| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Paleobotany |
Department of Environmental and Plant Biology, Ohio University, Athens, Ohio 45701 USA
Received for publication November 22, 2002. Accepted for publication March 11, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDESCRIPTIONDISCUSSIONLITERATURE CITED |
|---|
Key Words: Chesterian • Lyginopteridaceae • Mississippian • Pteridospermales • Trivena arkansana • wound response
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDESCRIPTIONDISCUSSIONLITERATURE CITED |
|---|
verification of the existence of seed ferns (Pteridospermae) based on the whole plant reconstruction of Lyginopteris. That plant was reconstructed by recognizing that stems, petioles, foliage, and ovules, which commonly co-occur in many fossil horizons, all share distinctive capitate epidermal glands and were therefore produced by the same plant. Prior to this publication, several authors (e.g., Stur, 1883
; Williamson, 1887
; Potonié, 1899
) recognized that a group of plants sharing characteristics of ferns and seed plants must have existed, but it was Oliver and Scott who finally confirmed the presence of this distinct group.
Five seed fern families are currently recognized from Carboniferous strata (Taylor and Taylor, 1993
): the Calamopityaceae, Buteoxylonaceae, Lyginopteridaceae, Medullosaceae, and Callistophytaceae, and these plants play a key role in many hypotheses of the evolution of seed plants. For example, it has been suggested that early seed ferns (Calamopityaceae and Lyginopteridaceae) evolved from aneurophytalean progymnosperms (Beck, 1960
, 1966
, 1976
; Banks, 1968
) and that conifers may have evolved by heterochronous change in shoot system structure from among a group of seed ferns (Rothwell, 1982
). In addition, similarities between the Medullosaceae and the Calamopityaceae suggest a close relationship between these two families (Stein and Beck, 1978
, 1992
; Mapes and Rothwell, 1980
; Galtier and Beck, 1992
). These and similar hypotheses underscore the importance of carefully analyzing the morphology and interrelationships of early seed ferns and accurately documenting their stratigraphic position.
The purpose of this study is to describe a new permineralized pteridosperm morphotaxon recovered from the marine black shales of the Fayetteville Formation near Fayetteville, Arkansas, USA. Specimens are recovered along with an abundant and diverse assemblage of marine and terrestrial fossil remains that facilitate accurate biostratigraphic dating and global correlation. The Fayetteville Formation is dated as middle Chesterian (Upper Mississippian), equivalent to the uppermost Lower Carboniferous, Pendelian (E1) Stage of the lower Namurian A, based on cephalopods, conodonts, and foraminifera (Meeks et al., 1997
) and miospores (Owens et al., 1979
). Specimens of the new pteridosperm stem display a relatively wide ratio of primary stele diameter to stem diameter and produce adventitious roots, Lyginorachis-type rachis bases, a distinct periderm, and a Dictyoxylon-type outer cortex: this combination of characters suggest these specimens are assignable to the Lyginopteridaceae (Galtier, 1988
; Stewart and Rothwell, 1993
; Taylor and Taylor, 1993
). Scrambling growth architecture and biostratigraphic data also support this assignment (Gordon, 1912
, 1938
; Baxter, 1949
; Beck, 1966
; Taylor and Millay, 1981
; Pigg et al., 1986
, 1987
). The new stems are eustelic like most species of Lyginopteris, but their mode of leaf trace production and rachis base vascularization is more like the supposedly primitive Heterangium subgenus Heterangium. This unique combination of characters reveals that the new specimens represent a new taxon, for which we propose the name Trivena arkansana gen. et sp. nov. Arthropod galleries filled with coprolites in the phloem of one of the specimens reveal wound response tissues, which further demonstrate that phytophagous arthropod and plant defense interactions had evolved by the Upper Mississippian.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDESCRIPTIONDISCUSSIONLITERATURE CITED |
|---|
) and were analyzed by serial cellulose acetate peels (Joy et al., 1956
) along with wafer sections and blocks cut as part of a previous study (Tomescu et al., 2001
). Peels were made at natural breaks in the specimen or after slicing the specimens into 1–2 cm sections (e.g., Fig. 1) on a Buehler Isomet Low Speed Rock Saw (Buehler, Evanston, Illinois, USA). Photo documentation was accomplished with a Phase One Digital Studio Camera (Phase One A/S, Frederiksberg, Denmark). Images were processed and plates were constructed using Adobe Photoshop 5.5, stored as TIFF and PSD files, and printed on a Shinko CH446i dye-sublimation printer (Shinko Electric, Tokyo, Japan). Specimens are reposited in the Ohio University Paleobotanical Herbarium (OUPH 13717–13763, 14503–15003).
|
|
SYSTEMATICS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDESCRIPTIONDISCUSSIONLITERATURE CITED |
|---|
Order
PTERIDOSPERMALES Sporne, 1974
Family
LYGINOPTERIDACEAE Takhtajan, 1953
Trivena
Dunn, Rothwell, and Mapes, gen. n.
Generic diagnosis
Pteridospermous axis with adventitious roots and decurrent, helically arranged leaves. Eustele with five sympodia located in the corners of the pentagonal pith. Secondary xylem with abundant parenchyma surrounded by vascular cambium, secondary phloem, inner cortex, periderm, middle cortex, Dictyoxylon-type outer cortex, and epidermis. Rachis bases vascularized by traces diverging radially from a single cauline bundle. Diverging leaf traces dividing in the middle cortex into three traces, each accompanied by secondary vascular tissues. Middle and Dictyoxylon-type outer cortex continuous with, and histologically similar to, ground tissues of rachis bases. Adventitious roots originate in the phloem-periderm region of the stem and produce secondary tissues in the middle cortex.
Etymology
Trivena: three-veined, for the three xylem bundles that vascularize the rachis base.
Type species
Trivena arkansana
Trivena arkansana
Dunn, Rothwell, and Mapes, sp. n.
Specific diagnosis
Stems approximately 30 mm in maximum diameter with internodes approximately 3.3 cm long; rachis bases decurrent for approximately 10 cm. Pith parenchymatous with resin-filled cells, sclerotic clusters absent from pith, xylem maturation marginally mesarch to exarch. Secondary xylem manoxylic with numerous wide, tall rays; phloem with rare fibers. Rachises produced helically: each base is vascularized by a trace diverging radially from a single cauline bundle. Diverging leaf trace accompanied by secondary vascular tissues forms a squat C shape and subsequently divides into three discrete bundles in the middle cortex: primary xylem of petiolar bundles mesarch and peripheral. Multilayered cortex of parenchymatous inner zone, periderm, parenchymatous middle zone with discontinuous rows of sclerotic clusters, and Dictyoxylon-type outer cortex. Epidermis/hypodermis up to several cells thick; epidermal capitate glands absent. Adventitious roots randomly produced.
Holotype
The holotype of Trivena arkansana consists of 302 cellulose acetate peels mounted on microscope slides and 25 sections of permineralized stem of specimen M2361 (OUPH 14503, 14504, 14508–14513, 14515–14517, 14522–14525, and 14528–14867).
Paratype
The paratype of Trivena arkansana consists of 115 cellulose acetate peels mounted on microscope slides and 22 sections of permineralized stem of specimen M1664 (OUPH 13732, 14507, 14514, 14518–14521, 14527, 13717–13731, 13733–13763, and 14815–14921).
Additional specimens
Eighteen cellulose acetate peels mounted on microscope slides of specimen M3157 (OUPH 14505, 14506, and 14975–14990) and 13 cellulose acetate peels mounted on microscope slides of specimen M1661 (OUPH 14991–15003).
Etymology
arkansana: for the state of Arkansas, from which these specimens were recovered.
Collection locality
White River (N/W1/4, Sec. 29, T14N, R28W, Durham 71/2' Quadrangle), Washington County, Arkansas.
Age and stratigraphy
Chesterian (Upper Mississippian), lower Namurian A or Lower Serpukhovian equivalent. In the lower shale unit Fayetteville Shale, Fayetteville Formation, Chester Series.
|
DESCRIPTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDESCRIPTIONDISCUSSIONLITERATURE CITED |
|---|
The vascular system of Trivena arkansana includes a eustele with five sympodia, abundant secondary xylem with numerous wide, tall rays, a vascular cambium and phloem. Ground tissues include a parenchymatous pith with resin-filled cells, a parenchymatous inner cortex, a periderm, a parenchymatous middle cortex with sclerotic clusters, and a Dictyoxylon-type outer cortex. Leaf traces arise by radial division from a single cauline protoxylem bundle. The resulting leaf trace is accompanied by secondary xylem and distally divides into three discrete bundles.
Pith
In cross section, the pith is pentagonal when not crushed and distorted (Fig. 6) and 3–4 mm in maximum diameter (mean = 3.75 mm). The pith is composed of loosely packed parenchyma and numerous cells with dark resin-like contents that may have been secretory in function (Fig. 7). Parenchyma cells are isodiametric in cross section and measure 45–128 µm (mean = 76 µm) in maximum length.
|
Secondary xylem
An extensively developed cylinder of loose wood surrounds the primary xylem and pith (Figs. 4 and 5). This wood ranges from 4 to 5.5 mm thick (mean = 4.75 mm) at the sides of the pentagonal pith and is only slightly thinner at the corners. Secondary xylem is composed of radial files of tracheids and interspersed xylem rays (Fig. 10). Tracheid files are 1–3 cells wide tangentially, with two being the most common (Figs. 10, 11, and 12). In cross section, tracheids are isodiametric and somewhat angular (Fig. 10), with diameters that range from 38 to 82 µm (mean = 54 µm). Pitting of the tracheid walls is alternate, with pits ranging from circular to oval (Fig. 13). Pits most commonly occur on radial walls, but occasionally are observed on all walls.
Xylem rays are wide and tall and are composed of radially aligned parenchyma cells. Rays are up to eight cells wide (Figs. 11 and 12) and up to 135 cells tall (Fig. 11). Ray parenchyma cells are thin walled, radially elongate (Fig. 10), and isodiametric to polygonal in longitudinal section (Fig. 12). Ray cells are 30–52 µm in diameter (mean = 37 µm) in longitudinal section and are up to 400 µm long when viewed in cross section.
Vascular cambium and secondary phloem
Vascular cambium is preserved as a narrow zone of cells at the periphery of the secondary xylem and represents fusiform and ray initials and their most recent derivatives (Fig. 10). This zone is 3–4 cells thick and is recognizable by position and the narrow radial dimension of the cells. Secondary phloem is represented by radial files of sieve cells that are oval to rectangular in cross section and phloem parenchyma cells that are generally crushed and have incompletely preserved, thinner walls (Figs. 8 and 10). Sieve cells are 81–48 by 60–36 µm in diameter (mean = 61 by 45 µm) (Fig. 10). No sieve areas have been observed. Toward the periphery of the stem, phloem rays are often dilated (Fig. 10), and the parenchyma cells of the rays become continuous with the parenchyma of the inner cortex. A small number of thick-walled cells that resemble phloem fibers are preserved in the rays (Fig. 8 at arrows). Clusters of thick-walled cells (= sclerotic clusters) filled with black resinous material are often found at the transition between rays and inner cortex (Fig. 10).
Cortex and leaf bases
Cortical ground tissues consist of a zone of inner cortex at the periphery of the secondary phloem, periderm, a middle cortex (Fig. 14), and a Dictyoxylon-type outer cortex (Figs. 17 and 18). Cells of the inner cortex are continuous with the parenchyma cells of the phloem rays and are generally radially compacted and poorly preserved (Fig. 10). This zone is often quite thin and may only be represented by compacted parenchyma cells flanking and exterior to the clusters of thick-walled, resin-filled sclerotic cells at the terminus of the phloem rays. Exterior to the inner cortex is a periderm that consists of up to six rectangular, thin-walled cells aligned in radial files (Fig. 16). These cells are tangentially elongate and up to 66 by 33 µm in maximum diameter.
|
; Pigg et al., 1987
). The Dictyoxylon-type outer cortex displays the characteristic "Roman numeral" configuration formed by bands of fibers that are radially aligned in cross section (Figs. 17 and 18) and that anastomose in longitudinal section (Fig. 19). Individual fibers are 30–68 µm (mean = 48 µm) in diameter. The bands of fibers are relatively ribbon-shaped in cross section, except near the epidermis where they commonly are dilated (Fig. 18). The bands are separated by thin-walled, isodiametric to polygonal parenchyma cells that are 38–112 µm (mean = 68 µm) in diameter (Figs. 17 and 18). At the exterior margin of the Dictyoxylon outer cortex is a zone one to several cells thick that may represent an epidermis and/or hypodermis (Fig. 17). The cells of this zone are consistently darker and have thicker cell walls than the underlying parenchyma cells. These possible epidermal cells are isodiametric and 57–96 µm in diameter (mean = 80 µm) with cell walls 9–12 µm (mean = 10 µm).
Leaf trace production
Leaf traces are produced in a 2/5 helix, and each arises by a radial division from a single protoxylem bundle located in one of the five corners of the pentagonal stele (Fig. 20). Distally, the leaf trace enlarges and becomes surrounded by secondary xylem as it extends through the stem secondary xylem (Figs. 21 and 25) and into the phloem and inner cortex (Fig. 22). As the leaf trace bundle passes through the periderm and into the middle cortex it assumes a squat C shape in cross section (Fig. 23). Subsequently, the arms of the C branch off at different levels of the stem, forming the characteristic three bundle rachis trace configuration of this genus (Fig. 24). Primary xylem of the foliar traces consists of a variable number of mesarch protoxylem strands at the outer margin of the metaxylem (Fig. 26). Protoxylem tracheids are 12–18 µm in diameter (mean = 14 µm) and metaxylem tracheids are 45–47 µm in diameter (mean = 57 µm). Secondary tissues of the foliar traces consist of radially aligned files of tracheids, xylem rays, and a poorly preserved cambio-phloic layer (Figs. 26 and 27). Files of secondary tracheids are 4–9 cells long and approximately 2–4 cells wide. Secondary xylem tracheids are 27–45 µm in diameter (mean = 34 µm): cellular details of ray cells have not been observed. Pitting of both metaxylem and secondary xylem tracheids is multiseriate, alternate with circular to oval apertures (Fig. 28).
|
|
Adventitious roots
Numerous adventitious roots were observed in the two larger specimens (M1664 and M2361). These roots are randomly produced, and in cross section, can be recognized by the presence of a bulge that forms at the periphery of the stem secondary xylem and by the funnelform shape these roots assume in the secondary xylem (Fig. 31). In longitudinal section these roots are circular to oval within the cauline secondary xylem (Fig. 32). Well-developed secondary tissues are preserved with some of the roots (Fig. 33); however, primary xylem configuration has not been observed.
Arthropod remains
Coprolites occur individually and as agglomerations in the pith, secondary xylem, phloem, and cortex of several of the specimens. They are isodiametric to ovoid and range in size from 72 to 135 µm by 54 to 105 µm (mean = 100 by 83 µm), with a rough textured surface. In the phloem these coprolites occur within galleries that exhibit distinct wound response (Fig. 34), suggesting the stem was living at the time of phytophagy. Wound response tissues consist of a layer of 3–4 parenchymatous cells that are generally flattened and form the wall of the gallery. Galleries are circular to slightly radially oval in cross section and are approximately 375 µm in tangential internal diameter. The walls of the galleries are 50–75 µm thick. No wound response can be observed in the other tissues of the stem.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDESCRIPTIONDISCUSSIONLITERATURE CITED |
|---|
, 1981
). These cavities are absent from Trivena. Additionally, medullosan stems produce leaf bases that are vascularized by traces that originate from several cauline protoxylem sympodia over the course of more than one node. With few exceptions, all other seed ferns produce leaf traces from one cauline protoxylem strand at a single node (Mapes and Rothwell, 1980
).
The Buteoxylonaceae is the most poorly understood seed fern family (Taylor and Taylor, 1993
) as it is known from only a few stem fragments from the upper Tournaisian of Scotland (Barnard and Long, 1973
, 1975
) and the Upper Devonian of Ireland (Matten et al., 1980
). The family is characterized as having a protostele surrounded by manoxylic wood with high narrow rays and rachises vascularized by a T-shaped bundle. However, some overlap of rachis morphology is reported between this family and the Lyginopteridaceae (Barnard and Long, 1973
; Meyer-Berthaud, 1990
).
The familial concepts of the Calamopityaceae and Lyginopteridaceae overlap in numerous characters (Table 1). For example, both groups include taxa that range from protostelic to eustelic, have primary xylem maturation that ranges from mesarch to exarch, and that produce sclerotic clusters in the parenchymatous cortex. Wood anatomy cannot be used to differentiate between the two families because the calamopityacean genera Calamopitys (Unger) sensu Galtier and Meyer-Berthaud (1989)
and Stenomyelon (Kidston and Gwynne-Vaughan) sensu Meyer-Berthaud and Stein (1995)
produce the same type of wood as the lyginopterid genera Lyginopteris Potonié and Heterangium Corda (Andrews, 1940
). Likewise, the presence of a Sparganum- or Dictyoxylon-type outer cortex is not a distinguishing character. The Calamopityaceae may produce either the Sparganum or Dictyoxylon type (Beck and Stein, 1987
) and in the lyginopterid genus Tetrastichia Gordon the presence of the Sparganum- or Dictyoxylon-type cortex is an ontogenetic feature (Gordon, 1938
).
|
; Galtier and Meyer-Berthaud, 1989
; Galtier and Beck, 1992
). However, this character must be used with caution because there is considerable overlap of petiole morphology between the two families (Beck and Stein, 1987
). For example, the petiole of Calamopitys is vascularized by at least four xylem bundles (Galtier and Meyer-Berthaud, 1989
) and the petiole of Heterangium is vascularized by up to 10 (Shadle and Stidd, 1975
). In addition, the genus Lyginorachis has never been formally diagnosed (Meyer-Berthaud, 1990
), and the generic concept has been expanded from petioles with adaxially concave xylem strands (Gordon, 1938
) to include petioles with W-, U-, or V-shaped traces, three distinct bundles, T-shaped traces, or papillonoid traces (Meyer-Berthaud, 1990
and references therein). Despite doubts as to the reliability of rachis morphology as a taxonomic character, this feature remains the primary criterion on which early pteridosperm systematics is based (e.g., Beck and Stein, 1987
; Galtier and Beck, 1992
). The presence of three xylem bundles, an outer cortex of radial bands of sclerenchyma in the leaf base of Trivena (Figs. 24 and 29), and the unattached rachis in the concretion along with specimen M1664 (Fig. 30) all fall within the range of variation for Lyginorachis-type rachises.
Of the 13 currently recognized species of Lyginorachis (Meyer-Berthaud, 1990
), two produce three distinct bundles; however, these two species are clearly distinct from the rachis produced by Trivena. Lyginorachis gordonii Galtier and Scott produces an outer cortex of the Sparganum type (Galtier and Scott, 1986
), unlike the Dictyoxylon-type outer cortex of Trivena, and in L. trinervis bundle trifurcation occurs in the rachis (Calder, 1935
), not in the stem cortex as in Trivena.
One character that may be diagnostic for distinguishing between the Lyginopteridaceae and the Calamopityaceae is the difference in the ratio of primary xylem diameter to stem diameter. It has been reported that calamopityacean stems have a small primary xylem to stem diameter ratio (averaging 1 : 10 and always exceeding 1 : 5), and lyginopterids have a large primary xylem to stem diameter ratio, ranging from 1 : 2 to 1 : 4 (Galtier, 1988
). As noted above, the mean ratio of primary xylem diameter to stem diameter in Trivena is 1 : 4.2 (range = 1 : 3.25–1 : 5.0). This is slightly lower than the ratio Galtier (1988)
reports for Lyginopteridaceae, but is clearly above that reported for Calamopityaceae.
A biological explanation for this difference in the ratio of pith to stem may be that calamopityaceans were primarily upright plants and that manoxylic lyginopterids were lianas or vine-like. Pycnoxylic lyginopterids such as Pitus Witham, Eristophyton Zalessky, and Stanwoodia Galtier and Scott were most likely upright plants; however, where a hypothesis of growth architecture is proposed, manoxylic lyginopterids are commonly reported as vine-like. Examples include Heterangium (Pigg et al., 1987
), Microspermopteris (Baxter, 1949
; Pigg et al., 1986
), Rhetinangium (Gordon, 1912
), Tetrastichia (Gordon, 1938
), and Lyginopteris (Taylor and Millay, 1981
; Gifford and Foster, 1989
; Stewart and Rothwell, 1993
; Tomescu et al., 2001
). Evidence of vine-like growth architecture for the lyginopterids includes the presence of adventitious roots and stems that are too small to support the large leaves produced by these plants. Adventitious roots are observed in Heterangium (Pigg et al., 1987
), Schopfiastrum Andrews (Stidd and Phillips, 1973
), Microspermopteris (Taylor and Stockey, 1976
), Laceya May and Matten (1983)
, and Rhetinangium (Gordon, 1912
). Among the Calamopityaceae, adventitious roots are rare. Vascular strands interpreted as rootlets are reported for Stenomyelon tuedianum (Kidston and Gwynne-Vaughan) Long by Meyer-Berthaud and Stein (1995)
, and one Stenomyelon compression specimen shows adventitious roots (Long, 1964
), but Long interprets Stenomyelon as a "straggling shrub." Additional evidence for a semi-self-supporting growth architecture for calamopityaceans includes a biomechanical analysis of Calamopitys (Rowe et al., 1993
) and the presence of a small procambium, limited wood production, absence of a periderm, a Sparganum outer cortex, and multifascicular rachis bundles in Stenomyelon and Calamopitys (Meyer-Berthaud and Stein, 1995
). It should be noted, however, that despite the absence of adventitious roots, a number of species of Calamopitys have been interpreted as possible vines based on small stem diameter and large leaf bases (Galtier, 1975
; Rowe and Galtier, 1988
). Like many lyginopterids, the presence of adventitious roots, small stems and large leaves suggests the growth architecture of Trivena was most likely prostrate, scrambling, or climbing.
Other features that may be used to distinguish the Lyginopteridaceae from the Calamopityaceae are stratigraphy and the presence or absence of a periderm. The Calamopityaceae are restricted to Famennian and Visean strata (Scott et al., 1984
), whereas the Lyginopteridaceae range from the Upper Devonian to the Upper Pennsylvanian (Phillips, 1980
; May and Matten, 1983
). As noted above, Trivena is from Namurian A strata and is therefore within the known range of lyginopterids but younger than the known range of calamopityaceans.
A periderm is produced by Lyginopteris (Williamson, 1890
), Heterangium (Pigg et al., 1987
), Schopfiastrum (Stidd and Philips, 1973
), and Microspermopteris (Pigg et al., 1986
) of the Lyginopteridaceae, but no calamopityaceans are reported to produce these tissues. A periderm was reported in Stenomyelon tuedianum (Kidston and Gwynne-Vaughan, 1912
) but a reinvestigation of the genus (Meyer-Berthaud and Stein, 1995
) reinterpreted this tissue as comparable to the peripheral parenchyma of Diichnia (Beck et al., 1992
) and Triichnia (Galtier and Beck, 1992
). No periderm is reported for Bostonia Stein and Beck or Galtiera Beck and Stein, and the absence of a periderm is a generic character for Calamopitys (Galtier and Meyer-Berthaud, 1989
). The absence of a periderm may be significant because it is part of a suite of characters that have been reported to suggest determinate growth in the Calamopityaceae (Hotton and Stein, 1994
; Meyer-Berthaud and Stein, 1995
).
Because no synapomorphies have been identified for the Lyginopteridaceae, Calamopityaceae, or Buteoxylonaceae, genera must be assigned to one of these families based on a combination of characters. Thus, Trivena is assigned to the Lyginopteridaceae based on the following: a mean ratio of primary stelar diameter to stem diameter of 1 : 4.2; scrambling or climbing growth architecture; the production of adventitious roots; leaf bases assignable to Lyginorachis; periderm; Dictyoxylon-type outer cortex; and the occurrence of the specimens in strata that are within the known stratigraphic range of the family.
There are currently 10 manoxylic lyginopterids to which Trivena can be compared (Table 1). Like Trivena, species of Lyginopteris are mesarch, produce adventitious roots, and have Dictyoxylon-type outer cortex, periderm (Bertram, 1989
), leaf bases assignable to Lyginorachis, and a stele to stem ratio of approximately 1 : 4 (Galtier, 1988
). In addition, Lyginopteris and Trivena overlap in biogeographic and biostratigraphic distribution (Tomescu et al., 2001
). However, Lyginopteris has sclerotic cluster in the pith (Bertram, 1989
), and the pith of Trivena has resin-filled parenchyma cells but no clusters (Fig. 7). More importantly, the two have distinctly different modes of leaf trace production. Traces of Lyginopteris originate as one bundle that bifurcates in the cauline secondary tissue and distally reunites and forms a U-, V-, or W-shape in the rachis base. Traces produced by Trivena are single in the stele and trifurcate in the stem cortex. Lyginopitys shares many of the above-mentioned characters with Lyginopteris including a similar rachis trace but is known only from the Tournaisian (Galtier, 1970
).
Tetrastichia, Tristichia, and Laceya have actinosteles ranging from three- to four-lobed, have no periderm, and produce leaf traces without secondary tissues that form a U or W shape in the rachis base (Gordon, 1938
; Long, 1961
; May and Matten, 1983
, respectively).
Schopfiastrum shares a Dictyoxylon-type outer cortex and the presence of a periderm with Trivena. Characters of Schopfiastrum not found in Trivena include a vitalized protostele, cortical canals rather than sclerotic clusters, and a distinctly different mode of leaf trace production. The leaves of Schopfiastrum are produced in an alternate-distichous phyllotaxy and are vascularized by two protoxylem poles that divide into several protoxylem poles and form an abaxially lobate bundle. In addition, Schopfiastrum is restricted to the Westphalian C-D of Iowa, Illinois, and Kentucky (Andrews, 1945
; Rothwell and Taylor, 1972
; Stidd and Phillips, 1973
).
The monospecific stem genus Rhetinangium is currently known only from the middle to upper Visean of Scotland and has a unique protostele with anastomosing groups of tracheids immersed in parenchymatous ground tissue (Gordon, 1912
). Additional features that can be used to distinguish this genus from Trivena include the following: exarch xylem maturation, an inner cortex with secretory canals, and a periderm that is absent. Phyllotaxy is helical like Trivena, and the rachis is vascularized by a trace with accompanying secondary tissues that originate from one sympodial cauline bundle, but in the rachis base the trace forms a unique bundle of many U-shaped xylem groups united in a corrugated band (Gordon, 1912
).
Microspermopteris is a seed fern that was originally described by Baxter (1949)
and subsequently reinvestigated by Taylor and Stockey (1976)
and Pigg et al. (1986)
. The known range of the genus is from the Visean of Scotland to the Westphalian C-D of North America (Bertram, 1989
). Stems have a characteristic protostele dissected into five wedges by longitudinal parenchymatous plates that radiate from the center of the stele. Protoxylem strands occur in pairs at the outer margin of the stele, with one strand on each side of a parenchyma plate. Rachis bases are vascularized by two traces, one from each strand of a pair, that fuse and form a C-shaped bundle as it passes through the cauline secondary tissues and cortex. In the petiole the trace assumes a squat V shape. No other taxa, including Trivena, produce this type of dissected stele and/or produce leaf traces by this mode.
One of the most extensively studied and widespread lyginopterid seed ferns is Heterangium, with as many as 22 permineralized species (Pigg et al., 1987
), ranging stratigraphically from the middle Visean to the Stephanian B (Bertram, 1989
). As originally established by Corda (1845)
, the generic concept of Heterangium is restricted to stems with a vitalized protostele of large metaxylem cells surrounded by smaller celled parenchyma. Additional characters of the genus include horizontal sclerenchyma plates in the inner cortex and either a Sparganum- or Dictyoxylon-type outer cortex (Stewart and Rothwell, 1993
). The genus exhibits variable modes of frond trace divergence and, based primarily on this character, was divided into two subgenera by Scott (1917)
. Subgenus Heterangium (= Euheterangium) is characterized by leaf traces that arise from a single protoxylem pole, and subgenus Polyangium has leaf traces that arise from two separate protoxylem poles. In addition, the protoxylem poles are more highly developed in the subgenus Polyangium (Scott, 1917
). Taylor and Millay (1981)
suggest these distinctions are generic rather than subgeneric, but to date, this has not been formalized. A third subgenus, Lyginangium (Scott, 1917
, 1923
), was established to accommodate two species that have centrifugal primary xylem surrounding a parenchymatous pith. This stelar morphology was thought to represent a transition between the protostele of Heterangium and the eustele of Lyginopteris (Kubart, 1914
); however, more recent analysis suggests that one of the species, H. andrei Kubart, should be placed in the genus Lyginopteris (Bertram, 1986
; Pigg et al., 1987
). The second species, H. intermedium Kubart, is poorly known (Pigg et al., 1987
). Regardless of whether the rachis trace originates from one or two protoxylem poles, where leaf base morphology of Heterangium is known, all species except one have rachis bases that are vascularized by few to several discrete bundles (Scott, 1917
; Pigg et al., 1987
). The lone exception, H. lintonii Stidd, is known from a single specimen and only tentatively assigned to Heterangium (Stidd, 1979
).
Leaf trace production in subgenus Heterangium shows obvious similarities to leaf trace production in Trivena. However, this is most likely the result of parallel evolution in two lyginopterid seed fern genera because morphological data suggest the evolution of a eustele from the heterangium group is unlikely. The subgenus Heterangium is considered primitive, with its cryptic protoxylem poles, and species of Polyangium are considered a highly derived endgroup because of their distinct protoxylem poles (Beck et al., 1982
) and stratigraphic position (Pigg et al., 1987
). Hirmer (1933)
analyzed the protoxylem architecture of H. kukuki Hirmer (subgenus Polyangium) and determined that like H. grievii Williamson of the subgenus Heterangium (Beck et al., 1982
), leaf traces are connected to a solid core of cauline xylem rather than sympodia as would be expected from the ancestor of a eustelic taxon. These data, along with the contemporaneous occurrence of primitive Heterangium species and Trivena, suggest the two taxa are distinct genera and not closely related.
Sclerotic clusters
Sclerotic clusters in the cortex of Trivena may have both taxonomic and ecological significance. Beck and Stein (1987)
considered two possible explanations for the presence of similar clusters found in association with the phloem of Galtiera: (1) they may represent a mechanism by which the cortex accommodates a diametric increase in secondary xylem, or (2) they may be a pathological response to a virus introduced by an insect vector. Evidence presented to support the second hypothesis includes that when viruses are introduced into plant tissues by insects, the virus is most commonly transmitted through the phloem and hyperplasia is a common histological plant response (Esau, 1956
). The histological response results in a mass of tissue similar to the sclerotic clusters of Galtiera (Beck and Stein, 1987
) and other seed ferns. Beck and Stein did not report direct evidence for the presence of insects in their specimen of Galtiera; however, such evidence is clear in Trivena (Fig. 34). As reported above, insect coprolites in galleries within the phloem of Trivena exhibit distinct wound response tissues, demonstrating that bark-burrowing arthropods were part of the Upper Mississippian ecosystem and that plants had evolved a reaction to such invasions: this gives support to the hypothesis that sclerotic clusters of some seed ferns may be a response to viral infections (Beck and Stein, 1987
). If this is true, sclerotic clusters in the phloem and randomly dispersed in the cortex are not a reliable taxonomic character in seed fern phylogeny. Of course, this would not apply to sclerotic clusters in discontinuous rows if they were produced by a discontinuous cambia. Evolutionarily, this suggests that plant-burrowing arthropods and co-evolutionary plant responses existed earlier than previously reported (e.g., Lesnikowska, 1990
).
Conclusion
The genus Trivena extends the range of variation and distribution of lyginopterid seed ferns and adds a new eustelic genus from a Mississippian non-coal swamp environment to the family. If, as mentioned above, seed plant evolution involves the pteridosperms, the evidence will very likely be found in a clastic swamp environment such as the Fayetteville flora, not from a coal swamp environment. That is because many workers suggest coal swamps were environments of evolutionary stasis for Paleozoic plants in general (DiMichele and Aronson, 1992
and references therein) and lyginopterids specifically (Taylor and Millay, 1981
). It is hoped that this, and ongoing studies of the fossil plants of the Fayetteville Formation, will provide valuable data in the continuing search for evidence of the phylogeny and relationships among the pteridosperms and ultimately all seed plants.
|
FOOTNOTES |
|---|
2 Author for reprint requests (e-mail: md369590{at}ohiou.edu)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSYSTEMATICSDESCRIPTIONDISCUSSIONLITERATURE CITED |
|---|
Andrews H. N. 1945 Contributions to our knowledge of American Carboniferous floras VII. Some pteridosperms from Iowa. Annals of the Missouri Botanical Garden 32: 323-361[CrossRef]
Banks H. P. 1968 The early history of land plants. In E. T. Drake [ed.], Evolution and environment, 73–107. Yale University Press, New Haven, Connecticut, USA
Barnard P. D. W. A. G. Long 1973 On the structure of a petrified stem and some associated seeds from the Lower Carboniferous rocks of East Lothian, Scotland. Transactions of the Royal Society of London 69: 91-108
Barnard P. D. W. A. G. Long 1975 Triradioxylon—a new genus of Lower Carboniferous petrified stems and petioles together with a review of the classification of early Pterophytina. Transactions of the Royal Society of Edinburgh 69: 231-250
Baxter R. W. 1949 Some pteridosperms and fructifications with particular reference to the Medullosae. Annals of the Missouri Botanical Garden 36: 287-352[CrossRef]
Beck C. B. 1960 The identity of Archaeopteris and Callixylon. Brittonia 12: 351-368[CrossRef]
Beck C. B. 1966 On the origin of Gymnosperms. Taxon 15: 337-338[CrossRef]
Beck C. B. 1976 Current status of the Progymnospermopsida. Review of Palaeobotany and Palynology 21: 5-23[CrossRef][ISI]
Beck C. B. J. Galtier W. E. Stein 1992 A reinvestigation of Diichnia Read from the New Albany Shale of Kentucky. Review of Palaeobotany and Palynology 75: 1-32
Beck C. B. R. Schmid G. W. Rothwell 1982 Stelar morphology and the primary vascular system of seed plants. Botanical Review 48: 691-815
Beck C. B. W. E. Stein 1987 Galtiera bostonensis, gen. et sp. nov., a protostelic calamopityacean from the New Albany Shale of Kentucky. Canadian Journal of Botany 65: 348-361[ISI]
Bertram U. 1986 On the evolution of stems of Lyginopteridaceae. L'Evolution des gymnosperms. Aproche biologique et paléobiologique. Resumes des communications, Colloque par la Fondation L. Emberger-C. Sauvage III Societie Botanique de France: 9–10
Bertram U. 1989 Untersuchungen an coal balls aus dem Namur A Von Ostrau unter Spezieller Berückichtigung der gattungen Heterangium, Lyginopteris, und Microspermopteris. Palaeontographica 214B: 125-224
Calder M. G. 1935 Further observations on the Genus Lyginorachis Kidston. Transactions of the Royal Society of Edinburgh 58: 549-559
Corda A. J. 1845 Flora Protogaea. Beiträge zur Flora der Vorwelt. S. Calvary, Berlin, Germany
DiMichele W. A. R. B. Aronson 1992 The Pennsylvanian-Permian vegetational transition: a terrestrial analogue to the Onshore-Offshore Hypothesis. Evolution 46: 807-825[CrossRef][ISI]
Esau K. 1956 An anatomist's view of virus diseases. American Journal of Botany 43: 739-748[CrossRef][ISI]
Galtier J. 1970 Recherches sur les végétaux á structure conservée du Carbonifère inférieur français. Paléoobiologique Continentale 1: 1-221
Galtier J. 1975 Variabilité anatomique et ramification des tiges de Calamopitys. Compte Rendu de l'Académie des Sciences, Paris 280: 1967-1970
Galtier J. 1988 Morphology and phylogenetic relationships of early pteridosperms. In C. B. Beck [ed.], Origin and evolution of gymnosperms, 135–176. Columbia University Press, New York, New York, USA
Galtier J. C. B. Beck 1992 Triichnia, a new eustelic calamopityacean from the Lower Carboniferous of France. Palaeontographica 224B: 1-16
Galtier J. B. Meyer-Berthaud 1989 Studies of the Early Carboniferous pteridosperm Calamopitys, a redescription of the type material from Saalfeld (GDR). Palaeontographica 213B: 1-36
Galtier J. A. C. Scott 1986 Lyginorachis gordonii, nouvelle Ptéridospermale probable du Carbonifère Inférieur d'Ecosse. Compte Rendu de l'Académie des Sciences, Paris 302: 251-256
Gifford E. M. A. S. Foster 1989 The morphology and evolution of vascular plants, 3rd ed. W. H. Freeman, New York, New York, USA
Gordon W. T. 1912 On Rhetinangium arberi, a new genus of Cycadofilices from the Calciferous Sandstone Series. Transactions of the Royal Society of Edinburgh 48: 813-825
Gordon W. T. 1938 On Tetrastichia bupatides, a Carboniferous pteridosperm from East Lothian. Transactions of the Royal Society of Edinburgh 59: 351-370
Hirmer M. 1933 Zur Kenntnis der strukturbietenden Pflanzenreste des jüngeren Paläozoikens. Palaeontographica 77B: 121-140
Hotton C. L. W. E. Stein 1994 An ontogenetic model for the Mississippian seed plant family Calamopityaceae. International Journal of Plant Science 155: 119-142[CrossRef]
Joy K. W. A. J. Willis W. S. Lacey 1956 A rapid cellulose acetate peel method in paleobotany. Annals of the Botanical Society of London 20: 635-637
Kidston R. D. T. Gwynne-Vaughan 1912 On the Carboniferous flora of Berwickshire. Part 1. Stenomyelon tuedianum Kidston. Transactions of the Royal Society of Edinburgh 48: 263-271
Kubart B. 1914 Über die Cycadofiliceen Heterangium und Lyginodendron aus dem Ostrauer Kohlenbecken. Oesterreichische botanische Zeitschrift 64: 8-19[CrossRef]
Lesnikowska A. D. 1990 Evidence of herbivory in tree-fern petioles from the Calhoun Coal (Upper Pennsylvanian) of Illinois. Palaios 5: 76-80
Long A. G. 1961 Tristichia ovensi gen. et sp. nov., a protostelic Lower Carboniferous pteridosperm from Berwickshire and East Lothian, with an account of some associated seeds and cupules. Transactions of the Royal Society of Edinburgh 64: 477-492
Long A. G. 1964 Some specimens of Stenomyelon and Kalymma from the Calciferous Sandstone Series of Berwickshire. Transactions of the Royal Society of Edinburgh 65: 435-447
Mapes G. G. W. Rothwell 1980 Quaestora amplecta gen. et sp. n. a structurally simple medullosan stem from the Upper Mississippian of Arkansas. American Journal of Botany 67: 636-647[CrossRef][ISI]
Matten L. C. W. S. Lacey B. I. May R. C. Lucas 1980 A megafossil flora from the uppermost Devonian near Ballyheigue, Co. Kerry, Ireland. Review of Palaeobotany and Palynology 29: 241-251[CrossRef][ISI]
May B. I. L. C. Matten 1983 A probable pteridosperm from the uppermost Devonian near Ballyheigue, County Kerry, Ireland. Botanical Journal of the Linnean Society 86: 103-123
Meeks L. K. A. L. Titus W. L. Manger 1997 Taphonomy and biostratigraphy of ammonoid cephalopods, Fayetteville Shale (middle Chesterian, Mississippian), Northern Arkansas. Proceedings of the XIII International Congress on the Carboniferous and Permian, Part 2 157: 311-317
Meyer-Berthaud B. 1990 Studies on a Lower Carboniferous flora from Kingswood near Pettycur, Scotland. III. Lyginorachis. Review of Palaeobotany and Palynology 63: 77-90[CrossRef][ISI]
Meyer-Berthaud B. W. E. Stein 1995 A reinvestigation of Stenomyelon from the late Tournaisian of Scotland. International Journal of Plant Sciences 156: 863-895[CrossRef]
Oliver F. W. D. H. Scott 1904 On the structure of the Paleozoic seed Lagenostoma lomaxi, with a statement of the evidence upon which it is referred to Lyginodendron. Royal Society of London, Philosophical Transactions B 197: 193-247
Owens B. S. Loboziak R. Coquel 1979 Late Mississippian-Early Pennsylvanian miospore assemblages from northern Arkansas. In P. K. Sutherland and W. L. Manger [eds.], Neuvième Congrès International de Stratigraphie et de Géologie du Carbonifère, Compte Rendu 2 377–384
Phillips T. L. 1980 Stratigraphic and geographic occurrences of permineralized coal-swamp plants—Upper Carboniferous of North America and Europe. In D. L. Dilcher and T. N. Taylor [eds.], Biostratigraphy of fossil plants, 25–92. Dowden, Hutchinson and Ross, Stroudsburg, Pennsylvania, USA
Pigg K. B. R. A. Stockey T. N. Taylor 1986 Studies of Paleozoic seed ferns: additional studies of Microspermopteris aphyllum Baxter. Botanical Gazette 147: 126-136[CrossRef][ISI]
Pigg K. B. T. N. Taylor R. A. Stockey 1987 Paleozoic seed ferns: Heterangium kentuckyensis sp. nov., from the Upper Carboniferous of North America. American Journal of Botany 74: 1184-1204[CrossRef][ISI]
Potonié H. 1899 Lehrbuch der Pflanzenpaleontologie mit besondere Rücksicht auf die Bedürfnisse des Geologen. Dümmber Berlin: 170–175, 230–233
Read C. B. 1936a Flora of the New Albany Shale; I. Diichnia kentuckyensis, a new representative of the Calamopityeae. U.S. Geological Survey Professional Paper 185: 149-161
Read C. B. 1936b Flora of the New Albany Shale; II. The Calamopityeae and their relationships. U.S. Geological Survey Professional Paper 186E: 81-104
Rothwell G. W. 1975 The Callistophytaceae (Pteridospermopsida): I. Vegetative structures. Palaeontographica 151B: 171-196
Rothwell G. W. 1981 The Callistophytales (Pteridospermopsida): reproductively sophisticated Paleozoic gymnosperms. Review of Palaeobotany and Palynology 32: 103-121
Rothwell G. W. 1982 New interpretations of the earliest conifers. Review of Palaeobotany and Palynology 37: 7-28[CrossRef][ISI]
Rothwell G. W. T. N. Taylor 1972 Carboniferous pteridosperm studies: morphology and anatomy of Schopfiastrum decussatum. Canadian Journal of Botany 50: 2649-2658[ISI]
