| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Paleobotany |
2Department of Biological Sciences, LSCB 124, University of South Alabama, Mobile, Alabama 36688 USA; 3Division of Paleobotany, Natural History Museum and Biodiversity Research Center, Lawrence, Kansas 66045 USA; 4Bayerische Staatssammlung für Paläontologie und Geologie, Funktionseinheit Paläontologie, Richard-Wagner-Straße 10, 80333 Munich, Germany; 5Department of Ecology and Evolutionary Biology and Natural History Museum and Biodiversity Center, University of Kansas, Lawrence, Kansas 66045 USA; 6Department of Paleobiology, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 205609 USA
Received for publication February 6, 2003. Accepted for publication May 20, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Carboniferous • cycads • Permian • Phasmatocyas • pteridosperms • Spermopteris • Taeniopteris
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDISCUSSIONLITERATURE CITED |
|---|
for fertile taeniopterid leaves from the Upper Carboniferous (Virgilian) Ireland Sandstone Member of the Lawrence Shale in eastern Kansas. These authors noted that the leaves conform to the foliage taxon Taeniopteris coriacea Göppert in all major features except for the attached ovules, and they proposed the combination Spermopteris coriacea (Göppert) Cridland and Morris ostensibly to create a more "natural" taxon. The elongate, fertile leaves assigned to Spermopteris coriacea are characterized by two longitudinal rows of closely appressed ovules, one row on each side of the inflated, multistranded midrib. The ovules occur in individual depressions and were interpreted as being attached to the abaxial surface of the lamina. Sterile taeniopterid leaves with broader laminae and denser venation were also found at the same locality (Figs. 5, 13), but Cridland and Morris (1960)
indicated that no definitive evidence linked them with the ovule-bearing leaves. The only evidence of articulated organs provided was one branch specimen with several attached leaves.
|
|
, 1973
, 1976
). Mamay's reconstruction of Phasmatocyas kansana based on material from the Lower Permian Wellington Formation of Kansas (Fig. 4 in Mamay, 1976
) showed a basal petiolar region bearing two lateral rows of ovules and a sterile distal lamina conforming to Taeniopteris coriacea. Although the fertile axes and taeniopterid laminae were not found in organic connection, they were hypothesized to be parts of the same leaf based on similar secretion cavities in both structures. This fertile leaf, consisting of a basal ovule-bearing portion and an expanded distal lamina, was considered an ancestral morphology to the megasporophyll of the extant cycad genus Cycas.
In a more detailed exposition of the taeniopterid ancestor theory of cycad origins, Mamay (1976)
proposed that the megasporophylls of cycads were ultimately derived from a Spermopteris-like ancestor, with intermediate morphologies represented by the Early Permian genera Archaeocycas and Phasmatocycas. This evolutionary transition series suggests the restriction of ovules on the Spermopteris-like ancestor to the proximal portions of the leaf and the phyletic shift of the point of ovule attachment from the abaxial lamina surface to the lateral edge of the midrib. Loss of the lamina from the proximal, ovule-bearing portion resulted in a structure like that of Phasmatocycas. Pinnulation of the distal lamina culminated in the megasporophyll structure of the putatively primitive cycad genus Cycas.
Although taeniopterid foliage is common in the Wellington Formation, the fertile axis on which the Mamay (1976)
Phasmatocycas reconstruction was based was determined from one short fragment. The putative secretion bodies on the fertile axis and sterile laminae provided convincing evidence that both organs were produced by the same plant; however, they were not proven to be parts of the same organ. Subsequently, Gillespie and Pfefferkorn (1986)
discovered more completely preserved specimens, which proved that the ovulate axes and taeniopterid laminae were indeed parts of the same organ based on unequivocal attachment evidence. However, in these specimens the ovules are not restricted to the elaminate petiolar region, but occur beneath the lamina where they are attached to the midrib. Therefore, the Gillespie and Pfefferkorn (1986
, fig. 3B) reconstruction of Phasmatocycas more closely resembled the material assigned to Spermopteris than previously realized. However, the major morphological difference remaining was the point of ovule attachment, i.e., the abaxial lamina surface in Spermopteris and the lateral edge of the midrib in Phasmatocycas.
Considering these major additions to our understanding of Phasmatocycas, it is remarkable that no critical reappraisal of the original Spermopteris specimens was undertaken to date. This lack is especially intriguing because the original description predates the suggestion that this fossil represents the earliest known progenitor of the cycads. In this paper, we reexamine the original specimens assigned to Spermopteris by Cridland and Morris (1960)
. We conclude that these fossils are more complete, articulated, and informative than originally indicated and that critical morphological features, including the points of ovule attachment, were misinterpreted. This paper also includes a reappraisal of several features of Phasmatocycas kansana based on reexamination of the Gillespie and Pfefferkorn (1986)
specimens and newly collected material from the Wellington Formation in Kansas. Our revised concepts suggest that the fertile leaves now known as Spermopteris coriacea and Phasmatocycas kansana are exceedingly similar, with both forms having ovules attached to the leaf midrib. The genus Spermopteris Cridland and Morris based on fertile material from North America is typified by the sterile foliage Taeniopteris coriacea Göppert. However, it cannot be proven that both S. coriacea and T. coriacea are conspecific. Regarding the large similarities of the fertile leaves previously described as S. coriacea with Phasmatocycas, we propose to describe this material as a new species of Phasmatocycas. Finally, some comments are offered on the implications of these findings to our understanding of the potential phylogenetic significance of Phasmatocycas.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDISCUSSIONLITERATURE CITED |
|---|
, which is housed in the Paleobotany Division of the Natural History Museum and Biodiversity Research Center of the University of Kansas in Lawrence, Kansas, USA. The specimens were collected in 1946 during building construction on the campus of Baker University in Baldwin City, Douglas County, Kansas. Although only portions of fertile and sterile leaves and one axis with attached leaf bases were figured in the original description, the collection actually consists of 33 slabs displaying many leaves and five specimens of axes with nearly complete leaves still in organic connection. The Baker University locality probably represented an unusually productive lens of siltstone in the otherwise mostly unfossiliferous Ireland Sandstone Member. The fossils are preserved as unstained to red/brown stained impressions in a buff, siltstone matrix (Figs. 1–13). Previous investigators performed substantial preparation on most of the specimens. This preparation was required because the fossils are not well exposed on the matrix surface due to the lack of distinct bedding planes. Some additional preparation was undertaken in the present study. Although Cridland and Morris (1960)
reported no cuticles present, poorly preserved specimens were discovered beneath the matrix covering two ovules on specimen #8129.
In addition to reexamining the Gillespie and Pfefferkorn (1986)
material of Phasmatocycas kansana housed at the National Museum of Natural History in Washington, D.C., supplementary material of this plant was collected from the Lower Permian (Leonardian) Wellington Formation exposures in roadside ditches along the county road (Deer Road) 4.0 km south of the town of Carlton, Dickinson County, Kansas. Although undoubtedly close to the Gillespie and Pfefferkorn (1986)
locality, the new specimens are apparently from a different horizon based on their preservation in a limestone rather than dolomite matrix and lack of algal laminations. Furthermore, the new collection lacks several other plant taxa reported by Gillespie and Pfefferkorn (1986)
, including Glenopteris and Sandrewia. Photography of specimens from both localities was challenging because of the uneven rock surfaces and lack of contrast.
Nomenclatural issues
The monotypic genus Spermopteris was erected to accommodate ovule-bearing leaves considered by Cridland and Morris (1960)
to represent the reproductive phase of the "form genus" Taeniopteris coriacea. However, rather than naming the fertile material from Kansas as a distinct morphotaxon with its own type specimen, the name Spermopteris coriacea was proposed as a new combination, with the European type material of Taeniopteris coriacea serving as the type for the fertile leaves from Kansas and presumably of all sterile material as well. This nomenclatural history is problematical because these fossils undoubtedly represent distinct morphospecies of uncertain relationships, and the name Taeniopteris coriacea is still in use as a morphospecies of primarily sterile leaves (Remy and Remy, 1975
). Furthermore, both Spermopteris coriacea and Phasmatocycas kansana are associated with sterile leaves resembling Taeniopteris coriacea. Thus, at least two distinct species of plant based on fertile material produced leaves otherwise conforming to Taeniopteris coriacea.
In the absence of fertile specimens from Europe, the sterile type material of Taeniopteris coriacea cannot be clearly linked with either North American ovule-bearing leaf type. We therefore propose that the fertile leaves described as Spermopteris coriacea, the fertile leaves of Phasmatocycas kansana, and Taeniopteris coriacea clearly represent distinct morphotaxa, if not distinct biological taxa, and ought to be separately named and typified. In particular, the name Spermopteris coriacea should be rejected. Because the genus Phasmatocycas was, in our opinion, validly established based on fertile specimens by Mamay (1973)
, we submit that Phasmatocycas is the best repository for all such Paleozoic taeniopterid leaves with abaxially situated ovules, including the material described by Cridland and Morris (1960)
as Spermopteris. Therefore, the fertile material formally known as Spermopteris coriacea will here be named as a new species of Phasmatocycas (i.e., P. bridwellii). The alternative of proposing the combination Phasmatocycas coriacea is not acceptable as the 2000 International Code of Botanical Nomenclature explicitly prohibits accepting a name while rejecting its type specimen (Art. 48.1).
|
SYSTEMATICS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDISCUSSIONLITERATURE CITED |
|---|
, emend. Gillespie et Pfefferkorn, 1986
.
Type species
Phasmatocycas kansana Mamay, 1973
.
Species
Phasmatocycas bridwellii nov. sp.
Synonymy
Spermopteris coriacea Cridland et Morris (1960
, Figs. 2–7, 9, 11–15).
|
Holotype
Slab #8129 (Figs. 2, 4, 6), inclusive of the fertile leaves, is here designated as the type specimen.
Paratypes
#8127 (Fig. 1), #8134 (Fig. 7), #8140 (Fig. 8), #8145 (Fig. 9).
Additional specimens
The entire collection consists of specimens #8127–8148.
Formation
Lawrence Shale (Ireland Sandstone Member).
Age
Upper Pennsylvanian (Virgilian).
Type locality
Baker University campus, Baldwin City, Douglas County, Kansas.
Repository
All specimens are housed at the Division of Paleobotany, Natural History and Biodiversity Research Center, University of Kansas, Lawrence, Kansas.
Etymology
The specific epithet bridwellii is proposed in recognition of Arthur Bridwell, who originally collected the specimens from Baldwin City, Kansas, in 1946.
Description and interpretation
Four slabs (#8128, #8134—slab with two axes, #8129, #8142) contain portions of axes 8–10 mm wide with attached leaves. Cridland and Morris (1960)
reported only one axis with attached leaves (#8142), even though the articulated specimens we describe here were parts of the original collection. Although the phyllotaxy is difficult to discern, a spiral leaf arrangement is most probable. The leaves are closely spaced on the axis as seen most clearly on specimen #8128 (Fig. 1), which shows a 6.3-cm-long axis fragment with at least 22 attached leaves.
No complete fertile leaves are known. The largest leaf fragments are up to 20 cm long, but were undoubtedly considerably longer (Figs. 1–2). The leaves are narrow, linear, and gradually taper to a sharply acute apex (Fig. 7). The petiole is short (
3–4 mm). In most specimens, the multistranded midrib appears flat and sometimes partly overlaps the abaxial side of the ovules as a narrow flange (Figs. 8–9). However, other specimens show a distinctly inflated midrib with no evidence of a flange (Fig. 6), suggesting that the flattened appearance of most of the midribs is an artifact caused by compression.
The leaf lamina is rather narrow, each side being 0.75x the width of the midrib (4 mm) in the leaf midregion. The lamina is laterally attached about halfway between the abaxial and adaxial sides of the midrib. This arrangement brings the ovules and lamina into contact, with the thickened midportion of the ovules producing persistent molds in the abaxial lamina surface (Figs. 3, 7). The venation density is relatively sparse (typically 10–15 veins/cm). Individual veins emerge from the midrib at acute angles, traverse the lamina unbranched or dichotomize only once near the midrib, and terminate at the lamina margin. Venation in the basal and apical regions of the megasporophyll are angled (up to 30°), but perpendicular to the midrib in the leaf midregion. Small pits, possibly representing former resin bodies based on their similarity to the known resin bodies of Phasmatocycas kansana, occur extensively on the lamina of many specimens. On several specimens, the lamina is variously degraded as evidenced by a faint, wrinkled appearance (Fig. 4). These degraded laminae are often completely obscured by the ovules, and in several specimens a lamina is not at all recognizable (Figs. 6, 8–9). Such specimens only display a midrib with attached ovules and are essentially indistinguishable from elaminate Phasmatocycas kansana material (Fig. 15). These specimens unequivocally show that the ovules are attached to the midrib rather than the lamina. Although it is possible to envisage that some fertile leaves originally were without a leaf lamina covering the ovules, the partially degraded specimens show that a lamina was originally present over the entire ovule-bearing portion, but perished during maturation of the ovules or were lost because of post-mortem decay.
Attached ovules, or evidence of the former position of ovules in the form of molds as described below, are not always present on the narrowly laminate leaves with sparse venation (Fig. 1). Nevertheless, such specimens are interpreted as ovulate leaves or secondarily derived from them, because when ovules or ovule molds are present, they occur only on leaves of this kind. Ovules never occur on the associated broader, more densely veined leaves described below (Figs. 2, 5, 10). Narrowly laminate specimens lacking ovules or molds probably represent the basal, ovule-free portions of fertile leaves.
Ovules occur along approximately the distal half of the leaves and are attached to both lateral sides of the midrib under the leaf lamina in elongate rows. The thickened midportion of each ovule is situated in a persistent, ovoid depression (mold) in the abaxial lamina surface (Figs. 3, 7). Cridland and Morris (1960)
interpreted the ovules as being attached to the leaf lamina within the molds. If this were the case, attachment scars should be visible within the molds, but none could be found. Although Cridland and Morris (1960
, Fig. 15) presented line drawings of specimen #8129 purporting to show four ovules directly attached to the leaf lamina, it is more likely that these ovules are detached from the leaf and were separated from the midrib during compression. This interpretation is supported by the observation that such ovules tend not to be in strict rows and are more or less randomly oriented (Fig. 3). Whenever ovules are attached, they are connected to the midrib and occur in rather straight rows (Figs. 4, 6, 8, 9).
Ovules are about 5 mm long x 2.5 mm wide, slightly flattened, and possess an apical cleft (Fig. 12). They are broadly attached by their bases to the lateral side of the midrib (Figs. 4, 6, 8, 9). Although Cridland and Morris (1960)
reported that cuticles were absent, dégagement of several ovules revealed cuticles that are somewhat smaller than the surrounding ovule impressions. Unfortunately, these cuticles lack cellular details. A thickened band, possibly representing the same type of thickened megaspore membrane as reported in Phasmatocycas kansana (Mamay, 1976
), is the only internal structure visible (Fig. 11). The cuticle does not show an apical cleft, but simply tapers toward the apex. Cridland and Morris (1960)
left the question open as to whether the apical cleft on the ovular impressions is a natural feature or a compressional artifact. In the description of Phasmatocycas kansana, which has identical ovules, Mamay (1976)
originally suggested that the cleft was a compressional artifact, because the cuticles do not display this structure, as we report here for Phasmatocycas bridwellii. It now seems unlikely, however, after considering the presence of such a cleft in taxa from different localities, that compressional forces strong enough to consistently split the apex of the ovules would not also cause corresponding splits in the cuticle. Therefore, we hypothesize that the apical cleft of both species was an original feature representing a noncutinized or weakly cutinized outer integumentary layer or extra-integumentary structure (e.g., a ridge, rudimentary wing, or aril-like structure) that left behind an impression in the matrix, but no preserved cuticle. This interpretation has the added benefit of accounting for the fact that the ovule cuticle of both forms is smaller than the ovule impression (i.e., the cuticle does not cover the part of the impression at the apex in which the cleft occurs). Such paired apical projections, which form a central cleft similar to that of Phasmatocycas, are a common feature of Paleozoic pteridosperm ovules such as those of Callistophyton (Rothwell, 1981
), Eremopteris (Delevoryas and Taylor, 1969
), and Nystroemia (Hilton and Li, 2003
).
Although Cridland and Morris (1960)
illustrated portions of the specimen here designated as the holotype (#8129) (Fig. 2) and based much of their description on it, the completeness and potential significance of this specimen for reconstructing the species was not made apparent. This slab contains 14 leaves (five sterile and nine ovulate leaves) on one surface. Five leaves (three sterile and two ovulate) are visible on the opposite surface, and dégaging has revealed that an indeterminate number of leaves remain hidden within the matrix. Unfortunately, the slab is broken off above the leaf bases; however, all of the leaves are radiating outward from a central point, which likely indicates that they were originally attached to a common axis. Based on this specimen, it is surprising that Cridland and Morris (1960)
did not conclude that the same plant probably produced the fertile and sterile taeniopterid leaves. The sterile leaves on the holotype are similar to the isolated, "larger Taeniopteris" leaves reported by those authors. They differ from the ovulate leaves in possessing a broader lamina (each side up to 5.5x the midrib width), less inflated midribs, and denser venation (up to 40 veins/cm) (Fig. 5). Several of the isolated sterile leaves are remarkably large (up to 18 cm long and 4 cm wide), but still incomplete (Fig. 13). However, one small cluster of complete immature sterile leaves occurs in the collection. They are more ovate with rounded apices relative to the ovulate leaves (Fig. 10). The sterile leaves often possess pits that probably represent secretion bodies (Fig. 13) based on their similarity to resin-containing structures on Phasmatocycas kansana.
It cannot be conclusively determined if the fertile and sterile leaves were produced in distinct clusters (i.e., "flushes"), but clusters are likely based on the specimens having leaves in organic connection to axes (Fig. 1), most of which exclusively have the narrow leaf morphology with relatively sparse venation indicative of fertile leaves. The immature sterile leaves also occur in a cluster with no fertile leaves (Fig. 10). The holotype has a mixture of both leaf types that were probably attached to a common axis, but due to the lack of the actual branch showing the attachment points, the relative arrangement of sterile leaves and fertile leaves cannot be discerned (Fig. 2). A suggested reconstruction of a Phasmatocyas bridwellii branch with attached fertile leaves is presented in Fig. 18.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDISCUSSIONLITERATURE CITED |
|---|
Phasmatocycas kansana
Based on our observations and recent reinvestigations by Gillespie and Pfefferkorn (1986)
, it is now clear that Phasmatocycas kansana and Phasmatocycas bridwellii are similar in general morphology, but differ in some features of lamina attachment, venation, and fertile/sterile leaf dimorphism (Table 1). Furthermore, clarification of the morphology of P. kansana has been made possible by comparisons with the relatively well-articulated material of P. bridwellii, reexamination of the P. kansana material in Gillespie and Pfefferkorn (1986)
, and newly discovered specimens from the Wellington Formation (Figs. 14–17).
|
collection refigured here (Figs. 16–17) clearly shows that the ovules were born beneath the leaf lamina rather than the elaminate petiolar region. However, in contrast to P. bridwellii, the lamina of P. kansana is attached very high on the midrib. Therefore, the lamina and ovules are not in close contact, and there are no corresponding ovule molds in the lamina surface. The lamina of the P. kansana fertile leaf (Figs. 16–17) is also broad (lamina on each side of midrib up to 1.5x midrib width) and has dense venation (about 30 veins/cm) relative to P. bridwellii. In these characters, P. kansana ovulate leaves are similar to the sterile leaves associated with P. bridwellii (Figs. 2, 5).
Other formerly problematic characters in Phasmatocycas kansana are clarified by comparisons with P. bridwellii. Of particular significance is the nature of the supposed flange emerging from the midrib and partly covering the ovules on their adaxial sides (Figs. 16–17). Gillespie and Pfefferkorn (1986)
proposed that this flange might be an artifact caused by the compression of an originally inflated midrib. However, they later argue against this interpretation by pointing out that other taxa in the same deposit (e.g., Callipteris) do not show evidence of midrib compression. The flange is included in their reconstruction of P. kansana (Fig. 1 in Gillespie and Pfefferkorn, 1986
). They also point out that dolostones and limestones can lithify quickly and therefore do not usually produce significant compression of fossils. A potential problem with this interpretation is the assumption that the midrib in the other taxa had similar characteristics to that of Phasmatocyas. We suggest instead that the broad, inflated midrib of Phasmatocycas may have been composed of significant amounts of parenchymatous tissue. The internal tissues of such a midrib may degrade relatively rapidly post-mortem and not resist even mild compressional forces. This interpretation is supported by the generally similar P. bridwellii specimens. Although the midribs of most specimens also have a flattened, flange-like structure (Figs. 8–9), a few display a distinctly inflated, terete midrib (Fig. 6). This clearly shows that the flattened specimens resulted from the compression of an originally inflated midrib.
Although Gillespie and Pfefferkorn (1986)
showed that the ovules of Phasmatocycas kansana were typically covered by a leaf lamina, they observed that some specimens consisted of an axis with attached ovules only (Fig. 14), even when the counterpart was carefully examined, suggesting that at least some parts of the leaves may have been truly elaminate. The better-articulated material of Phasmatocycas bridwellii indicates that the lamina often degrades considerably and is easily lost or obscured by the ovules (Fig. 4). This was probably a general feature of Phasmatocyas, and we suggest that all ovules of both species were originally overlain by a lamina, which sometimes perished during ovule maturation.
A significant feature of Phasmatocycas kansana, which Mamay (1973
, 1976
) used as the primary evidence that the ovulate axes and taeniopterid laminae belonged to the same plant, is the presence of spherules associated with glandlike structures. Such structures sometimes occur on P. bridwellii and associated sterile leaves (Fig. 13). Identical structures were considered as possible sporangia by Sellards (1901)
and as iron oxide globules by Cridland and Morris (1960)
. Mamay (1976)
hypothesized that they may have been involved in the production of food rewards for pollinating insects. However, new specimens of P. kansana from the Wellington Formation sometimes have small spheres of semi-transparent, amber-colored material associated with the pits, supporting the interpretation of these structures as resin bodies (Fig. 15). If this interpretation is correct, then such bodies perhaps functioned to repel foliage feeders rather than attracting or rewarding pollinators.
A significant difference between Phasmatocycas kansana and P. bridwellii concerns the nature of the fertile/sterile leaf dimorphism. As noted by Gillespie and Pfefferkorn (1986)
, the sterile leaves associated with P. kansana (Fig. 15) are rather similar to the fertile leaves of the same plant (Figs. 16–17). The primary difference is that the lamina of the associated sterile leaves is attached lower on the midrib (Gillespie and Pfefferkorn, 1986
). The sterile leaves associated with P. bridwellii have broader laminae, denser venation, and narrower midribs than the ovulate leaves (Figs. 5, 10).
Mamay (1976)
described other axes from the Lower Permian of Texas with laterally attached ovules as ?Phasmatocycas spectabilis. This material differs from elaminate specimens of P. kansana and P. bridwellii only in being much larger. Because no associated taeniopterid lamina was found, it cannot be determined if the overall morphology of the leaf was similar. Additional material of this taxon is required to more fully assess ?P. spectabilis.
Other relevant taxa
A complete review of all taeniopterid taxa implicated as possible early cycads or cycad ancestors is beyond the scope of this contribution. However, a few forms have received particular attention as potential members of the "pteridosperm"–cycad transition, and these are briefly considered.
Mamay (1973
, 1976
) described Archaeocycas from the Permian of Texas as morphologically intermediate between Spermopteris sensu Cridland and Morris (1960)
and Phasmatocycas. This leaf consists of an expanded distal lamina and a narrower basal portion. Molds on the basal portion were interpreted as probable sites of ovule attachment on the lamina surface. Unfortunately, the material of Archaeocycas is rather poorly preserved, and better material is needed to verify the precise position of the ovules.
Kerp (1983)
suggested that Sobernheimia jonkeri from the Lower Permian of Germany might represent an intermediate form between Spermopteris sensu Cridland and Morris (1960)
and Phasmatocycas. This fossil consists of a robust axis with lateral structures interpreted as leaflets. Between the leaflets are thickened, ovoid bodies interpreted as ovules. This morphological interpretation would indicate that pinnate leaves occurred early in the "pteridosperm"–cycad transitional series. However, S. jonkeri is known only from one poorly preserved specimen, and the interpretations of this fossil are therefore tentative.
Leary (1990)
described what he interpreted as fertile specimens of Lesleya from the Lower Pennsylvanian (Namurian B or C) of Illinois. Structures termed ovulate receptacles occur laterally along the petiole. This morphological arrangement is similar to the original Mamay (1973
, 1976
) Phasmatocycas kansana reconstruction, thus pushing the supposed origin of cycads back about 35 million years. However, based on the more recent reconstructions of Phasmatocycas (Fig. 1 in Gillespie and Pfefferkorn, 1986
) and the present study, the proposed morphological correspondences between Lesleya and Phasmatocycas are greatly reduced. Furthermore, the supposed ovulate petiolar region of Lesleya has only raised areas interpreted as "ovulate receptacles." That these areas actually represent sites of ovule attachment has not, in our opinion, been definitively demonstrated.
The only Paleozoic taeniopterid leaf with possible ovules attached to the lamina of which we are aware is Eophyllogonium cathayense from the Permian of China (Mei et al., 1992
). The numerous small structures interpreted as ovules are situated along the crenulate leaf margin. This species has several unusual morphological features relative to other Paleozoic taeniopterids, including reticulate venation and an inverted, conelike micropylar area on the supposed ovules. It is therefore unlikely that Eophyllogonium is closely related to Phasmatocycas or the cycads. A similar morphology has been documented in the putative gigantopterid remains from China described by Li and Yao (1983)
.
Cycad origins
Although the morphology of Phasmatocycas is becoming better known, its position as an early cycad is now less certain. The reconstructions of Phasmatocycas in Gillespie and Pfefferkorn (1986)
and the present study are distinctly less Cycas-like than originally believed (Mamay, 1973
, 1976
). Although the Cycas-like gross morphology of the original Phasmatocycas concept is generally considered the primary evidence for a cycadalean affinity, this reconstruction was proposed only after cuticular analysis of the ovules, which Mamay deemed to be the primary evidence for a cycadalean relationship. In particular, the cuticles of Phasmatocycas ovules may be similar to those of the putative cycad Beania from the Yorkshire Jurassic (Harris, 1964
). The arrangement of the ovules in two rows along the leaf midrib was also regarded as Cycas-like. However, a reconsideration of these features suggests that they may be problematic as clear indicators of cycadalean affinity, especially when seen in the light of recent discoveries. For example, Harris (1964)
previously pointed out that Beania ovules have no features unambiguously distinguishing them from those of conifers and ginkgoes. Only specimens attached to Beania megastrobili can be attributed with any confidence to the Cycadales. Furthermore, some pinnate pteridosperm leaves (e.g., Dicksonites pluckenetii) may have had ovules situated beneath the lamina along both sides of the midvein (Langiaux, 1986
; Galtier and Bethoux, 2002
) as in Phasmatocycas. Therefore, it may be that such an ovule arrangement was widespread among some Paleozoic pteridosperm groups and may not be an unambiguous indicator of cycadalean affinity unique to the taeniopterids.
In addition, the phylogenetic position of Phasmatocycas as the ultimate progenitor of cycads has been challenged by the discovery of indisputable Paleozoic cycad remains. For example, fossils described by Zhu and Du (1981)
and Gao and Thomas (1989)
show that cycad megasporophylls very similar to those of extant Cycas species were already present in the Early Permian of China. However, there are numerous taeniopterid and mixed taeniopterid/pinnate leaf taxa from China that have, in fact, been proposed as early cycads based primarily on leaf morphology and/or cuticular features, such as Procycas densinervioides, Cladotaeniopteris shaanxiensis, and Lepingia emarginata (Zhang and Mo, 1981
; Liu and Yao, 2002
). Although these forms are provocative as possible cycads, fertile specimens will be required before their systematic position can be demonstrated.
There is little question that the newly emerging concept of Phasmatocycas is decidedly less cycad-like than previously believed. Our reinvestigation also removes the primary evidence for the supposed phyletic shift of ovule position from laminar in the "Spermopteris"-like ancestor to the leaf midrib in Phasmatocycas and Cycas. The P. bridwellii morphology proves that the ovules were already attached to the rachis in some of the earliest taeniopterid fossils. Although the morphology of Phasmatocycas is becoming progressively better known, its true phylogenetic position is more problematic than ever because many critical aspects of its morphology and anatomy remain unknown. For example, nothing is known regarding the male reproductive structures. Information on the internal anatomy of the stems is also needed because cycads typically display distinctive anatomical features (e.g., girdling leaf traces). Perhaps continued prospecting in the Late Paleozoic formations of eastern Kansas, a region known to produce important, anatomically preserved plant fossils, will eventually produce such material (Rothwell and Mapes, 1988
).
|
FOOTNOTES |
|---|
7 Telephone: 1-251-460-7528; e-mail: baxsmith{at}jaguar1.usouthal.edu
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDISCUSSIONLITERATURE CITED |
|---|
Delevoryas T. T. N. Taylor 1969 A probable pteridosperm with eremopterid foliage from the Allegheny Group of northern Pennsylvania. Postilla 133: 1-14
Galtier J. O. Béthoux 2002 Morphology and growth habit of Dicksonites plunkenetii from the Upper Carboniferous of Graissessac (France). Geobios 35: 525-535
Gao Z. B. A. Thomas 1989 Occurrence of earliest cycads in the Permian of China and its bearing on their evolution. Chinese Science Bulletin 34: 766-769[ISI]
Gillespie W. H. H. W. Pfefferkorn 1986 Taeniopterid lamina on Phasmatocycas megasporophylls (Cycadales) from the Lower Permian of Kansas, U.S.A. Review of Palaeobotany and Palynology 49: 99-116
Harris T. M. 1964 The Yorkshire Jurassic flora II. Caytoniales, Cycadales & pteridosperms. British Museum (Natural History), London, UK
Hilton J. C.-S. Li 2003 Reinvestigation of Nystroemia pectiformis Halle, an enigmatic seed plant from the Late Permian of China. Palaeontology 46: 29-51
Kerp J. H. F. 1983 Aspects of Permian palaeobotany and palynology. I. Sobernheimia jonkeri nov. gen., nov. sp., a new fossil plant of cycadalean affinity from the Waderner Gruppe of Sobernheim. Review of Palaeobotany and Palynology 38: 173-183[CrossRef]
Langiaux J. 1986 Découverte d'une spécimen exceptionnel de fructification femelle de Dicksonites sterzeli Zeiller 1888 dans le Stéphanien du Massif Central française. Comptes Rendus de l'Académie des Sciences de Paris Série II 302: 755-760
Leary R. L. 1990 Possible Early Pennsylvanian ancestor of the Cycadales. Science 249: 1152-1154
Li X. Z. Yao 1983 Fructifications of gigantopterids from South China. Palaeontigraphica B 185: 11-26
Liu L.-J. Z.-Q. Yao 2002 Lepingia: a new genus of probable cycadalean affinity with taeniopterid lamina from the Permian of south China. International Journal of Plant Sciences 163: 175-183[CrossRef]
Mamay S. H. 1969 Cycads-fossil evidence of late Paleozoic origin. Science 164: 295-296
Mamay S. H. 1973 Archaeocycas and Phasmatocycas—new genera of Permian cycads. Journal of of Research, U.S. Geological Survey 1: 687-689
Mamay S. H. 1976 Paleozoic origin of the cycads. U.S. Geological Survey Professional Paper 934: 1-48
Mei M.-T. D. L. Dilcher Z.-H. Wan 1992 A new seed-bearing leaf from the Permian of China. Palaeobotanist 41: 98-109
Remy W. R. Remy 1975 Beiträge zur Kenntnis des Morpho-Genus Taeniopteris Brongniart. Argumenta Palaeobotanica 4: 31-37
Rothwell G. W. 1981 The Callistophytales (Pteridospermopsida): reproductively sophisticated Paleozoic gymnosperms. Review of Palaeobotany and Palynology 32: 103-121
Rothwell G. W. G. Mapes 1988 Vegetation of a Paleozoic conifer community. In G. Mapes and R. Mapes [eds.], Regional geology and paleontology of Upper Paleozoic Hamilton Quarry area in southeastern Kansas. Kansas Geological Survey Guidebook Series 6, 213–223. Kansas Geological Survey, Lawrence, Kansas, USA
Sellards E. H. 1901 Permian plants—Taeniopteris of the Permian of Kansas. Kansas University Quarterly 10: 1-12
Zhang S.-Z. Z.-G. Mo 1981 On the occurrence of cycadophytes with slender growth habit in the Permian of China. GSA Special Paper 187: 237-242.
Zhu J.-N. X.-M. Du 1981 A new cycad—Primocycas chinensis gen. et. sp. nov. from the Lower Permian in Shanxi, China and its significance (in Chinese). Acta Botanica Sinica 23: 401-404
This article has been cited by other articles:
|
|
 |
|
  E. J. Hermsen, T. N. Taylor, E. L. Taylor, and D. Wm. Stevenson Cataphylls of the Middle Triassic cycad Antarcticycas schopfii and new insights into cycad evolution Am. J. Botany, May 1, 2006; 93(5): 724 - 738. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Patterns of segregation and convergence in the evolution of fern and seed plant leaf morphologies Paleobiology, January 1, 2005; 31(1): 117 - 140.
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
