Abstract
A group of equisetalean shoots with distinctive small leaves occurs widely in the upper Bashkirian and lower Moscovian coal-bearing deposits of Euramerica. They have often been named Asterophyllites grandis and Asterophyllites charaeformis in the past, but the use of these names is illegitimate for these species. In this study, these shoots have been assigned to five fossil species: Asterophyllites delicatulus, Asterophyllites parvulus, Asterophyllites gracilis, Asterophyllites taylorianum, and Asterophyllites lubnensis. Two of the species are associated with Calamostachys strobili, while the other three are associated with Palaeostachya strobili. The distribution of the shoot species partly reflects differences in elevation and substrate conditions of where the parent plants grew.
INTRODUCTION
Major changes occurred to terrestrial biotas during Pennsylvanian (late Carboniferous) times, including significant taxonomic diversifications of plants (Cleal & Cascales-Miñana, 2014, 2021; Cleal et al., 2012), insects (Condamine et al., 2016; Prokop et al., 2023) and tetrapod vertebrates (Sahney et al., 2010). These diversifications were mainly centred on paralic palaeotropical habitats that had formed through falling sea-levels caused by the Late Palaeozoic Ice Age (e.g., Fielding et al., 2008; Montañez & Poulsen, 2013). These palaeotropical lowlands were also subjected to landscape disruption caused by the orogenic growth of the nearby Central Pangaean Mountains following the suturing of the Gondwana and Laurasia continental plates—the Variscan Orogeny (e.g., Gayer et al., 1993; Schulmann et al., 2014).
These newly exposed lowlands witnessed the development of an extensive wetland biome, often referred to as the coal swamps, which at its maximum extent spanned approximately 10⁶ km² between present-day central USA and the Donets in Ukraine (Cleal & Thomas, 2005). There has been considerable debate about the extent to which the coal swamps interacted with the changes in climate and landscapes (e.g., Cleal & Thomas, 1999; Cleal et al., 2009, 2011; Gastaldo et al., 1996; Phillips & DiMichele, 1992). Key to understanding this relationship is the floristic evolution of the coal swamp vegetation, which in turn depends on a proper understanding of the taxonomy of the plant fossils.
The Carboniferous coal swamp vegetation was dominated by five major groups of plants: the arborescent lycopsids, the equisetaleans (calamitacean and sphenophyllacean), ferns (marattialean and non-marattialean), ‘pteridosperms’ (mainly lyginopteridalean and medullosalean) and pinopsids (cordaitalean) (Cleal, 2021a, 2021b). This article will focus on a distinctive group of calamitaceans characterised by whorls of unusually small leaves that have previously been referred to various species, including Asterophyllites grandis (Sternberg) Lindley and Hutton and Asterophyllites charaeformis (Sternberg) Göppert. Although widespread, these leaf fossils have been the subject of considerable taxonomic confusion. As a first step in resolving this confusion, this article will, for the first time, photographically document some of the key nomenclatural type specimens. These types do not, of course, define the species; types define the species names, not the species themselves. However, they provide important empirical fixed points around which the species concepts can be constructed. As a preliminary attempt to revise these species, a review has been made of the published records of the species names represented by these types, based on listings in the Fossilium catalogus (Jongmans, 1914; Jongmans & Dijkstra, 1969; van Amerom, 1995) and the Smithsonian Institution’s Biodiversity Heritage Library (Gwinn & Rinaldo, 2009—www.bioodiversitylibrary.org). This will inevitably be constrained by what is recorded in the literature and will no doubt need to be modified through further study of museum collections and fieldwork. Nevertheless, it will hopefully provide a starting point for the development of an improved species classification for these fossils.
The stratigraphical distribution of the taxa will be mainly given in the context of the Heerlen Regional Chronostratigraphy, as this is still widely used for the deposits of the Carboniferous coal swamps (Wagner, 1974; Wagner & Winkler Prins, 2016), although, where possible, reference will also be made to the IUGS Global Chronostratigraphy (Wagner, 2017). The macrofloral biostratigraphy is as summarised by Opluštil et al. (2022).
TAXONOMIC FRAMEWORK
Many non-palaeobotanists (and even some palaeobotanists) lack a clear understanding of how plant fossils are classified and the resulting taxa are named. Since the pioneering work of Sternberg (1820, 1821) and Brongniart (1822, 1828), it has been recognised that understanding past vegetation is best achieved by naming and classifying plant fossils as intrinsic entities, rather than by trying to classify the original, now-extinct plants that produced the fossils. This did not mean that the resulting palaeobotanical taxonomy does not reflect the taxonomy of the original plants; rather, it was an attempt to produce a scheme that reflected both the systematics of the original plants (as far as possible) and the constraints introduced by taphonomy (for a historical discussion see Cleal & Thomas, 2010, 2021). The currently accepted approach as outlined in the ICN (International Code of Nomenclature—Turland et al., 2018, Art. 1.2) classifies plant fossils as fossil-taxa (mainly species and genera, but in principle any rank) whose names are based on a type that is a fossil specimen, and comprise ‘… the remains of one or more parts of the parent organism, or one or more of their life-history stages, in one or more preservational states’.
In the present study, separate sets of fossil genera and fossil species are recognised for the foliage and reproductive structures (strobili) of these equisetopsids. Where possible, the two sets of taxa have been correlated so as better to reflect the systematic relationships of the parent plants. Eventually, it may prove possible to fully integrate the two taxonomies to approach a whole-organism taxonomy, but this will currently be premature; we still lack sufficient evidence regarding the correlation between foliage and strobili in these Palaeozoic equisetopsids. An example of the problems that can arise from prematurely integrating taxa for different plant parts is the case of a whole-plant reconstruction (or, at least, a combined taxon for foliage, strobili, and stems) of a Permian equisetopsid, named by Liu et al. (2021) as Palaeostachya. This widely used fossil genus was formerly restricted to strobili. As pointed out by Cleal and Thomas (2021), such a compound fossil genus brings little real benefit, other than providing a label to affix to an illustration of the reconstructed plant in textbooks, and will produce substantial disruption to the traditional taxonomy of the fossil strobili, which still has considerable practical value to palaeobotanists. At least for the time being, retaining separate fossil taxa for the strobili and foliage of these plants is preferable.
There are also fossil taxa for the stems, such as the fossil genus Calamites, described by Sternberg in 1820. The Calamites fossil species are primarily defined by the branch scars on the main stems (Crookall, 1969; Jongmans, 1911; Kidston & Jongmans, 1917; Weiss, 1884), and these scars tend to reflect the overall architecture of the plants (Thomas, 2014). At present, however, there are few cases where the stem taxa have been correlated with the foliage and fructification taxa. The stem taxa will not be discussed further in this context.
MATERIALS
This study originated from work on the types of several Asterophyllites species, from the middle Westphalian of the Central and Western Bohemia Basin, Czech Republic (A. grandis, A. charaeformis, Asterophyllites delicatulus), in the collections of the National Museum, Prague; from South Wales, United Kingdom (Asterophyllites taylorianum) in the collections of the Natural History Museum, London; from Georgia, USA (Asterophyllites gracilis) in the National Museum of Natural History, Smithsonian Institution; and from the Ruhr Coalfield, Germany (Asterophyllites roehlii) in the collections of GeoSphere Austria, Vienna.
SYSTEMATIC PALAEONTOLOGY
Although the species and genera dealt with in this analysis are all fossil taxa, for brevity, they will be referred to here as species and genera. Only the foliage taxa will be formally dealt with, although, where relevant, the taxa of the associated strobili will also be mentioned. The morphological features used to differentiate the species are summarised in Figure 1.
Main morphological features used to differentiate the species of small-leafed Asterophyllites delt with in this article.
Order Equisetales de Candole ex Berchtold & Presl, 1820
Family Calamitaceae Unger, 1840
Remarks. This plant family consists mainly of the late Carboniferous arborescent equisetaleans in which the strobili have alternating whorls of sporangiophores and sterile bracts. The family has sometimes been named Calamostachyaceae Meyen, 1978, based on the mistaken belief that a family name must be rooted in the name of a fructification genus. However, this is contrary to accepted nomenclatural practice as outlined in the International Code of Nomenclature (Turland et al., 2018). The earliest available family name whose type can be included within this family is Calamitaceae, and so the latter name must take precedence. Good (1975) questioned whether these Palaeozoic equisetaleans represented a family separate from the extant Equisetaceae, Richard ex Lamarck and de Candolle, 1805; however, a phylogenetic analysis by Elgorriaga et al. (2018) has resolved them as quite distinct.
Fossil-genus Asterophyllites Brongniart 1828, nom. cons.
Type species. Asterophyllites equisetiformis (Schlotheim ex Sternberg) Brongniart, 1828 (≡ Bornia equisetiformis Schlotheim ex Sternberg, 1825).
Remarks. The species discussed here were initially assigned to the genus Bechera Sternberg, 1825. However, that name is illegitimate for two reasons: Sternberg had also included the type of an earlier published genus name of charophycean alga Gyrogonites Lamarck, 1801, and so his Bechera was nomenclaturally superfluous (Doweld, 2017); and it was also a later homonym of the angiosperm genus name Bechera Turton, 1802. The presently discussed species are now usually assigned to Asterophyllites, although that genus name has also been the subject of some confusion (Cleal & Thomas, 2018). In summary, the name was first used by Brongniart (1822) for a rather different type of foliage with flatter whorls of leaves that are now assigned to Annularia Sternberg. However, following a proposal by Vogellehner (1967), Asterophyllites has been conserved with A. equisetiformis as type, making it compatible with how the name has been normally used, for calamitacean foliage with cup-shaped whorls of slender, linear leaves (Abbott, 1958; Cleal & Shute, 2016; Cleal & Thomas, 1994).
A. delicatulus (Sternberg) Brongniart, emend. nov.
Plates 1 and 2
1825 Bechera delicatula Sternberg, p. xxxi, pl. 49, Figure 2.
Asterophyllites delicatulus (Sternberg) Brongniart; 1, 2. Holotype of A. delicatulus, National Museum Prague, Specimen E 44; 3, 4. Holotype of Asterophyllites charaeformis (Sternberg) Göppert; National Museum Prague, Specimen E 50; 5–7. Calamostachys strobili associated with A. delicatulus foliage; National Museum Prague, Specimen E 4735. All specimens from Duckmantian (upper Bashkirian) Whetstone Horizon, Svinna. Scale bars = 20 mm (Figures 1, 3, and 6) or 5 mm (Figures 2, 4, 5, and 7).
Asterophyllites delicatulus (Sternberg) Brongniart; Svinná, Kladno–Rakovnik Basin, Czech Republic; Duckmantian (upper Bashkirian). Holotype of Asterophyllites grandis (Sternberg) Lindley and Hutton; National Museum Prague, Specimen E 4375. Scale bars = 20 mm (Figure 1), 5 mm (Figure 2).
1825 Bechera grandis Sternberg nom. illegit., p. xxx, pl. 49, Figure 1.
1825 Bechera charaeformis Sternberg, p. xxx, pl. 55, Figures 3 and 5.
1828 A. delicatulus (Sternberg) Brongniart, p. 159.
1844 A. charaeformis (Sternberg) Göppert, p. 198.
1854 A. grandis (Sternberg) Geinitz nom. illegit., p. 35 (non pl. 14, Figure 15—to A. grandis Lindley & Hutton, 1832).
Holotype. National Museum (Prague) Specimen E 44 (Plate 1, Figures 2 and 3). Provenance Svinná mines, near Radnice, Plzeň region, Czech Republic; Brousek ash layer (‘Whetstone Horizon’) between the lower and upper Radnice coals (upper Duckmantian Substage), Radnice Member, Kladno Formation (Opluštil et al., 2007).
Emended diagnosis. Foliage adpressions of ultimate shoots with slender stems ca. 1 mm wide with fine longitudinal ribs. Isophyllous whorls of four slender leaves borne at 3–4 mm intervals along most of the shoot, reducing to 1.5 mm in the distal part of the shoot, attached typically at 90°. Leaves on ultimate shoots are typically up to 5–7 mm long, reducing to ca. 1.5 mm in the distal part of the shoot. Leaves are needle-like, typically gently arched so that distal parts are at 30°–45° to the supporting stem, becoming more arched in the distal part of the shoot, so the end is near parallel to the stem.
Description of B. delicatula holotype (Plate 1, Figures 1 and 2). Three fragments of leafy shoots, two of which are penultimate shoots. Penultimate stems are slender, 1 mm wide, with longitudinal ribs. Ultimate shoots spaced at 4.5‒6 mm intervals, the longest shoot 18 mm long; whorls of leaves at 2‒3 mm intervals. Leaves up to 5 mm long, attached to the stem at right angles, extend out in a straight line for about half to a third of their length, then arch broadly so that at their distal end they are c. 45° to stem.
Description of B. grandis holotype (Plate 2). Stem preserved for 175 mm, 11 mm wide, with 0.5 mm wide longitudinal ribs. Nodes spaced at 29‒36 mm intervals; stem thickened at nodes by 1‒2 mm. Two penultimate leafy shoots were preserved near the nodes, although not in attachment. Penultimate shoots preserved for lengths of 75 mm and 95 mm, but neither is complete. Stems of penultimate shoots 2.5‒3 mm wide, with minor surface detail preserved. Nodes are spaced at 13 mm in the proximal part of the shoot, reducing to 11 mm in more distal positions. Ultimate leafy shoots are attached to whorls at 50°–80°; the number of ultimate shoots per node is impossible to determine. Most extended ultimate shoot preserved for a length of 33 mm, but still incomplete, ca. 1 mm wide. Leaves attached at intervals of 3‒4 mm; the number of leaves per whorl is challenging to determine, but in one part, there is a suggestion that there were four. Leaves up to 7 mm long, reducing to 4 mm in the distal parts of shoots, linear with an acute apex. Largest leaves attached at about right angles, arch broadly so the end of the leaf is at c. 30°‒45° to the stem.
Description of B. charaeformis holotype (Plate 1, Figures 2 and 3). A tapered terminal fragment of a shoot with whorls of very slender leaves up to 2 mm long. Leaves attached at about right angles to the stem, then arch broadly so lie parallel to the stem in the distal part. Leaves spaced at intervals of c. 1.5 in proximal part of stem. On the reverse of the specimen is another similar-sized shoot, but parallel-sided, which suggests that there were just four leaves per whorl.
Associated strobili. Another specimen preserved in a similar matrix to that of the A. delicatulus type has small associated strobili (Plate 1, Figures 5–7). Although not in direct attachment, the intimate association of the strobili with this foliage and their general similarity to the form of the terminal shoots make it almost certain that they were produced by the same plants. The strobili are 12–15 mm long, 4–6 mm wide, parallel-sided in their proximal part but markedly tapered in their distal part. Slender bracts, probably arranged in whorls, ca. 5 mm long initially attached at about right angles in the proximal part of the strobilus and then arch along the whole length so their distal part is parallel to the strobilus axis; more distal bracts are straighter and start at a narrower angle to strobilus. A whorl of what appear to be large sporangiophores occurs between bract whorls.
Similar-shaped (partly tapered) but more elongate strobili were described by Němejc (1953, pl. 9, Figures 1–5) from the Kladno mines in the same basin (refigured here as Plate 3). These strobili are 24–42 mm long, 4–5 mm wide; larger strobili parallel-sided, shorter strobili somewhat tapered along length; bract whorls spaced at 2 mm intervals; bracts 3 mm long, strongly arced so the distal part is parallel to the strobilus axis. Spores similar to Calamospora breviradiata Kosanke have been described from these strobili (Bek, 2021; Bek & Opluštil, 1998; Bek & Votočková Frojdová, 2024). These strobili were found closely associated with abundant A. delicatulus foliage. Němejc (1953) named these strobili as Calamostachys grandis Jongmans, but that name was not valid: Jongmans (1911, p. 312) was simply referring to ‘Calamostachys von A. grandis Sternberg’ and not establishing a new species; moreover, the strobili described by Jongmans were larger and more consistently cylindrical than these Kladno specimens (up to 50 mm long, 7 mm wide).
Calamostachys strobili associated with Asterophyllites delicatulus shoots; Kladno, Czech Republic; Duckmantian (upper Bashkirian). 1, 2. National Museum Prague, Specimen E 1187; 3. National Museum Prague, Specimen E 1184. Scale bars 10 mm (Figure 1), 5 mm (Figures 2 and 3).
Remarks. The earliest published epithets that could be used for this foliage species were based on B. grandis, B. delicatula and B. charaeformis. The types of all three species names originated from the upper Duckmantian Whetstone Horizon at Svinná and have very similar-shaped leaves.
The B. grandis type is by far the most complete and representative of this species, and under normal circumstances it would be sensible to use that epithet combined with Asterophyllites, as had been proposed by Geinitz (1854). However, the combination A. grandis had already been published by Lindley and Hutton (1832, p. 57, pl. 17) based on an entirely different holotype with much larger leaves; Lindley & Hutton were clearly not basing their A. grandis on B. grandis Sternberg, which they referred to and described quite separately (Lindley & Hutton, 1832, p. 61, pl. 19, Figure 1). A. grandis (Sternberg) Geinitz was therefore an illegitimate later homonym of A. grandis (Lindley & Hutton, 1832) and so must be rejected.
The type of B. delicatula is less complete than that of B. grandis, but nevertheless has whorls of very similar slender, gently curved leaves. There can be little doubt that they belong to the same species (compare Plate 1, Figure 2 with Plate 2). Sternberg somewhat enigmatically referred to B. delicatula as a Species Dubiae, but from his description, it seems that this was mainly because he was uncertain as to whether it belonged to Bechera, rather than being uncertain that it was a legitimate species. The epithet may therefore be regarded as validly published and was later published as A. delicatulus by Brongniart (1828).
Sternberg’s epithet charaeformis has been widely used in the literature in the combination A. charaeformis for a species with relatively smaller, less needle-like and more curved leaves, and where there are often more leaves in each whorl. However, this is due to a misinterpretation of Sternberg’s types. Sternberg wrote that these types originated from a locality he named as Walsch in Bohemia, but this was an error, as the label of the sole existing type gives its origin as Svinná. The matrix is the same as for the other specimens from that locality. Although the leaves of the one existing type (Plate 1, Figures 2 and 3) are smaller and more curved than the types of B. grandis and B. delicatula, this is clearly because they are from the distal part of a shoot. Since the B. charaeformis type occurs in close association in the Svinná flora with the B. grandis and B. delicatula types, and there are no examples there of more proximal shoots showing the smaller, more curved leaves that most other authors have named as A. charaeformis, there can be little doubt that type B. charaeformis belongs to the same species as the B. delicatula type, and so the epithets are synonyms. The species that has been widely misnamed as A. charaeformis is here renamed Asterophyllites parvulus (see below).
Of the three synonyms used for this fossil species by Sternberg, it is therefore evident that one was illegitimate (B. grandis) and another was atypical and has been widely misused for a quite different species (B. charaeformis). It is consequently proposed that the species should be named using the third epithet delicatula, that is, A. delicatulus.
The type of Bechera diffusa Sternberg, 1825 (Sternberg, 1821, pl. 19, Figure 3) is also almost certainly a small-leaved Asterophyllites, this time from the Whetstone Horizon at Radnitz. However, the type is now lost (Kvaček et al., 2021), and the preservation is evidently poor; the species name must currently be regarded as unusable.
Distribution. This species has been extensively reported in the literature, usually under the name A. grandis. However, these records are mainly unillustrated, and where there are illustrations, they do not show the whorls of four, gently curved, needle-like leaves that characterise this species. The only records that agree with this species as interpreted here are from the Whetstone Horizon (lower Paripteris linguefolia Zone) of the Kladno–Rakovnik Basin, from Svinná (the nomenclatural types), Rakovnik and Žebnice (Feistmantel, 1874, pl. 12, Figure 4; pl. 13, Figure 3) and Otvovice (Hofmann & Ryba, 1899, pl. 2, Figure 7).
A. parvulus Dawson emend. nov.
Plate 4
Asterophyllites parvulus Dawson. 1, 2. No 13 Callender Pit, Falkirk Glen, Stirlingshire, UK; Langsettian (upper Bashkirian); 3, 4. Standard Colliery, Gynshir, Glamorgan, UK; Bolsovian lower Pennant Formation (lower Moscovian). British Geological Survey, Keyworth, Kidston Collection; Kidst.5103 (Figures 1 and 2); Kidst.4043 (Figures 3 and 4). Reproduced with permission, British Geological Survey. Scale bars = 20 mm (Figures 1 and 3) or 5 mm (Figures 2 and 4).
1840 B. charaeformis Sternberg; Morris in Prestwich (non-Sternberg), pl. 38, Figure 2.
1844 A. charaeformis (Sternberg) Göppert (non-Sternberg), p. 198.
1861 Asterophyllites parvula Dawson, p. 168, Figure 6.
1906 A. parvulus Dawson; Matthew, p. 122, pl. 6, Figures 1–3.
Neotype. The original types described by Dawson (1861) consisted of small fragments of shoots exhibiting whorls of small, curved, rigid leaves, found at Hartt’s Bed No. 2, Lancaster, and the Fern Ledges, New Brunswick, Canada. However, these specimens now appear lost. Subsequently, Matthew (1906) figured more complete material from the same locality, the counterpart of one of which is designated here as neotype (photographically refigured by Álvarez-Vázquez & Wagner, 2017, Figure 13): Specimen NMMG 3435, New Brunswick Museum, Saint John, Canada. These deposits are probably upper Langsettian (Falcon-Lang & Miller, 2007) and so of the Calymmotheca hoeninghausii Zone.
Emended diagnosis. Foliage adpressions of ultimate shoots with slender stems typically <1 mm wide with fine longitudinal ribs. Ultimate shoots with cup-shaped, isophyllous whorls of typically 5–6 small, rigid leaves borne at 1.0–2.0 mm intervals, whorls of up to 10 leaves on lower order shoots. Leaves slender, falcate with acute apex, typically 2–3 mm long; leaves attached at 90° to the stem and curved along most of their length so that the distal parts lie nearly parallel to the supporting stem and extend to near the next whorl.
Description of neotype (see Álvarez-Vázquez & Wagner, 2017, Figure 13). Fragment of penultimate shoot 35 mm long, ca. 1 mm wide, with fine longitudinal ribs, and internodes at 2–3 mm intervals. Each node has 8–10 slender, tapered, slightly curved leaves up to 3 mm long, and at least two ultimate shoots. Ultimate shoots with slender stems < 1 mm wide with nodes at 2–3 mm intervals, each with a whorl of 5–6 leaves; each leaf up to 2.5 mm long, slender, falcate, strongly curved so the distal end is nearly parallel to the supporting stem.
Associated strobili (Bell, 1944). Slender, cylindrical strobili 5 mm wide and preserved up to 20 mm long. The strobilus axis bears whorls of 2–3 mm long bracts at 2–3 mm intervals; bracts attached at right angles to the axis, initially straight, then curve rapidly at the midpoint so the distal part is nearly parallel to the axis. Sporangiophores are attached to the strobilus axis about midway between the bract whorls. Weiss (1884) named these strobili Paracalamostachys williamsonianum Weiss.
Remarks. This is the species that has been widely misnamed in the literature as A. charaeformis based on records by Morris (in Prestwich, 1840) and Göppert (1844), which resulted from a misinterpretation of the type of that name. As shown earlier, the A. charaeformis type has fewer leaves per whorl, and the leaves are normally thinner and less arched than in A. charaeformis auct. and is in fact a distal fragment of an A. delicatulus shoot. An alternative name is therefore needed, the best candidate being A. parvulus Dawson, whose types have leaf-whorls indistinguishable from A. charaeformis auct. (see also Stopes, 1914).
Although this species is widespread, it is usually recorded as fragments of leafy shoots in the adpression record. It has also been reported in lower Langsettian coal balls of northern England, described by Thomas (1911) as ‘Type II (C. grandis Sternb.)’, with whorls of up to 10 small, 3–4 mm long, falcate leaves. They are associated with small, cylindrical strobili (Thomas, 1909) that are often referred to as Calamostachys binneyana Williamson. These shoots are quite different from A. grandis auct. (i.e., A. delicatulus), which usually has only four very slender leaves per whorl and is much more similar to A. parvulus shoots. Carruthers (1911) described several other Asterophyllites fragments with small leaves but slightly different anatomies, but which he suggested may represent variation within a biological species.
Distribution. A. parvulus appears to be widespread in the lowland adpression floras of palaeotropical Euramerica, mainly as relatively fragmentary remains. There are numerous records of A. charaeformis auct. from the Appalachians (Blake et al., 2002), Canadian Maritimes, notably Nova Scotia and New Brunswick (Álvarez-Vázquez & Wagner, 2017), the United Kingdom (Cleal, 2005, 2007; Crookall, 1969; Jongmans, 1940), the Ruhr, Germany (Josten, 1991), Upper Silesia (Kotasowa & Migier in Zdanowski & Żakowa, 1995) and the Dobrudzha in Bulgaria (Tenchov, 1987). They seem to be most abundant in the Lonchopteris rugosa Zone (Duckmantian Substage), but also reportedly range into the C. hoeninghausii Zone (Langsettian Substage) and Paripteris lingueofolia Zone (Bolsovian Substage). There are also records of foliage and the associated strobili (C. binneyana) from the Langsettian coal balls of the United Kingdom, Belgium, and Germany (Galtier, 1997). There are records of A. charaeformis auct. from the intra-montane Svoge Basin of Bulgaria, associated with Calamostachys strobili (Tenchov, 1977, 1987); also, from Central and Western Bohemia, Czech Republic (Opluštil et al., 2020), although in the latter area, they are not part of the autochthonous peat substrate vegetation. There are no unequivocal records from the intra-montane Saar-Lorraine basin.
A. gracilis Lesquereux emend. nov.
Plates 5 and 6
Asterophyllites gracilis Lesquereux; Dade County, Georgia, USA; Pottsville Formation (upper Bashkirian). 1, 2. Neotype, Smithsonian Institution specimen USNM PAL 18117 (photo by Nathan Kitto); 3. Smithsonian Institution specimen USNM PAL 18119 (photo by Frederic Cochard). Scale bars 10 mm (Figures 1 and 3), 5 mm (Figure 2). Images courtesy of the Smithsonian Institution.
Asterophyllites roehlii Stur; spoil tip of Ritterburg Mine, near Bochum, Germany. Lectotype, collections of GeoSphere Austria, Vienna, Stur Collection Specimen 1887 02 47a. 1. Whole specimen, scale bar = 20 mm; 2. Copy of illustration in Stur (1887, pl. 14, Figure 11); 2, 3. Close-ups of shoots with contrast enhanced to show details of leaf whorls. Scale bars = 5 mm. Copyright Festi & Di Franco, GeoSphere Austria.
1860 A. gracilis Lesquereux, p. 310, pl. 2, Figure 4.
1879 A. gracilis Lesquereux, pl. 2, Figures 4 and 5.
1880 A. gracilis Lesquereux, p. 42.
1884 A. gracilis Lesquereux, pl. 93, Figures 3–7.
1887 A. roehlii Stur, p. 209, pl. 14, Figures 11–13.
1887 Asterophyllites paleaceus Stur nom. nudum, p. 71.
(?)1888 Asterophyllites lycopodioides Zeiller, p. 380 (types Zeiller, 1886, pl. 59, Figures 1 and 2).
Neotype. The syntypes figured by Lesquereux (1860) originated from the Pottsville Group of Arkansas, USA. These specimens, which were reportedly very fragmentary, now appear lost. Lesquereux (1879, 1880, 1884) later documented other specimens under this name from similar-aged strata in Alabama and Georgia, USA. Of these specimens, USNM P.18117 in the collections of the USA National Museum of Natural History, Smithsonian Institution, Washington, DC, USA, is designated as neotype (Pl. 5, Figures 1 and 2). The specimen originated from the Pottsville Formation of Dade County, Georgia (USA). The specimen was previously illustrated as a photograph by Abbott (1958, p. 48, Figure 85).
Emended diagnosis. Foliage adpressions of ultimate shoots with slender stems ca. 1 mm wide with fine longitudinal ribs. Isophyllous whorls of four slender leaves borne at 1–4 mm intervals along the shoot, attached typically at 90°. Leaves on ultimate shoots are typically 2–4 mm long. Leaves are needle-like, typically broadly arched so that the distal parts are more or less parallel to the supporting stem.
Description of A. gracilis neotype (Plate 5, Figures 1 and 2). Leafy shoot with at least three orders of branching. Stems of penultimate shoots 1–2 mm wide, with fine longitudinal ribs. Ultimate shoots are borne obliquely (30°–40°) at intervals of 8–10 mm, in a planar arrangement. Stems of ultimate shoots 0.5–1 mm wide, up to at least 30 mm long, parallel-sided for most of their length. Whorls of leaves borne at 1.0–1.5 mm intervals on ultimate stems; leaves 2–3 mm long, needle-like, curving along most of their length, the distal part is parallel to the stem.
Description of A. roehlii types (Plate 6). Fragments of leafy shoot ca. 100 mm long, preserved on dark matrix. Penultimate stem 3–4 mm wide, longitudinally-ribbed, with ultimate stems at 20 mm intervals; penultimate stem slightly swollen where ultimate shoots attached. Ultimate shoots up to 30 mm long with whorls of four needle-like leaves at 1.0–1.5 mm intervals; leaves 2–3 mm long, attached at near right angles, arch broadly so the distal end is near right angles to the supporting stem.
Associated strobili (Plate 5, Figure 3). Small, cylindrical to somewhat tapered strobili, 8 mm wide and preserved up to 30 mm long. The strobilus axis bears whorls of 2–4 mm long bracts at 2–3 mm intervals; bracts attached at right angles to the axis, initially straight, then curve rapidly at the midpoint so the distal part is near parallel to the axis. Sporangiophores are attached to the strobilus in the axil of the bract.
Remarks. This species has often been misidentified as A. charaeformis auct. (i.e., A parvulus) (e.g., Álvarez-Vázquez & Wagner, 2017; Crookall, 1969) but differs from that species in having slenderer, needle-like leaves and being associated with Palaeostachya strobili. It also differs from A. delicatulus, A. taylorianum and Asterophyllites lubnensis in having smaller leaves (Figure 1).
White (1900) suggested that A. gracilis Lesquereux was an illegitimate name as it was a later homonym of a different species, A. gracilis (Sternberg) Brongniart, and so proposed a replacement name, Asterophyllites arkansanum White. However, there is no evidence that the name A. gracilis (Sternberg) Brongniart was ever validly published, and so A. arkansanum is redundant.
Stur (1887) described specimens from Duckmantian strata in the Ruhr Coalfield in Germany as whorls of small, needle-like that are indistinguishable from A. gracilis, and which are also associated with Palaeostachya strobili. Stur appears to have been unaware of Lesquereux’s work (at least, he did not refer to it) and made these German specimens the types of a new species, A. roehlii. As the protologue of the latter postdates the publication of A. gracilis, the latter takes precedence.
Stur (1887, pl. 11, Figures 2 and 3) also figured very similar leafy shoots from Belgium associated with stems that he named Calamites paleaceus Stur. Stur (1887, p. 71) also introduced the name Asterophyllites paleaceus, presumably referring to these leafy shoots, but without providing a diagnosis or designating a type. Although the name A. paleaceus has occasionally been used (e.g., Jongmans, 1911), it has never been validly published. Jongmans suggested that A. paleaceus leaves were borne on stems with small scales or papillae, but this has never been fully documented.
A specimen from the lower Duckmantian Sulzbach Formation of Saar-Lorraine was figured as A. charaeformis by Laveine (1989, pl. 9, Figure 2). However, this has much smaller (2–3 mm long) and more slender leaves than A. charaeformis auct. (i.e., A. parvulus) and fewer leaves per whorl (4–5). This specimen is similar to the types of A. gracilis.
Asterophyllites lycopodioides Zeiller, originally documented from the lower Asturian of Nord-Pas-de-Calais, France (Zeiller, 1888), also has whorls of small (2 mm long) arched leaves. Zeiller suggests that there are 4–6 leaves per whorl, but this is not clear from the original illustration (Zeiller, 1886, pl. 59, Figures 1 and 2). The leaves of the A. lycopodiodes type appear to be closely adpressed to the supporting stem, and many authors have regarded this as a diagnostic feature (e.g., Crookall, 1969; Jongmans, 1911). However, this species has only been reported in a few incompletely preserved specimens, and this feature may be merely taphonomic.
Distribution. This is a rare species known from the lowland paralic coalfields, notably the Ruhr in Germany; for instance, in addition to the types, specimens from here have been figured by Hofmann and Ryba (1899, pl. 2, Figure 7) and Jongmans and Kukuk (1913, pl. 18, Figure 5; pl. 19, Figures 1–3); also, from Nord-Pas-de-Calais and Belgium. Also, from Saar-Lorraine (Laveine, 1989).
A. taylorianum Stur
Plate 7
Asterophyllites taylorianum Cleal and Shute foliage with associated Palaeostachya wagneri Cleal and Shute strobili; Hirwaen, near Ebbw Vales, South Wales, United Kingdom; Middle Coal Measures, Duckmantian (Bashkirian); Natural History Museum, London. 1, 2. Holotype of A. taylorianum, specimen NHM V.68601b; 3, 4. P. wagneri (holotype), specimen NHM V.68607. Scale bars 5 mm (Figures 1 and 3), 10 mm (Figures 2 and 4).
2016 A. taylorianum Cleal & Shute, p. 29, Figures 2 and 3.
Holotype. Natural History Museum (London) specimen V 68601 a (Plate 7, Figures 1 and 2). Provenance Ebbw Vales, South Wales, UK; Duckmantian Middle Coal Measures Formation
Diagnosis. Foliage adpressions of ultimate shoots with slender stems ca. 1 mm wide with fine longitudinal ribs. Isophyllous whorls of four slender leaves borne at 1–2 mm intervals along shoot, attached typically at 90°. Leaves on ultimate shoots are typically up to 5–7 mm long. Leaves are needle-like, typically broadly arched so that the distal parts are more or less parallel to the supporting stem.
Description. For a description of the types, see Cleal & Shute (2016).
Associated strobili (Plate 7, Figures 3 and 4). Cleal and Shute (2016) described well-preserved strobili with these shoots as Palaeostachya wagneri Cleal and Shute. These strobili are 25–35 mm long, 5–7 mm wide, with slender, spreading bracts, and the distal ends of the bracts are oblique to the strobilus axis.
Remarks. This species resembles A. delicatulus in typically having four leaves in each whorl, with the leaves generally being of similar size and shape. However, the A. taylorianum leaves are usually more curved, with their distal end lying nearly parallel to the axis. Also, A. taylorianum is associated with Palaeostachya strobili, in contrast to the Calamostachys strobili found with A. delicatulus.
Very similar leafy shoots have been documented from the Langsettian and Duckmantian of the Ruhr Coalfield, Germany, under the names A. delicatulus (Roehl, 1869, pl. 2, Figure 6; pl. 3, Figures 1–3; pl. 4, Figure 1) and A. grandis (Josten, 1991, pl. 30, Figures 1 and 2). Josten stated that they bore Calamostachys strobili, but this was not based on evidence from the Ruhr but rather was merely paraphrasing the remarks of Zeiller (1888), who had misidentified A. parvulus shoots associated with Calamostachys strobili as A. grandis (i.e., A. delicatulus).
Distribution. This is a rare species, mainly known from the Lonchopteris rugosa Zone (Duckmantian) lowland paralic coalfields, notably in South Wales, United Kingdom (type locality) and the Ruhr, Germany.
A. lubnensis Cleal sp. nov.
Plate 8
Asterophyllites lubnensis sp. nov.; Rako Mine, Lubná, Rakovnik, Czech Republic; Upper Lubná coal, lower Bolsovian upper Radnice Member (lower Bolsovian). 1, 2. Holotype, National Museum (Prague) Specimen E 1144; 3, 4. Paratype, National Museum (Prague) Specimen 25136. Scale bars 20 mm (Figures 1 and 3), 5 mm (Figures 2 and 4).
Holotype. National Museum (Prague) Specimen E 1144 (Plate 8, Figures 1 and 2). Provenance Rako Mine, Lubná, Rakovnik, Czech Republic; whitish sandy tuffitic interlayer of the Upper Lubná coal, upper Radnice Member (lower Bolsovian, Kladno Formation (Opluštil & Vízdal, 1995; Opluštil et al., 2007).
Diagnosis. Foliage adpressions of ultimate shoots with slender stems ca. 1 mm wide with fine longitudinal ribs. Isophyllous whorls of up to 10 slender leaves borne at 1.5–2.5 mm intervals along the shoot, reducing to 1.5 mm in the distal part of the shoot, attached typically at 90°. Leaves on ultimate shoots are up to 5–7 mm long. Leaves are slender-falcate with acute apex, typically straight for half their length, then bending sharply so the distal part is near a right angle to the stem.
Description of holotype (Plate 8, Figures 1 and 2). Penultimate ribbed axes up to 10 mm wide with nodes at 10–20 mm intervals. Each node has two or four ultimate shoots (the number is difficult to verify due to compression) attached at varying but mainly obtuse angles, and numerous elongate, falcate, gently curved leaves, 17‒20 mm long.
Ultimate shoots up to 77 mm long, usually 6‒7 mm wide, increasing to 12 mm wide in more proximal positions. The axes of ultimate shoots are slender, ca. 1 mm wide, with nodes at up to 5 mm intervals, becoming more closely spaced at 1 mm intervals in the distal part of the shoot. Each node has a whorl of leaves; the proximal parts of the shoot can have up to 16 leaves per whorl, while the more distal parts have as few as four leaves per whorl. However, most typically, there are 8–10 leaves per whorl in the middle parts of shoots. Leaves are slender-falcate with an acuminate apex, typically 5–7 mm long, attached at right angles to the stem or deflected slightly adaxially. Leaves in the middle and distal parts of shoots are straight along for about half their length and then bend so their distal parts are about parallel to the supporting stem; leaves from more proximal parts of the shoots tend to be curved along most of their length, with their distal part at 20°‒30° to the stem.
Associated strobili. Alongside (but not attached) to shoots of this type are strobili, 10 mm wide and preserved for a length of 45–70 mm, with whorls of bracts at ca. 5 mm intervals. The bracts are attached at near right angles to nodes of the strobilus axis, straight for the first two-thirds of their length and then curve. The distal end of each bract is parallel to the strobilus axis and nearly touches the base of the next whorl of bracts. The sporangiophores are not well preserved but were evidently in a Palaeostachya mode of attachment.
Němejc (1953) compared these strobili with Palaeostachya elongata (Presl) Weiss. However, that species is difficult to interpret as the holotype is only known from the drawing illustrated in Presl (1837, pl. 1); this shows it is clearly a Palaeostachya, but other features are only diagrammatically reproduced. Provisionally, therefore, the strobili being dealt with here will be referred to as Palaeostachya aff. elongata (Presl) Weiss.
Remarks. Leafy shoots of this type were documented by Němejc (1953, Figure 1). They are clearly within the A. grandis general group of species characterised by small whorls of slender leaves, but differ in having more leaves in each whorl: typically, 8–10, rather than just 4 in A. delicatulus, A. taylorianum and A. gracilis, or 5–6 in A. parvulus. A. lubnensis also differs from A. gracilis and A. parvulus in the leaves being somewhat larger, and from A. delicatulus in the leaves being more curved.
A possible specimen of this species from the lower Sulzbach Formation (Duckmantian) of Saar-Lorraine was figured by Laveine (1989, pl. 9, Figure 1). It was named A. grandis. Although the leaves are of similar size, they are much more curved and more falcate than in that species (i.e., A. delicatulus) and occur in whorls of 8–10 (rather than 4) leaves. However, this is the only known possible example of this species from Saar-Lorraine, and its presence in this flora needs verifying.
Specimens documented by Tenchov (1977) from the Svoge Basin in Bulgaria, identified as A. grandis, were described as shoots with whorls of 8–10 leaves, each leaf being slender, curved, and 8 mm long. Also found here are strobili similar to those of Palaeostachya aff. elongata (Němejc, 1943). These are very similar to the types of A. lubnensis.
Distribution. This species is only known from the Paripteris linguefolia Zone (Bolsovian Substage) of West and Central Bohemia, Czech Republic (the types; see also Ettingshausen, 1854, p. 2, Figures 2, 3; Trapl, 1927, pl. 1, Figure 5), Zwickau, Germany (Geinitz, 1855, pl. 17, Figures 4 and 6), Saar-Lorraine Laveine (1989). Laveine (1989, pl. 9, Figure 1) and Svoge, Bulgaria (Tenchov, 1977).
DISCUSSION
The five species summarised here are the remains of a distinctive group of small-leafed equisetaleans found in the C. hoeninghausii, Lonchopteris rugosa and lower Paripteris linguaefolia macrofloral zones (sensu Opluštil et al., 2022) of palaeotropical Euramerica. These biozones indicate a Langsettian to middle Bolsovian (late Bashkirian—earliest Moscovian) age, corresponding mainly to the interval of global climatic cooling and relatively low sea-levels known as Glacial Phase C3 (sensu Fielding et al., 2023) of the Late Palaeozoic Ice Age. The species survived the brief C3–C4 interglacial that caused extensive flooding and habitat disruption during the middle Bolsovian (near the Bashkirian—Moscovian boundary) but then disappeared from the fossil record shortly after.
The five species are differentiated mainly on relatively minor features of leaf size and curvature, and the number of leaves per whorl on the shoots (Figure 1). Given that the fossils are often relatively fragmentary, it must be asked if these differences are real or merely variation within one or a smaller number of species. It is obviously difficult to be sure, but the view taken here is that the differences do in fact reflect ‘natural’ species, as there is a broad consistency in the distribution of each species (whether they occur in lowland or intra-montane settings) and their associated strobili (Table 1).
Distribution and associated strobili.
It has been suggested that calamitaceans favoured both peat and clastic substrate settings within the coal swamps (e.g., DiMichele, 2014), and A. delicatulus remains occur in both settings in West and Central Bohemia (Opluštil et al., 2020). A. parvulus also seems to have occurred in both habitat types in lowland areas, although it appears to be absent from peat substrates in West and Central Bohemia. However, A. taylorianum and A. roehlii are only known from clastic substrate vegetation (as are other calamitaceans with larger-leafed foliage such as A. spinulosa Sternberg–Rößler & Noll, 2007; Cleal, 2023). Evidently, further work is needed to determine the habitat preferences of these Palaeozoic equisetaleans.
This study demonstrates the importance of trying to incorporate evidence of the associated strobili when classifying calamitacean foliage. The leaves are morphologically simple and on their own provide relatively few characters on which to base a classification; adding evidence from the strobili enlarges the suite of available taxonomic characters. Eventually, we may be able to integrate the fossil taxa of the foliage and leafy shoots to make something approaching whole-plant taxa. In addition to the examples documented here, it has been suggested that Annularia spinulosa shoots bore Calamostachys tuberculata (Sternberg) Jongmans strobili, Annularia sphenophylloides (Zenker) Gutbier shoots bore Calamostachys calathifera Weiss strobili, and A. equisetiformis Brongniart shoots bore Calamostachys germanica Weiss strobili (e.g., see discussions in Jongmans, 1911; Crookall, 1969). However, there are still many shoot/strobilus correlations that have not been verified, and a complete integration of the taxonomies may cause more confusion than provide insight into the natural relationships of the parent plants; see discussion on analogous problems with anatomically preserved calamitacean remains (Good, 1975). For the time being, retaining separate taxonomies for the shoots and strobili is the safest option.
CONCLUSIONS
Five species of calamitacean equisetaleans (‘horsetails’) with distinctive, small leaves, belonging to the genus Asterophyllites, occurred in the palaeotropical wetlands of Euramerica during the late Bashkirian Glacial Phase C3 of the Late Palaeozoic Ice Age. Key morphological characters for differentiating the species of leafy shoots include the size, curvature, and width of the leaves, as well as the number of leaves per whorl. Two of the species of leafy shoots are associated with Calamostachys strobili, and the other three with Palaeostachya strobili. They provide evidence for the diversity of these calamitaceans in Pennsylvanian times, with species distributions being partly controlled by substrate type and the elevation of where the parent plants grew.
Footnotes
Acknowledgements
The author thanks Prof. Jiři Kvaček and Dr Milan Libertín for facilitating access to the collections at the National Museum (Prague, Czech Rep.) during a visit funded by the European Union SYNTHESYS programme (CZ–TAF–6452), and for providing photographs of the types of A. grandis, A. charaeformis, A. delicatulus and A. lubnensis. Thanks also go to Dr Daniela Festi for providing photographs of the type of A. roehlii in the collections of GeopSphere Austria (Wien); to Jonathan Wingerath for providing photographs of the neotypes of A. gracilis in the collections of the National Museum of Natural History, Smithsonian Institution, Washington DC (USA); to Dr Peta Hayes and the late Cedric Shute for providing photographs of the type of A. taylorianum in the collections of the Natural History Museum, London (UK); and to Dr Michael Howe for providing photographs of specimens in the Kidston Collection, British Geological Survey, Keyworth (UK).
Data Availability
All specimens illustrated and described in this article can be accessed in public museums.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Ethics Statement
Full permission has been obtained to illustrate the specimens shown in this article.
Funding
The author received no financial support for the research, authorship and/or publication of this article.
