Land-plant diversity and the end-Permian mass extinction
P. McAllister Rees*
Department of the Geophysical Sciences, University of Chicago, 5734 South Ellis Avenue, Chicago, Illinois 60637, USA
ABSTRACT
The Permian and Triassic represent a time of major global climate change from
icehouse to hothouse conditions and significant (;258) northward motion of landmasses
amalgamated in essentially one supercontinent, Pangea. The greatest of all mass extinctions occurred around the Permian-Triassic boundary (251 Ma), although there is no
consensus regarding the cause(s). Recent studies have suggested a meteor impact and
worldwide die-off of vegetation, on the basis of sparse local observations. However, new
analyses of global Permian and Triassic plant data in a paleogeographic context show
that the scale and timing of effects varied markedly between regions. The patterns are
best explained by differences in geography, climate, and fossil preservation, not by catastrophic events. Caution should be exercised when extrapolating local observations to
global-scale interpretations. At the other extreme, global compilations of biotic change
through time can be misleading if the effects of geography, climate, and preservation bias
are not considered.
Keywords: mass extinctions, Permian, Triassic, biodiversity, paleogeography, paleoclimate.
INTRODUCTION
The Permian and Triassic represent a time of global climate
change from icehouse to hothouse conditions, possibly caused by a rise
in atmospheric CO2 to around five times present levels (e.g., Berner
and Kothavala, 2001). The Pangean supercontinent (Fig. 1) typifies
Permian and Triassic geography. It migrated ;258 northward through
these intervals, such that the Early Permian north and south polar regions were ocean and land, respectively (Fig. 1B), but the opposite was
true by the Late Triassic (Fig. 1C). A number of microcontinents (e.g.,
North and South China) traversed low latitudes during the Permian,
but most of these collided with the main Pangean landmass by the Late
Triassic.
Permian terrestrial climate zones (biomes) have been determined
from fossil floras and climate-sensitive sedimentary rocks (e.g., Ziegler,
1990; Gibbs et al., 2002; Rees et al., 2002). Climates were well differentiated from the equator to poles, although extensive continental
ice sheets were absent after the Early Permian and did not redevelop
until the late Cenozoic (e.g., Markwick and Rowley, 1998), indicating
a latitudinal temperature gradient less severe than at present. A complete spectrum of temperate biomes and subtropical deserts existed in
both hemispheres, but equatorial coals and tropical rainforests were
only well developed throughout the Permian on the Chinese microcontinents. Farther west, in Euramerica (Fig. 1), climate conditions
changed gradually during the Early Permian such that the main coalforming plants were replaced by ones adapted to drier conditions. Geography and climates of the Late Triassic were similar to those of the
Early Jurassic, which has been the subject of modeling and data based
global climate studies (e.g., Chandler et al., 1992; Rees et al., 2000).
In contrast to the Permian, all equatorial regions were only seasonally
wet while extensive forests and coal swamps occupied the temperate
regions.
Recent studies have suggested an extraterrestrial meteor impact
and worldwide catastrophic die-off of vegetation at the PermianTriassic boundary (Ward et al., 2000; Becker et al., 2001; Kaiho et al.,
2001). However, the evidence for a meteor impact is equivocal (see
discussion by Farley et al., 2001), as is that for a global loss of vegetation. Ward et al. (2000) documented a change from meandering to
*E-mail: rees@geosci.uchicago.edu.
Figure 1. Global paleogeographic maps (modified from Ziegler
et al., 1997; Rees et al., 2000). A: Late Permian (ca. 255 Ma);
major geographic regions are highlighted. B: Early Permian (ca.
285 Ma). C: Early Jurassic (ca. 190 Ma).
q 2002 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or editing@geosociety.org.
Geology; September 2002; v. 30; no. 9; p. 827–830; 4 figures; Data Repository item 2002096.
827
Figure 3. Correlation of genus diversity and plant localities for
each Permian (P) and Triassic (T) stage (from data shown in
Fig. 2A).
Figure 2. Global compilations of plant genera for 16 Permian and
Triassic stages. A: Temporal patterns of diversity and plant localities. B: Spatial-diversity patterns, scaled in six size categories
within 108 latitudinal bins (largest squares, .
.75 genera; smallest,
1–15 genera).
braided river systems across the Permian-Triassic boundary in South
Africa, which they ascribed to a loss of vegetation. Their local and
regional interpretations of increasing aridity causing loss of vegetation
in large areas of Gondwana may be correct, but extrapolation to a truly
global scenario is based on lithological comparison with a single
boundary section in England. Twitchett et al. (2001) documented vegetation change across the boundary in Greenland, which they attributed
to a more localized ‘‘ecosystem collapse.’’ Here I assess patterns of
land-plant diversity throughout the Permian and Triassic in the context
of changing geography and climate. My results show that effects were
regional and can be explained without invoking catastrophic
mechanisms.
GLOBAL PATTERNS
Figure 2 shows my global stage-level analyses of Permian and
Triassic floras. The data set was compiled from a comprehensive survey of paleobotanical literature1 and comprises leaf genera, with the
addition of lycopsid and sphenopsid stem and root genera (preserved
more commonly than foliage in these plant groups). One approach to
assessing biotic change is to plot global summations of diversity for
successive time intervals (Fig. 2A). A marked decrease is apparent at
the Permian-Triassic boundary, with an initial decrease between the
1GSA Data Repository item 2002096, Data set references, genus diversity,
and locality counts, is available on request from Documents Secretary, GSA,
P.O. Box 9140, Boulder, CO 80301-9140, USA, editing@geosociety.org, or at
www.geosociety.org/pubs/ft2002.htm.
828
last two stages of the Permian. This two-step decline is compatible
with observations of the marine faunal record (e.g., Erwin, 1994; Knoll
et al., 1996; Hallam and Wignall, 1999; Jin et al., 2000), but reveals
nothing about possible causes. One of these may be preservation bias,
suggested by the strong correlation of diversity with the number of
plant localities (Figs. 2A and 3).
A more informative approach is to calculate diversity in each stage
as a function of paleolatitude (Fig. 2B). The effects of northward motion of Pangea (Fig. 1) are also evident in Figure 2B: the blank areas
at top left and bottom right represent ocean. There is also a climate
component, with a trend from maximum diversity at low latitudes in
the Early Permian (the last remnants of the Carboniferous equatorial
rainforest biome) toward maximum diversity at middle to high latitudes
(and near absence of vegetation at low latitudes) by the Late Triassic.
This finding is consistent with the effects of a gradual transition from
global icehouse to hothouse climates and corresponding floristic changes (e.g., Knoll, 1984; Niklas et al., 1985; Meyen, 1987; DiMichele and
Hook, 1992; Rees et al., 2000, 2002). The shape of the diversity curve
(Fig. 2A) is due to between-stage differences in the total latitudinal
range as well as the number and location of latitudinal bins containing
plant remains. These reflect the different ranges and types of biomes
represented, and the decrease across the Permian-Triassic boundary
should be viewed in this context.
REGIONAL PATTERNS
Regional analyses (Figs. 1 and 4, A–E) show that trends in diversity and the number, location, and latitudinal range of localities are
generally consistent within each region. Significantly, patterns are different between regions, such that relative levels are rarely the same in
any given stage. Diversity increases across the Permian-Triassic boundary in Angara (Fig. 4A), but remains low in Euramerica (Fig. 4B) from
the Middle Permian into the Middle Triassic. Diversity decreases across
the boundary in Gondwana, North China, and South China (Fig. 4, C–
E). The boundary decrease in the unassigned terranes (Fig. 4F) is an
artifact, because all earliest Triassic localities were assignable to the
five regions.
Regional diversity histories reflect sampling and latitudinal position, such that major differences are not confined to the PermianTriassic boundary. In Gondwana, diversity levels are low in Permian
stage 5 (Roadian), similar to those in Triassic stage 1 (Induan), even
though Roadian localities are relatively numerous. Diversity levels are
higher in adjacent Permian stages 4 (Kungurian) and 6 (Wordian). Wordian levels reflect, in part, the greater number of localities, but the
Kungurian has fewer localities, yet higher diversity. This difference
can be explained by referring to the spatial diversity plot (Fig. 2B).
Roadian (P5) levels are low between 408S and 708S compared to the
adjacent stages; it is the decrease in sample size within this temperate
high-diversity biome (Rees et al., 2002) that causes the regional-level
decrease. Increased latitudinal range relative to the Kungurian (P4)
does not compensate for this, because diversities are low in the Roadian
higher latitude localities. Elsewhere, Roadian diversity is relatively
high in Angara and North China, but low in Euramerica and South
GEOLOGY, September 2002
Figure 4. A–E: Regional stage-level compilations of plant genera. F: Remaining data for terranes unassignable to five regions (geographic
units were delimited using tectonic and paleomagnetic evidence). Sequence and explanation of plots for each region as in Figure 2, except
that paleolatitudinal plots show locality distributions instead of within-bin diversity. Additional plot shows relative occurrences of genera
in each region, grouped into four categories: paleophytic (black: cordaites, gigantopterids, glossopterids, lycopsids, pteridosperms), mesophytic (dark gray: cycadophytes, ferns, ginkgophytes, peltasperms, pinales), sphenopsids (light gray), and unassigned (white; stages
that are completely white [in B, E, F] have no plant record).
China. In Angara, diversity levels are lowest in Triassic stage 3 (Anisian) owing to the relatively low number and latitudinal range of localities. Diversity levels increase during the Anisian in Euramerica,
Gondwana, and North China, while they decrease slightly in South
China from already low levels.
As with the global diversity pattern (Fig. 2A), each regional one
can be understood in terms of variations in sampling, although this
provides only a partial explanation. It is also necessary to consider the
effects of regional-level geographic and climatic changes on what were,
in reality, different plants growing in different biomes. The third plot
in each part of Figure 4 shows stage-level variations in relative occurrences between genera grouped into four categories: typically paleophytic (black), typically mesophytic (dark gray), sphenopsids (light
gray), and unassigned (white). Overall changes between the paleophytic and mesophytic proportions are gradual in each case, although the
details vary among regions. In Angara, the diversity increase across
the Permian-Triassic boundary was accompanied by a decrease in the
proportion of arborescent plants (e.g., cordaite trees) relative to herbaceous ones such as ferns and cycadophytes. The Siberian flood basalts have been dated at or near the boundary (Renne et al., 1995;
Bowring et al., 1998), and their effect on vegetation in northern temperate biomes may explain why herbaceous pioneer communities developed more readily than arborescent climax ones. The volcanism
provides a possible mechanism to explain the marine mass extinction
(Renne et al., 1995; Bowring et al., 1998; Kozur, 1998; Jin et al., 2000),
but does not account for the vegetation patterns seen in other regions.
Euramerica was subject to progressively drier climates during the Early
GEOLOGY, September 2002
and middle Permian (Knoll, 1984; Meyen, 1987; DiMichele and Hook,
1992; Rees et al., 2002). Plants such as gigantopterids, pteridosperms,
and sphenopsids disappeared or declined in abundance while others
such as ginkgophytes and conifers increased. This pronounced regional
change commenced well before and continued beyond the PermianTriassic boundary. There is some evidence for a similar transition from
wetter to drier climates in Gondwana (Retallack, 1995; Retallack and
Krull, 1999; Ward et al., 2000), although this occurred around the
Permian-Triassic boundary. Sphenopsids and glossopterids (the main
southern temperate coal-forming plants in the Permian) declined in
abundance, while cycadophytes, ginkgophytes, and peltasperms increased. In North China, there are no post-Permian cordaite and gigantopterid records, but conifers appeared in the last stage and increased in abundance during the Early Triassic. The change in relative
proportions of gigantopterids and conifers is reminiscent of the pattern
in Euramerica during the Early and middle Permian. The record in
South China is poor across the boundary but, as in North China, there
are no post-Permian records of cordaites and gigantopterids. Cycadophytes and sphenopsids increased in abundance, but the relative proportion of ferns remained constant. The regional patterns support the
view that the global ‘‘coal gap’’ in the Early Triassic represents a loss
of peat-forming plants or their environments in each region (Retallack
et al., 1996), not a worldwide die-off of vegetation.
The contrasting regional patterns can be explained by differences
in geography and climate. South China was positioned in a tropical
ever-wet biome in the Permian (Rees et al., 2002), but was seasonally
dry in the Early and Middle Triassic and warm temperate in the Late
829
Triassic and Early Jurassic (Rees et al., 2000), reflecting its northward
motion through climate zones (Figs. 1 and 4E). The patterns in Angara,
Euramerica, and North China also reflect northward motion (Figs. 1
and 4, A, B, D), combined with global warming. High levels of diversity in the Late Triassic are due primarily to the position of these
regions in warm-temperate biomes in a hothouse world (Rees et al.,
2000). In Gondwana, the gradual loss of interior lakes and seaways
during the Late Permian and Early Triassic may have been caused by
continued uplift of the circumpolar mountain chain (Fig. 1A), with
consequent loss of oceanic moisture sources and major effects on regional vegetation (e.g., Ziegler et al., 1997; Ward et al., 2000). Diversity increased later in the Triassic, but did not reach Permian levels
(Fig. 4C). Northward motion of Pangea resulted in the presence of open
ocean poleward of ;708S by the Late Triassic (Figs. 1C and 2B). The
cool-temperate biome that had been so widespread in Gondwana during
the Permian (Rees et al., 2002) was effectively absent by the Late
Triassic, and the warm-temperate biome was restricted in this region
relative to the seasonally dry and desert biomes (Rees et al., 2000).
CONCLUSIONS
Clearly, major changes occurred in terrestrial vegetation during
the Permian and Triassic (e.g., Knoll, 1984; Meyen, 1987; DiMichele
and Hook, 1992; Retallack et al., 1996; Rees et al., 2002). My results
show that the scale and timing varied between regions and affected
different plant groups, as pointed out by Knoll (1984). Admittedly, my
results are derived from stage-level analyses and cannot capture shortterm, millennial-scale changes (e.g., Twitchett et al., 2001). However,
they provide a truly global documentation of land-plant diversity in
the context of changing geography and climate. Caution should be
exercised when extrapolating local observations, including those at the
Permian-Triassic boundary (e.g., Ward et al., 2000; Becker et al., 2001;
Kaiho et al., 2001; Twitchett et al., 2001), to catastrophic global-scale
interpretations. At the other extreme, global compilations of biotic
change through time (e.g., Sepkoski, 1993; Niklas, 1997) can be misleading if effects such as geography, climate, and preservation bias are
not considered (e.g., Raup, 1972; Knoll, 1984; Allison and Briggs,
1993; Rees et al., 2000, 2002; Jablonski, 2001).
ACKNOWLEDGMENTS
Fred Ziegler compiled the initial floral data set and David Rowley wrote
the plate-rotation program that enabled paleolatitudinal determinations. I thank
them, as well as Wolfgang Kiessling, Alistair McGowan, David Jablonski, Derek Briggs, David Sunderlin, and Scott Wing, for discussions. Reviews by Andrew Knoll and Richard Twitchett further improved this manuscript. Supported
by National Science Foundation grant ATM 00-00545.
REFERENCES CITED
Allison, P.A., and Briggs, D.E.G., 1993, Paleolatitudinal sampling bias, Phanerozoic
species diversity, and the end-Permian extinction: Geology, v. 21, p. 65–68.
Becker, L., Poreda, R.J., Hunt, A.G., Bunch, T.E., and Rampino, M., 2001,
Impact event at the Permian-Triassic boundary: Evidence from extraterrestrial noble gases in fullerenes: Science, v. 291, p. 1530–1533.
Berner, R.A., and Kothavala, Z., 2001, GEOCARB III: A revised model of
atmospheric CO2 over Phanerozoic time: American Journal of Science,
v. 301, p. 182–204.
Bowring, S.A., Erwin, D.H., Jin, Y.G., Martin, M.W., Davidek, K., and Wang,
W., 1998, U/Pb zircon geochronology and tempo of the end-Permian mass
extinction: Science, v. 280, p. 1039–1045.
Chandler, M.A., Rind, D., and Ruedy, R., 1992, Pangaean climate during the
Early Jurassic: GCM simulations and the sedimentary record of paleoclimate: Geological Society of America Bulletin, v. 104, p. 543–559.
DiMichele, W.A., and Hook, R.W., 1992, Paleozoic terrestrial ecosystems, in
Behrensmeyer, A.K., et al., eds., Terrestrial ecosystems through time: Evolutionary paleoecology of terrestrial plants and animals: Chicago, Illinois,
University of Chicago Press, p. 205–325.
Erwin, D.H., 1994, The Permo-Triassic extinction: Nature, v. 367, p. 231–235.
Farley, K.A., Mukhopadhyay, S., Isozaki, Y., Becker, L., and Poreda, R.J., 2001,
An extraterrestrial impact at the Permian-Triassic boundary?: Science,
v. 293, p. 2343.
830
Gibbs, M.T., Rees, P.M., Kutzbach, J.E., Ziegler, A.M., Behling, P.J., and Rowley, D.B., 2002, Simulations of Permian climate and comparisons with
climate-sensitive sediments: Journal of Geology, v. 110, p. 33–55.
Hallam, A., and Wignall, P.B., 1999, Mass extinctions and sea-level changes:
Earth-Science Reviews, v. 48, p. 217–250.
Jablonski, D., 2001, Lessons from the past: Evolutionary impacts of mass extinctions: National Academy of Sciences Proceedings, v. 98,
p. 5393–5398.
Jin, Y.G., Wang, Y., Wang, W., Shang, Q.H., Cao, C.Q., and Erwin, D.H., 2000,
Pattern of marine mass extinction near the Permian-Triassic boundary in
South China: Science, v. 289, p. 432–436.
Kaiho, K., Kajiwara, Y., Nakano, T., Miura, Y., Kawahata, H., Tazaki, K., Ueshima, M., Chen Zhongqiang, and Shi, G.R., 2001, End-Permian catastrophe
by a bolide impact: Evidence of a gigantic release of sulfur from the
mantle: Geology, v. 29, p. 815–818.
Knoll, A.H., 1984, Patterns of extinction in the fossil record of vascular plants,
in Nitecki, M., ed., Extinctions: Chicago, Illinois, University of Chicago
Press, p. 21–68.
Knoll, A.H., Bambach, R.K., Canfield, D.E., and Grotzinger, J.P., 1996, Comparative Earth history and Late Permian mass extinction: Science, v. 273,
p. 452–457.
Kozur, H.W., 1998, Some aspects of the Permian-Triassic boundary (PTB) and
of the possible causes for the biotic crisis around this boundary: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 143, p. 227–272.
Markwick, P.J., and Rowley, D.B., 1998, The geological evidence for Triassic
to Pleistocene glaciations: Implications for eustasy, in Pindell, J.L., and
Drake, C.L., eds., Paleogeographic evolution and non-glacial eustasy,
northern South America: SEPM (Society for Sedimentary Geology) Special Publication 58, p. 17–43.
Meyen, S.V., 1987, Fundamentals of palaeobotany: London, Chapman and Hall,
432 p.
Niklas, K.J., 1997, The evolutionary biology of plants: Chicago, Illinois, University of Chicago Press, 449 p.
Niklas, K.J., Tiffney, B.H., and Knoll, A.H., 1985, Patterns in vascular land
plant diversification: An analysis at the species level, in Valentine, J.W.,
ed., Phanerozoic diversity patterns: Profiles in macroevolution: Princeton,
New Jersey, Princeton University Press, p. 97–128.
Raup, D.M., 1972, Taxonomic diversity during the Phanerozoic: Science,
v. 177, p. 1065–1071.
Rees, P.M., Ziegler, A.M., and Valdes, P.J., 2000, Jurassic phytogeography and
climates: New data and model comparisons, in Huber, B.T., et al., eds.,
Warm climates in Earth history: Cambridge, Cambridge University Press,
p. 297–318.
Rees, P.M., Ziegler, A.M., Gibbs, M.T., Kutzbach, J.E., Behling, P.J., and Rowley, D.B., 2002, Permian phytogeographic patterns and climate data/model
comparisons: Journal of Geology, v. 110, p. 1–31.
Renne, P.R., Zichao, Z., Richards, M.A., Black, M.T., and Basu, A.R., 1995,
Synchrony and causal relations between Permian-Triassic boundary crises
and Siberian flood volcanism: Science, v. 269, p. 1413–1416.
Retallack, G.J., 1995, Permian–Triassic life crisis on land: Science, v. 267,
p. 77–80.
Retallack, G.J., and Krull, E.S., 1999, Landscape ecological shift at the PermianTriassic boundary in Antarctica: Australian Journal of Earth Sciences,
v. 46, p. 785–812.
Retallack, G.J., Veevers, J.J., and Morante, R., 1996, Global coal gap between
Permian–Triassic extinction and Middle Triassic recovery of peat-forming
plants: Geological Society of America Bulletin, v. 108, p. 195–207.
Sepkoski, J.J., 1993, Ten years in the library: New data confirm paleontological
patterns: Paleobiology, v. 19, p. 43–51.
Twitchett, R.J., Looy, C.V., Morante, R., Visscher, H., and Wignall, P.B., 2001,
Rapid and synchronous collapse of marine and terrestrial ecosystems during the end-Permian biotic crisis: Geology, v. 29, p. 351–354.
Ward, P.D., Montgomery, D.R., and Smith, R., 2000, Altered river morphology
in South Africa related to the Permian–Triassic extinction: Science, v. 289,
p. 1740–1743.
Ziegler, A.M., 1990, Phytogeographic patterns and continental configurations
during the Permian Period, in McKerrow, W.S., and Scotese, C.R., eds.,
Palaeozoic palaeogeography and biogeography: Geological Society [London] Memoir 12, p. 363–379.
Ziegler, A.M., Hulver, M.L., and Rowley, D.B., 1997, Permian world topography and climate, in Martini, I.P., ed., Late glacial and post-glacial environmental changes—Quaternary, Carboniferous-Permian and Proterozoic: New York, Oxford University Press, p. 111–146.
Manuscript received February 5, 2002
Revised manuscript received May 24, 2002
Manuscript accepted May 29, 2002
Printed in USA
GEOLOGY, September 2002