ArticlePDF Available

Abstract and Figures

Glyptostrobus Endlicher is well represented in early Early Cretaceous to Pleistocene deposits in the middle to high latitudes of North America and Eurasia. Although the taxonomy and nomenclature of the genus is complicated, the fossil record indicates Glyptostrobus was represented by a small number of species. The genus first appears in Aptian age deposits from western Canada and Greenland, and achieved a wide distribution early in its evolutionary history. Exchange of Glyptostrobus between Asia and North America occurred across the Spitsbergen and Beringian corridors, which were functional about 110 and 100 million years ago, respectively. The Late Cretaceous fossil record of Glyptostrobus shows that the genus had spread into Russia, China and the shores of the Turgai Strait. By the early Tertiary, Glyptostrobus was a prominent constituent of the polar broad-leaved deciduous forests. Paleocene age deposits across western Canada and the United States indicate the genus was present in great abundance in the lowland warm temperate and subtropical forests east of the Rocky Mountains. The broad distribution in North America and Russia during the Paleocene and Eocene indicates that Glyptostrobus grew and reproduced under a diverse range of climatic and environmental conditions, including the cold and unique lighting conditions of the polar latitudes. The presence of Glyptostrobus in Europe indicates the North Atlantic land bridges that extended between North America and Eurasia (Fennoscandia) and Europe during the early Tertiary were used. In Europe, extensive Glyptostrobus dominated swamps occupied the Central European Depression during the late Tertiary. Increasing global aridity and cooling, as well as landscape stabilization together with increasing competition for resources and habitat by representatives of the Pinaceae, seem to have forced the genus out of North America, Europe and most of Asia during the Miocene and Pliocene. In Japan, Glyptostrobus persisted until the early Pleistocene. After the early Pleistocene extinction in Japan, Glyptostrobus reappeared in southeastern China. Details of the taxonomic and biogeographic history of Glyptostrobus are examined.
Content may be subject to copyright.
Metasequoia:
Back from the Brink?
An Update
Proceedings of the Second
International Symposium on
Metasequoia and Associated Plants
August 6–10, 2006
Edited by
Hong Yang and Leo J. Hickey
Bulletin of the
Peabody Museum of Natural History
Yale University
Volume 48, Issue 2 • 31 October 2007 ISSN 0079-032X
Introduction
Glyptostrobus Endlicher (Endlicher 1847) is one
of seven monospecific genera of the Taxodiaceae,
which is now subsumed under the Cupressaceae,
with Glyptostrobuspensilis(Staunton exD. Don)
K. Koch occurring in small, restricted and pur-
portedly natural populations in South Vietnam
and Fujian, Guangdong, Guangxi, Hunan and
Yunnan provinces in southeast China (Fu and Jin
1992; Luu and Thomas 2004; Farjon 2005; Li and
Xia 2005). However, the extent of the presumed
natural populations in China and Vietnam re-
mains debatable (Farjon 2005; Farjon et al. 2005).
In China, 14 populations of G. pensilis have been
identified from the five aforementioned
provinces (Li and Xia 2005). In the broad-leaved
evergreen rain forests of Fujian and Guangdong
provinces, G. pensilis is reported to be wide-
spread in the lowland swamps, floodplains and
The Taxonomy and Biogeographic History
of Glyptostrobus Endlicher (Cupressaceae)
Ben A. LePage
Academy of Natural Sciences, 1900 Benjamin Franklin Parkway, Philadelphia, PA 19103 USA
URS Corporation, 335 Commerce Drive, Fort Washington, PA 19034-2623 USA
email: Ben_LePage@URSCorp.com
Abstract
Glyptostrobus Endlicher is well represented in early Early Cretaceous to Pleistocene deposits in
the middle to high latitudes of North America and Eurasia. Although the taxonomy and nomen-
clature of the genus is complicated, the fossil record indicates Glyptostrobus was represented by
a small number of species. The genus first appears in Aptian age deposits from western Canada
and Greenland, and achieved a wide distribution early in its evolutionary history. Exchange of
Glyptostrobus between Asia and North America occurred across the Spitsbergen and Beringian
corridors, which were functional about 110 and 100 million years ago, respectively. The Late
Cretaceous fossil record of Glyptostrobusshows that the genus had spread into Russia, China and
the shores of the Turgai Strait. By the early Tertiary, Glyptostrobus was a prominent constituent
of the polar broad-leaved deciduous forests. Paleocene age deposits across western Canada and
the United States indicate the genus was present in great abundance in the lowland warm tem-
perate and subtropical forests east of the Rocky Mountains. The broad distribution in North
America and Russia during the Paleocene and Eocene indicates that Glyptostrobus grew and re-
produced under a diverse range of climatic and environmental conditions, including the cold
and unique lighting conditions of the polar latitudes. The presence of Glyptostrobus in Europe
indicates the North Atlantic land bridges that extended between North America and Eurasia
(Fennoscandia) and Europe during the early Tertiary were used. In Europe, extensive Glypto-
strobus dominated swamps occupied the Central European Depression during the late Tertiary.
Increasing global aridity and cooling, as well as landscape stabilization together with increasing
competition for resources and habitat by representatives of the Pinaceae, seem to have forced the
genus out of North America, Europe and most of Asia during the Miocene and Pliocene. In
Japan, Glyptostrobus persisted until the early Pleistocene. After the early Pleistocene extinction
in Japan, Glyptostrobus reappeared in southeastern China. Details of the taxonomic and biogeo-
graphic history of Glyptostrobus are examined.
Keywords
China, Cretaceous, Eurasia, Japan, land bridges, migration, North America, orogeny, paleogeog-
raphy, taxonomy, Tertiary, Turgai Strait.
Bulletin of the Peabody Museum of Natural History 48(2):359–426, October 2007.
© 2007 Peabody Museum of Natural History, Yale University. All rights reserved. — http://www.peabody.yale.edu
river deltas (Wang 1961; Farjon 2005; Farjon et
al. 2005). The species is heliophilous, intolerant
of competition and of saline and alkaline soils
and water; it seems to be well adapted to growing
in a variety of soil types found on river flood-
plains, riparian corridors, deltas and uplands at
elevations that range from sea level to 1980 m
above sea level (Fu and Jin 1992; Farjon 2005; Li
and Xia 2005).
In Vietnam, G. pensilis grows in Dac Lac
Province at elevations of 550 and 750 m in satu-
rated, basalt-derived soils in two small swamp
forest communities that are dominated by Dipte-
rocarpus Gaertner, Pterocarpus Jacquin and
members of the Myrtaceae (Luu and Thomas
2004). These swamp forest communities experi-
ence a tropical monsoon climate with a mean an-
nual temperature of 20 to 23 °C and an annual
rainfall of 1300 to 1800 mm. There are approxi-
mately 220 mature individuals of G. pensilis in a
50 ha reserve at Earal, while 34 individuals occur
in a 100 ha reserve at Trap Kso (Luu and Thomas
2004). Continued human encroachment, fire
and the inability of Glyptostrobus to regenerate
naturally in these populations threaten G. pen-
silis,so that a 25% reduction in the number of in-
dividuals of Glyptostrobus over the next 30 to 40
years is expected (Luu and Thomas 2004).
Glyptostrobus is currently listed as a threat-
ened species because the populations are small,
severely fragmented or known to exist at no
more than five locations; also, the extent of their
geographic distribution and the number of loca-
tions or subpopulations within their current dis-
tribution is expected to continue to decline
(Farjon et al. 2005). Despite the grave situation
associated with the survival of the extant popula-
tions, the occurrence of fertile and vegetative re-
mains of Glyptostrobus in the plant fossil record
indicates that the genus was once distributed ex-
tensively from the subtropical to the polar re-
gions of the Northern Hemisphere for over a
hundred million years (Florin 1963).
The distribution of Glyptostrobus in space
and in time further indicates that representatives
of the genus grew under a diverse range of cli-
matic and environmental conditions. As such,
the climate under which G. pensilis grows today
should not be considered indicative of the tem-
peratures or environmental conditions under
which ancient representatives of the genus grew,
or that G. pensilis is incapable of growing under
much colder conditions. Laboratory experiments
have shown that the leaves, buds and twigs are
frost resistant to temperatures as low as –18 to
–20 °C (Sakai and Larcher 1987). For example,
trees of G. pensilis growing at the United States
National Arboretum in Washington, DC, experi-
ence temperatures that are well below 0 °C for
days at a time when winter cold fronts push
through the region (World Climate 1996).
This paper presents a brief overview of the
taxonomy of Glyptostrobusand discusses a more
detailed review of the biogeographic history of
the genus throughout Mesozoic and Cenozoic
time in light of the tectonic and climatic history
of the Northern Hemisphere. This analysis of the
fossil record of Glyptostrobus reviewed nearly
400 published reports (Appendices A and B), ap-
proximately 70% of the published records in
which Glyptostrobusis reported, and is in no way
meant to be exhaustive. As such, the issue associ-
ated with circumscription of the fossil record of
the genus has not been performed and will be the
focus of another paper. Nevertheless, the spa-
tially and temporally representative sample of the
fossils considered here indicates that most of the
fossils conform morphologically to the living
species G. pensilis and that, according to a de-
tailed taxonomic assessment of the fossil re-
mains, Glyptostrobus seems to be another
spectacular example of morphological stasis over
a geologically long period.
Taxonomy of Fossil Glyptostrobus
Brongniart (1833) was the first to describe Glypt-
ostrobus fossils. He assigned the Miocene age
leaves and seed cones of Glyptostrobus from the
island of Iliodroma (now called Alonnisos) in
Greece to TaxodiumeuropaeumBrongniart. The
fossil material was compared with extant conifer
species known and assigned at the time to Cu-
pressusL., ThujaL., JuniperusL., CallitrisVente-
nat and Taxodium Richard. Comparison of the
seed cones and leaves of the fossil material with
representatives of these extant genera led Brong-
niart to conclude the fossils were those of an ex-
tinct species of Taxodium. Although Endlicher
(1847) described and erected the modern genus
Glyptostrobus several years later, the fossils de-
scribed by Brongniart (1833) were not recog-
Bulletin of the Peabody Museum of Natural History 48(2) – October 2007
360
nized by plant systematists of the time, including
Brongniart, as being those of Glyptostrobus and
were transferred to the fossil genus Taxodites
Unger (Endlicher 1847; Brongniart 1849; Göp-
pert 1850; Unger 1850a). It was not until Unger
(1850b) was able to compare the existing fossil
material with the newly described living species
Glyptostrobus heterophyllus (Brongniart)
Endlicher (= G. pensilis) described by Endlicher
(1847) that Unger recognized the fossils de-
scribed by Brongniart were indeed those of
Glyptostrobus. He created the new combination,
Glyptostrobus europaeus (Brongniart) Unger,
which is the correct epithet despite the wide-
spread use of G. europaeus (Brongniart) Heer in
the literature.
Since then, hundreds of reports of Glypto-
strobusfossils have been published (Appendix A;
Jongmans and Dijkstra 1973). The lack of signif-
icant morphological diversification observed be-
tween living and fossil Glyptostrobus foliage and
seed cones leaves little doubt of the accuracy of
most of the fossil identifications, especially when
the foliage is associated with seed cones. More
importantly though, the fossil record of Glypto-
strobus indicates that many of the species de-
scribed are synonyms and that there are
considerably fewer than the more than 30 species
of fossil Glyptostrobus described to date. In fact,
most of the fossils described in the literature are
reported by several of the authors as being more
or less identical to living G. pensilis. This review
of the literature did not show significant differ-
ences between many of the fossil species that
were erected and existing fossils and G. pensilis.
The creation of new species was commonly
based on slight differences in the size and shape
of the fossil remains or the geologic age of the
fossils.
Although there have been several excellent
reviews of the fossil record of the genus (Berry
1916, 1924; Brown 1936, 1962; Christophel
1976), Brown (1936) and Christophel (1976)
point out that Glyptostrobus has cupressoid,
cryptomeroid and taxodioid foliage with inter-
mediates of the three prominent leaf morpholo-
gies. This together with a complicated taxonomic
history has made interpretation of the evolution-
ary history of Glyptostrobus difficult. Reassess-
ment of the taxonomy of the genus is in progress
and, despite being incomplete, the data suggest
the number of fossil Glyptostrobus species is
smaller than that reported in the literature. This
situation is not unique among the redwoods
given that there are only three species of fossil
Metasequoia Miki currently recognized over its
110-million-year-old evolutionary history (see
Stockey et al. 2001; LePage et al. 2005). As is the
case with Metasequoia (Stockey et al. 2001; Le-
Page et al. 2005), there is no reason why there
should be so few recognizable species of fossil
Glyptostrobus over the entirety of its evolution-
ary history. While the gross morphological fea-
tures of the leaves and seed cones at present seem
to provide little useful phylogenetic information
for species level segregation, hopefully further
detailed anatomical and morphological study of
existing and new fossil Glyptostrobusremains, as
well as living G. pensilis, will reveal that the
genus had a more speciose fossil history than that
indicated in the fossil record.
The list of fossil Glyptostrobus and pertinent
locality data compiled for this report are pre-
sented in Appendices A and B. No attempt was
made to revise the taxonomy of fossil Glyptostro-
bus, as this effort was outside of the scope of this
paper. Nevertheless, the totality of these fossils is
important for understanding and interpreting
the biogeographic history of the genus. The diffi-
culties associated with providing a reliable and
meaningful assessment of the number of species
throughout the fossil record of the genus limits
the usefulness of interpreting the biogeographic
history of the genus at the species level. There-
fore, the biogeographic history of the genus is
best seen as that of a single entity.
Phytogeography and Climate
The fossil record of Glyptostrobus indicates the
genus was distributed widely throughout the
Northern Hemisphere from the Early Cretaceous
until the early Pleistocene (Appendix B; Figures
1 to 12; Florin 1963). Based on fertile and vege-
tative remains, Glyptostrobus is first recorded in
the Aptian Kootenai Formation from Alberta,
Canada, the Aptian Kootanie Formation from
Montana, USA, and the Aptian–Albian Kome
Formation from Pattorfik (syn: Pagtofik) and
Ekkorfat (syn: Ikorfat), Nûgssuaq Peninsula,
Greenland (see Figure 1; Heer 1874; Dawson
1885; Ward 1905; Koch 1964). Note that there are
The Taxonomy and Biogeographic History of Glyptostrobus Endlicher (Cupressaceae) • LePage 361
several putative species of Glyptostrobus from
the Potomac Group of Virginia, Washington,
DC, and Maryland, USA.
Specimens of G. expansus (Fontaine) Ward,
G. fastigiatus (Fontaine) Ward and Taxodium
(Glyptostrobus)denticulatum Fontaine from the
Aptian Patuxent Formation from several loca-
tions in Virginia, USA, were described by
Fontaine (1889) and later synonomized by Berry
(1911) under Sphenolepis kurriana (Dunker)
Schenk. However, no explanation or justification
for the transfer of these specimens was provided.
Two other species and one variety, G. brookensis
(Fontaine) Ward, G. ramosus (Fontaine) Ward
and G. brookensis var. angustifolius (Fontaine)
Ward from the Albian Patapsco Formation of
Virginia, Washington, DC, and Maryland, also
described by Fontaine (1889), were synonomized
by Berry (1911) under Widdringtonites ramosus
(Fontaine) Berry. Berry (1911:429) argued that
the lack of Glyptostrobus seed cones in the Po-
tomac and “extremely close resemblance” of the
leafy branches and leaves to the younger W. re-
ichii (Ettingshausen) Heer from the Raritan
(Cenomanian–Turonian) and Magothy (Conia-
cian–Santonian) Formations warranted assign-
ment of these taxa to W. ramosus. However, seed
cones of W. ramosus are not known and the only
known fertile material of W. reichii at the time
was from Europe (Berry 1911). While the ratio-
nale to support the transfer of these specimens to
either S. kurriana or W. ramosus is questionable,
detailed reexamination of them is necessary to
determine the accuracy of the identifications. If
these specimens were indeed those of Glypto-
strobus, the reports would provide valuable in-
formation on the early evolutionary history of
the genus.
Distribution of the three Aptian reports of
Glyptostrobus ranged from about 45° N to about
55° N paleolatitude, which places them in the
transition zone between Vakhrameev’s (1991)
subtropical Indo-European floristic region and
temperate Siberian–Canadian region. In North
America, the Indo-European floristic region ex-
tended from about 50° to 55° N paleolatitude.
Vakhrameev (1991) suggested that precipitation
in this region along the Atlantic Coastal Plain
and the western interior of North America was
abundant and that the temperature was moder-
ate. Megafloral and dispersed cuticle assemblages
Bulletin of the Peabody Museum of Natural History 48(2) – October 2007
362
Figure 2. Generalized paleogeographic reconstruc-
tion of the Northern Hemisphere in polar projection
during the early Early Cretaceous (Albian, ca. 102
Ma), showing the Beringian Corridor, the Spitsber-
gen Corridor, possible land bridges (?) between
North America, Eurasia and Asia, and the distribu-
tion of fossil Glyptostrobus. The dots indicate Aptian
to Cenomanian. Abbreviations: B, Beringian Corri-
dor; NA, North America; SC, Spitzbergen Corridor;
WIS, Western Interior Seaway. Modified from Le-
Page and Basinger (1995) and LePage et al. (2005)
and the references therein.
Figure1. Generalized paleogeographic reconstruction
of the Northern Hemisphere in polar projection during
the Early Cretaceous (Aptian, ca. 110 Ma), showing the
Spitsbergen Corridor and the distribution of fossil
Glyptostrobus. The dots indicate Aptian to Albian. Ab-
breviations: SC, Spitzbergen Corridor; TU, Turgai
Strait. Modified from LePage and Basinger (1995) and
LePage et al. (2005) and the references therein.
from the Indo-European floristic region provide
strong evidence for evergreen vegetation and few
to no freezing temperatures, while megafloral as-
semblages from the Siberian–Canadian region,
which was located north of 50° to 55° N paleolat-
itude, were characterized by predominantly de-
ciduous taxa that were presumably adapted to
freezing temperatures, winter darkness, or both
(Upchurch 1993). Living G. pensilis is faculta-
tively deciduous and though it may shed its
leaves today as a result of water stress due to
drought, the fossil record of Glyptostrobus indi-
cates representatives of the genus first appeared
in a mesothermal (sensu Wolfe 1979) transition
zone located between the subtropical Indo-Euro-
pean and temperate Siberian–Canadian regions.
This suggests early representatives of Glyptostro-
bus were also probably deciduous and that the
ability to shed leaves in response to periods of en-
vironmental stresses today is a retained pleisio-
morphic character.
The Spitsbergen Corridor between Asia and
North America was functional during the Aptian
(LePage and Basinger 1995), but apparently
Glyptostrobus did not use this land bridge, given
the lack of Glyptostrobus fossils in Asia at this
time. However, by Albian–Cenomanian time the
genus is presumed to have migrated from North
America through the Spitsbergen Corridor to
Russia (see Figure 2). During the Aptian, the cli-
mate in Asia was warm, dry and unsuitable for
mesic taxa, but cooler and more humid condi-
tions were prevalent by Albian time (Vakhra-
meev 1991). Similarly, in North America climate
was cooler, more humid and wetter during Al-
bian and Cenomanian time (Vakhrameev 1991).
Although a warm and dry climate might account
for the absence of Glyptostrobus in Asia during
the Aptian, unsuitable habitat, physical barriers,
ecophysiological limitations and the inability to
thrive in a polar environment could have further
contributed to the absence of Glyptostrobus in
Asia during Aptian time.
Given the lack of floral and faunal data from
around the Spitsbergen Corridor, it is uncertain
whether this land bridge was functional during
the Late Cretaceous (see Figure 3; LePage and
Basinger 1995). Nevertheless, biotic interchange
between Asia and North America continued to
occur through the Beringian Corridor, which
was established by Albian–Cenomanian time.
The fossil record indicates Glyptostrobus is pre-
sent in the northern regions of North America
and Asia and in southern Eurasia during the
Cenomanian and Turonian (see Figure 3; Etting-
shausen 1867; Dawson 1882; Frič and Bayer
1900; Hollick 1930; Baikovskaya 1956; Svesh-
nikova 1967; Lebedev 1982; Knobloch and Mai
1986). However, functional land bridges between
North America and Eurasia, or between Asia and
Eurasia, during Albian to Turonian time have not
been documented. Temperature estimates of the
regions where Glyptostrobus is recorded during
the Albian to Turonian range from megathermal
to microthermal. During the latest Albian to
middle Cenomanian, the transition zone be-
tween megathermal and mesothermal vegetation
occurred at about 40° N paleolatitude and the
mean annual temperature of this transition zone
is estimated to have been 20 °C, while the transi-
tion zone between mesothermal and microther-
mal vegetation occurred at about 70° N
paleolatitude with an estimated mean annual
temperature of 13 °C (Upchurch 1993).
Figure 3. Generalized paleogeographic reconstruc-
tion of the Northern Hemisphere in polar projection
during the early Late Cretaceous (Turonian, ca. 92
Ma), showing the Beringian Corridor, possible land
bridges (?) between North America, Eurasia and Asia,
and the distribution of fossil Glyptostrobus. The Spits-
bergen Corridor (?) may not have been functional at
this time. The dots indicate Cenomanian to Turonian;
squares are Coniacian. Abbreviations: B, Beringian
Corridor; NA, North America; WIS, Western Interior
Seaway. Modified from LePage and Basinger (1995)
and LePage et al. (2005) and the references therein.
The Taxonomy and Biogeographic History of Glyptostrobus Endlicher (Cupressaceae) • LePage 363
The Albian to Turonian distribution pattern
of Glyptostrobus illustrates three important
points. First, the physical and ecophysiological
barriers that precluded the movement of Glypto-
strobusfrom North America into Asia during the
Aptian had been overcome by the Albian. Sec-
ond, following the presumed demise of the Spits-
bergen Corridor by Albian–Cenomanian time,
Glyptostrobus probably continued to migrate
from North America to Asia through the Berin-
gian Corridor. However, the paleolatitude of the
Beringian Corridor was between 75° N and
nearly 90° N and at a considerably higher paleo-
latitude than the Spitsbergen Corridor. As such,
ecophysiological adaptation to a polar environ-
ment and its unique light regime, in which trees
would have experienced prolonged periods of
complete darkness and daylight, were essential
for survival above the Arctic Circle. Representa-
tives of the genus had evolved physiologically to
grow and reproduce under these unique condi-
tions. Being deciduous would have certainly fa-
cilitated migration of the genus through the
polar regions. Third, despite an estimated mean
annual temperature of 13 °C at a paleolatitude of
approximately 70° N during the early Late Creta-
ceous, the polar winters were cold and probably
experienced freezing temperatures during the
dark winter months throughout the Cretaceous
and early Tertiary (see LePage 2003a and refer-
ences therein). This indicates the genus was able
to tolerate a wide range of temperature extremes
early in its evolutionary history with no apparent
effect on the morphology of the seed cones and
leaves; it contradicts the notion that Glyptostro-
bus is intolerant of cold or freezing conditions.
Except for the absence of Glyptostrobus in
Eurasia and Europe during late Late Cretaceous
(see Figure 4; Santonian to Maastrichtian), the
distribution of the genus in Asia and North
America remained more or less the same as that
of the Cenomanian to Coniacian (see Figures 3
and 4). Reasons for its absence in Eurasia at this
time are unknown, but high or fluctuating sea
levels during the Late Cretaceous (Haq et al.
1988) may have eliminated suitable wetland
habitats and so extirpated the genus from Eura-
sia. Campanian and Maastrichtian age floras
from eastern Kazakhstan and the Kuznets Basin
near the Turgai Strait show that the climate was
cooling and becoming more mesic (Baikovskaya
1956; Shilin and Romanova 1978; Vakhrameev
1991). Temperature estimates for the late Ceno-
manian to the late Maastrichtian indicate the
boundary between megathermal and mesother-
mal vegetation was located at 40° to 50° N paleo-
latitude and the boundary between mesothermal
and microthermal vegetation was at 65° to 75° N
paleolatitude (Upchurch 1993). The distribution
of Glyptostrobus during the Cenomanian to
Maastrichtian shows that representatives of the
genus were prevalent in the mesothermal and
microthermal vegetation zones that experienced
abundant precipitation and a wide range of tem-
perature and environmental conditions (see Fig-
ures 3 and 4). The distribution of Glyptostrobus
in Asia and North America during the latter part
of the Late Cretaceous further illustrates the sen-
sitivity of the genus to aridity (see Figure 4). In
Asia, as the climate was becoming cooler and
more humid, Glyptostrobus responded by ex-
tending further south to approximately 45° N pa-
leolatitude, while in North America the climate
was becoming drier in the mid-latitudes and the
genus retreated from a paleolatitude of approxi-
mately 50° N to approximately 65° N (see Figure
Figure 4. Generalized paleogeographic reconstruc-
tion of the Northern Hemisphere in polar projection
during the Late Cretaceous (Maastrichtian, ca. 71
Ma), showing the Beringian Corridor and the distrib-
ution of fossil Glyptostrobus. Triangles indicate San-
tonian; plus signs are Campanian; dots are Maas-
trichtian. Abbreviations: B, Beringian Corridor; Eu,
Europe; NA, North America; WIS, Western Interior
Seaway. Modified from LePage and Basinger (1995)
and LePage et al. (2005) and the references therein.
Bulletin of the Peabody Museum of Natural History 48(2) – October 2007
364
4; Wolfe and Upchurch 1986, 1987; Vakhrameev
1991; Upchurch 1993).
A major reorganization of the vegetation
zones in North America occurred at the Creta-
ceous–Tertiary boundary as a result of increased
precipitation and warming (Wolfe and Upchurch
1986, 1987). The broad-leaved deciduous vegeta-
tion that occupied regions north of 65° N paleo-
latitude during the Late Cretaceous now
extended south to approximately 50° N paleolat-
itude (Upchurch 1993). South of paleolatitude
48° N precipitation increased significantly and
the open canopy woodland forests of the West-
ern Interior and Gulf Coastal Plain gave way to
multistratal megathermal rainforests (Wolfe and
Upchurch 1986; Upchurch and Wolfe 1987).
Range expansion of Glyptostrobus in North
America and Europe at this time is also coinci-
dent with the creation of large expanses of
mesothermal lowland swamps, floodplains, river
deltas and mesic microthermal upland habitats
that formed in response to increased tectonism,
precipitation and orogenies (see Figures 5 and 6).
Glyptostrobusfirst achieves its maximum ge-
ographic and latitudinal range in the late Pale-
ocene–early Eocene and maintains this until
Oligocene time. Reports of Glyptostrobus from
the late Paleocene–early Eocene Wilcox Group
in Texas, the Eocene Holly Springs Sand and Lis-
bon Formation in Mississippi, and the Lagrange
Formation in Tennessee, USA (Berry 1916, 1924,
1930, 1941) indicate that the genus responded to
increased regional precipitation and the creation
of suitable wetland habitat by extending south to
approximately 30° N paleolatitude (see Figure 6).
Although it seems to have always been restricted
to wet lowland habitats associated with distur-
bance regimes, the range of distribution (30° to
80° N paleolatitude) at this time indicates Glypt-
ostrobus extended from the subtropical to cool
temperate regions in North America and Asia
and grew under a wide range of climatic and en-
vironmental conditions.
The Turgai Strait populations of Glyptostro-
busthat were present during the Late Cretaceous
are notably absent from this region during the
Paleocene (see Figure 5). However, this absence
may be an artifact created by the lack of suitable
deposits or deposits of suitable age in the region.
Reports of Glyptostrobusfrom the Paleocene and
Eocene of Europe indicate the genus was reintro-
Figure 5. Generalized paleogeographic reconstruc-
tion of the Northern Hemisphere in polar projection
during the Paleocene (ca. 60 Ma), showing the distri-
bution of fossil Glyptostrobus. Note the spread of
Glyptostrobus into the westernmost part of Eurasia by
way of the southern Thulian land bridge. Abbrevia-
tions: B, Beringian Corridor; D, DeGeer Route; Eu,
Europe; T, Thulian Route. Modified from LePage and
Basinger (1995) and LePage et al. (2005) and the ref-
erences therein.
Figure 6. Generalized paleogeographic reconstruc-
tion of the Northern Hemisphere in polar projection
during the middle Eocene (ca. 45 Ma), showing the
distribution of fossil Glyptostrobus. By the end of the
Eocene and the early Oligocene, the DeGeer and Thu-
lian routes between North America and Europe are
broken and the geographic extent of Glyptostrobus
reaches its maximum latitudinal extent, which persists
into the Oligocene. Abbreviations:B, Beringian Corri-
dor; D, DeGeer Route; T, Thulian Route. Modified
from LePage and Basinger (1995) and LePage et al.
(2005) and the references therein.
The Taxonomy and Biogeographic History of Glyptostrobus Endlicher (Cupressaceae) • LePage 365
duced to Europe at this time (Gardner
1883–1886; Laurent 1912; Seward 1919; Johnson
and Gilmore 1921; Knobloch 1969; Mai and
Walther 1985; Boulter and Kvaček 1989; Mai
1995; Fairon-Demaret et al. 2003). Two land
bridges were available to flora and fauna for ex-
change between North America and Europe dur-
ing the early Tertiary (see Figures 5 and 6).
Throughout the Paleocene and Eocene the
DeGeer Route (McKenna 1972a, 1983a) linked
North America and Eurasia (during the Pale-
ocene and Eocene Eurasia is also referred to in
the literature as Fennoscandia), while the more
southern Thulian Route existed intermittently or
as a series of islands between southern Green-
land and Europe (McKenna 1972a, 1975; Thiede
and Eldholm 1983). Sea floor spreading in the
Norwegian–Greenland Sea at the Eocene–Oligo-
cene boundary effectively destroyed these land
bridges, terminated all terrestrial communica-
tion between these two regions and isolated the
European Glyptostrobus population from the
North American populations (McKenna 1972b;
Dawson et al. 1975, 1976; West et al. 1977; West
and Dawson 1978; Hoch 1983). Glyptostrobus
seems to have preferred the southern Thulian
Route rather than the more northern DeGeer
Route that linked North America and Eurasia
(see Figures 6 and 7).
The movement of polar broad-leaved decid-
uous forest elements, including Glyptostrobus,
from Asia and Europe into the West Siberian
Plain and Eurasia during the Oligocene is coinci-
dent with several major events (see Figure 7; Le-
Page and Basinger 1991, 1995; LePage 2001,
2003b, 2003c). First was the drying of the Turgai
Strait, a shallow epicontinental seaway that ex-
tended from the Arctic Ocean to the Tethyan
Sea, separating Eurasia and Asia and effectively
precluding floral and faunal exchange between
these two regions (McKenna 1972b, 1983b;
Tiffney 1985; LePage and Basinger 1991, 1995).
Second, the effects of the Himalayan orogeny
may have created suitable habitat for Glyptostro-
bus along the southern margin of the
Asian–Eurasian coastline. Since Glyptostrobus
seems to thrive in regions where disturbance as-
sociated with orogenic activity is prevalent, the
Asian–Eurasian coastline could have provided
suitable lowland habitat and an extensive mar-
itime migratory corridor between Asia and Eura-
Figure 8. Generalized paleogeographic reconstruc-
tion of the Northern Hemisphere in polar projection
during the Miocene (ca. 16 Ma), showing the distri-
bution of fossil Glyptostrobus. Note the Glyptostrobus
populations in Eurasia, Asia and North America
occur in areas of prior disturbance (e.g., Turgai
Strait) and regions of active orogeny. Abbreviations:
B, Beringian Corridor; WIS, Western Interior Sea-
way. Modified from LePage and Basinger (1995) and
LePage et al. (2005) and the references therein.
Bulletin of the Peabody Museum of Natural History 48(2) – October 2007
366
Figure 7. Generalized paleogeographic reconstruc-
tion of the Northern Hemisphere in polar projection
during the Oligocene (ca. 29 Ma), showing the dis-
tribution of fossil Glyptostrobus. Note the spread of
Glyptostrobus from Asia into eastern Eurasia and the
West Siberian Plain following regression of the Tur-
gai Strait. The dark gray region in west-central Asia
indicates the location of the remnants of the Turgai
Strait. Abbreviations: B, Beringian Corridor; Eu, Eu-
rope; WIS, Western Interior Seaway. Modified from
LePage and Basinger (1995) and LePage et al. (2005)
and the references therein.
sia. Finally, a major global climatic cooling at the
Eocene–Oligocene boundary (about 34 Ma) that
Wolfe (1985) called the Terminal Eocene Event
caused 25% to 40% of the genera that grew in Eu-
rope, western North America and Alaska during
the Eocene to go extinct afterward (Wolfe 1997).
In the area that is now Kazakhstan, the elimina-
tion of the subtropical Eocene elements and the
replacement of the flora with more temperate el-
ements began during the Rupelian and Chattian
(early to late Oligocene), so that by Aquitanian
(early Miocene) time the subtropical elements
had been completely eliminated (Zhilin 1989).
By the Oligocene, Glyptostrobus was a con-
stituent of the thermophilic and mesophytic Tur-
gai floras that appeared in Kazakhstan and the
West Siberian Plain (see Figure 7; Zhilin 1989).
Because of the lack of suitable Eocene age de-
posits in central Asia and Eurasia (see Figure 6),
it is not possible to determine with certainty the
source of the Kazakhstan and West Siberian
Glyptostrobus populations. However, consider-
ing the biogeographic history of Metasequoia
(LePage et al. 2005) and the fact that Metase-
quoia and Glyptostrobus were commonly co-oc-
curring constituents of the lowland swamp forest
communities during the Paleocene and Eocene,
it is likely that the source of Glyptostrobus was
from Asia and that migration of the genus pro-
ceeded east to west from Asia into Kazakhstan
and the West Siberian Plain.
Paleobotanical and magnetostratigraphic
data from the West Siberian Plain indicate that
the thermophilic and mesophytic Turgai flora of
the late Oligocene and early Miocene (pre-Chron
C5D) shifted to a much cooler and drier forest
steppe flora by the late early and early late Mio-
cene (Gnibidenko et al. 1999). The cooling and
drying experienced in the West Siberian Plain is
consistent with the global trend towards cooler
and drier climates that began in response to the
buildup of the East Antarctic ice sheet at about
the middle Miocene (Potts and Behrensmeyer
1992).
The report of Glyptostrobusfrom the late Ol-
igocene–early Miocene Li Formation in north-
ern Thailand is of particular interest because its
latitude of approximately 17°50N is the south-
ernmost record for the genus (Endo 1964, 1966).
Although Endo (1964) suggested the associated
flora grew in a warm temperate climate, a tropi-
cal climate is likely, given its location. One possi-
ble explanation for this discrepancy is that the
flora grew at higher elevations where tempera-
tures were cooler. Such a situation exists today in
Vietnam, where Glyptostrobusoccurs at latitudes
of 13°01N at Earal and 13°09N at Trap Kso, but
grows at elevations of 550 and 750 m, respec-
tively (Luu and Thomas 2004).
The Miocene distribution of Glyptostrobus
indicates the genus was prevalent throughout
Europe, Japan and portions of central Asia and
North America (see Figures 8 and 9). Glyptostro-
bus disappeared from the West Siberian Plain
and Kazakhstan as climate continued to become
cooler and drier throughout the Miocene. Zhilin
(1989) indicates that Glyptostrobus is absent
from late Miocene floras throughout Kaza-
khstan. In Europe, extensive Glyptostrobus-
dominated swamps were prevalent throughout
the Central European Depression, which ex-
tended approximately 1000 km east of the North
Sea and 300 km in a north–south direction
(Schneider 1990). For example, Kovar-Eder et al.
(2004) report that Glyptostrobus was a common
Figure 9. Distribution of Glyptostrobus in Japan and
the Asian mainland during the late Miocene. Note the
land bridge between Japan and the Asian mainland.
Modified from Minato (1965) and Yang (1991).
The Taxonomy and Biogeographic History of Glyptostrobus Endlicher (Cupressaceae) • LePage 367
element of the late early to early middle Miocene
Parschlug flora of Austria. Based on the autecol-
ogy of the fossil assemblage, they suggest that
Glyptostrobus formed oligotypic wetland gallery
forests together with Myrica lignitum (Unger)
Saporta, Liquidambar europaea A. Braun and
possibly Zelkova zelkovifolia (Unger) Bůžek &
Kotlaba along the lake shores. Analysis of the
vegetation in the early Miocene Most Basin in
northern Bohemia indicates Glyptostrobus was a
dominant tree in the peat-forming mires (Kvaček
et al. 2004), while Givulescu and Ticleanu (1983)
report Glyptostrobus formed extensive forested
moors in Romania, especially during Pannonian
time.
The plant fossil record indicates Glyptostro-
bus was being eliminated from regional floras
throughout the world during the later part of the
Miocene and Pliocene (see Figure 10). Glypto-
strobus remained in China until the Miocene,
but it was a relatively minor floristic element
(Guo 1985). In North America, Russia and Eu-
rope Glyptostrobus persisted in the middle to
high latitude regions until the Pliocene (see Fig-
ure 10; Laurent and Marty 1923; Smith 1938;
Kryshtofovich 1946; Wolfe et al. 1966). In
Greece, climate shifted from being warm and
humid during the late Miocene to arid or semi-
arid during the Pliocene (Ioakim et al. 2005). It is
difficult to precisely determine whether the
demise of Glyptostrobus during the latter part of
the Tertiary is due to increasing aridity, decreas-
ing global temperature or increased competition
for resources from the Pinaceae (LePage et al.
2005), or some combination thereof, but given
the ability of Glyptostrobus to tolerate cold tem-
peratures, increasing aridity coupled with in-
creasing competition for space and resources
likely led to range reduction and ultimately the
extinction of the genus from North America and
Europe.
The late Miocene distribution of Glyptostro-
bus in Japan indicates the genus was well repre-
sented in Hokkaido and Honshu (see Figure 9).
At this time Japan was connected to China, en-
abling the free exchange of flora and fauna be-
tween these areas. Glyptostrobus disappeared
from Hokkaido by the Pliocene, because of habi-
tat partitioning due to climate and environmen-
tal change (Momohara 1997, 2005), but the pop-
ulations in the south (present day Honshu and
Bulletin of the Peabody Museum of Natural History 48(2) – October 2007
368
Figure 10. Generalized paleogeographic reconstruc-
tion of the Northern Hemisphere in polar projection
during the Pliocene (ca. 3 Ma), showing the Beringian
Corridor (B) and the distribution of fossil Glyptostro-
bus. The last large populations of Glyptostrobus are in
Europe and Japan at this time. Modified from LePage
and Basinger (1995) and LePage et al. (2005) and the
references therein.
Figure 11. Distribution of Glyptostrobus in Japan and
the Asian mainland during Pliocene time. Note that
the land connection between Japan and the mainland
that was present during the late Miocene is now bro-
ken and the mainland and northern Japanese popula-
tions of Glyptostrobus have nearly disappeared. Mod-
ified from Minato (1965) and Yang (1991).
Kyushu) remained (see Figure 11). The land con-
nection that existed between Japan and China
during the late Miocene was severed during the
Pliocene, isolating the Japanese populations.
During the late Pliocene and early Pleistocene
the land bridge between Japan and the China was
re-established (see Figure 12). According to Mi-
nato (1965) and Yang (1991), the land bridge be-
tween China and Japan was broken and re-estab-
lished several times during Pliocene–Pleistocene
time. Glyptostrobus remained a constituent of
the lowland forests in Japan until the early Pleis-
tocene, about 1.1 to 0.8 Ma (Momohara et al.
1990; Momohara 1994).
The paucity of Glyptostrobusfossils in China
during the Pliocene and Pleistocene and its reap-
pearance in present day South Vietnam and in
Fujian, Guangdong, Guangxi, Hunan and Yun-
nan provinces in southeast China is of consider-
able interest from the standpoint of the origin of
the extant populations, evoking several ques-
tions. Did Glyptostrobus survive undetected in
China throughout the Pliocene and Pleistocene?
Did Glyptostrobus become extinct in China at
the end of the Miocene and become re-estab-
lished from the Japanese populations sometime
before the connection between China and Japan
was broken permanently during the early Pleis-
tocene? Alternatively, did the Thai population
give rise to the Vietnamese and Chinese popula-
tions or just the Vietnamese elements? Though it
is interesting to speculate on the origin of the ex-
tant populations in Asia, there is simply not
enough data at this time to address either ques-
tion. Such answers will require further paleob-
otanical research in southeast Asia focused on
Pliocene, Pleistocene and Quaternary deposits,
along with genomic studies on the individuals
from the extant Glyptostrobus populations, so
that the vegetation history of these regions, as
well as the origin of the extant Glyptostrobus
populations, can be reconstructed in more detail.
Summary
Review of the data shows that the taxonomy of
Glyptostrobus is in dire need of revision if we are
to better understand the evolutionary history of
the genus and species level phylogenetic relation-
ships. Currently, more than 30 species of fossil
Glyptostrobus are recognized in the literature.
Based on the diagnoses that accompany these
taxa, most of the fossil species are nomenclatural
synonyms. Therefore, the number of fossil
species that likely existed over the approximately
110-million-year history of the genus is probably
small. The lack of significant morphological dif-
ferences between the fossil remains described to
date and the living G. pensilis further support
this conclusion. While Glyptostrobus may be an-
other example of a taxon that has shown spectac-
ular morphological stasis over an incredibly long
period of geologic time, the mechanisms respon-
sible for such prolonged morphological stasis are
poorly understood.
We know from the plant fossil record that
Glyptostrobuswas widely distributed throughout
the Northern Hemisphere from the Early Creta-
ceous until the early Pleistocene. Distribution of
the genus extended from the middle to high lati-
tudes and encompassed subtropical to cool tem-
perate climates. Clearly, representatives of the
genus were well adapted to a wide range of cli-
matic and environmental conditions. However,
the lowland swamp, floodplain, riparian and
paludal habitats associated with the alluvial and
The Taxonomy and Biogeographic History of Glyptostrobus Endlicher (Cupressaceae) • LePage 369
Figure 12. The distribution of Glyptostrobus in Japan
and the Asian mainland during the Pliocene and
Pleistocene. Note that the connection between Japan
and the mainland is re-established. Modified from
Minato (1965) and Yang (1991).
deltaic environments that formed in areas where
tectonism, orogeny and sea level change were
prevalent seem to have been the preferred habi-
tats of Glyptostrobus.
Perhaps the single most important observa-
tion of this review of the biogeography of Glypt-
ostrobus is that its distribution pattern is
coincident with periods of environmental distur-
bance, especially disturbances associated with
orogenic events. Glyptostrobus is known to be
intolerant of competition (Farjon 2005), except
in habitats characterized by wet conditions.
Therefore, the habitats in which Glyptostrobus
grows today are probably representative of the
physical setting under which past representatives
grew. The lowland swamps, floodplains and river
deltas that formed in response to active moun-
tain building and areas subject to fluctuations in
sea level throughout the Late Mesozoic and
Cenozoic would have provided the optimal con-
ditions necessary for Glyptostrobus to survive.
Global cooling and drying together with a net re-
duction in areas of active mountain building
contributed to a decline in suitable habitat and,
ultimately, the distribution of Glyptostrobus.
Given the biogeographic history of the genus and
its current restricted distribution, additional
questions emerge. Is Glyptostrobusin the process
of going extinct? Ar