See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/240563906

A New Definition and a Lectotypification of the Genus Cooksonia Lang 1937

Article  in  International Journal of Plant Sciences · February 2010

DOI: 10.1086/648988

CITATIONS READS

28 2,933

2 authors, including:

P. Gerrienne
University of Liège
162 PUBLICATIONS   1,599 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Diversity of Devonian euphyllophytes View project

Annual meeting on recent research about palynology and all connected aspect View project

All content following this page was uploaded by P. Gerrienne on 31 May 2014.

The user has requested enhancement of the downloaded file.

Int. J. Plant Sci. 171(2):199–215. 2010.
Ó 2010 by The University of Chicago. All rights reserved.
1058-5893/2010/17102-0007$15.00 DOI: 10.1086/648988

A NEW DEFINITION AND A LECTOTYPIFICATION

OF THE GENUS COOKSONIA LANG 1937
Paul Gonez1 and Philippe Gerrienne
Paleobotany, Palynology, Micropaleontology (PPM); Département de Géologie, Université de Liège,
Bâtiment B18, Sart Tilman, 4000 Liège 1, Belgium

The genus Cooksonia Lang 1937 includes some of the earliest land plants. Specimens of Cooksonia pertoni
Lang 1937 are considered the earliest Eutracheophytes. The definition of the genus is thus central to the
delineation of the clade. However, the generic diagnosis is problematic. It is not restrictive enough, and most of
the few diagnostic characters are plesiomorphic. Observations on new specimens of Cooksonia paranensis
Gerrienne et al. 2001, a species very close to C. pertoni, considered along with a compilation of the Cooksonia
literature, allow us to propose more precise diagnostic characters. An allometric study was performed on more
than 100 specimens of C. paranensis. This study allows discrimination of true morphological variations from
growth stages. The growth habit of Cooksonia is discussed. An emended diagnosis including apomorphic
characters is given for the genus, as well as a lectotypification of the genus and the type-species.

Keywords: Cooksonia, diagnosis, early land plants, taxonomy.

Online enhancement: appendix tables.

Introduction (ii) propose a taxonomy that tends to be closer to the biologi-

The processes involved in the colonization of land by
plants remain unclear, as is their timing. Most ancient evi-
dence of land plants is spores from Middle Ordovician strata
(at ;470 Ma; Strother et al. 1996). Most ancient phytodebris,
sporangial fragments, are recovered from the Upper Ordovi-
cian (;455 Ma; Wellman et al. 2003). However, putative spo-
rangia were recently found in the Cambrian (;500 Ma) by
Strother (2008). Those fossils provide very fragmentary infor-
mation on the morphologies and systematic affinities of the
presumably terrestrial organisms that produced such biologi-
cal structures.
The earliest aerial shoot in the fossil record is from the Wen-
lock strata of Ireland (;425 Ma); it is a sporophyte of Cookso-
nia Lang 1937 (Edwards et al. 1983). Cooksonia is thus
believed to be one of the earliest land plants. Exceptional pres-
ervation of some specimens of Cooksonia pertoni Lang 1937
allowed demonstration of anatomical features, including simple
tracheids (Edwards et al. 1986, 1992). Those specimens are
considered the most basal Eutracheophytes. The genus, in its
current acceptance, is distributed worldwide (Edwards 1990).
Nevertheless, the delineation of the genus Cooksonia, and
more generally of early land plants, is inaccurate. Numerous
genera were created before the ‘‘cladistic revolution.’’ Most
have not been revised since; as a result, descriptions and di-
agnoses are often incompatible with modern phylogenetic
systematics, which tries to identify shared derived characters.
Revising taxonomy of those early land plants is necessary to
(i) permit a more secure attribution of a specimen to a taxon,
1 Author for correspondence; e-mail: paul.gonez@hotmail.fr.
Manuscript received May 2009; revised manuscript received August 2009.
cal reality (i.e., Linnean species), and (iii) provide a more
solid basis for phylogenetic studies.
We offer here a detailed study of the species Cooksonia para-
nensis. This study consists of a detailed redescription of the
plant and of allometric studies of the major quantitative charac-
ters (branch length and sporangium height and width). It pro-
vides additional characters to define the genus and highlights
the importance of several characters at a generic level. This
study is used to propose hypotheses about the paleobiology of
Cooksonia. We also include a review of the Cooksonia litera-
ture, as it offers characters suitable for an accurate diagnosis.
Taxonomic Background
Taxonomy
The original diagnosis is ‘‘Dichotomously branched, slender,
leafless stems, with terminal sporangia that are short and wide.
Epidermis composed of elongate, pointed, thick-walled cells.
Central vascular cylinder consisting of annular tracheids’’ (Lang
1937, p. 288; fig. 1). Sporangia contain tetrads of spores (Lang
1937). Lang (1937) also discovered fragments of cuticle in situ.
As this genus was erected before January 1, 1958, the desig-
nation of types was not mandatory (McNeill et al. 2006, art.
37.1). Thus, there is neither a type species assigned for the ge-
nus nor holotypes for the two species described by Lang. The
genus and the species are therefore in need of lectotypification.
The Cooksonia Species
Cooksonia pertoni Lang 1937. Characters of the vege-
tative parts are as described above. They are shared by all

199
200 INTERNATIONAL JOURNAL OF PLANT SCIENCES

between the axis and the sporangium is more gradual, and the
sporangium seems to be sunken into the axis.
Cooksonia banksii Habgood et al. 2002. Cooksonia
banksii is morphologically very close to C. paranensis. The
axis/sporangium transition is gradual, and the sporangium is
sunken into the axis. Despite those similarities, a new species
was erected because of the two different preservation modes:
C. paranensis was created on the basis of compression fossils,
and C. banksii was described from permineralizations. Ac-
cording to the authors (Habgood et al. 2002), this forbade
valuable comparison between the two types. Recognizing
a given taxon fossilized in a variety of ways is a recurrent is-
sue in paleobotany (Galtier 1986).

Cooksonia in the Phylogeny

Most recent phylogenies (Kenrick and Crane 1997a, their
fig. 4.31) identify C. pertoni and C. caledonica as stem-group
Eutracheophyte. These Cooksonia species represents basal
members of the clade, which encompasses all fossil and extant
Fig. 1 Gross morphology of Cooksonia, redrawn from the holotype vascular plants. Another species, C. cambrensis, is placed in
of Cooksonia paranensis Gerrienne et al. 2001. The proximal part of the Lycophytina stem group (Kenrick and Crane 1997a, their
the plant is not known. fig. 4.31), indicating that the genus in its present definition is
polyphyletic. Furthermore, in Kenrick and Crane’s (1997a)
analysis, Cooksonia shares many character states with the
the species described below. Sporangia are considerably outgroup members, which shows that those character states
stretched horizontally. Their morphology (fig. 2A) is homoge- are plesiomorphic in Cooksonia. Consequently, a better delin-
nous, but there is a great variability in size. Lang (1937) ob- eation of the genus is essential to understand (i) the plesiomor-
served trilete spores in situ. Several types of spores have later phic state of characters within the Eutracheophytes, (ii) the
been isolated. This led to the erection of four subspecies: relationships between fossil and extant Eutracheophytes, and
Cooksonia pertoni ssp. pertoni, C. pertoni ssp. synorispora, (iii) the innovations that permitted the land plant explosion dur-
C. pertoni ssp. apiculispora (Fanning et al. 1988), and C. per- ing the Early Devonian (Kenrick and Crane 1997b).
toni ssp. reticulispora (Habgood et al. 2002).
Cooksonia hemisphaerica Lang 1937. This species (fig.
Cooksonia: A Genus Characterized by
2B) was described from the type locality. It differs from C.
Nonrestrictive Characters
pertoni in having globose, hemispherical sporangia. Sporan-
gia are approximately as high as they are wide, and the sub- The diagnosis cited above is problematic because of the
tending axis widens just below the sporangium (Lang 1937). lack of restrictive characters. Many early land plants show
They contain trilete spores (Edwards 1979). dichotomous leafless axes and terminal sporangia. The genus
Cooksonia crassiparietilis Yurina 1969. This species is
known from three specimens only. The sporangium is reni-
form and shows a thick distal dehiscence line (Edwards
1970) dividing the sporangium into two equal parts.
Cooksonia caledonica Edwards 1970. This species (fig.
2C) was included in the genus on the basis of the ‘‘wider-than-
high sporangium’’ character. This species differs from C. perto-
ni in the sporangial construction, as Edwards (1970) identified
a distal dehiscence line, opening the sporangium into two equal
valves.
Cooksonia cambrensis Edwards 1979. The characters
of the species (fig. 2D) are a spherical or subspherical sporan-
gium and no widening of the axis below the sporangium.
Cooksonia bohemica Schweitzer 1980. Only one spec-
imen is known. Vegetative parts of this species present origi-
nal characters: axes are profusely branched and more robust
than in other species (Schweitzer 1980). Sporangia are badly
preserved and present various shapes due to compression.
Cooksonia paranensis Gerrienne et al. 2001. This spe-
cies (fig. 2E) is close to C. pertoni. A few differences are percep- Fig. 2 Sporangium morphology of Cooksonia species. A, C. pertoni;
tible: axes are more slender than in C. pertoni, the transition B, C. hemispaerica; C, C. caledonica; D, C. cambrensis; E, C. paranensis.
 GONEZ & GERRIENNE—A NEW DEFINITION OF COOKSONIA 201

Cooksonia can thus encompass a large number of morphol-

ogies with different potential affinities. Controversial attribu-
tions of fossils to the genus Cooksonia are numerous. For
instance, Cooksonia downtonensis Heard (1939) was infor-
mally attributed to Steganotheca in Edwards (1970); Cookso-
nia sp. in Croft and Lang (1942, pl. XI, fig. 43) was renamed
Uskiella by Shute and Edwards (1989). The most restrictive
part of the diagnosis is the wider-than-high sporangium char-
acter (Edwards 1970), but as the sporangia are often com-
pressed in various ways, it needs to be more explicitly defined.
Absence of apomorphic character in the original diagnosis
is problematic too. Isotomous branching, elongated epider-
mal cells, and lack of leaves are characters shared by many
early land plants. The concept of Cooksonia as defined by
Lang (1937) is therefore hardly usable in cladistic studies
(Kenrick and Crane 1997a). Clearly, apomorphic characters
have to be defined for the genus. This would allow producing
more accurate phylogenetic studies. The elucidation of phylo-
genetic relationships among basal Embryophytes is of great
importance as it is becoming increasingly clear that most of the Fig. 3 Illustration of the distances representing sporangium width
extant groups diverged very early in the history of land plants. (a), sporangium height (b), terminal branch length (c).

Material and Methods
Material
Cooksonia paranensis. Specimens of C. paranensis be-
long to a flora that was collected from the Paraná Basin, Bra-
zil, from nine localities distributed along the northeastern
and southeastern borders of the basin, in a coastal environ-
ment. The fossiliferous beds are early Lochkovian in age, as
indicated by the simultaneous occurrence of the spores Aneu-
rospora geiki Wellman and Richardson, Synosporites ver-
rucatus Richardson and Lister, and Dictyotriletes granulatus
Steemans (Gerrienne et al. 2001). Cooksonia paranensis has
been recovered from five localities (Gerrienne et al. 2001).
Brazilian fossils are preserved as adpressions, that is, com-
pressions that lost more than 80% of their initial volume
(Bateman 1991). Material is housed in the paleobotany col-
lections of the University of Liège, Belgium.
Type material of Cooksonia pertoni. Type material was
originally collected from Perton Quarry, Wales. Fossils are
conserved as adpressions (Bateman 1991) in a fine-grained
sandstone from ‘‘Downtonian’’ (approximately equivalent to
the Pridoli) strata, in a proximal marine environment (Lang
1937). The type material is composed of 12 specimens, three
of which possess a counterpart. It is housed in the paleobot-
any collections of the Natural History Museum of London.
Methods
All specimens were freed from remains of sediment cover-
ing them by dégagement (Fairon-Demaret et al. 1999). Digital
photographs of the specimens were taken with a Zeiss Stemi
2000C stereomicroscope equipped with a CCD camera, and
measurements were taken using AxioVision digital image pro-
cessing software.
Measurements were taken according to a standardized
method (fig. 3). Sporangium width corresponds to maximal
diameter of the sporangium. Sporangium height was more
difficult to estimate. The basis of the sporangium was arbi-
trarily set as the point where the edges of the axis diverge.
The length of the terminal axis was measured with the spo-
rangium included, in order to avoid the uncertainty of the
proximal limit of the sporangium. Deformed or folded speci-
mens were not measured.
Descriptive statistical values were computed for each pa-
rameter and will be referred to as follows: x stands for mean,
V(x) is the variance, and N is the number of measures for one
parameter. We performed only uni- and bivariate analyses. The
measures were calculated with Microsoft Office Excel 2007.
For bivariate analyses, that is, allometric study, the relationship
between the two parameters was determined by the ordinary
least squares regression. The lack-of-fit test (a ¼ 0:01; Burns
and Ryan 1983) was performed, and each regression passed it.
Therefore, the probability that the correlation observed actu-
ally exists is more than 99% in each case.
Results
Brazilian Material: Cooksonia paranensis
Axis description. Cooksonia axes are smooth and devoid
of epidermal outgrowths. Axes branch dichotomously, most
divisions being strictly isotomous. Two specimens show one
trichotomy (fig. 4A). Two large specimens show three levels of
branching (fig. 4A, 4B), as do the holotype (Gerrienne et al.
2001, pl. I: 1, 2) and the probably complete specimen il-
lustrated by Gerrienne et al. (2006, pl. I); none of the other
specimens show more than two levels of branching. Most
specimens are proximally incomplete. On one specimen, the
preserved part of the most proximal axis reaches 47 mm (fig.
4B). This specimen is 69 mm long (fig. 4B) and is the longest
observed in our collection. One specimen shows a small trian-
gular emergence at the branching point; it is ;0.25 mm long
and located in the continuity of the subtending axis (fig. 4C,
4D).
Axis width is constant along a branch, but changes accord-
ing to the level of branching. The width of a branch resulting
202 INTERNATIONAL JOURNAL OF PLANT SCIENCES

Fig. 4 Gross views of Cooksonia paranensis. A, ULg 13541, showing three levels of branching, including a trichotomy. B, ULg 13542, the
longest specimen of the collection, showing three levels of branching; a long proximal axis is preserved. C, ULg 13543; the arrow points to an
emergence at the dichotomizing point, interpreted as an aborted trichotomy. D, ULg 13543, detail of the emergence of the specimen.

from a dichotomy decreases on average by 26% (N ¼ 33) as
compared with the width of the branch below the branching
point.
The length of the branches depends on the branching level.
Distalmost branch (which will be referred to as d-level branch)
length shows a wide range of variation (1.68–18.94 mm; x ¼
9.18 mm; V ðxÞ ¼ 24:41; N ¼ 40). The ratio between the d-level
branch length and the sporangium diameter is presented in fig-
ures 5 and 6. Lower branching level axis (d-1- and d-2-level
branches) length varies much less: from 2.63 to 7.04 mm;
x ¼ 4:7 mm, V ðxÞ ¼ 2:01, and N ¼ 13.
One some specimens, the compression preserved the epi-
dermal cells imprint of a sporangium (fig. 7A, 7B). The cells
appear to be elongated (;10 mm wide and 60–80 mm long)
proximally and more or less isodiametric (5–10 mm across)
distally.
Interpretation. A strong linear correlation exists between
sporangium width (which likely reflects its growth stage) and
the d-level branch length (fig. 5). This correlation suggests that
the sporangium is already differentiated at or just after the divi-
sion of the axis and that sporangium growth and maybe matu-
ration are contemporaneous with the axis elongation (fig. 6).
 GONEZ & GERRIENNE—A NEW DEFINITION OF COOKSONIA 203

Fig. 5 Relationship between sporangium width and terminal branch length. R ¼ 0:7582; a ¼ 1%.

Alternatively, maturation could also occur only after axis elon-
gation. Our record of specimens showing a strong variation of
the terminal axis length is thus interpreted as the record of vari-
ous ontogenetic stages of the terminal branches and of the spo-
rangium. The growth is probably carried out by cell-stretching,
as shown by the cellular pattern of the epidermis (fig. 7A).
Weak variance in the d-1 and d-2 branches length suggests
that their growth is completed. Length of d-1 and d-2 branches
for mature specimens (i.e., with long d-level branches and large
sporangia) is always much shorter compared with d-level
branches length (40% on average; N ¼ 13). This seems to be a
diagnostic character for C. paranensis. The fact that the growth
occurs only in terminal branches suggests that it is performed
by an apical meristematic structure, as for all extant Polyspo-
rangiophytes. Four growth stages of C. paranensis are schema-
tized at fig. 6.
The presence of occasional trichotomies suggests that the
branching pattern of the genus is more variable than cur-
rently recognized. The structure indicated by an arrow in fig.
4C and enlarged in fig. 4D is interpreted as an aborted tri-
chotomy.
Description of sporangia in apico-basal view. Most of the
sporangia preserved in apico-basal view are circular in outline
(fig. 7C); some are elliptical (fig. 7D). Diameter ranges from
0.58 to 3.84 mm (x ¼ 2:05 mm; N ¼ 135). On two specimens
(fig. 7E, 7F), the three-dimensional shape of the sporangium
is preserved. One (fig. 7E) shows the external morphology,
plateau-shaped; the other (fig. 7F) exhibits the internal mor-
phology of the sporangium, a funnel-shaped pit, but most speci-
mens are flat (fig. 7C, 7D, 7G, 7H). On some specimens
(fig. 7C, 7D, 7G, 7H), the border of the sporangium is black.
Border width ranges from 0.09 to 0.19 mm (x ¼ 0:14 mm;
N ¼ 7).
Interpretation. The outline of the sporangium top is circu-
lar; elliptic shapes probably result from taphonomical condi-
tions. This is interpreted as a diagnostic character, probably
at the generic level. The plateau-shaped specimens suggest
the presence of an operculum on the top of the sporangium.
The border (which marks the place of the sporangial wall)
encloses a cavity that contained sporogenous tissue. The pro-
gressive darkening toward the sporangial periphery is inter-
preted as indicating the presence of the sterome described
by Edwards et al. (1986) or most likely a concentration of or-
ganic matter toward the sporangial periphery that occurred
during the infill of the sporangial cavity by sediment. The
dark area in the center of some specimens is presumably the
point of attachment of the subtending axis.
Description of laterally compressed sporangia. Most later-
ally compressed sporangia are elongated. The subtending
axis width increases over a more or less short distance to form
the sporangium. Diameter ranges from 0.58 to 3.49 mm,
x ¼ 1:70 mm; height ranges from 0.38 to 3.76 mm, x ¼ 1:64
mm; N ¼ 89. Lateral compression of the apical plateau results
in three types of preservation. (i) The apical plateau is seen in
profile. It thus appears as a line, perpendicular to the subtend-
Fig. 6 Illustration of four growth stages of Cooksonia paranensis.
 Fig. 7 Sporangia of Cooksonia paranensis. A, ULg 13544, showing a cellular pattern at its surface. B, ULg 13544, general view. C–H,
Sporangia in apical views. C, ULg 13545, sporangium with circular outline. D, ULg 13546, sporangium with elliptical outline. E, ULg 13547,
sporangium with apical plateau. F, ULg 13548, sporangium with the external shape of its cavity preserved. G, ULg 13517, sporangium with
a darkened peripheral zone. H, ULg 13549, sporangium with a clearly individualized peripheral ring. I–L, Laterally compressed sporangia. I, ULg
13551, sporangium with horizontal apical plateau. J, ULg 13552, sporangium with apical plateau tilted toward the observer. K, ULg 13553,
sporangium with apical plateau tilted toward the sediment. L, ULg 13554, three-dimensionally preserved sporangium, showing the shape of the
flared internal funnel.
 GONEZ & GERRIENNE—A NEW DEFINITION OF COOKSONIA
ing axis (fig. 7I). (ii) The apical plateau is tilted toward the
observer (fig. 7J). This configuration is the most frequent. (iii)
The apical plateau is tilted in the opposite direction or folded
up (fig. 7K). This configuration is the least frequent. One spec-
imen is three-dimensionally preserved: the sporangium is
funnel-shaped (fig. 7L). On some specimens, a lighter zone is
discernible. Its basal limit is well marked and curved (fig. 7K).
Sporangia present a wide range of morphological variation:
some characteristic examples are presented and described
(fig. 8; table 1).
Interpretation. The fact that the apical plateau is rarely
folded suggests that the sporangium walls are stiff. The clear-
cut, curved, lighter zone on some specimens may represent
the sporangial cavity. It extends more or less deeply in the
subtending axis. Variation of the concavity of the sporangial
edges is most probably of taphonomic origin. This character
should therefore not be taken into account in estimating the
morphological variability.
Sporangium width and height are significantly correlated
(fig. 9A) and show a continuous variation. This suggests that
this variation occurs at an intraspecific level. A specimen
showing two sporangia of clearly different morphologies (fig.
8H) strengthens the statistical data. This shows that the cate-
gories of sporangium presented in table 1 do not have any
taxonomical value. The range of sporangial shapes of the
Brazilian material can most probably be explained by within-
population variation coupled with taphonomic biases. This
conclusion is confirmed by the normal bell curve distribution
of the sporangium width (fig. 9B).
Type Material: Cooksonia pertoni
Description. Most fossils are not very well preserved. This
results in difficulties with measurements, especially of sporan-
gium height. Several illustrated specimens (Lang 124C, 92E,
239C, 1061, 1021) do not show any sporangia. Their attribu-
tion to the genus Cooksonia is therefore speculative. Those
fossils will not be taken into consideration. No indisputable
specimens of Cooksonia (i.e., showing at least one sporan-
gium) show more than one level of branching. Most axes are
slender (fig. 10A). Only one specimen shows more robust axes
(fig. 10B).
One specimen shows a pointed emergence on the sporan-
gium (fig. 10C). Some other sporangia, preserved as apico-
basal compressions, bear such emergences (fig. 10D). They
are not mentioned by Lang.
Sporangia preserved as lateral compressions are more
abundant. Transition between axis and sporangium is gradual.
The subtending axis width increases over a short distance to
form the sporangium. Sporangium is shaped like a flat-sided
trumpet (fig. 10A). Some specimens show much flattened spo-
rangia: they are disk shaped (fig. 10B, 10E). The top of the
sporangium is often distorted. A more important quantity of
organic matter is preserved on the top (fig. 10E). A few spo-
rangia are preserved as apico-basal compressions (fig. 10F).
These structures have been described by Lang (1937, p. 279)
as ‘‘circular bodies, incertae sedis.’’ Their diameter ranges
from 1 to 1.5 mm.
Interpretation. The sporangial shape suggests that its con-
struction in C. paranensis and C. pertoni is strongly similar.
205
Distortion of the sporangia can be explained by a combina-
tion of two circumstances. (i) Sporangia are fossilized in a
marine facies. They probably underwent a longer transporta-
tion than the Brazilian specimens. (ii) Distal part of the spo-
rangia, very flared, causes the sporangium to be disk shaped
(fig. 10F). This disk constitutes a more breakable zone. It is
torn on one specimen (fig. 10B). This distal flattening of the
sporangium is a specific character discriminating C. pertoni
from C. paranensis: none of the Brazilian specimens show it.
Specimens bearing emergences do not belong to the genus
Cooksonia. They are part of a close genus, Pertonella Fanning
et al. 1991. Spines are not taken into account in the circum-
scription of the concept of Cooksonia because we consider
that presence of spines is a character to be considered of a ge-
neric level rather than specific at this grade of organization.
Fanning et al. (1990, p. 727), for instance, justify the creation
of the genus Caia by ‘‘the shape of the sporangial apex and the
presence of sporangial emergences.’’ This decision is taken to
emphasize the diversity present in the earliest plant macrofos-
sil assemblages.
Discussion
Toward a New Diagnosis of the Genus
Compilation of characters. Several morphological and an-
atomical characters can now be used to define the genus (fig.
11). Five characters are related to the axis: (i) occasional tri-
chotomies (in Cooksonia paranensis [this article]); (ii) a cuti-
cule (in Cooksonia pertoni; SEM observation [Edwards et al.
1986]); (iii) stomata with reniform guard cells and appar-
ently no modified subsidiary cells, present in the distal part
of an axis and on the axis/sporangium transition (in C. pertoni;
SEM observation [Edwards et al. 1986, 1992]; in Cooksonia
banksii; SEM observation [Habgood et al. 2002]); (iv) a ste-
rome, that is, a peripheral thickened cellular zone that helps
keep the plant erect (in a transverse section of an axis of C.
pertoni; SEM observation [Edwards et al. 1986]); (v) a cen-
tral strand of C-type tracheids, 2–11 mm in diameter, with
ring-shaped or annular secondary thickenings (in C. pertoni;
SEM observation [Edwards et al. 1992; Edwards 1993, 2003]).
These characters are not apomorphic for Polysporangio-
phytes, but they have an importance in the circumscription
of the genus, as they are particularly restrictive when discov-
ered on such old material.
Apomorphic characters can also define the Cooksonia spo-
rangium: (i) sporangium in direct continuity with the sub-
tending axis that widens gradually so that the limit between
the axis and the sporangium is transitional (in C. pertoni and
in C. paranensis [this article]; in C. banksii [Habgood et al.
2002]); (ii) sporangium shaped like a flat-sided trumpet (in
C. pertoni [Edwards et al. 1992] and in C. paranensis (this
article)]; (iii) top of sporangium circular in apico-basal view
(in C. pertoni [Edwards et al. 1992] and in C. paranensis;
[this article]); and (iv) top of the sporangium flat and com-
posed of one layer of isodiametric cells (in C. pertoni; SEM
observation [Edwards et al. 1986]) and additional papillate
cells (in C. pertoni; SEM observation [Edwards et al. 1992]).
At the periphery, this one-cell-thick layer is attached to the
most distal part of the sporangium wall. Figure 1a from Ed-
 Fig. 8 Morphological diversity of the sporangia of Cooksonia paranensis. A, ULg 13549, sporangium with straight edges. B, ULg 13555, long
sporangium with inward curved edges. C, ULg 13556, short sporangium with inward curved edges. D, ULg 13557, short sporangium with
outward curved edges. E, ULg 13543, long sporangium with outward curved edges. F, ULg 13532, sporangium showing two levels of flaring. G,
ULg 13558, sporangium with one outward curved edge and one inward curved edge. H, ULg 13559, specimen showing two attached sporangia
with different shapes.
 GONEZ & GERRIENNE—A NEW DEFINITION OF COOKSONIA 207

Various Categories of Sporangial Morphologies of Cooksonia paranensis Observed in the Paraná Basin Flora
Sporangium characters
Sporangium with straight edges
Sporangium edges are curved inward:
Long
Short
Sporangium edges are curved outward:
Short
Long
Double concavity
Sporangium with one edge curved
outward and the other inward
Table 1
Description Frequency index Illustration
Axis/sporangium transition is angular to subangular; all 3 Figure 8A
sporangia of this type are elongated
Sporangium is outlined by a gentle flaring of the subtending 2 Figure 8B
axis; on some specimens, edges of the sporangium become
distally perpendicular to the axis.
Axis splays abruptly to form the sporangium 1 Figure 8C
Sporangium is wine glass shaped 1 Figure 8D
Edge curving is slight: edges are almost straight; one some 3 Figure 8E
specimens, edges of the sporangium become distally
perpendicular to the axis
Axis exhibits two flaring levels 1 Figure 8F
On some specimens, edges of the sporangium become distally 1 Figure 8G
perpendicular to the axis,

wards et al. (1992) clearly shows this layer, the peripheral
part of which forms a ring slightly raised as compared to the
center of the sporangium. This is probably the result of com-
paction and/or dehydration of the sporangial cavity content.
The sporangial cavity is shallow in C. pertoni (D. Edwards,
personal communication, 2008).
No clear dehiscence feature has been detected on the sporan-
gia of C. paranensis. Some specimens of C. pertoni (see Edwards
et al. 1992, their fig. 1a) suggest that the operculum was sim-
ply torn (probably by desiccation) and then let the spores out.
The equatorial crassitude of in situ Cooksonia spores is con-
sidered to be a character of genus-level importance. Spores
with diverse morphological features are found in Cooksonia
sporangia. Table 2 is based on the review published by Ed-
wards and Richardson (1996) and presents a synthesis of dif-
ferent spore types found in Lower Devonian sporangia. The
upper part of the table shows different spores taxa identified
in Cooksonia sporangia; spores from Cooksonia caledonica
and close plants are presented in the middle part of the table;
the lower part presents the spore taxon found in Pertonella,
a taxon morphologically close to Cooksonia. The spores from
C. pertoni, Cooksonia cambrensis, and C. banksii might ap-
pear heterogeneous, but they all present an equatorial crassi-
tude. The spores from cf. C. caledonica and close taxa form
a homogeneous group (Apiculiretusispora and Retusotriletes
are close genera; both are retusoid, that is, with prominent
contact areas and curvatures). It should be noted that Cooksonia
and Pertonella, that are morphologically very close, produce
different spores, which is another demonstration of the cryptic
diversity emphasized by Fanning et al. (1988).
Comparison of the species of Cooksonia to the updated
concept of the genus. Although Lang (1937) did not explic-
itly designate C. pertoni as the type species, it is clear that
this taxon has to be considered as the type of the genus (see
McNeill et al. 2006, art. 37.3; Lang 1937, p. 253): it is the
first species described in the protologue (Lang 1937, p. 250).
It is also the best-known species.
One of us (P. Gonez) examined the type material of C. cam-
brensis. The compressed sporangia are not trumpet shaped.
Preservation of the sporangial morphology in C. cambrensis
Edwards prevents a secure tridimensional interpretation of the
compression. It is even possible that some specimens of C.
cambrensis could be badly preserved or unusually compressed
C. pertoni specimens (D. Edwards, personal communication,
2008). Thus, characteristic morphology of C. cambrensis
might be the result of taphonomic biases, so that its specific
status will not to be reconsidered here. Firstly, its preservation
does not allow a tridimensional reconstruction of the sporan-
gium and accurate comparisons. More material is needed to
come to such conclusions. Secondly, its specific status was
confirmed by morphological evidence. Fanning’s univariate
analysis (1987), involving C. pertoni, C. hemisphaerica, and
C. cambrensis, shows discontinuities in the distribution of the
values of several quantitative characters that correspond to
the species involved in the study.
We also examined the type material of C. hemisphaerica
Lang. The compressed sporangia are not trumpet shaped.
Two specimens show features suggestive of a dehiscence line.
The material is most probably heterogeneous. Such features
were also observed by Fanning et al. (1992). However, we de-
cided not to exclude C. hemisphaerica from the genus for two
reasons. (1) Neither Edwards nor we could decide whether
the features observed are a true dehiscence line. (2) Cooksonia
hemisphaerica is a species known from very poorly preserved
compressions that prevent a secure three-dimensional interpre-
tation of the sporangial shape. Some specimens of the type ma-
terial show sporangia that could have been trumpet shaped in
life (P. Gonez, personal observation). The spores of C. hemi-
sphaerica are devoid of an equatorial crassitude: this is the only
demonstrative feature that suggests that the species could be-
long to another genus. Nonetheless, we think that this is insuffi-
cient to justify the exclusion of C. hemisphaerica from the
genus Cooksonia. More information is needed to make a deci-
sion concerning the status of that species.
We also examined the type material of C. caledonica Ed-
wards. The sporangia of that plant are clearly different from
those of C. pertoni: they are more or less reniform and bi-
valved, and they possess a specialized subdistal structure for
208 INTERNATIONAL JOURNAL OF PLANT SCIENCES

Fig. 9 Cooksonia paranensis sporangium population: A, Relationship between sporangium width and height, R ¼ 8349; a ¼ 1%. B, Distri-
bution of the values of sporangium width for the 219 measured specimens.

dehiscence. This is sufficient to consider that C. caledonica
should be included in another genus (Edwards et al. 2001).
From high-definition images of the type material (courtesy
of the Lomonosov Moscow State University), the sporangia
of Cooksonia crassiparietilis Yurina are bivalved and show a
thick structure for dehiscence. Although much larger, they are
very close to that of C. caledonica (Edwards 1970). Cooksonia
crassiparietilis should accordingly be also included in another
genus.
From high-definition images of the holotype (courtesy of the
Natural History Museum of Sweden), Cooksonia bohemica
Schweitzer appears as a tiny bushy plant; one sporangium is
undoubtedly attached to the vegetative parts. Bad preserva-
tion prevents secure three-dimensional interpretation of the
sporangia, but one is interpreted as being trumpet shaped.
No character of the fossil is clearly inconsistent with the ge-
neric updated diagnosis proposed here.
Cooksonia paranensis Gerrienne et al. is an indisputable
member of the genus as it shows almost exactly the same
sporangial construction as the type species. It can be distin-
guished from C. pertoni by its slender axes and the more
gradual transition between axis and sporangium.
Cooksonia banksii Habgood et al. undoubtedly belongs to
the genus Cooksonia too, as it also shows almost exactly the
same sporangial construction as the type species. As with C.
paranensis, C. banksii can be distinguished from C. pertoni
by its more gradual transition between subtending axis and
the sunken sporangial cavity. Cooksonia banksii is on the
 Fig. 10 Type material of Cooksonia pertoni. A, V58011a (Lang no. 1242a), the best-preserved specimen of the type material. B, V58007 (Lang
no. 124X), showing robust axes and a torn sporangium. C, V58006 (Lang no. 124F), showing an emergence (indicated by the arrow); specimen
attributed to the genus Pertonella. D, V58011b (Lang No 1242b), sporangium in apical view, showing emergences; specimen attributed to
Pertonella. E, V58008 (Lang no. 124Yb), sporangium with a disk-shaped apex. F, V58011b (Lang no. 1242b), sporangium in apical view.
210 INTERNATIONAL JOURNAL OF PLANT SCIENCES
Fig. 11 Cooksonia. Hypothetic longitudinal section of a sporangium and its subtending axis, showing morphological and anatomical
apomorphic characters.
contrary hardly distinguishable from C. paranensis: their spo-
rangial morphologies are identical. One might thus consider
C. banksii as a junior synonym of C. paranensis. Habgood
et al. (2002) argue that different type of preservation (i.e.,
compression for C. paranensis and permineralization for C.
banksii) prevents any meaningful comparison of the two fos-
sil materials. This is not necessarily true; permineralization
fossils and compression fossils are commonly given the same
name. For example, permineralized specimens were attrib-
uted by Edwards et al. (1994) to the genus Salopella Edwards
and Richardson 1974, created for compressed specimens. In
this case, a slight difference between C. paranensis and C.
banksii is nevertheless perceptible. The epidermis of the spo-
rangium wall in C. banksii includes elongate cells proximally
and large (20–30 mm across) isodiametric cells distally; those
isodiametric cells are very small (5–10 mm) in C. paranensis.
This appears to be sufficient to keep C. paranensis and C.
banksii as separate species.
Several records were left as Cooksonia sp. Most of them
were given that name because sporangia are incomplete (Ed-
wards et al. 2001) or because of the bad preservation (Edwards
and Rogerson 1979; Janvier et al. 1987). Poor preservation
prevents a secure three-dimensional interpretation of the com-
pressed sporangia morphology and implies absence of any re-
markable feature except overall morphology. Therefore, those
specimens do not add any information about the construction
or the morphological variability of the sporangia of Cookso-
nia. The name of Cooksonia was even probably given to those
fossils because of the wide range of morphologies that is en-
compassed by Lang’s (1937) diagnosis.
Yet three records of fossils identified as Cooksonia sp. are
worth discussing. (1) Cooksonia-type sporangia from the
Wenlock strata of Ireland (Edwards et al. 1983, their figs. 1,
2): the illustrations shows compressed sporangia of a flared
trumpet shape. Fossils are said to resemble C. pertoni. The
specific name was not attributed with certainty because of
preservation, but those specimens undoubtedly belong to
Cooksonia and represent the first occurrence of the genus.
However, this specimen was not demonstrated to be vascular-
ized. (2) Cooksonia sp. (Edwards et al. 2004): This Cookso-
nia shows terminal branches decreasing in diameter toward
the apex, with a distalmost gentle flaring of the axis form-
ing the sporangium. This growth morphology is unusual in
Cooksonia records, but sporangial architecture is the same as
other Cooksonia (i.e., consisting in a progressive flaring of
the axis). It is therefore clearly a member of the genus. As
the authors pointed out, more material is needed either to
give a specific attribution to this fossil or to erect a new spe-
cies. (3) Cooksonia sp. (Schultka 2003): This record includes
several complete or nearly complete plant remains. The distal
structures interpreted as sporangia are variously shaped and
lobed. According to the author (Schultka 2003, p. 7), the fos-
sils show an organization strongly analogous to that of the
gametophyte Sciadophyton. We suggest that they should bet-
ter be attributed to that genus.
Phylogeny of the updated Cooksonia. The relationships of
Cooksonia within basal Polysporangiophytes were estimated
through two cladistic analyses based on the work of Kenrick
and Crane (1997a, chapter 4) to which we added the new
data on the sporangium morphology of Cooksonia. The 33
characters and the character states of our analyses are listed
in table A1 in the online edition of the International Journal
of Plant Sciences. The characters used for this analysis are
those selected by Kenrick and Crane (1997a). The character
state ‘‘trumpet-shaped sporangium’’ (coded 6) was added to
character 27 ‘‘sporangium shape,’’ and the character state
 Table 2
In Situ Spores Found in Some Sporangia from Lower Devonian Taxa
Parent plant species, age Spore construction and sculpture Taxon of dispersed spore Size range (mm) Author of spore identification
Cooksonia pertoni:
Pridoli: Laevigate, crassitate Ambitisporites 22–43 Fanning et al. 1988
Murornate, crassitate Synorisporites verrucatus 15–30
Apiculate, crassitate Streelispora newportensis Aneurospora 18–30
Crassitate, reticulate distal ornament, and
laevigate contact faces cf. Chelinospora 13–15 Habgood et al. 2002
Cooksonia hemispherica:
Lochkovian Curvaturate, microgranulate Apiculiretusispora ? new taxon ? 23–32 Fanning and Edwards 1992
Cooksonia cambrensis:
Pridoli Microgranulate, crassitate New taxon ? 9–22 Fanning et al. 1991a
Cooksonia banksii:
Lochkovian Laevigate, crassitate cf. Ambitisporites avitus 13–23 Habgood et al. 2002
Cf. Cooksonia caledonica/Renalia:
Lochkovian Azonate, curvature coincident with equator, Apiculiretusispora sp. 12–16 Fanning et al. 1992
distally granulate, proximally microgranulate
Renalia hueberi:
Emsian Azonate, fine ornament or laevigate, curvaturate Retusotriletes/Apiculiretusispora 46–70 Gensel 1976
Cooksonia crassiparietilis:
Emsian Azonate, partly curvaturate, small grana Apiculiretusispora cf. plicata 50–60 McGregor 1973
Resilitheca salopensis:
Lochkovian Curvaturate, azonate Retusotriletes cf. philipsii-regulatus 22–39 Edwards et al. 1995
Sporathylacium salopense:
Lochkovian Crassitate, noncurvaturate, microgranulate New taxon ? ?;30? Edwards et al. 2001
ornament, possible distal verrucate/murornate
structure
Pertonella dactylethra:
Pridoli Distally laevigate, curvaturate and with tectal Retusotriletes coronadus 34–45 Fanning et al. 1991b
sculpture
 Fig. 12 Cladogram of relationships among basal Polysporangiophytes (modified from Kenrick and Crane 1997a). All the Cooksonia species
are included; 50% majority rule consensus tree. See tables A1 and A2 in the online edition of the International Journal of Plant Sciences for
character list, character states, and data matrix.

212
 Fig. 13 Cladogram of relationships among basal Polysporangiophytes (modified from Kenrick and Crane 1997a). Only Cooksonia caledonica,
Cooksonia paranensis, and Cooksonia pertoni are included. Strict consensus tree. See tables A1 and A2 in the online edition of the International
Journal of Plant Sciences for character list, character states, and data matrix.

213
214 INTERNATIONAL JOURNAL OF PLANT SCIENCES
‘‘breaking up of operculum’’ (coded 5) was added to charac-
ter 29 ‘‘sporangium dehiscence.’’ We first analyzed a data
matrix including all the taxa considered by Kenrick and
Crane (1997a) to which C. banksii, C. bohemica, C. crassi-
parietilis, C. hemisphaerica, and C. paranensis were added
(table A2 in the online edition of the International Journal of
Plant Sciences). Haplomitrium and Sphaerocarpos were des-
ignated as outgroups. Trees were generated using the heuris-
tic search routine of PAUP (ver. 4.0b10; Swofford 1998). The
analysis yielded more than 2,300,000 equally parsimonious
trees of 74 steps with consistency indexes of 0.7568. The
50% majority rule consensus tree (fig. 12) confirms that the
genus sensu Lang (1937) is polyphyletic. The position of C.
banksii, C. paranensis, and C. pertoni is unresolved relative
to the Euphyllophytina and the Lycophytina; C. caledonica
and C. crassiparietilis are placed in the Lycophytina stem
group. The second analysis was performed with only the three
best-known Cooksonia species, C. pertoni, C. caledonica, and
C. paranensis. The analysis yielded 1134 equally parsimoni-
ous trees of 73 steps with consistency indexes of 0.767. The
50% majority rule consensus tree is presented at fig. 13. It
confirms that C. pertoni and C. paranensis are closely related
species. They are in a sister group relationship with the Lyco-
phytina and related taxa rather than with the Eutracheo-
phytes. Cooksonia caledonica is more closely related to the
Lycophytina, which again confirms the polyphyletic status of
the genus Cooksonia sensu Lang (1937) and the need to in-
clude the species caledonica in another genus.
Systematics
In the original publication of the genus, Lang did not clearly
identify a type species for his genus. Moreover, he did not as-
sign type specimens for the species he described. However, be-
fore 1958, designation of type specimen was not mandatory
(McNeill et al. 2006; art. 37.1). Lang’s descriptions are there-
fore valid. Nevertheless, a lectotype has to be designated
(MacNeill et al. 2006; art. 9.2).
Cooksonia Lang 1937
Emended diagnosis. Erect plant, up to 7 cm high; axes
branching up to three times isotomously, with occasional tri-
chotomies. Axes and sporangia covered by a cuticle including
stomata with reniform guard cells. Axis with peripheral ster-
ome and central vascular strand of C-type tracheids. Gradual
subtending axis/sporangium transition; sporangium terminal,
trumpet shaped. Top of the sporangium flat, circular in apico-
Literature Cited
Bateman RM 1991 Palaeoecology. Pages 34–116 in CJ Cleal, ed.
Plant fossils in geological investigation: the Palaeozoic. Ellis
Harwood, New York.
Burns DA, TA Ryan Jr 1983 A diagnostic test for lack of fit in
regression models. Pages 286–290 in Proc Stat Comput Sect.
American Statistical Association, Washington DC, USA.
Croft WN, WH Lang 1942 The Lower Devonian flora of the Senni
Beds of Monmouthshire and Breconshire. Philos Trans R Soc B 231:
131–164.
basal view, and composed of one layer of isodiametric cells.
Spores with a regular equatorial crassitude.
Type species. Cooksonia pertoni Lang 1937.
Species excludenda. Cooksonia caledonica Edwards 1970;
Cooksonia crassiparietilis Yurina 1969.
Other species. Cooksonia hemisphaerica Lang 1937 (doubt-
ful because based on badly preserved specimens); Cooksonia
cambrensis Edwards 1979 (doubtful because based on badly
preserved specimens); Cooksonia bohemica Schweitzer 1980
(doubtful because based on badly preserved specimens); Cook-
sonia paranensis Gerrienne et al. 2001; Cooksonia banksii
Habgood et al. 2002.
Cooksonia pertoni Lang 1937
Lectotype. Lang no. 1242/V58011 (Lang 1937, pl. VIII,
fig. 8; fig. 10A of this article). This is the first specimen illus-
trated by Lang and the best preserved in the original material
collection.
Syntypes. Lang no. 124Y/V58008 (Lang 1937, pl. VIII, fig.
7; fig. 10E in this article); Lang no. 124X/V58007 (Lang 1937,
pl. VIII, fig. 12; fig. 10B in this article); Lang no. 261A/V58027
(Lang 1937, pl. IX, fig. 27).
Excludenda. Lang no. 124F/V58006 (Lang 1937, pl. VIII,
fig. 6; fig. 10C in this article), which belongs to the genus Per-
tonella; Lang no. 124 C (Lang 1937, pl. VIII, fig. 4); Lang 92
G/V58004 (Lang 1937, pl. IX, fig. 22); Lang no. 92E (Lang
1937, pl. IX, fig. 21); Lang no. 239C/V58025 (Lang 1937,
pl. IX, fig. 24). All those specimens do not show any diagnostic
feature (i.e., at least one sporangium).
Acknowledgments
We thank Professor D. Edwards, FRS, for her cordial wel-
come of P. Gonez in Cardiff and helpful explanations about
Cooksonia; Dr. C. Berry, Dr. L. Axe, Dr. S. Stewart, and Dr.
P. Hayes for their help and cordial welcome of P. Gonez; Dr.
M.-A. C. Rodrigues, Dr. E. Pereira, and Dr. S. Bergamaschi
from Rio de Janeiro State University for having given access
to the material and for their help during fieldwork; Dr. S.
MacLoughlin from the Natural History Museum of Sweden
for pictures of C. bohemica; Dr. O. Orlova from Lomonosov
Moscow State University for pictures of C. crassiparietilis;
Dr. P. Steemans for his help with Cooksonia spores; and Dr.
C. Crônier for her critical reading of the statistical analysis.
P. Gonez holds a FRIA grant; P. Gerrienne is an FRS-FNRS
research associate.
Edwards D 1970 Fertile Rhyniophytina from the Lower Devonian of
Britain. Palaeontology 13:451–461.
——— 1979 A Late Silurian flora from the Lower Old Red
Sandstones of south-west Dyfed. Palaeontology 22:23–52.
——— 1990 Constraints on Silurian and Early Devonian phyto-
geographic analysis based on megafossils. Pages 233–242
in WS MacKerrow, CR Scotese, eds. Palaeozoic palaeogeog-
raphy and biogeography. Geol Soc Mem 12. Geological Society,
London.
 GONEZ & GERRIENNE—A NEW DEFINITION OF COOKSONIA
——— 1993 Cells and tissues in the vegetative sporophytes of early
land plants. New Phytol 125:225–247.
——— 2003 Xylem in early tracheophytes. Plant Cell Environ 26:
57–72.
Edwards D, HP Banks, SJ Ciurca Jr, RS Laub 2004 New Silurian
Cooksonias from dolostones of north-eastern North America. Bot J
Linn Soc 146:399–413.
Edwards D, KL Davis, L Axe 1992 A vascular conducting strand in
the early land plant Cooksonia. Nature 357:683–685.
Edwards D, U Fanning, JB Richardson 1986 Stomata and sterome in
early land plants. Nature 323:438–440.
——— 1994 Lower Devonian coalified sporangia from Shropshire:
Salopella Edwards & Richardson and Tortilicaulis Edwards. Bot J
Linn Soc 116:89–110.
——— 1995 Exceptional preservation in Lower Devonian coalified
fossils from the Welsh Borderland: a new genus based on reniform
sporangia lacing thickened borders. Bot J Linn Soc 117:223–254.
Edwards D, J Feehan, DG Smith 1983 A late Wenlock flora from Co.
Tipperary, Ireland. Bot J Linn Soc 86:19–36.
Edwards D, EM Morel, F Parades, DG Ganuza, A Zúñiga 2001 Plant
assemblages from the Silurian of southern Bolivia and their
palaeogeographic significance. Bot J Linn Soc 135:229–250.
Edwards D, JB Richardson 1974 Lower Devonian (Dittonian) plants
from the Welsh Borderland. Palaeontology 17:311–324.
——— 1996 Review of in situ spores in early land plants. Pages
391–407 in J Jansonius, DC MacGregor, eds. Palynology: principles
and applications. Vol 1. American Association of Stratigraphic
Palynologists Foundation, College Station, TX.
Edwards D, CW Rogerson 1979 New records of fertile Rhyniophy-
tina from the late Silurian of Wales. Geol Mag 116:93–98.
Fairon-Demaret M, J Hilton, CM Berry 1999 Surface preparation of
macrofossils (dégagement). Pages 33–35 in TP Jones, NP Rowe, eds.
Fossil plants and spores: modern techniques. Geological Society,
London.
Fanning U 1987 Late Silurian–Early Devonian plant assemblages from
the Welsh Borderland. PhD diss. Cardiff University, Cardiff, Wales.
Fanning U, D Edwards, JB Richardson 1990 Further evidence for
diversity in late Silurian land vegetation. J Geol Soc Lond 147:
725–728.
——— 1991a A review of in situ spores in Silurian land plants. In S
Blackmore, SH Barnes, eds. Pollen and spores, patterns of
diversification. Syst Assoc Spec Vol 44:25–47.
——— 1991b A new rhyniophytoid from the late Silurian of the
Welsh Borderland. Neues Jahrb Geol Palaeontol Abh 183:37–47.
——— 1992 A diverse assemblage of early land plants from the
Lower Devonian of the Welsh Borderland. Bot J Linn Soc 109:
161–188.
Fanning U, Richardson JB, D Edwards 1988 Cryptic evolution in an
early land plant. Evol Trends Plants 2:13–24.
Galtier J 1986 Taxonomic problems due to preservation: comparing
compression and permineralized taxa. Pages 1–16 in RA Spicer, BA
Thomas, eds. Systematic and taxonomic approaches in palaeo-
botany. Syst Assoc Spec Vol 31. Clarendon, Oxford.
View publication stats
215
Gensel P 1976 Renalia hueberi, a new plant from the Lower Devonian
of Gaspé. Rev Palaeobot Palynol 22:19–37.
Gerrienne P, S Bergamaschi, E Pereira, M-AC Rodrigues, P Stee-
mans 2001 An Early Devonian flora, including Cooksonia, from
the Paranà Basin (Brazil). Rev Palaeobot Palynol 116:19–38.
Gerrienne P, DL Dilcher, S Bergamaschi, I Milagres, E Pereira, M-AC
Rodrigues 2006 An exceptional specimen of the early land plant
Cooksonia paranensis, and a hypothesis on the life cycle of the
earliest Eutracheophytes. Rev Palaeobot Palynol 142:123–130.
Habgood KS, D Edwards, L Axe 2002 New perspectives on Cook-
sonia from the Lower Devonian of the Welsh Borderland. Bot J Linn
Soc 139:339–359.
Heard A 1939 Further notes on Lower Devonian plants from South
Wales. J Geol Soc Lond 95:223–229.
Janvier P, A Blieck, P Gerrienne, TD Thanh 1987 Faune et flore de la
Formation de Sika (Devonien inferieur) dans la presqu’ile de Do Son
(Vietnam). Bull Mus Hist Nat Paris 9:291–301.
Kenrick P, PR Crane 1997a The origin and early diversification of
land plants: a cladistic study. Smithsonian Institution, Washington
DC.
——— 1997b The origin and early diversification of plants on land.
Nature 389:33–39.
Lang WH 1937 On the plant-remains from the Downtonian of
England and Wales. Philos Trans R Soc B 227:245–291.
MacGregor DC 1973 Lower and Middle Devonian spores of eastern
Gaspé, Canada. II. Biostratigraphy. Palaeontogr Abt B 163:111–142.
McNeill J, FR Barrie, HM Burdet, V Demoulin, DL Hawksworth,
K Marhold, DH Nicolson, et al., eds. 2006 International code
of botanical nomenclature (Vienna Code). Adopted by the 17th
International Botanical Congress, Vienna, July 2005. Gantner,
Ruggell, Liechtenstein.
Schultka S 2003 Cooksonia: zur Morphologie einer fruhen Land-
pflanze aus dem Unterdevon der Eifel. Cour Forschungsinst
Senckenb 241:7–17.
Schweitzer HJ 1980 Die Gattungen Renalia Gensel und Psilophyton
Dawson im Unterdevon des Rheinlandes. Bonn Palaeobot Mitt 6:
1–34.
Shute C, D Edwards 1989 A new rhyniopsid with novel sporangium
organization from the Lower Devonian of South Wales. Bot J Linn
Soc 100:111-137.
Strother PK 2008 Spores and plant fragment from Cambrian rocks.
Page 270 in Proceedings of the 8th International Organisation of
Palaeobotany conference and 12th International Palynological
Congress, Bonn. University of Bonn.
Strother PK, S Al-Hajiri, A Traverse 1996 New evidence for land
plants from the lower Middle Ordovician of Saudi Arabia. Geology
24:55–58.
Swofford, D L 1998 PAUP*: phylogenetic analysis using parsimony
and other methods. Version 4.0 beta. Sinauer, Sunderland, MA.
Wellman CH, PL Osterloff, U Mohiuddin 2003 Fragments of the
earliest land plants. Nature 425:282–285.
Yurina AL 1969 The Devonian flora of central Khazachstan. Mater
Geol Sentral Kazakh 8:14–19.