Systematics, ontogeny, and phylogenetic implications of exceptional anatomically preserved cycadophyte leaves from the Middle Jurassic of Bearreraig Bay, Skye, northwest Scotland.
I. Abstract II. Introduction III. Locality and Materials IV. Systematic Paleobotany V. Discussion VI. Acknowledgments VII. Literature Cited
The Cycadales (early Permian to Present) and the Bennettitales (late Permian to late Cretaceous) are collectively known as the "cycadophytes," as they show strikingly similar vegetative morphology (e.g., Delevoryas, 1968, 1982; Pant, 1987; Stevenson, 1990; Norstog & Nicholls, 1997). They have strongly contrasting, distinct reproductive structures, however, the Bennettitales being widely regarded as the relatively derived sister group to the angiosperm-gnetalean clade (Doyle & Donoghue, 1986; Crane, 1988; Stewart & Rothwell, 1993; Taylor & Taylor, 1993). The fossil record of cycadophyte leaves is dominated by adpressions that do not preserve internal anatomy, so that the few anatomically preserved Mesozoic floras provide especially important insights. Well-known adpression floras include the Rhaetic flora of Scoresby Sound, east Greenland (Harris, 1937), the Aalenian-Bajocian flora of Yorkshire (Harris, 1961, 1964, 1969), the largely Kimmeridgian flora of the French Jura (Barale, 1981), and the Kimmeridgian flora of Brora, Scotland (Seward, 1911; van Konijnenburg-van Cittert & van der Burgh, 1989). Japan (Lower Jurassic: Kimura & Tsujii, 1982) and Mexico (Middle Jurassic: Person & Delevoryas, 1982) have yielded floras geographically more distant from Skye. The only comparable three-dimensionally preserved Jurassic leaf flora occurs in the Rajmahal Hills of India (Bose, 1953; Rao & Achuthan, 1967; Bose & Kasat, 1972a; Sukh-Dev & Zeba-Bano, 1975). Among the adpression floras, those of Yorkshire, Oxfordshire, and Lincolnshire are closest to Skye in both age and geography (Harris, 1964, 1969; Oldham, 1976).
Bearreraig Bay, on the east coast of Skye, northwest Scotland, has yielded an allochthonous, shallow-marine assemblage of plant fragments showing relatively low species-level diversity (Bateman & Morton, 1994; Bateman et al., 2000). About 15 organ species represent at least 11 whole-plant species from the Equisetales, Marattiales, Filicales, Coniferales, Cycadales, and Bennettitales. Leaf, stem, wood, and reproductive fragments are preserved within calcite-rich nodules, showing varying degrees of anatomical preservation.
This article focuses on the cycadophytes from Bearreraig. The material described consists of a small number of substantial and remarkably well preserved leaf segments assigned to the cycad Nilssonia cf. tenuinervis Seward and the bennettites Otozamites mortonii sp. nov. and Otozamites sp.
III. Locality and Materials
The locality at Bearreraig Bay is situated on the east coast of the Isle of Skye, about 10 km north of Portree on the Trottemish Peninsula. The cycad and bennettite fragments came from several horizons in the section, extending from the top of the Dun Caan Shale Member to the middle of the Holm Sandstone Member of the Bearreraig Sandstone Formation (Morton, 1984). This represents about 10 my of deposition, spanning the well-documented Aalenian-Bajocian boundary at the site. The flora was deposited 25-30 km offshore, in fully marine sediments accumulating in the Middle Jurassic Hebrides Basin (Bateman et al., 2000). Decay and dissolution of abundant marine invertebrate tests, principally of ammonites and bivalves, and subsequent reprecipitation of calcite generated nodules that petrified entrained plant debris (Bateman, 1999). This flora represents an unusual sample of the source community, as the plant material preserved does not represent the most resistant part of the biomass that would be expected from the highly allochthonous nature of this flora. Delicate fusainized fronds of the ferns Phlebopteris and Hausmannia, and apparently immature non-fusainized bennettite leaves, occur in a flora that had traveled an estimated 25-30 km prior to deposition. To explain the preservation of such delicate structures, Bateman et al. (2000) suggested that the plant remains floated as a raft before sinking to the sea floor at the zone of freshwater/seawater inversion. Because of this bias in preservation, and the limited size of the flora, reconstructions of the source communities remain speculative.
The cycadophytic material consisted of several leaf fragments preserved as calcite permineralizations in three nodules, designated BE10/1, BE75/8, and BE86/1, collected from Bearreraig by Nicol Morton and RMB between 1963 and 1995. Processed materials will be housed at the Natural History Museum, London.
Serial acetate peels were made using the classic method of Joy et al. (1956), with 10% HCI, and specimens were prepared for petrographic thin-sections and SEM examination using standard paleobotanical techniques (e.g., Jones & Rowe, 1999). Submergence of hand specimens under alcohol enhanced contrast for photography.
Comparisons with extant cycads were made using slide collections of the University of Edinburgh, as well as living specimens cultivated at the Royal Botanic Garden Edinburgh and the New York Botanical Garden.
IV. Systematic Paleobotany
Genus: Nilssonia Brongniart 1825
Brongniart, A., 1825. Observations sur les vegetaux fossiles renfermes dans les Gres de Hoer en Scanie. Ann. Sci. Nat. (Paris) 4: 210.
Nilssonia cf. tenuinervis Seward
Seward, A. C., 1900. The Jurassic Flora. I. The Yorkshire Coast. Catalogue of Mesozoic Plants in the Department of Geology, British Museum (Natural History) 3: 241, text-fig. 41.
Figures: Figs. 1-7, specimen BE10/1.
[Figures 1-7 OMITTED]
Description: The largest foliar fragment from Bearreraig is part of the rachis, with a planar fragment of lamina extending laterally to a maximum of 12 mm on one side and 2 mm on the other (Fig. 1). Parallel unbranched veins emerge from the rachis at a frequency of 18-19 veins per cm and angles of 81-89[degrees], intersecting the entire margin at an angle of 45-55[degrees]. The lamina is glabrous, and the cuticles are only poorly preserved.
The rachis is 1.8 mm wide and 0.8 mm high in transverse section. Pyrite deposition obscures the arrangement of vascular tissues, but an oval outer cortex of sclerenchyma is preserved. The indifferently preserved epidermis reveals extensions of collenchyma and parenchyma that form triangular ridges, similar to those in winged stems of some extant Leguminosae. These run along the rachis; they extend 0.5 mm laterally below, but are independent of, the lamina (Fig. 2).
The lamina is thin (0.2 mm in the intercostal regions), and the upper epidermal cells and cuticle are poorly preserved. The palisade mesophyll constitutes about half the total thickness of the lamina, and the subtending spongy mesophyll of isodiametric cells (ca. 33/[micro]m) is 2-3 cells thick. The lower epidermal cells have nonsinuous anticlinal walls (Fig. 3) and are equidimensional in intercostal regions (16 [micro]m in the anticlinal dimension) but more elongated below the veins (up to 50 [micro]m parallel to the veins). Within the veins, the phloem is abaxial to the xylem tracheids, which show annular and scalariform thickenings (Fig. 4); both are surrounded by a sheath of sclerenchyma. The lamina is narrower in the intercostal regions.
Series of oval cavities occur between the veins on the upper surface of the lamina. These cavities are 110 x 140 [micro]m and 100 [micro]m deep, lined with dense cells, together constituting a composite structure ca. 280 [micro]m across. Large oval cavities with dense surrounding tissues were observed within the spongy mesophyll layer of some sections (lumen 148 x 115 [micro]m: Fig. 5). Stomata on the abaxial surface are unevenly scattered in the intercostal areas and are generally oriented perpendicular to the veins. They are haplocheilic (sensu Krassilov, 1976) and slightly sunken (Fig. 6).
Discussion: Descriptions of the morphologically convergent, more or less entire-leaved genera Nilssonia Brongniart (Cycadales) and Nilssoniopteris Nathorst (Bennettitales) differ in only three characters: stomatal subsidiary cells, anticlinal cell-wall patterns, and rachis structure. The subsidiary cells of the Bearreraig specimen are multiple, corresponding with the description of Nilssonia. Although Nilssoniopteris was described by Harris (1969) as "usually" having sinuous walls, implying some ambiguity, the cell walls of this specimen are clearly cycadalean. The lamina is continuous over the rachis in Nilssonia (Watson & Sincock, 1992) but discontinuous in Nilssoniopteris; again, the Bearreraig specimen is consistent with Nilssonia.
The cavities in the adaxial surface of the leaves from Bearreraig are arranged in a similar pattern as, and are equal in size to, the resin bodies observed in Nilssonia tenuinervis Harris (Harris, 1964), a species not previously reported outside Yorkshire. The description of N. orientalis Heer, previously recorded at Sutherland, Scotland and Bohemia, Germany (Kimmeridgian-Cenomanian), correlates well with the Bearreraig Nilssonia, except that resin bodies are not mentioned (cf. Harris, 1961; van Konijnenburg-van Cittert & van der Burgh, 1989; Kvacek, 1995). Similar resin bodies characterize the adaxial surface of the lamina of Triassic Glandulataenia Pant (Pant, 1990). Although the surface of the lower epidermis of the Bearreraig specimen is poorly preserved, the trichome bases and occasional papillae typical of N. tenuinervis are apparent.
Stopes (1910) described the anatomy of a Cretaceous Nilssonia orientalis lamina that lacked differentiation of the mesophyll within the leaf and possessed large marginal resin ducts and sunken stomata, but, sadly, rachis anatomy was not discussed. Nilssonia tenuinervis has a much thinner lamina, with stomata that are not sunken. Oval cavities in the upper and lower surfaces of the lamina of the Bearreraig leaf are smaller and of a different shape from the resin bodies of other species and may be comparable to large storage cells or to idioblasts described by Rao and Achuthan (1967) for Ptilophyllum cutchense Morris (Fig. 7).
Harris (1964) described Nilssonia tenuinervis as a linear leaf, up to 4 cm wide, typically with an entire, thin lamina containing 100 [micro]m-wide resin bodies between the veins. The surface of the lamina is glabrous and not undulating, and the margins and midrib are both depressed. The veins are rarely fewer than 30 per cm and arise from the midrib at an angle of 85 [degrees]. The cuticles are very delicate, with impressions of straight-walled cells elongated on the upper epidermis and over the veins of the lower epidermis (where there are frequent oval trichome bases) but isodimensional between the veins. Stomata are scattered in broad zones between the veins on the abaxial surface only. The guard cells are only slightly sunken, and the irregular groups of subsidiary cells occasionally bear papillae.
Transfusion tissue (connective, thick-walled cells associated with the vascular bundles of all modern cycads; Esau, 1953; Fahn, 1967; Pant, 1987) is absent from these specimens. The rachis structure seen in the Bearreraig Nilssonia specimens is not found elsewhere among fossil or modern cycads, although it is comparable to the abaxial extensions of the rachis reconstructed for the sporophyll Phasmatocycas (Gillespie & Pfefferkorn, 1986). In terms of overall leaf morphology, when compared with living specimens, Nilssonia cf. tenuinervis is perhaps most similar to the leaves of Stangeria T. Moore, although the central vein is single rather than multiple.
Genus: Otozamites Braun
Braun, C. F. W., 1843. In Munster, Beitr. Petrefacten-Kunde 6: 36.
Otozamites mortonii Dower, R. Bateman and D. Stevenson, sp. nov.
Otozamites penna Harris simulans foliolis latoribus 5.5 mm usque latis abaxialiter ornatus, venulis promulis 13-18 per cm ornatis diversa.
Figures: Figs. 8-17, 20.
[Figures 8-17 OMITTED]
Diagnosis: Pinnate leaf 7 to 27 mm in width with rachis 2.5 mm wide. Pinnae linear-lanceolate, falcate, 5.5 mm wide, tapering to a blunt apex. Smaller leaves with shorter, imbricate, recurved, suboval pinnae. Basiscopic margin contracted, acroscopic margin slightly enlarged to auriculate. About 13 veins across pinnae, to 18 per cm. Adaxial cuticle poorly preserved; abaxial cuticle densely papillate with impressions of sinuous cell walls. Syndetocheilic stomata are transversely oriented in bands between veins, ca. 80 per [mm.sup.2]. Vascular system of rachis composed of several vascular bundles, almost completely joined to form an outer horseshoe-shaped amphivasal bundle and an inner cylindrical amphicribal bundle. Centrally located phloem of the inner bundle contains dispersed sclerenchyma, in the form of phloem fibers.
Etymology: The species is named in honor of Nicol Morton, who collected specimen BE75/8 and has devoted much of his career to studying the Jurassic stratigraphy of Bearreraig Bay.
Age: Aalenian-Bajocian boundary, Middle Jurassic.
Type locality: Bearreraig Bay, Skye, northwest Scotland.
Discussion: Two sizes of leaves represent this species at Bearreraig Bay. The larger leaves are shown in Figs. 8-9, and the smaller leaves are best represented by that shown in Fig. 10. The smaller leaves, reconstructed from serial sections, have pinnae that are small, oval, strongly recurved, and imbricate with about 40% overlap (Figs. 11-12). The larger leaves have longer, linear-lanceolate leaves that are also recurved but are not imbricate. As the anatomy and cuticular characters of the leaves are identical, they are interpreted as leaves from the same species that represent different developmental stages, the imbricate leaves being immature.
The pinnae are alternate, emerging at an angle of 60-55[degrees] to the rachis. Initially perpendicular to upper surface of the rachis, they rapidly curve to the horizontal (Figs. 10, 13) and taper from a maximum width of 5.5 mm near the base to a blunt apex. The lamina is ca. 0.8 mm thick, thickening slightly near the margins. The bases of the pinnae are almost symmetrical, the acroscopic margin being only slightly enlarged. The veins emerge from the whole area of attachment, diverging initially but soon becoming parallel, branching at several points but not anastomosing (Fig. 12). They are embedded in the thick lamina of the pinnae and end near the apex.
The rachis consists of concentric tissues surrounding a horseshoe-shaped stele (Fig. 13). The thick (3.7 [micro]m) cuticle covers a thick-walled epidermis. The hypodermis or outer cortex consists of 5-6 layers of thick-walled isodiametric (26 [micro]m) collenchyma (often filled with dark content, possibly diagenetic pyrite). The inner cortex is composed of parenchymatous ground tissue. The vascular bundle sheath consists of thick-walled cells (35 [micro]m in radial dimension) and 5-6 layers of sclerenchyma. In the outer vascular band, the phloem has been diagenetically lost, leaving lacunae that form an almost continuous ring. The xylem is arranged in radial files of 5-7 cells; the tracheids are about 14 x 16 [micro]m in transverse section, with annular and pitted wall thickenings. Additional sclerenchyma and parenchyma cells form ground tissue between the outer and inner vascular regions. The inner tissues are arranged in reverse order, with sclerenchyma, xylem with protoxylem poles on the inside, next to an irregular central area of phloem lacunae interspersed with sclerenchyma. The inner vascular band is amphivasal, a condition in which the phloem is surrounded by the xylem. The pinna veins emerge from the regions of the stele in which the outer and inner bundles merge.
Tissues of the pinnae are unusually well preserved (Fig. 14). Cells of the upper epidermis are small and slightly columnar (2.5 [micro]m anticlinally x 12 [micro]m periclinally), with cuticle preserved in places (2.5 [micro]m thick). A layer of elongated cells (24 [micro]m, isodiametric in transverse section but length unknown) oriented parallel to the veins, extends over the whole pinna immediately beneath the epidermis. This resembles cycadalean transfusion tissue, but the latter occurs between the vascular bundles of extant cycad leaves (Gifford & Foster, 1989; Norstog & Nicholls, 1997). The rachis has a smooth, irregular surface. Occasional swellings in the epidermis, which influenced the hypodermis development directly below, may be glands or fungal fruiting bodies (M. Krings, pers. comm., 1999; Fig. 15). The palisade cells average 145 x 25 x 25 [micro]m, and the underlying spongy mesophyll encloses the vascular bundles. A sheath of sclerenchyma cells surrounds each vascular bundle, which contains several large xylem tracheids concentrated adaxially. The lower epidermis has strongly sinuous, anticlinal cell walls and is densely papillate (Fig. 16). The papillae are large, irregular, and often flattened but rarely bilobed (47 [micro]m long); the enclosing cuticle thickens to 6.2 [micro]m. The stomata are syndetocheilic and dispersed evenly over the abaxial surface of the pinnae, with their apertures aligned across the pinnae. The stomata and indistinct subsidiary cells are obscured by the papillae (Fig. 17).
Pinna morphology is used to distinguish among the widespread Jurassic leaf-genera Otozamites, Zamites Brongniart, and Ptilophyllum Morris (Harris, 1969). An acroscopic auricle, characteristic of Otozamites but absent from Ptilophyllum and Zamites, is deemed to be present when the first acroscopic vein arises at an angle to the rachis greater than 90[degrees] or when the first basiscopic vein terminates farther toward the apex than the first acrosopic vein (Watson & Sincock, 1992). All specimens from Bearreraig show veins rapidly diverging from the point of attachment of the pinnae. The basiscopic margins of Otozamites and Zamites are contracted, whereas the basiscopic margins of Ptilophyllum are decurrent and only occasionally straight (Harris, 1969; Watson & Sincock, 1992). The Bearreraig leaves have contracted basiscopic margins and slightly expanded acroscopic margins. Thus, leaf morphology suggests close affinities with Otozamites.
The vascular arrangement in the rachis is comparable to that of the Cretaceous Ptilophyllum cutchense (Rao & Achuthan, 1967) and Zamites guptai (Sharma) Dower, R. Bateman, and D. Stevenson, comb. nov., from India (Sharma, 1967), although the lacunae and xylem in the outer vascular band are more continuous in the Bearreraig leaf. The Otozamites sp. described by Tidwell et al. (1987) has vascular bundles randomly placed in the rachis. This is strikingly different from the highly organized rachis of the Skye material. The Skye specimens show some similarities with petioles of the extant cycad Bowenia spectabilis Hook. ex Hook. f. (Poole, 1923). The omega-shaped vascular arrangement is the most common arrangement in the primary rachises of extant cycads and is interpreted as the more primitive state, the exceptions being Stangeria and Bowenia, with a ring of vascular bundles and up to three concentric rings of bundles, respectively (Crane, 1988; Stevenson, 1990).
Trichome bases are absent from the lower epidermis of the Bearreraig specimens, which are more densely papillate than are several comparable species, including Zamites wendyellisae from the English Wealden (Harris, 1969; Oldham, 1976; Watson & Sincock, 1992) and Otozamites penna, Ptilophyllum pectinoides, and Z. gigas from Yorkshire (Harris, 1969). Bilobed papillae characterize P hirsutum, but the Bearreraig specimens resemble Otozamites penna Harris most closely in gross morphology. Harris (1969) described O. penna as a pinnate leaf up to 4 cm wide. The pinnae are a maximum of 25 mm long and 3 mm wide, with vein densities of 40-50 per cm. The lower epidermis bears little ornamentation, although some specimens show small papillae overlapping the stomata and a few trichomes. In comparison with O. penna, however, the specimens from Skye have longer, wider pinnae and far lower vein density, as well as a more complex lower epidermis. It is on the basis of these characters that we assign these leaves to a new, anatomically preserved organ-species, Otozamites mortonii Dower, R. Bateman, and D. Stevenson.
Otozamites cf. mortonii
Figures: Figs. 18-19, specimen BE86/1.
[FIGURES 18-19 OMITTED]
Description: This is a single, large specimen of a leaf that occurred lower in the Bearreraig sequence and was not preserved within a well-formed calcite nodule, therefore yielding little anatomical information. The apex of the frond is absent, but the specimen is 8 cm long, and the total width of the leaf is 12 cm. The pinnae are straight, alternate to subopposite, and emerge at ca. 60[degrees] to the rachis; they are on average 6 cm long and 4.5 mm wide, with near-parallel sides that gradually taper to a blunt apex. The pinna attachment is 2 mm wide, and veins arise from the whole area, diverging rapidly into an expanded acroscopic auricle so that the lamina extends slightly over the rachis and contracts the basiscopic margin. The parallel veins show a density of 30 per cm, branch only rarely, and appear to terminate at the margins. The rachis is 2 mm in diameter, longitudinally striated, and slightly expanded to 2.5 mm toward the basal end.
Discussion: This specimen broadly resembles Otozamites mortonii. Although the pinnae are twice as long as those of O. mortonii, they are of similar width and basal morphology; thus, this leaf could be a larger and less well preserved organ of the same whole-plant species. This leaf could represent a ecophenotypic variation of O. mortonii, perhaps a larger shade leaf from a plant not receiving direct sunlight (Fig. 1).
Anatomically preserved Zamites
Two foliar organ species described by Sharma (1967) from the Rajmahal Hills of India as Ptilophyllum guptai and P. sparsifolium show closer affinities with the genus Zamites than with Ptilophyllum. As described above, Harris (1969) stated that the basiscopic margins of the pinnae of Ptilophyllum are decurrent. The leaves described by Sharma have contracted basiscopic margins, such that the bases and their venation are symmetrical and rounded. In contrast, the genus Otozamites shows auriculate acroscopic margins and asymmetrical venation. Based on this (admittedly typological) character, both of these leaves are herewith reassigned to the organ genus Zamites, which is characterized by symmetrical bases of the pinnae. The cuticular characters of the Rajmahal Hills leaves are also consistent with this taxonomic revision.
Zamites guptai (Sharma) Dower, R. Bateman, and D. Stevenson, comb. nov.
Basionym: Ptilophyllum guptai Sharma, Palaeontographica Abt. B. 120: 143.
Zamites sparsifolium (Sharma) Dower, R. Bateman, and D. Stevenson, comb. nov.
Basionym: Ptilophyllum sparsifolium Sharma, Palaeontographica Abt. B. 120: 145.
The Cycadales and Bennettitales have been distinguished on the basis of reproductive characters in most phylogenetic treatments (e.g., Donoghue et al., 1989; Doyle & Donoghue, 1992; Nixon et al., 1994; Rothwell & Serbet, 1994). In contrast, their vegetative organs appear very similar, and the paucity of diagnostic foliar characters has made identification of isolated leaves problematic. Leaf identification therefore uses the overall morphology of the leaves and pinnae, preferably supported by their cuticular characters.
The nature of the anticlinal cell-wall outlines has been used most frequently in the fossil record to distinguish between the Bennettitales and the Cycadales: Those of cycad leaves are straight to curved, but those of bennettites are typically sinuous. In extant cycads, however, sinuous cell walls characterize the leaves of Stangeria eriopus (Kunze) Baillon, Cycas micholitzii Dyer, and several species of Macrozamia (Pant & Nautiyal, 1963). Although this character is not linked to water stress in cycads, it is environmentally influenced in some angiosperms, and Pant and Nautiyal (1963) reported that, in some Macrozamia species, the cell walls appear either straight or sinuous, depending on whether the internal or external surface of the cuticle is observed. It would be interesting, therefore, to investigate the variability of this character in modern cycads that are known to show sinuous cell walls and to relate this to variation seen in fossils such as Nilssoniopteris--a desirable test of the validity of this character in classification at the ordinal level.
Stomatal morphology is also widely viewed as diagnostic: Cycads have haplocheilic stomata, whereas bennettites have syndetocheilic stomata. Florin (1933) first used these terms to describe the ontogenetic origins of subsidiary cells, but practical problems of assessing ontogeny in fossils have meant that these terms are now used to describe static morphology when applied to fossils (Krassilov, 1976). Although these terms play a vital role in the classification of the Bennettitales (Harris, 1969; Delevoryas, 1982; Crane, 1988; Watson & Sincock, 1992), they are still not sufficiently clearly defined in the literature. In the fossil record, syndetocheilic stomata are generally taken to have paracytic (i.e., a pair of) subsidiary cells, as opposed to the multiple subsidiary cells of haplocheilic stomata. Syndetocheilic stomata are found in the Bennettitales, Welwitschia, Gnetum, and some angiosperms (Gifford & Foster, 1989).
Similarities have been noted between the epidermis of these Nilssonia specimens and those of extant cycads such as Macrozamia. The outward appearances of the stomata and the arrangement of subsidiary and guard cells are remarkably similar, considering the intervening 180 million years. However, as the details of the Nilssonia are obscured by imperfect preservation, comparison can only be tentative. A similar comparison between the Tertiary Lepidozamia hopeites (Cookson) L. Johnson and the extant Lepidozamia hopei Regel, accompanied by similar macroscopic leaf morphology, resulted in their recognition as congeneric (Greguss, 1965; Jones, 1993). Fortunately, the gross morphologies of Nilssonia and Macrozamia are sufficiently distinct to separate these leaves.
The ornamentation of Otozamites mortonii is more dense than has been described for most other bennettite leaves. Ornamentation is often regarded as an environmentally plastic feature of plant leaves. Trichomes and papillae are known to vary with leaf age, predation, ultraviolet light intensity, or other environmental perturbations such as volcanic ashfall (Stevenson, 1981; Archangelsky et al., 1995; Barbacka & van Konijnenburg--van Cittert, 1998). Thus, the cuticles of O. mortonii may indicate stressed conditions. Papilla occurrence and morphology have been used as taxonomic characters by several authors, but the descriptive terms are not always clearly defined, and the considerable likelihood of infraspecific and environmentally plastic variation has rarely been adequately addressed.
The pinnae of Otozamites mortonii are thick relative to their surface area, perhaps as an ontogenetic feature of incompletely expanded leaves, as in modern cycads, but more likely in response to dry conditions (Stevenson, 1981). The smaller leaves of O. mortonii are interpreted as immature because of the imbricate, recurved shape of the pinnae, although their anatomy, venation density, and cuticular characters are identical to those of the fully expanded leaves.
Some characteristics of Otozamites mortonii, including thick pinnae and dense papillae, are reminiscent of those seen in modern plants that are adapted to low water availability (Bold et al., 1987; Stevenson et al., 1996). In this highly allochthonous flora, however, knowledge of the source communities can only be tentatively inferred. There is evidence for at least two environment types: The presence of a portion of Equisetites stem indicates a riparian or aquatic habitat, whereas the cycadophytes and ferns suggest drier habitats, the ferns being consistently burned (Bateman et al., 2000).
The detail preserved in these leaves suggests that the enclosing calcite nodules formed very rapidly after entrainment (Bateman, 1999) and that the leaves were lost from the plant by trauma rather than by seasonal abscission. Contemporaneous degradative fungal hyphae observed within the mesophyll of these remarkable leaves (Fig. 20) are similarly mineralized and indicate a short period of slight decay prior to burial. Interestingly, Rao and Achuthan (1967) noted spherical fungal fruiting bodies (but no hyphae) in specimens of Ptilophyllum cutchense.
The topologies of recent morphological phylogenetic analyses clearly demonstrate the distant relationship between the Cycadales and the relatively derived Bennettitales. The Bennettitales have consistently been placed near the angiosperms plus the Gnetales (Crane, 1985b; Doyle & Donoghue, 1987), and the Cycadales occur as a more basally divergent group, below the conifers and sometimes below Glossopteris and the Caytoniales (Doyle & Donoghue, 1986, 1992; Nixon et al., 1994; Rothwell & Serbet, 1994). These analyses, however, rely unduly heavily on reproductive characters, and often "Nilssoniopteris or taeniopterid foliage" is included only as a single aggregate foliar character.
Foliar characters require particular attention when comparing fossil cycadophytes with their nearest living relatives in "high-level" phylogenetic analyses of seed plants. Of the foliar characters treated in such analyses, only a few have been scored differently for cycads and bennettites, and their behavior is ambiguous in different analyses. Supposed differences include the aforementioned stomatal structure and the pitting of secondary xylem tracheids. Doyle and Donoghue (1986), Loconte and Stevenson (1990), and Nixon et al. (1994) also included nodal anatomy, girdling leaf bases, and axillary buds as vegetative characters that differ between cycads and bennettites, but regrettably these cannot be scored for isolated leaves.
We believe that leaf characters have been underused in "high-level" phylogenetic analyses. This may reflect concerns about potentially greater ecophenotypic variation among leaves (such as the shade and sun leaves discussed above) and unwarranted fears of anatomical and morphological convergence (cf. Crane, 1985a; Thomas & Spicer, 1987; Stewart & Rothwell, 1993). Bateman and Simpson (1998) detected only slightly greater homoplasy in vegetative rather than reproductive organs across the plant kingdom but noted, in their specific case study of the fossil rhizomorphic lycopsids, that the leaves were unusually homoplastic. However, they also observed a positive correlation between the complexity of a plant organ and the strength of its phylogenetie signal. High phenotypic complexity (often in turn a correlate of multifunctionality) increased the probability of correctly identifying homologous characters a priori. Thus, the simple microphyllous leaves of the rhizomorphic lycopsids contrast radically with the morphologically and anatomically complex megaphyllous leaves of the cycadophytes, which play a much stronger role in light capture, gaseous exchange, and physical and chemical protection from herbivory (cf. Phillips & DiMichele, 1992; Jones, 1993). Moreover, the microphylls of rhizomorphic lycopsids were rapidly discarded during ontogeny, whereas extant cycad megaphylls are less ephemeral, usually being retained for several years prior to death.
More generally, there is no a priori reason for supposing that any suite of characters is more highly evolutionarily conserved than any other, and certainly no valid a priori argument that cycadophyte leaves offer weaker phylogenetic signals than do the corresponding reproductive structures. At present, the main impairments to employing leaves in phylogenetic analyses of cycadophytes are the relative paucity of comparative data and the tendency to perpetually focus on a few supposedly optimally diagnostic characters long used in cycadophyte taxonomy. Thus, we believe that cycadophyte leaves merit further anatomical study and ontogenetic observation, with the aim of genuinely optimizing their contribution to morphological phylogenies of these important clades. This is best achieved via the a posteriori congruence test of homology that is inherent in cladogram construction, preferably applied simultaneously to both extant and extinct species.
Sadly, the five years that followed submission of this manuscript have not, as far as we can determine, added further to the sparse data available on anatomically preserved cycadophyte leaves. The empirical effort in descriptive paleobotany has waned in response to changes in research emphasis driven by shifts in the transient enthusiasms of funding bodies.
Rather, the most notable event has been the sudden, molecularly inspired collapse of the previously popular (and still morphologically credible) "anthophyte" hypothesis of angiosperm origins, which identified a single origin for the flower by placing gnetaleans as sister group to angiosperms and bennettites as sister to both (e.g., Crane, 1985, 1988; Doyle & Donoghue, 1986, 1987, 1992; Nixon et al., 1994; Rothwell & Serbet, 1994). The collapse occurred when a series of molecular phylogenetic studies, based on varying combinations of the plastid genes rbcL, atpB, psaA, and psbB, the mitochondrial regions mtSSU, coxl, and atpA, and nuclear 18S rDNA, placed extant gnetaleans as a member of a modern coniferophyte clade rather than as sister to modern angiosperms (cf. Chaw et al., 2000; Soltis et al., 2002). These molecular phylogenetic studies place with approximately equal (and relatively weak) confidence the gnetaleans in two locations: either as sister to all other extant gymnosperms or as sister to Pinaceae and thus embedded deep within conifers--a placement that gains some additional support from recent studies of key developmental genes (well reviewed by Theissen et al., 2002). In this hypothesis Gingko is viewed as sister to the conifers, either alone or weakly attached to the cycads. Consequently, several previously compelling synapomorphies are being reinterpreted as potentially being remarkable parallelisms.
If the unravelling of the anthophyte hypothesis is justified, its implications are profound, especially when considered in the context of fossil as well as extant taxa. In particular, divergence between extant angiosperms and gymnosperms would be placed at ca. 300 Ma, thereby reinvigorating older "heretical" paleobotanical hypotheses that the angiosperms originated directly from one of several candidate groups of pteridosperms (cf. Thomas, 1925; Meeuse, 1961). Our comments in this article on the potential phylogenetic value of foliar anatomical characters should therefore be viewed in the context of the subsequent phylogenetic instability caused by molecular studies that sample many genie regions but (of necessity) few major lineages. Indeed, these more recent revelations require a further thorough revision of the phylogenetic placement of the bennettites relative to other groups, not least the cycads (e.g., W. Crepet, pers. comm., 2001). We cannot presently exclude the possibility that the remarkable foliar "parallelisms" between cycads and bennettites explored in this and other articles are in fact blatant synapomorphies, particular given the "blurring" observed by us of some of the traditional, and supposedly discrete, diagnostic characters of the two cycadophyte groups.
We thank Nicol Morton for providing specimen BE75/8, Paula Rudall for critically reading the manuscript, and Euan Clarkson, Mike Hall, Frieda Christie, and Ajit Subramanian for technical assistance. Discussions with Mike Krings, Ben LePage, and Walt Cressler were most helpful. We also thank the University of Edinburgh and the Royal Botanic Garden Edinburgh for funding the later stages of this project as an undergraduate thesis for BLD, and the Botanical Research Fund for supporting fieldwork by RMB.
VII. Literature Cited
Archangelsky, A., R. Andreis, S. Archangelsky & A. Artabe. 1995. Cuticular characters adapted to volcanic stress in a new Cretaceous cycad leaf from Patagonia, Argentina. Rev. Paleobot. Palynol. 89:213-233.
Barale, G. 1981. La paleoflore Jurassique du Jura Francais. Documents de Laboratoire de Geologie Lyon, 81.
Barbacka, M. & J. H. A. van Konijnenburg-van Cittert. 1998. Sun and shade leaves in two Jurassic species of pteridosperms. Rev. Paleobot. Palynol. 103: 209-221.
Bateman, R. M. 1999. Bulk geochemistry as a guide to provenance and diagenesis. Pp. 169-173 in T. P. Jones & N. P. Rowe (eds.), Fossil plants and spores: Modern techniques. Geological Society, London.
-- & N. Morton. 1994. New petrified Middle Jurassic floras from nearshore marine sediments at Bearreraig, Skye. Amer. J. Bot. 81:88 (abstract).
-- & N. J. Simpson. 1998. Comparing phylogenetic signals from reproductive and vegetative organs. Pp. 231-253 in S. J. Owens & P. J. Rudall (eds.), Reproductive biology in systematics, conservation and economic botany. Royal Botanic Gardens, Kew.
--, N. Morton & B. L. Dower. 2000. Early Middle Jurassic communities in northwest Scotland: Paleoecological and paleoctimatic significance. Pp. 501-511 in R. Hall, P. Smith & T. Poulton (eds.), GeoResearch Forum 6 (Proceedings of the Fifth International Symposium on the Jurassic System). TransTech Publications, Zurich.
Bold, H. C., C. J. Alexopoulos & T. Delevoryas. 1987. Morphology of plants and fungi. Harper & Row, New York.
Bose, M. N. 1953. Ptilophyllum amarjolense, sp. nov. from the Rajmahal Hills, Bihar. Proc. Nat. Inst. Sci., India 19: 605-612.
-- & M. L. Kasat. 1972a. The genus Ptilophyllum in India. Palaeobotanist 19: 115-143.
-- & --. 1972b. On a petrified specimen of Dictyozamites from the Rajmahal Hills, India. Palaeobotanist 19: 248-252.
Chaw, S.-M., C. L. Parkinson, Y. Cheng, T. M. Vincent & J. D. Palmer. 2000. Seed-plant phylogeny inferred from all three plant genomes: Monophyly of extant gymnosperms and origins of Gnetales from conifers. Proc. Natl. Acad. Sci. USA 97: 4086-4091.
Crane, P. R. 1985a. Phylogenetic analysis of seed plants and the origin of angiosperms. Ann. Missouri Bot. Gard. 72: 716-793.
--. 1985b. Morphology and relationships of the Bennettitales. Pp. 162-175 in R. A. Spicer & B. A. Thomas (eds.), Systematic and taxonomic approaches in palaeobotany. Systematics Association Special Publication 31. Oxford University Press, New York
--. 1988. Major clades and relationships in "higher" gymnosperms. Pp. 218-272 in C. B. Beck (ed.), Origin and evolution of gymnosperms. Columbia Univ. Press, New York.
Delevoryas, T. 1962. Morphology and evolution of fossil plants. Holt, Rinehart, and Winston, New York.
--. 1968. Some aspects of cycadeoid evolution. Bot. J. Linn. Soc. 61: 137-146.
--. 1982. Perspectives on the origin of cycads and cycadeoids. Rev. Paleobot. Palynol. 37:115-132.
Donoghue, M. J., J. A. Doyle, J. Gauthier, A. G. Kluge & T. M. Rowe. 1989. The importance of fossils in phylogeny reconstruction. Ann. Rev. Ecol. Syst. 20:431-460.
Doyle, J. A. & M. J. Donoghue. 1986. Seed plant phylogeny and the origin of angiosperms: An experimental cladistic approach. Bot. Rev. 52: 321-431.
-- & --. 1987. The importance of fossils in elucidating seed plant phylogeny and macroevolution. Rev. Paleobot. Palynol. 50: 63-95.
-- & --. 1992. Fossils and seed plant phylogeny reanalyzed. Brittonia 44: 89-106.
Esau, K. 1953. Plant anatomy. Wiley, New York.
Fahn, A. 1967. Plant anatomy. Pergamon Press, New York.
Florin, R. 1933. Studien tiber Cycadales des Mesozoicums nebst Erorterungen tiber die Spaltoffnungsapparate der Bennettitales. K. svenska. Vetensk Akad. Handl., Tredje Ser. 12: 1-134.
Gifford, E. M. & A. S. Foster. 1989. Morphology and evolution of vascular plants. Ed 3. W. H. Freeman, New York.
Gillespie, W. H. & H. W. Pfefferkorn. 1986. Taeniopterid lamina on Phasmatocycas megasporophylls (Cycadales) from the Lower Permian of Kansas, U.S.A. Rev. Paleobot. Palynol. 49:99-116.
Greguss, P. 1965. The relationships of Cycadales on the basis of their xylotomy, branching and leaf epidermis. Palaeobotanist 14: 94-101.
Harris, T. M. 1937. The fossil flora of Scoresby Sound, East Greenland. Medd, Gronland, Khobenhavn, 112: 1-114, pl. 1.
--. 1961. The fossil cycads. Palaeontology 4: 313-323.
--. 1964. The Yorkshire Jurassic flora, II. Caytoniales, Cycadales and Pteridospermales. British Museum (Natural History), London.
--. 1969. The Yorkshire Jurassic Flora, III. Bennettitales. British Museum (Natural History), London.
Jones, D. L. 1993. Cycads of the world. Smithsonian Institution Press, Washington, DC.
Jones, T. & N. P. Rowe (eds.). 1999. Fossil plants and spores: Modern techniques. Geological Society, London.
Joy, K. W., A. J. Willis & W. S. Lacey. 1956. A rapid cellulose peel technique in palaeobotany. Ann. Bot. 20: 635-637.
Kimura, T. & M. Tsujii. 1982. Early Jurassic plants of Japan. Part 4. Trans. Proc. Palaeont. Soc. Japan, N.S. 125: 259-276.
Krassilov, V. A. 1976. Bennettitalean stomata. Palaeobotanist 25:179-184.
Kvacek, J. 1995. Cycadales and Bennettitales leaf compressions of the Bohemian Cenomanian, Central Europe. Rev. Paleobot. Palynol. 84: 389-412.
Loconte, H. & D. W. Stevenson. 1990. Cladistics of the Spermatophyta. Brittonia 42:197-211.
Magallon, S. & M. J. Sanderson. 2002. Relationships among seed plants inferred from highly conserved genes: Sorting conflicting phylogenetic signals among ancient lineages. Amer. J. Bot. 89: 1991-2006.
Meeuse, A. D. J. 1961. The Pentoxylales and the origin of the monocotyledons. Proc. Koninklijke Needed. Akad. Wentsensch. 82: 343-369.
Morton, N. 1984. Aalenian-Bajocian boundary at Bearreraig, Isle of Skye, Scotland. Pp. 341-352 in J. Michelsen and A. Zeiss (eds.), International Symposium on Jurassic Stratigraphy. Geological Survey of Denmark, Copenhagen.
Nixon, K. C., W. L. Crepet, D. Stevenson & E. M. Friis. 1994. A reevaluation of seed plant phylogeny. Ann. Missouri Bot. Gard. 81: 484-533.
Norstog, K. & T. Nicholls. 1997. The biology of the cycads. Cornell Univ. Press, Ithaca, NY.
Oldham, T. C. B. 1976. Flora from the Wealden plant-debris beds of England. Palaeontology 19: 1-437.
Pant, D. D. 1987. The fossil history and phylogeny of the Cycadales. Geophytology, 17: 125-162.
--. 1990. On the genus Glandulataenia from the Triassic of Nidpuri, India. Mem. New York Bot. Gard. 57: 186-199.
-- & D. D. Nautiyal. 1963. Cuticle and epidermis of recent Cycadales: Leaves sporangia and seeds. Senck. Biol. 44:257-348.
Person, C. P. & T. Delevoryas. 1982. The Middle Jurassic flora of Oaxaca, Mexico. Palacontographica Abt. B 180: 82-119.
Phillips, T. L. & W. A. DiMichele. 1992. Comparative ecology and life-history biology of arborescent lycopsids in Late Carboniferous swamps of Euramerica. Ann. Missouri Bot. Gard. 79: 560-588.
Poole, J. P. 1923. Comparative anatomy of leaves of cycads, with reference to the cycadofilicales. Bot. Gaz. 76: 203-214.
Rao, A. R. & V. Achuthan. 1967. Further contribution to our knowledge of Ptilophyllum. Palaeobotanist 16: 249-257.
Rothwell, G. W. & R. Serbet. 1994. Lignophyte phylogeny and the evolution of spermatophytes: A numerical cladistic analysis. Syst. Bot. 19: 443-482.
Seward, A. C. 1911. The Jurassic flora of Sutherland. Trans. R. Soc. Edinburgh 47:643-709.
Sharma, B. D. 1967. Investigations on the Jurassic flora of Rajmahal Hills, India. Palaeontographica Abt. B. 120: 139-150.
Soltis, D. E., P. S. Soltis & M. J. Zanis. 2002. Phylogeny of seed plants based on evidence from eight genes. Amer. J. Bot. 89: 1670-1681.
Stevenson, D. W. 1981. Observations on ptyxis, phenology, and trichomes in the Cycadales and their systematic implications. Amer. J. Bot. 68:1104-1114.
--. 1990. Morpholgy and systematics of the Cycadales. Mere. New York Bot. Gard. 57:8-55.
--, K. J. Norstog & D. V. Molsen. 1996. Midribs of cycad pinnae. Brittonia 48: 67-74.
Stewart, W. N. & G. W. Rothwell. 1993. Paleobotany and the evolution of plants. Cambridge Univ. Press, Cambridge.
Stopes, M. C. 1910. The internal anatomy of "Nilssonia orientalis." Ann. Bot. 14:389-393.
Suhk-Dev & Zeba-Bano. 1975. Three species of Ptilophyllum from Basna, Madhya Pradesh. Palaeobotanist 24: 161-189.
Taylor, T. N. & E. L. Taylor. 1993. The biology and evolution of fossil plants. Prentice Hall, Englewood Cliffs, NJ.
Theissen, G., A. Becker, K.-U. Winter, T. Munster, C. Kirchner & H. Saedler. 2002. How the land plants learned their floral ABCs: The role of MADS-box genes in the evolutionary origin of flowers. Pp. 173-205 in Q. C. B. Cronk, R. M. Bateman & J. A. Hawkins (eds.), Developmental genetics and plant evolution. Taylor & Francis, London.
Thomas B. A. & R. A. Spicer. 1987. The evolution and palaeobiology of land plants. Croom Helm, London and Sydney.
Thomas, H. H. 1925. The Caytoniales: A new group of angiospermous plants from the Jurassic rocks of Yorkshire. Phil. Trans. R. Soc. Lond. B 213:299-363.
Tidwell, W. D., J.-H. Kim & T. Kimura. 1987. Mid-Mesozoic leaves from near Ida Bay, southern Tasmania, Australia. Pap. Proc. R. Soc. Tasmania 121: 159-170.
Van Konijnenburg-van Cittert, J. H. A. & J. van der Burgh. 1989. The flora from the Kimmeridgian (Upper Jurassic) of Culgower, Sutherland, Scotland. Rev. Paleobot. Palynol. 61: l-51.
Watson, J. & C. A. Sincock. 1992. Bennettitales of the English Wealden. Monograph of the Palaeontological Society, London.
BEATRICE L. DOWER
Land Use Consultants
Glasgow, G12 8JJ, Scotland
RICHARD M. BATEMAN
Department of Botany
The Natural History Museum
London, SW7 5BD, England
DENNIS WM. STEVENSON
New York Botanical Garden
Bronx, NY 10458, U.S.A.
|Printer friendly Cite/link Email Feedback|
|Author:||Dower, Beatrice L.; Bateman, Richard M.; Stevenson, Dennis Wm.|
|Publication:||The Botanical Review|
|Date:||Apr 1, 2004|
|Previous Article:||New discoveries of cycads and advancement of conservation of cycads in China.|
|Next Article:||Two new petrified cycad stems, Brunoa gen. nov. and Worsdellia gen. nov., from the Cretaceous of patagonia (Bajo de Santa Rosa, Rio Negro Province),...|