusions and outlook

This study detected on a regional scale the shift in pollen-producing flora from monosaccate-dominated

in the Lower Coal Seam to bisaccate-dominated in the Upper Coal Seam coinciding with the appearance

of striate bisaccate pollen as seen in other studies in the Witbank Coalfield (Falcon 1984; Götz and

Ruckwied 2014). This floral change represents a climatic shift from a post-glacial cold to a fluctuating

cool-temperate climate as a result of the northward movement of Gondwana from the late Carboniferous

(Pennsylvanian) to the Middle Jurassic preceding continental breakup. The palynological data also

highlights shifts in the vegetational composition in the highlands and lowlands and the relationship of

certain floral compositions with specific palaeoenvironments. The dry highland vegetation shifts from

coniferous in the Lower Coal Seam to a more diverse gymnosperm- and pteridosperm-dominated

community in the Upper Coal Seam. The lowland vegetation of the Lower Coal Seam is dominated by

ferns but includes horsetail ferns and lycopods as well as cordaitalean trees which may be limited to the

palaeovalleys at the edge of the broad swamplands. In the Upper Coal Seam, Glossopteris-

Gangamopteris vegetation becomes dominant along with a diverse community of cycads, ferns,

lycopods and seed ferns. Palaeoenvironmental analysis indicates a stable Lower Coal Seam which was

mainly swamp-dominated. The intraseam parting is not present at every locality but in the Witbank

Coalfield it consists of sandstone suggesting a proximal depositional environment. The

palaeoenvironment of the Upper Coal Seam varies at each locality and is likely controlled by the warmer

climate stimulating input of fresh meltwater from the regressing glaciers in the highlands and the uneven

palaeotopography. Future work on the Witbank coals should focus on the identification of fossil

charcoal as an indicator of palaeo-wildfires. Such findings could add further information on the

palaeoecology and palaeoenvironment of these post-glacial Gondwanan deposits.

Acknowledgements This study was partly funded by the DST-NRF Centre of Excellence for Integrated

Mineral and Energy Resource Analysis (CIMERA). Additional funding was provided by the National

Research Foundation of South Africa (NRF), grant no. 94589. BHP Billiton kindly provided access to

the core samples. The comments of Paulo Fernandes (Universidade do Algarve) and an anonymous

reviewer are gratefully acknowledged.

References

Anderson JM (1977) The biostratigraphy of the Permian and Triassic. Part 3. A review of Gondwana

palynology with particular reference to the northern Karoo Basin, South Africa. Mem Bot Surv S

Afr 4:1–33

Backhouse J (1991) Permian palynostratigraphy of the Collie Basin, Western Australia. Rev Palaeobot

Palynol 67:237–314

Balme BE (1980) Palynology and the Carboniferous‐Permian boundary in Australia and other

Gondwana continents. Palynology 4:43–55

Balme BE (1995) Fossil in situ spores and pollen grains: An annotated catalogue. Rev Palaeobot Palynol

87:81–322

Cadle AB, Cairncross B, Christie ADM, Roberts, DL (1993) The Karoo Basin of South Africa: type

basin for the coal-bearing deposits of southern Africa. Int J Coal Geol 23(1):117–157

Cairncross B (1989) Paleodepositional environments and tectonosedimentary controls of the postglacial

Permian coals, Karoo Basin, South Africa. Int J Coal Geol 12(1):365–380

Cairncross B, Cadle AB (1988) Palaeoenvironmental control on coal formation, distribution and quality

in the Permian Vryheid Formation, East Witbank Coalfield, South Africa. Int J Coal Geol 9:343–

370

Caputo MV, Crowell JC (1985) Migration of glacial centers across Gondwana during Paleozoic Era.

GSA Bull 96:1020–1036

Catuneanu O, Hancox PJ, Rubidge BS (1998) Reciprocal flexural behaviour and contrasting

stratigraphies: a new basin development model for the Karoo retroarc foreland system, South Africa.

Basin Res 10:417–39

Catuneanu O, Wopfner H, Eriksson PG, Cairncross B, Rubidge BS, Smith RMH, Hancox PJ (2005)

The Karoo basins of south-central Africa. J Afr Earth Sci 43:211–253

Cazzulo-Klepzig M, Mendonça Filho JG, Guerra-Sommer M, Menezes TR, Simas MW, Mendonca JO,

Degani-Schmidt I (2009) Effect of volcanic ash-fall on a Permian peat-forming environment, on the

basis of palynology, palynofacies and paleobotany (Faxinal Coalfield, Brazil). Rev Brasil Paleont

12(3):179–194

Degani-Schmidt I, Guerra-Sommer M, Mendonça JO, Mendonça Filho JG, Jasper A, Cazzulo-Klepzig

M, Iannuzzi R (2015) Charcoalified logs as evidence of hypautochthonous/autochthonous wildfire

events in a peat-forming environment from the Permian of southern Paraná Basin (Brazil). Int J Coal

Geol 146:55–67

Diessel CFK (1992) Coal-Bearing Depositional Systems, Springer, Berlin, pp 721

Falcon RMS (1986) A brief review of the origin, formation and distribution of coal in Southern Africa.

In: Anhaeusser CR, Maske S (ed) Mineral Deposits of Southern Africa. Geological Society of South

Africa 2:1879–1898

Falcon RMS (1989) Macro and micro-factors affecting coal-seam quality and distribution in southern

Africa with particular reference to the No. 2 seam, Witbank Coalfield, South Africa. Int J Coal Geol

12:681–731

Falcon RMS, Pinheiro H, Sheperd P (1984) The palynobiostratigraphy of the major coal seams in the

Witbank Basin with lithostratigraphic, chronostratigraphic and palaeoclimatic implications.

Comunicações dos Serv Geol Portugal 70:215–243

Gastaldo RA (1994) The genesis and sedimentation of phytoclasts with examples from coastal

environments. In: Traverse A (ed) Sedimentation of organic particles. Cambridge University Press,

pp 103–127

Götz AE, Ruckwied, K (2014) Palynological records of the Early Permian postglacial climate

amelioration (Karoo Basin, South Africa). Palaeobio Palaeoenv 94(2):229–235

Götz AE, Ruckwied K, Littke R, Hartkopf-Fröder C (2012) Facies, sedimentary organic matter content

and coal petrography of the No. 2 coal seam, northern Witbank Basin (South Africa). Proceedings

of the 34th International Geological Congress, Abstract 3383 (CD-ROM, ISBN 978-0-646-57800-

2)

Goldberg K (2004) Floral diversity in the assessment of paleoclimate in the Paraná Basin, southern

Brasil. J Geol 112:719–727

Hancox JP, Götz AE (2014) South Africa’s coalfields – a 2014 perspective. Int J Coal Geol 132:170–

254

Hart GF (1967) Micropalaeontology of the Karroo deposits in South and Central Africa. In: I.U.G.S.

Secretariat (ed) Reviews of the first symposium on Gondwana stratigraphy, Haarlem, Netherlands,

pp 161–172

Iannuzzi R, Souza PA, Holz M (2007) Lower Permian post-glacial succession in the southernmost

Brazilian Paraná Basin: stratigraphy and floral (macro and micro) record. In: Díaz-Martinez E,

Rábano I (ed) 4th European Meeting on the Palaeontology and Stratigraphy of Latin America.

Cuadernos del Museo Geominero 8:207-212

Iannuzzi R, Souza PA, Holz M (2010) Stratigraphic and paleofloristic record of the Lower Permian

postglacial succession in the southern Brazilian Paraná Basin. GSA Spec Papers 468:113–132

Isbell JL, Cole DI, Catuneanu O (2008) Carboniferous-Permian glaciation in the main Karoo Basin,

South Africa: Stratigraphy, depositional controls, and glacial dynamics. GSA Spec. Papers 441:71–

82

Jasper A, Uhl D, Guerra-Sommer M, Bernardes-de-Oliviera MEC, Machado NTG (2011) Upper

Paleozoic charcoal remains from South America: Multiple evidences of fire events in the coal

bearing strata of the Paraná Basin, Brazil. Palaeogeogr Palaeoclimatol Palaeoecol 306:205–218

Jasper A, Uhl D, Agnihotri D, Tewari R, Pandita SK, Wanderley Benicio JR, Fabbrin Pires E, Stock

Da Rosa ÁA, Bhat GD, Pillai SSK (2016) Evidence of wildfires in the Late Permian

(Changsinghian) Zewan Formation of Kashmir, India. Current Science 110(3):419–423

Johnson MR, Van Vuuren CJ, Visser JNJ, Cole DI, Wickens H de V, Christie ADM, Roberts DL (1997)

The foreland Karoo Basin, South Africa. In: Selley RC (ed) Sedimentary Basins of Africa. Elsevier,

Amsterdam, pp 269–317

Kauffmann M, Jasper A, Uhl D, Meneghini J, Osterkamp IC, Zvirtes G, Fabbrin Pires E (2016)

Evidence for palaeo-wildfire in the Late Permian palaeotropics – Charcoal from the Motuca

Formation in the Parnaíba Basin, Brazil. Palaeogeogr Palaeoclimatol Palaeoecol 450:122–128

Lindström S (1995) Early Permian palynostratigraphy of the northern Heimefrontfjella mountain range,

Dronning Maud Land, Antarctica. Rev Palaeobot Palynol 89:359–415

López-Gamundí OR (1997) Glacial-postglacial transition in the late Paleozoic basins of southern South

America. In: Martini IP (ed) Late Glacial and Postglacial Environmental Changes. OUP, Oxford, pp

147–168

MacRae CS, Aitken GA (1997) Contributions in the field of palaeopalynology at the Bernard Price

Institute, past, present and future. Palaeont Afr 33:37–40

Manfroi J, Uhl D, Guerra-Sommer M, Francischini H, Guillermo Martinelli A, Bento Soares M, Jasper

A (2015) Extending the database of Permian palaeo-wildfire on Gondwana: Charcoal remains from

the Rio do Rasto Formation (Paraná Basin), Middle Permian, Rio Grande do Sul State, Brazil.

Palaeogeogr Palaeoclimatol Palaeoecol 436:77–84

Ruckwied K, Götz AE, Jones, P (2014) Palynological records of the Permian Ecca Group (South

Africa): Utilizing climatic icehouse–greenhouse signals for cross basin correlations. Palaeogeogr

Palaeoclimatol Palaeoecol 413:167–172

Ryan PJ (1968) Stratigraphy of the Ecca Series and Lowermost Beds (Permian) in the Great Karoo

Basin of South Africa. PhD Thesis, University of the Witwatersrand, Johannesburg, pp 1–285

Seward AC, Leslie TN (1908) Permo-Carboniferous Plants from Vereeniging, Transvaal. Quarterly

Journal of the Geological Society 64(1-4):109–126

Snyman CP (1998) Coal. In: Wilson MGC, Anhaeusser CR (ed) The Mineral Resources of South

Africa. Council for Geoscience Handbook 16:136–205

Stephenson MH (2008) A review of the palynostratigraphy of Gondwanan Late Carboniferous to Early

Permian glacigene successions. GSA Spec Pap 441:317–330

Stephenson MH, McLean D (1999) International correlation of Early Permian palynofloras from the

Karoo sediments of Morupule, Botswana. S Afr J Geol 102:3–14

Stephenson MH, Leng MJ, Vane CH, Osterloff PL, Arrowsmith C (2005) Investigating the record of

Permian climate change from argillaceous sediments, Oman. J Geol Soc London 162:641–651

Stephenson MH, Angiolini L, Leng MJ (2007) The Early Permian fossil record of Gondwana and its

relationship to deglaciation: A review. In: Williams M, Haywood AM, Gregory FJ, Schmidt DN

(eds) Deep-Time Perspectives on Climate Change: Marrying the Signal from Computer Models and

Biological Proxies. TMS Spec Publ, Geol Soc London, pp 169–189

Traverse A (2007) Paleopalynology, 2nd ed, Springer, pp 487

Tyson RV (1995) Sedimentary Organic Matter: Organic Facies and Palynofacies. London: Chapman &

Hall.

Vidal G (1988) A palynological preparation method. Palynology 12:215–220

Visser JNJ (1986) Lateral lithofacies relationships in the glacigene Dwyka Formation in the western

and central parts of the Karoo Basin. Transactions of the Geological Society of South Africa 89:373–

383

Wheeler A, Götz AE (2016) Palynofacies patterns of the Highveld coal deposits (Karoo Basin, South

Africa): Clues to reconstruction of palaeoenvironment and palaeoclimate. Acta Palaeobot 56:3–15

Wopfner H (1999) The Early Permian deglaciation event between East Africa and northwestern

Australia. J Afr Earth Sci 29:77–90

Figure Captions

Fig. 1. Climate reconstruction of the Main Karoo Basin based on pollen-producing flora (from Hancox

and Götz 2014; modified after Falcon 1986).

Fig. 2. Distribution of the five main stratigraphic groups (Dwyka, Ecca, Beaufort, Stormberg,

Drakensberg) of the Main Karoo Basin and the location of this study (modified after Catuneanu et al.

1998).

Fig. 3. Map of sampling locations in the Witbank Coalfield and positions of the five major palaeovalleys

within the coalfield (Grootvlei, Vischkuil, Coronation, Bank and Arnot valleys). Dashed line marks the

Smithfield Ridge, separating the Witbank and Highveld coalfields (modified from Hancox and Götz

2014).

Fig. 4. Relative abundances of palynomorphs, AOM and DOM at sampling localities Palesa,

HWH1418, BHS14 and ALBN11. AOM – Amorphous Organic Matter. DOM – Degraded Organic

Matter. Palaeoenvironment and palaeoclimate interpretation of the No. 2 Coal Seam inferred from

changes in relative abundances of palynomorphs, AOM, DOM, and freshwater algae.

Fig. 5. Spore/pollen, opaque/translucent phytoclast and equidimensional/blade-shaped phytoclast ratios

at sampling localities Palesa, HWH1418, BHS14 and ALBN11. Palaeoenvironment and palaeoclimate

interpretation of the No. 2 Coal Seam inferred from changes in ratios of spores and pollen grains, and

phytoclasts.

Fig. 6. Schematic reconstruction of the floral composition of the Witbank Coalfield. Ferns, Cordaites

and hinterland conifers in the Lower Coal Seam (a) give way to a diverse Glossopteris-Gangamopteris

flora in the Upper Coal Seam (b).

Fig. 7. Palynofacies of the No. 2 Coal Seam, Witbank Coalfield. (a) Lower Seam (HWH1418). High

amount of opaque phytoclasts and spores. op = opaque phytoclast, sp = spore, bs = bisaccate pollen

grain. (b) Upper Seam (BHS14). High amount of large, opaque blade-shaped phytoclasts and striate

bisaccate pollen grains. opb = opaque blade-shaped phytoclast, sbs = striate bisaccate pollen grain. (c)

Upper Seam, lake setting (Palesa). Phytoclasts of various sizes and shapes. ope = opaque

equidimensional, opb = opaque blade-shaped, tre = translucent equidimensional, trb = translucent blade-

shaped. (d) Intra seam, fluvial setting (ALBN11). High amount of opaque equidimensional phytoclasts.

ope = opaque equidimensional phytoclast, bs = bisaccate pollen grain. (e) Lower Seam (BHS14). High

amount of cuticles, spores, and pollen grains (monosaccate and bisaccate). bs = bisaccate pollen grain,

ms = monosaccate pollen grain, cu = cuticle, sp = spore. (f) Upper Seam (BHS14). High amount of

opaque phytoclasts and striate bisaccate pollen grains. ope = opaque equidimensional, opb = opaque

blade-shaped, sbs = striate bisaccate pollen grain.

Table 1. Palynofacies countings of No. 2 Coal Seam samples from localities Palesa, HWH1418, BHS14

and ALBN11 (Witbank Coalfield).

Figures

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

a

b

Figure 6

Figure 7

Locality Sample Phytoclasts Palynomorphs

AOM DOM Algae

Total

Counts opaque translucent Pollen Grains Spores

equidim. n-shaped equidim. n-

shaped cuticles monosac. bisac. non-t. bisac. t. monolete trilete

Pale

sa

1 268 76 28 13 9 35 26 0 2 12 12 20 0 501 2 285 83 18 5 4 42 21 0 4 10 16 15 0 503 3 280 70 17 6 3 38 33 0 1 18 15 24 0 505 4 279 65 10 6 3 45 38 1 1 14 19 22 0 503 5 158 36 148 32 1 21 53 10 0 1 2 30 9 501 6 276 45 18 8 5 10 68 37 0 8 12 20 0 507 7 238 56 15 3 8 25 80 42 0 9 15 18 0 509 8 126 10 175 14 2 12 72 56 0 1 1 32 6 507 9 102 15 185 24 1 4 68 58 0 1 2 35 10 505

10 110 12 161 28 1 0 85 65 0 1 1 30 8 502

HWH1

418

1 265 50 53 24 8 10 10 0 10 38 25 10 0 503 2 251 69 46 21 12 15 7 0 6 42 18 15 0 502 3 271 58 35 28 5 11 13 0 3 52 16 9 0 501 4 276 60 39 17 8 7 10 0 5 49 22 12 0 505 5 343 28 24 8 1 8 20 0 1 12 0 58 1 504 6 363 24 20 10 0 5 18 2 2 7 0 52 0 503 7 335 15 31 10 1 5 27 2 2 10 0 67 1 506 8 153 84 168 35 2 1 20 13 0 1 3 18 9 507 9 143 57 152 43 5 3 37 31 0 0 1 25 5 502

10 150 50 177 28 1 1 45 24 1 4 1 16 12 508 11 138 42 186 35 7 1 31 26 1 1 5 20 9 502 12 125 65 162 52 5 0 47 22 1 1 1 21 6 508 13 110 70 165 38 1 0 55 38 0 3 2 15 10 507 14 121 19 190 37 1 0 63 38 0 2 1 23 11 506

BHS1

4

1 209 53 115 38 12 15 1 0 7 28 18 12 0 508 2 211 69 90 25 8 20 8 0 4 41 25 10 0 511 3 225 85 60 19 7 12 13 0 8 48 19 6 0 502 4 200 128 51 24 8 18 9 0 10 37 11 8 0 504 5 248 88 43 26 5 10 9 0 10 35 18 15 0 507 6 207 120 45 30 8 10 7 2 5 42 16 12 0 504 7 150 24 182 53 1 3 18 21 0 3 2 38 11 506 8 238 75 50 20 6 1 35 36 1 12 20 7 0 501 9 241 83 63 12 9 1 28 38 0 5 18 8 0 506

10 200 115 47 18 10 0 47 35 0 2 23 8 0 505

ALBN

11

1 180 48 194 54 3 13 1 0 6 6 0 13 0 518 2 211 52 217 52 2 14 5 0 0 28 0 10 0 591 3 139 49 263 79 1 15 9 0 2 7 0 19 0 583 4 125 29 163 58 7 20 16 3 8 18 5 52 0 504 5 143 36 205 55 1 18 21 0 11 19 1 19 0 529 6 189 46 143 46 1 1 29 2 0 25 3 16 0 501 7 155 44 199 53 2 4 14 0 4 18 1 8 0 502 8 145 37 217 72 1 0 29 1 7 16 0 13 0 538

Table 1