Fungal abundance spike and the Permian–Triassic boundary in the Karoo Supergroup (South Africa)

Abstract

The most severe mass extinction of marine species and terrestrial vertebrates and plants is associated with the Permian–Triassic boundary (∼251 Ma). The extinction interval is also marked by the disappearance of most Late Permian gymnosperm palynomorphs at a layer containing solely the abundant remains of fungi. This ‘fungal spike’ apparently represents widespread devastation of arboreous vegetation. Stratigraphic and palynological study of the Carlton Heights section in the southern Karoo Basin of South Africa revealed a 1-m-thick fungal spike zone that occurs simultaneously with the last appearance of typically Late Permian gymnosperm pollen. The plant extinction and fungal spike zone are found above the last occurrence of Late Permian mammal-like reptiles of the Dicynodont Zone at other Karoo sections. Using the fungal event as a time line in marine and non-marine sections allows placement of the marine extinctions and the extinction of terrestrial plants and reptiles within a brief crisis interval of less than about 40 000 years at the end of the Permian.

Full text

Fungal abundance spike and the Permian^Triassic boundary
in the Karoo Supergroup (South Africa)
Maureen B. Steiner a, Yoram Eshet b;c, Michael R. Rampino d;e;,
Dylan M. Schwindt d
aDepartment of Geology and Geophysics, University of Wyoming, Laramie, WY 82071, USA
bGeological Survey of Israel, Jerusalem 95501, Israel
cTel Hai Academic College, Tel Hai 12210, Israel
dEarth and Environmental Science Program, New York University, New York, NY 10003, USA
eNASA, Goddard Institute for Space Studies, New York, NY 10025, USA
Received 22 March 2002; accepted 31 December 2002
Abstract
The most severe mass extinction of marine species and terrestrial vertebrates and plants is associated with the
Permian^Triassic boundary (V251 Ma). The extinction interval is also marked by the disappearance of most Late
Permian gymnosperm palynomorphs at a layer containing solely the abundant remains of fungi. This ‘fungal spike’
apparently represents widespread devastation of arboreous vegetation. Stratigraphic and palynological study of the
Carlton Heights section in the southern Karoo Basin of South Africa revealed a 1-m-thick fungal spike zone that
occurs simultaneously with the last appearance of typically Late Permian gymnosperm pollen. The plant extinction
and fungal spike zone are found above the last occurrence of Late Permian mammal-like reptiles of the Dicynodont
Zone at other Karoo sections. Using the fungal event as a time line in marine and non-marine sections allows
placement of the marine extinctions and the extinction of terrestrial plants and reptiles within a brief crisis interval of
less than about 40 000 years at the end of the Permian.
 2003 Elsevier Science B.V. All rights reserved.
Keywords: Permian^Triassic boundary; extinction; fungal spike; vertebrates ; South Africa
1. Introduction
The end-Permian mass extinction eliminated
more than 90% of marine species (Raup, 1979 ;
Jin et al., 2000). Terrestrial biota also su¡ered
dramatically: an estimated 70% of terrestrial ver-
tebrate families were eradicated (Maxwell, 1992),
insects su¡ered a major loss of taxa (Labandiera
and Sepkoski, 1993), and more than 90% of Late
Permian gymnosperm species died out (Retallack,
1995; Visscher et al., 1996; Looy et al., 1999).
The plant extinction is evidenced by the disap-
pearance of almost all Late Permian gymnosperm
pollen at a horizon containing only fungal re-
mains and woody debris (Visscher et al., 1996).
This fungal abundance event was followed by ap-
pearance of an Early Triassic palyno£ora domi-
nated by lycopod spores and bisaccate gymno-
0031-0182 / 03 / $ ^ see front matter  2003 Elsevier Science B.V. All rights reserved.
doi: 10.1016/S0031-0182(03)00230-X
* Corresponding author. Fax: +1-212-995-4015.
E-mail address: mrr1@nyu.edu (M.R. Rampino).
PALAEO 3036 28-4-03
Palaeogeography, Palaeoclimatology, Palaeoecology 194 (2003) 405^414
www.elsevier.com/locate/palaeo

sperm pollen (Eshet et al., 1995; Visscher et al.,
1996). The widespread plant extinction and sub-
sequent £ood of fungal remains has been inter-
preted as indicating destruction of terrestrial veg-
etation and accumulation of decaying organic
debris (Ouyang and Utting, 1990; Eshet et al.,
1995; Visscher et al., 1996; Poort et al., 1997).
This fungal proliferation event has been ob-
served both in terrestrial and shallow marine se-
quences. In marine sections, the brief interval rich
in fungal remains is found close to the level
marked by the mass extinction of marine organ-
isms (Visscher and Brugman, 1986; Visscher et
al., 1996; Twitchett et al., 2001). An abrupt neg-
ative shift in carbon isotopes in the oceans also
occurs close to the time of the reduction in gym-
nosperms (Looy et al., 2000), and the zone
marked exclusively by abundant fungal spores
(Visscher et al., 1996; Wignall et al., 1996). In
some Permian^Triassic sections enrichment in
fungal remains occur at other levels (commonly
the tops of regressive subtidal cycles; Cirilli et
al., 1998), but these are not associated with the
major disappearance of Late Permian pollen, and
the concentration of fungal remains does not
reach the V100% levels seen in the end-Permian
abundance spike.
The widespread end-Permian fungal spike could
provide a timeline for correlating the marine and
non-marine records. A good place to establish this
correlation is the Upper Permian and Lower Tri-
assic Beaufort Group of the Karoo Supergroup of
South Africa, which is well known for its record
of the succession of mammal-like reptiles across
the Permian^Triassic (P^T) boundary (Kitching,
1977; Rubidge, 1995). The demise of the herbiv-
orous dicynodonts of the Dicynodon assemblage
zone, and their replacement by the Lystrosaurus
assemblage fauna has served in the past as the
approximate de¢nition of the P^T boundary in
the Karoo, although the ¢rst occurrence of Ly-
strosaurus is now known to precede the last oc-
currence of Dicynodon (e.g. Smith, 1990, 1995;
Smith and Ward, 2001).
The Beaufort Group consists of an apparently
uninterrupted succession of alluvial sedimentation
(Catuneanu and Elango, 2001). Sediments shed
from the Cape Fold Belt region produced a £uvial
network which prograded across the Karoo Basin
during Late Permian time. Lacustrine mudstones,
£uvial overbank mudstones and channel sand-
stones make up a sedimentary sequence up to
6 km thick in the Karoo Basin in South Africa.
A change from predominantly green mudstone
and sandstone deposited by high sinuosity river
systems to multistoried channel and sheet sand-
stones intercalated with maroon mudstones, typi-
cal of deposition by braided streams, occurs close
to the P^T boundary (as de¢ned by the vertebrate
assemblages) (Smith, 1990, 1995; Ward et al.,
2000; Smith and Ward, 2001).
We studied the stratigraphy and palynology of
the well-exposed Carlton Heights section, located
along the Graa¡-Reinet^Colesburg highway be-
tween Middleburg and Neupoort, South Africa
(Fig. 1). Samples were collected in gullies to the
southeast and below the main highway (on the
B.P. Erasmus farm, ‘Beskuitfontein’), and along
the main highway itself (the road is at the 57 m
level in our section) about 500 m north of the
Carlton Heights railway stop (GPS : 31‡13.03PS,
24‡56.96PE).
2. Stratigraphy of the Carlton Heights
section
At the Carlton Heights locality (Fig. 2), £at
lying, predominantly greenish mudstone, siltstone
and thin tabular sandstones dominate the lower
part of the section. These were previously mapped
as Upper Permian (Balfour Formation) (Keyser,
1977) based on lithology and stratigraphic posi-
tion relative to the subsequent widespread change
from mudstone to dominantly sandstone facies. A
Late Permian age for these deposits is also sug-
gested by our discovery of a large hip bone and
other skeletal elements possibly belonging to Di-
cynodon sp. in the lower part of the section (be-
tween 2.8 and 10.5 m above the base of the sec-
tion) (Fig. 2).
At about 33.5 m above the base of the Carlton
Heights section, the green mudstone/sandstone se-
quence is overlain by V15 m of laminated to
massive maroon mudstone with occasional thin
sands (Fig. 3; well-exposed in the railroad cut
PALAEO 3036 28-4-03
M.B. Steiner et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 194 (2003) 405^414406

just to the south of our section), capped by a 4^
5 m-thick grayish^green sand unit. Lystrosaurus
¢rst occurs in our section at 38 m above the
base. Based on lithology and stratigraphic posi-
tion, we correlate this laminated maroon mud-
stone with a similar unit recently identi¢ed by
Smith and Ward (2001) at the Bethulie and Loots-
berg Pass sections in the Karoo. An increased
number of Lystrosaurus fossils were noted near
the top of this unit at Carlton Heights (V51 m
above the base of the section).
At V56 m above the base of the Carlton
Heights section, the maroon mudstone and sand-
stone unit grades upward into thin-bedded alter-
nating green and red siltstones and ¢ne sand-
stones showing abundant sub-horizontal cylindri-
cal burrows (Fig. 4). The thin-bedded burrowed
unit is overlain (at 58.5 m above the base of the
section) by a thin (V5-cm-thick), very ¢ne-
grained clay-rich layer. This laterally continuous
(on outcrop scale) layer is heavily burrowed, and
is marked by red and yellow^brown alteration
products (Fig. 4). Clay-mineral analysis by stan-
dard semi-quantitative X-ray di¡raction methods
(Brindley and Brown, 1980) shows that the layer
is composed predominantly of illite and illite^
smectite. The layer also contains quartz, low al-
bite, gypsum, chlorite, mica, and jarosite, and
thus has a heavily weathered detrital signature.
About 0.5 m above this marker layer (at 59 m
above the base of the section), the ¢rst laterally
widespread multistoried sandstone with an ero-
sional base containing lenses of mud-pebble and
carbonate-nodule conglomerates (identi¢ed with
the base of the Katberg Formation) is encoun-
tered (Figs. 2 and 4).
The Katberg Formation at Carlton Heights
(s270 m in total thickness) is represented by a
facies dominated by gray and white ¢ne- to
coarse-grained sandstone that is typically multi-
storied and laterally extensive (Fig. 4). The sands
commonly show scoured bases with lenses of in-
tra-formational mud-pebble and pedogenic car-
bonate-nodule conglomerates, horizontal strati¢-
cation, and large-scale trough cross strati¢cation
(Smith, 1995). The thick multistoried sands are
interbedded with thin red mudstone units showing
desiccation features, such as sand-¢lled mud-
cracks.
3. Palynology
We sampled a total of 73 m at 0.5^3-m intervals
Fig. 1. Location map showing the Carlton Heights locality, Karoo Basin, South Africa.
PALAEO 3036 28-4-03
M.B. Steiner et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 194 (2003) 405^414 407

(Fig. 2). Palynological slides were prepared for
microscopic study using standard procedures
(Doher, 1980). Out of the 29 samples that were
analyzed for palynomorphs, only seven were bar-
ren. The palynomorph species distribution within
the sampled section is shown in Fig. 5. Three
palynological assemblage zones were identi¢ed in
the Carlton Heights section:
Fig. 2. Stratigraphic section across the P^T boundary in Karoo Supergroup strata at Carlton Heights, South Africa. The upper-
most 11 m in the Katberg Formation sandstones studied are not shown, as they were barren of palynofossils. Asterisks indicate
samples analyzed for pollen and spores. Dicynodon sp.? indicates skeletal material possibly belong to Dicynodon sp.
PALAEO 3036 28-4-03
M.B. Steiner et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 194 (2003) 405^414408

(1) The Late Permian Klausipollenites schauber-
geri Zone, dominated by taxa of the form genera
Protohaploxypinus and Falcisporites.
(2) An interval composed almost entirely of
fungal cell remains (Reduviasporonites or its junior
synonyms Chordecystia or Tympanicysta)(Vis-
scher et al., 1996) and abundant recycled woody
material ^ the fungal spike zone (Fig. 5). The fungal
spike interval is only V1 m thick (57.6^58.6 m
above the base of the section) (Fig. 2). (We note
that some researchers have interpreted Tympani-
cysta as a green alga (Afronin et al., 2001), but
most workers agree on the fungal interpretation
(Visscher et al., 1996.)
(3) The fungal spike is followed by the Early
Triassic Kraeuselisporites^Lunatisporites Zone,
dominated by species of the lycopod Kraeuseli-
sporites and the bisaccate pollen Lunatisporites
and Platysaccus.
The interval from V49 m to 51.2 m in the
section was found to be barren of palynomorphs,
and eight of the 17 Late Permian palynomorph
taxa that we identi¢ed last occur at or below
this barren zone. The other nine Late Permian
taxa have last occurrences at or just below the
base of the fungal spike zone (Fig. 5). The last
occurrences just prior to the lower barren zone
could represent an initial decrease in plant diver-
sity as part of a crisis period extending over some
tens of thousands of years. On the other hand, the
Fig. 3. Laminated to massive maroon mudstone with thin siltstone interbeds along the railway cut northeast of the Carlton
Heights railway stop (person for scale) (see Fig. 2). This lithologic unit is believed to be correlative with the P^T event beds of
Smith and Ward (2001) described from the Bethulie and Lootsberg Pass sections in the Karoo.
PALAEO 3036 28-4-03
M.B. Steiner et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 194 (2003) 405^414 409

reduction in palynomorphs could be the result of
poor preservation of pollen and spores in sandy
deposits.
4. The fungal spike and the end-Permian mass
extinction
The characterization of the P^T boundary in-
terval by a severe land-plant extinction and an
abrupt, short-lived £ood of fungal remains is
now apparent from numerous studies around
the world (Visscher and Brugman, 1986; Eshet
et al., 1995; Visscher et al., 1996). Previous stud-
ies of Karoo Supergroup rocks produced a paly-
nozonation scheme for this time span that showed
evidence of a major turnover or extinction of
palynomorphs at or near the P^T boundary (as
de¢ned by the vertebrate assemblages) (Anderson,
1977; Stapleton, 1978; Utting, 1979; Nyambe and
Utting, 1997). In sub-Equatorial Africa, a fungal
abundance spike has been reported in sections
spanning the P^T boundary from Kenya (Hankel,
1992) and Madagascar (Wright and Askin, 1987).
The widespread fungal proliferation near the
P^T boundary has been interpreted by a number
of workers as re£ecting loss of arboreous vegeta-
tion on a large scale, a major decrease in standing
biomass, and the build-up of decaying vegetation
on land (Visscher and Brugman, 1986; Visscher et
al., 1996). In the same interval, the plant macro-
fossil record shows the extinction of the Glosso-
pteris £ora across Gondwana, and the disappear-
ance of related Vertebraria root traces (Retallack,
1995). Recovery from the extinction and renewed
diversi¢cation in land-plants were relatively slow,
taking about 4 Myr (Eshet et al., 1995 ; Looy et
al., 1999).
The £ood of fungal remains was apparently a
short-lived event. In the Carlton Heights se-
quence, the zone marked exclusively by fungal
remains spans only V1 m of the sedimentary rec-
KATBERG
SANDSTONE
clay-rich layer
Fungal Spike
BALFOUR
FORMATION
Fig. 4. The fungal spike zone just below the base of the Katberg Formation on the Graa¡-Reinet^Colesburg highway at Carlton
Heights. The fungal spike zone, showing thin-bedded strati¢cation and sub-horizontal burrowing, is V1 m thick (from 57.5 to
58.5 m in the section). The fungal spike zone is capped by a clay-rich layer with red to yellow^brown alteration. The one meter
white bar is for scale.
PALAEO 3036 28-4-03
M.B. Steiner et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 194 (2003) 405^414410

ord (Figs. 2 and 5). At minimum estimated accu-
mulation rates for the sediments of the Balfour
Formation, this would mean 92000 years dura-
tion for the episode, with burrowing and the
mixed depositional regimes probably making this
an upper limit. Sedimentologic evidence of the
dramatic loss of terrestrial vegetation ^ the
marked lithologic change to the ¢rst thick, multi-
storied channel and sheet sandstones typical of
braided streams (Katberg Formation) (Ward et
al., 2000) ^ occurs about 50 cm (estimated as
91000 years) above the fungal spike zone at
Carlton Heights (Fig. 4).
The results of several recent studies allow us to
compare the timing of the extinction of terrestrial
plants and reptiles with the marine extinctions.
Twitchett et al. (2001) studied a marine section
in Greenland that contained abundant and well-
preserved marine fauna as well as terrestrial paly-
nomorphs. The sediments also recorded the neg-
ative excursion in N13C in marine carbonate and
organic carbon. Based on estimated sedimentation
rates, Twitchett et al. (2001) concluded that the
marine and terrestrial ecosystem collapse occurred
over the same stratigraphic interval and took just
a few tens of thousands of years. The faunal and
£oral extinctions were apparently coeval with the
initial negative shift in N13C. The rapid N13C shift
could be a result of the rapid loss of primary
productivity in the oceans and on land (Caldeira
and Rampino, 1993), and the enhanced delivery
of light carbon (including terrestrial plant debris)
to the ocean £oor (Broecker and Peacock, 1999 ;
Sephton et al., 2002).
Woody debris
Densoisporites complicatus
Densoisporites playfordii
Falcisporites stabilis
Falcisporites zapfei
Klausipollenites schaubergeri
Protohaploxypinus limpidus
Protohaploxypinus microcorpus
Protohaploxypinus richteri
Protohaploxypinus samoilovichii
Punctatosporites sp.
Reticuloidosporites warchianus
Triplexisporites playfordii
Guthoerlisporites cancellosus
Protohaploxypinus varius
Vittatina ovalis
Laevigatosporites callosus
Protohaploxypinus jacobii
Lueckisporites singhii
Limitisporites sp.
Fungal cells
Lunatisporites noviaulensis
Kraeuselisporites cuspidus
Kraeuselisporites sp.
Lunatisporites pellucidus
Lunatisporites transversundatus
Platysaccus leschickii
Platysaccus quenslandi
Platysaccus papilionis
Lundbladispora brevicula
Lunatisporites sp.
72.3
71.9
59.5
59.04
59 ||
58.95 || | |
58.8 ||||
58.75 |||
58.7 |
58.64
58.59
58.3
58.2
58.1
57.6
55.5
54.34 | |
52.64 | | | |
51.2 | | | | | | | | |
50.7 | | | | | | | | |
50.3 | | | | | | | | |
49 | | | | | | | | |
48.5 | | | | | | | |
38.3 | | | | | |
33.35 | |||| | | | |
26.6 | | | |||| | ||
23.6 | || | | ||||||
10.6 | | | | | | | |
0.35
0 || |||||
Sample Depth (m)
Zone
Age
Kraeuselisporites-
Lunatisporites spp.
Fungal
Spike
Klausipollenites
schaubergeri
Early Triassic
P-T
Boundary
Z
one
Late Permian
BARREN (sandstone)
BARREN
1.0
Fungal Spike Zone
Thickness (m)
72.3
Fig. 5. Distribution chart (not to scale) of palynomorphs identi¢ed in the Carlton Heights stratigraphic section. The fungal spike
interval, marked exclusively by fungal remains and woody debris, is V1 m in thickness.
PALAEO 3036 28-4-03
M.B. Steiner et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 194 (2003) 405^414 411

A similarly negative N13C isotope shift has been
reported from some terrestrial sections (Morante,
1996; Krull and Retallack, 2000), and from mo-
lecular fossils in land-plant leaf cuticles deposited
in marine sediments (Sephton et al., 2002), most
likely re£ecting a synchronous atmospheric car-
bon isotope shift. Recently, a negative excursion
in carbon isotope ratios has been reported from
carbonate soil nodules and bone material from
the Bethulie section in the Karoo Basin (MacLeod
et al., 2000). Thus, based on the results of Twitch-
ett et al. (2001), the negative excursion in N13Cin
the Karoo section should be very close to the time
of the marine extinction and the devastation of
terrestrial ecosystems.
At Bethulie, the initiation of the N13C excursion
at about 45 m in the section coincides with the
local ¢rst appearance of Lystrosaurus; the N13 C
values begin to return to their former levels after
the last appearance of Dicynodon at 57 m (Mac-
Leod et al., 2000). The carbon isotope anomaly
corresponds to the base of the laminated maroon
mudstone event beds in which Smith and Ward
(2001) place the P^T boundary. At average sedi-
mentation rates for Karoo deposits (50 cm/1000
years) the interval of the laminated event beds
and the overlap of Dicynodont sp. and Lystrosau-
rus would be about 25 000 years.
At the Carlton Heights locality, the base of the
fungal spike zone is about 20 m above the base of
the maroon mudstone event beds (Fig. 2). At
average sedimentation rates for Karoo deposits,
the base of the P^T event beds could be about
40 000 years prior to the fungal spike but these
calculations are somewhat uncertain.
5. Conclusions
Palynological study at Carlton Heights identi-
¢ed a 1-m-thick zone containing only abundant
fungal remains and woody debris coincident
with the last appearance of typically Late Permian
gymnosperm palynomorphs. The zone apparently
represents proliferation of fungi upon large vol-
umes of decaying plant matter. This fungal spike
occurs just below the ¢rst Katberg Sandstone,
which signi¢es a basin-wide change to braided
stream patterns, probably related to the wide-
spread loss of vegetation (Ward et al., 2000).
The latest Permian £ood of fungal remains
might serve as a widespread marker bed of brief
duration in marine and non-marine deposits. In
marine sections, the fungal spike has been esti-
mated to be coeval with the marine extinction
and the negative shift in carbon isotopes that oc-
curred at the end of the Permian (Twitchett et al.,
2001). The discovery of the fungal spike in the
classic fossiliferous Karoo sequence, within the
interval of dramatic faunal turnover in terrestrial
vertebrates and land-plants, allows correlation of
the terrestrial and marine mass extinctions at the
P^T boundary.
The stratigraphy at Carlton Heights, when
combined with the recent work of Smith and
Ward (2001) on other Karoo sections, suggests
that the disappearance of typically Late Permian
vertebrates in the Karoo Basin and the ¢nal gym-
nosperm plant extinction and fungal spike zone
took place over an interval of less than 40 000
years.
Acknowledgements
We thank John Hancox, Isabel Montanez, and
Neil Tabor for help in the ¢eld, Robert C. Rey-
nolds at Dartmouth College for X-ray di¡raction
mineral analyses, Henk Brinkhuis, Roger M.H.
Smith and Peter D. Ward for helpful discussions
and information, and Simonetta Cirilli and Eve-
lyn Krull for critical reviews. M.R.R. was sup-
ported in part by a New York University Re-
search Challenge Grant. We are grateful to the
B.P. Erasmus and G. Van Zyl families for permis-
sions and assistance in the Karoo.
References
Afronin, S.A.E., Barinova, S., Krassilov, V.A., 2001. A bloom
of Tympanicysta Balme, 1980 (green algae of zygnematalean
a⁄nities) at the Permian^Triassic boundary. Geodiversitas
23, 481^487.
Anderson, J.M., 1977. The biostratigraphy of the Permian and
Triassic: Part 3, A review of Gondwana Permian palynology
PALAEO 3036 28-4-03
M.B. Steiner et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 194 (2003) 405^414412

with particular reference to the northern Karoo Basin,
South Africa. Mem. Bot. Surv. S. Afr. 41, 1^300.
Brindley, G.W., Brown, G. (Eds.), 1980. Crystal Structures of
Clay Minerals and Their X-Ray Identi¢cation. Mineralogi-
cal Society, London.
Broecker, W.S., Peacock, S., 1999. An ecologic explanation for
the Permo^Triassic carbon and sulfur isotope shifts. Glob.
Biogeochem. Cycles 13, 1167^1172.
Caldeira, K., Rampino, M.R., 1993. Aftermath of the end-
Cretaceous mass extinction: Possible biogeochemical stabi-
lization of the carbon cycle and climate. Paleoceanography
8, 515^525.
Catuneanu, O., Elango, H.N., 2001. Tectonic control on £u-
vial styles: The Balfour Formation of the Karoo Basin,
South Africa. Sediment. Geol. 140, 291^313.
Cirilli, S., Radrizzani, C.P., Ponton, M., Radrizzani, S., 1998.
Stratigraphical and palaeoenvironmental analysis of the
Permian^Triassic transition in the Badia Valley (Southern
Alps, Italy). Palaeogeogr. Palaeoclimatol. Palaeoecol. 138,
85^113.
Doher, L., 1980. Palynomorph preparation procedures cur-
rently used in the paleontology and stratigraphy laborato-
ries. U.S. Geological Survery Circular 830, 29 pp.
Eshet, Y., Rampino, M.R., Visscher, H., 1995. Fungal event
and palynological record of ecological crisis and recovery
across the Permian^Triassic boundary. Geology 23, 967^
970.
Hankel, O., 1992. Late Permian to Early Triassic micro£oral
assemblages from the Maji Ya Chumvi Formation, Kenya.
Rev. Palaeobot. Palynol. 72, 129^147.
Jin, Y.G., Wang, Y., Wang, W., Shang, Q.H., Cao, C.Q.,
Erwin, D.H., 2000. Pattern of marine mass extinction near
the Permian^Triassic boundary in South China. Science 289,
432^436.
Keyser, N., 1977. Geological Map of the Republic of South
Africa and the Kingdoms of Lesotho and Swaziland. South
African Council for Geoscience, Johannesburg.
Kitching, J.W., 1977. The distribution of the Karroo verte-
brate fauna. Mem. Bernard Price Inst. Palaeontol. Res. (Jo-
hannesburg) 1, 1^131.
Krull, E.S., Retallack, G.J., 2000. Delta C-13 depth pro¢les
from paleosols across the Permian^Triassic boundary: Evi-
dence for methane release. Geol. Soc. Am. Bull. 112, 1459^
1472.
Labandiera, C.C., Sepkoski, J.J., Jr., 1993. Insect diversity in
the fossil record. Science 261, 310^315.
Looy, C.V., Brugman, W.A., Dilcher, D.L., Visscher, H.,
1999. The delayed resurgence of equatorial forests after
the Permian^Triassic ecologic crisis. Proc. Natl. Acad. Sci.
USA 96, 13857^13862.
Looy, C.V., Twitchett, R.J., Dilcher, D.L., Van Konijnenburg-
Van Cittert, J.H.A., Visscher, H., 2000. Life in the end-Per-
mian dead zone. Proc. Natl. Acad. Sci. USA 98, 7879^7883.
MacLeod, K.G., Smith, R.M.H., Koch, P.L., Ward, P.D.,
2000. Timing of mammal-like reptile extinctions across the
Permian^Triassic boundary in South Africa. Geology 28,
227^230.
Maxwell, W.D., 1992. Permian and Early Triassic extinction of
nonmarine tetrapods. Paleontology 35, 571^583.
Morante, R., 1996. Permian and early Triassic isotopic records
of carbon and strontium in Australia and a scenario of
events about the Permian^Triassic boundary. Hist. Biol.
11, 289^310.
Nyambe, I.A., Utting, J., 1997. Stratigraphy and palynostra-
tigraphy, Karoo Supergroup (Permian and Triassic), mid-
Zambezi Valley, southern Zambia. J. Afr. Earth Sci. 24,
563^583.
Ouyang, S., Utting, J., 1990. Palynology of Upper Permian
and Lower Triassic rocks, Meishan, Changxing County,
Zhejiang Province, China. Rev. Palaeobot. Palynol. 66,
65^103.
Poort, R.J., Clement-Westerhof, J.A., Looy, C.V., Visscher,
H., 1997. Aspects of Permian palaeobotany and palynology.
17. Conifer extinction in Europe at the Permian^Triassic
junction: Morphology, ultrastructure and geographic/strati-
graphic distribution of Nuskoisporites dulhuntyi (prepollen of
Ortiseia,Walchiaceae). Rev. Palaeobot. Palynol. 97, 9^39.
Raup, D.M., 1979. Size of the Permo^Triassic bottleneck and
its evolutionary implications. Science 206, 217^218.
Retallack, G.J., 1995. Permian^Triassic life crisis on land. Sci-
ence 267, 77^80.
Rubidge, B.S. (Ed.), 1995. Biostratigraphy of the Beaufort
Group (Karoo Supergroup). Council for Geoscience, Geo-
logical Survey of South Africa, SACS Biostratigraphic Ser-
vices, Johannesburg, 1, 46 pp.
Sephton, M.A., Looy, C.V., Veefkind, R.J., Brinkhuis, H., De
Leeuw, J.W., Visscher, H., 2002. A synchronous record of
N13C shifts in the oceans and atmosphere at the end of the
Permian. Geological Society of America Special Paper.
Smith, R.M.H., 1990. A review of stratigraphy and sedimen-
tary environments of the Karoo Basin of South-Africa.
J. Afr. Earth Sci. 10, 117^137.
Smith, R.M.H., 1995. Changing £uvial environments across
the Permian^Triassic boundary in the Karoo Basin, South-
Africa and possible causes of tetrapod extinctions. Palaeo-
geogr. Palaeoclimatol. Palaeoecol. 117, 81^104.
Smith, R.M.H., Ward, P.D., 2001. Pattern of vertebrate ex-
tinctions across an event bed at the Permian^Triassic bound-
ary in the Karoo Basin of South Africa. Geology 29, 1147^
1150.
Stapleton, R.P., 1978. Micro£ora from a possible Permo^Tri-
assic transition in South Africa. Rev. Palaeobot. Palynol. 25,
253^258.
Twitchett, R.J., Looy, C.J., Morante, R., Visscher, H.,
Wignall, P.B., 2001. Rapid and synchronous collapse of ma-
rine and terrestrial ecosystems during the end-Permian biotic
crisis. Geology 29, 351^354.
Utting, J., 1979. Pollen and spore assemblages from the Upper
Permian of the North Luangava Valley, Zambia. Proceed-
ings of the 4th International Palynology Conference, 2,
Moscow, pp. 165^174.
Visscher, H., Brugman, W.A., 1986. The Permian^Triassic
boundary in the Southern Alps : A palynological approach.
Mem. Soc. Geol. Ital. 34, 121^128.
PALAEO 3036 28-4-03
M.B. Steiner et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 194 (2003) 405^414 413

Visscher, H., Brinkhuiss, H., Dilcher, D.L., Elsik, W.C., Eshet,
Y., Looy, C.V., Rampino, M.R., Traverse, A., 1996. The
terminal Paleozoic fungal event: Evidence of terrestrial eco-
system destabilization and collapse. Proc. Natl. Acad. Sci.
USA 93, 2155^2158.
Ward, P.D., Montgomery, D.R., Smith, R., 2000. Altered river
morphology in South Africa related to the Permian^Triassic
extinction. Science 289, 1740^1743.
Wignall, P.B., Kozur, H., Hallam, A., 1996. On the timing of
palaeoenvironmental changes at the Permo^Triassic (P/Tr)
boundary using conodont biostratigraphy. Hist. Biol. 12,
39^62.
Wright, R.P., Askin, R.A., 1987. The Permian^Triassic bound-
ary in the Southern Morondava Basin of Madagascar as
de¢ned by plant microfossils. Geophys. Monogr. 41, 157^
166.
PALAEO 3036 28-4-03
M.B. Steiner et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 194 (2003) 405^414414