Articles
Reassignment of vertebrate ichnotaxa from the Upper Carboniferous ‘Fern Ledges’, Lancaster Formation, Saint John, New Brunswick, Canada
ABSTRACT
Vertebrate ichnotaxa described by George Frederic Matthew in 1910 from the Upper Carboniferous (Lower Pennsylvanian) ‘Fern Ledges’ of Saint John, New Brunswick, were dismissed as dubious trackways by previous authors. Thus, three new ichnospecies Matthew described appeared in the 1975 Treatise on Invertebrate Paleontology as “unrecognized or unrecognizable” and were mostly forgotten by vertebrate ichnologists. These traces include Hylopus (?) variabilis, Nanopus (?) vetustus and Bipezia bilobata. One ichnospecies, Hylopus (?) variabilis, here is retained as a valid tetrapod footprint ichnotaxon and reassigned to the ichnogenus Limnopus as a new combination, together with other poorly preserved specimens Matthew labeled, but never described. Nanopus (?) vetustus and Bipezia bilobata named by Matthew in the same paper, have been reexamined and remain as nomina dubia.
RÉSUMÉ
Les ichnotaxons vertébrés de Fern Ledges du Carbonifère supérieur (Pennsylvanien inférieur) de Saint John, au Nouveau-Brunswick, décrits par George Frederic Matthew en 1910 avaient été rejetés par des auteurs précédents les considérant comme des séries de traces douteuses. Trois nouvelles ichnoespèces décrites par Matthew sont ainsi apparues dans le Treatise on Invertebrate Paleontology (Traité sur la paléontologie des invertébrés) de 1975 à titre d’espèces « non identifiées ou non identifiables » et ont principalement été laissées aux ichnologues des vertébrés. Les traces en question comprennent des traces d’Hylopus (?) variabilis, de Nanopus (?) vetustus et de Bipezia bilobata. L’une de ces ichnoespèces, l’Hylopus (?) variabilis, est retenue ici comme spécimen valide d’ichnotaxon se rapportant à une empreinte de pas de tétrapode et a été rattachée à nouveau à l’ichnogenre Limnopus en tant que nouvelle combinaison, conjointement avec d’autres spécimens mal conservés auxquels Matthew avait donné un nom, mais qu’il n’avait jamais décrits. Le Nanopus (?) vetustus et le Bipezia bilobata, cités par Matthew dans le même article, ont été réexaminés et leurs désignations nomina dubia.
[Traduit par la redaction]
INTRODUCTION
1 In the early 1900s, George Frederic Matthew (1837–1923) identified and reviewed tetrapod footprints and trackways from the Maritimes region of Canada in an attempt to provide some organization to the classification of vertebrate ichnotaxa (Matthew 1903a–e, 1904, 1905). In the process, Matthew became a pioneer in the study of Carboniferous tetrapod ichnology (Leonardi 1987; Cotton et al. 1995; Haubold et al. 2005). The specimens he examined and described were mostly collected at Joggins (Cotton et al. 1995; Calder 2006; Stimson et al. 2012) and at other Carboniferous localities in Nova Scotia. In his final paper on tetrapod ichnofossils, Matthew (1910) described three new ichnospecies, Hylopus (?) variabilis, Nanopus (?) vetustus and Bipezia bilobata, from the Upper Carboniferous Lancaster Formation at ‘Fern Ledges’, Saint John, New Brunswick (Fig. 1). He published the first two ichnotaxa with the question marks suggesting he recognized the poor preservation of the fossils. Matthew stated in the paper that tetrapod tracks had been previously recognized in situ on the surfaces of hard sandstone beds in the ‘Little River group’ (Lancaster Formation), but that they were not collected, nor were any specimens described. In this study we reexamine the published specimens that Matthew did collect. We also document additional poorly preserved footprints found among plant fossils in the collections of the Natural History Society of New Brunswick that were discovered by Matthew, C.F. Hartt (1840–1878), W.J. Wilson (1857–1920) and others. Although Matthew recognized Hylopus and Nanopus as Devonian–Carboniferous ichnogenera, by the early 1900s Matthew was convinced that at ‘Fern Ledges’ the rocks were Silurian (Matthew 1910), older than the Devonian age he and others originally proposed. The ‘Fern Ledges’ site is known mostly for its plant and invertebrate assemblages (Dunlop and Miller 2007; Falcon-Lang and Miller 2007a), and for a late nineteenth to early twentieth century controversy concerning the age of the rocks (Falcon-Lang and Miller 2007b), which was ultimately resolved by Marie Stopes in her classic palaeobotanical memoir on the Upper Carboniferous plants of the site (Stopes 1914).
Display large image of Figure 1
2 All three of Matthew’s ichnospecies have remained misunderstood orphans in tetrapod ichnology, in part because Matthew’s description was published just a few years prior to the demise of the Natural History Society of New Brunswick, where Matthew conducted most of his research activities. The Society collections became difficult to access, even after they were acquired by the New Brunswick Museum in 1932. By 1917, the Society had ceased publication of its Bulletin, and by 1921 Matthew had retired and moved to New York to live with his son William Diller Matthew (Colbert 1992). Matthew’s collection was largely inaccessible when Abel (1935) reviewed Hylopus (?) variabilis and Nanopus (?) vetustus. Even though he had not seen the specimens, Abel concluded that they are not vertebrate tracks. Both ichnotaxa appeared in Part W of the Treatise on Invertebrate Paleontology dealing with Trace Fossils and Problematica (Häntzschel 1975; Miller 1996), where they were treated as ‘unrecognized and unrecognizable’ based on Abel’s (1935, p. 78) judgment. Thus, subsequent workers have also not regarded them as tetrapod track ichnotaxa (e.g., Sundberg et al. 1990). Bipezia bilobata was also reported in Part W of the Treatise (Häntzschel 1975) as a dubious tetrapod trace based on Glaessner (1957), who considered it a possible arthropod trace synonymous with Isopodichnus.
3 Matthew’s specimens of all three ichnotaxa were relocated in 1987 during re-organization of the New Brunswick Museum palaeontology collection (Miller 1988). Falcon-Lang and Miller (2007a) published the fossils as (?) Hylopus variabilis and (?) Nanopus vetustus and concluded they were possibly poorly preserved vertebrate undertracks, but that assignment to Hylopus or Nanopus was questionable. The fossils are re-evaluated and re-illustrated here as part of a project to examine tetrapod tracks from New Brunswick and are accordingly reassigned. Hereafter we simply refer to the species as Hylopus variabilis and Nanopus vetustus. Also included in our study are previously unrecognized examples of Hylopus variabilis and other putative tetrapod tracks found in the New Brunswick Museum collections that had not been previously described.
4 In summary, the purpose of this paper is to reassess the validity of Matthew’s three tetrapod ichnotaxa from the Lancaster Formation. We demonstrate that Häntzschel (1975) was correct in his assessment that Nanopus vetustus and Bipezia bilobata are indeed nomina dubia. We also reinstate Matthew’s ichnospecies Hylopus variabilis as a tetrapod footprint and reassign the ichnospecies to the ichnogenus Limnopus; the only ichnogenus treated systematically herein.
GEOLOGIC SETTING
5 The ‘Fern Ledges’ section belongs to the Upper Carboniferous Lancaster Formation of the Cumberland Group. The Lancaster Formation is found along the Bay of Fundy coast of southern New Brunswick from Saint John to Lepreau (Fig. 1A). It is exposed along coastal sections, much of it in the intertidal zone. In the type locality at ‘Fern Ledges’ to Duck Cove, the Lancaster Formation consists of grey to greenish-grey quartzose sandstone and dark grey, greenishgrey or black shale. The Formation also includes units of coarse-grained sandstone, quartz-pebble conglomerate, minor red mudstone, and fine-grained sandstone. Falcon-Lang and Miller (2007a) reviewed the palaeoenvironments and palaeoecology of the ‘Fern Ledges’ section, with the emphasis on the invertebrate fauna and ichnofossils. A review of the Lancaster Formation flora is underway, with revisions of several taxa completed (Wagner 2001, 2005a, b). An Early Carboniferous (Langsettian) age is currently assigned to the formation (Falcon-Lang and Miller 2007a).
6 Matthew (1910) recorded the tetrapod ichnofossils as coming from the ‘Lower Cordaite’ shale at ‘Fern Ledges’. C.F. Hartt made the first comprehensive collections at the site (Fig. 1B) and used the term ‘Cordaite Shale’ (in Bailey 1865, p. 131; in Dawson 1868, p. 514, 516) probably to describe the plant-rich shale beds, which he divided into plant beds 2-8 (in Bailey 1865, p. 134-140; in Dawson 1868, p. 517-523; Matthew 1906). More precise information was not recorded with the fossils. Almost all of the historical fossil collections from the Lancaster Formation at ‘Fern Ledges’, including diverse plant compressions and remains of terrestrial and aquatic fauna, were likely derived from Facies 3 of Falcon-Lang and Miller (2007a). They described Facies 3 as lenticular, upward-coarsening sheets of sediment 0.12–2.97 m thick. Mudstone beds, 0.02–0.71 m thick, grade upward into fine- to medium-grained, horizontally bedded sandstone 0.10–2.26 m thick, with symmetrical ripple marks. Some sandstone beds may show low-angle, mutually erosive cross-beds, or localized channel fills. Upward-coarsening units are usually capped by palaeosols. Fossils are most abundant in the mudstone beds overlying palaeosols. Facies 3 was interpreted as either representing wave-reworked mouth-bar deposits associated with small channels, or with linear clastic shorelines in which rhythmic bedding in mudstones may result from weak tidal influence. Falcon-Lang and Miller (2007a) concluded that the ‘Fern Ledges’ sediments were deposited on a tectonically influenced coastal plain where braided channels drained into a shallow brackish bay.
ICHNOTAXONOMIC ASSESSMENT OF HYLOPUS VARIABILIS
7 A fossil found by William Logan in 1841 from rocks at Horton Bluff, Nova Scotia, was the first Mississippian tetrapod footprint specimen to be discovered in North America (Logan 1842; Dawson, 1882). Logan’s specimen was not given much attention until it was named and described as Hylopus hardingi by Dawson (1882). Fillmore et al. (2012) revised the diagnosis of Hylopus, in the process reviewing all three known ichnospecies of that genus. Hylopus was considered monospecific with the only valid ichnospecies being Hylopus hardingi Dawson 1882. Hylopus variabilis was not considered in the revision by Fillmore et al. (2012), presumably as it was considered a nomen dubium by Abel (1935) and in the Treatise (Häntzschel 1975). In their emended diagnosis, Fillmore et al. (2012, p. 58) described Hylopus as follows:
8 Hylopus variabilis does not conform to this ichnogenus concept for Hylopus. It differs from the emended diagnosis in that the manus is smaller than the pes. The digits on the manus of Hylopus variabilis are of subequal length, increasing only slightly in length from digits I–III from 4 mm to 6.2 mm. However, digit IV in Hylopus hardingi is much longer than the other digits in the manus. The digits on the pes of Hylopus hardingi are subequal whereas in Hylopus variabilis they increase in length from I–IV, with digit IV being the longest. Given these differences, Hylopus variabilis requires reassignment.
9 Lucas and Dalman (2013) re-examined Pennsylvanian trackways described by King (1845) from western Pennsylvania. In the process, King’s ichnospecies, Thenaropus heterodactylus, was reassigned to the ichnogenus Limnopus Reassignment of vertebrate ichnotaxa from the Upper Carboniferous ‘Fern Ledges’, Lancaster Formation, Saint John, N.B., Canada and considered to be the senior subjective synonym of the type ichnospecies, Limnopus vagus. They thus concluded that Limnopus is monospecific and that Limnopus heterodactylus is the valid name for the single species, all other names being synonymous with Limnopus heterodactylus. Lucas and Dalman (2013, p. 236) emended the genus concept, providing a diagnosis of Limnopus for quadrupedal tetrapod tracks consisting of as follows:
10 Hylopus variabilis does not easily conform to any other known ichnotaxon, but is not sufficiently well preserved to justify a new ichnogenus. Hylopus variabilis most closely resembles Limnopus. However, that genus was described by Lucas and Dalman (2013) as having thick digits with blunt or rounded digit terminations, whereas Hylopus variabilis has thin and short digits. The morphology of Hylopus variabilis likely reflects shallow underprints, and for reasons discussed below should not at this time be considered a junior synonym of Limnopus heterodactylus. Thus, Matthew’s form is here transferred to Limnopus but retained as a separate ichnospecies, at least until better specimens are discovered to either confirm it as a separate ichnospecies or more confidently establish its synonymy with Limnopus heterodactylus. We suggest a revision of the ichnogeneric concept of Limnopus to better accommodate Matthew’s species.
SYSTEMATIC ICHNOLOGY
Ichnogenus Limnopus Marsh 1894
- 1894 Limnopus Marsh, p. 82, pl. 2 (2), pl. 3 (2).
- 1912 Permomegatherium n. g.; Delage, p. 241.
- 1963 Opisthopus n. g; Heyler & Lessertisseur, p. 177, figs. 25–31.
- 1973 Limnopus Marsh; Haubold, p. 8; fig. 13.
- 1973 Opisthopus Heyler & Lessertisseur; Haubold, p. 30.
- 1973 Permomegatherium Delage; Haubold, p. 30.
- 1973 Limnopus Marsh; Haubold & Sarjeant, p. 900, fig. 2 (4), pls. 3–4.
- 2005 Limnopus Marsh; Voigt p. 73, figs. 21 D–F, 22 H–N, 24.
- 2013 Limnopus Marsh; Lucas and Dalman, p. 236, figs. 3, 6, 11.
12 EMENDED DIAGNOSIS: Quadrupedal tetrapod tracks consisting of a pentadactyl pes and a tetradactyl manus; pes digits increase in length from I–IV, with digit V subequal in length to digit II; commonly with relatively thick manus and pes digits with blunt to rounded distal tips, but may be short and slender; broad rounded sole imprints; manus just in front of pes; manus digits I–III obliquely set toward the trackway midline; and manus digits that increase in length from I–III, with digit IV subequal in length to digit II (Modified from Lucas and Dalman 2013).
13 REMARKS: Prior to this study Limnopus has been considered monospecific. Voigt (2005) provided a comprehensive review of Limnopus with all previous species included in the genus and species concept for Limnopus vagus Marsh. Limnopus was reviewed by Lucas and Dalman (2013) who reassigned all known species to Limnopus heterodactylus, considering the genus to be monospecific. We have expanded the generic concept to accommodate minor morphological variation at the species level, thus including Hylopus variabilis (discussed below).
Limnopus heterodactylus (King 1845)
- 1844 Batrachia; King, p. 179.
- 1845 Thenaropus heterodactylus n. sp.; King, p. 348–352, figs. 7–9.
- 1873 unnamed; Mudge, p. 71.
- 1894 Limnopus vagus n. sp.; Marsh, p. 82, pl. 2 (2), pl. 3 (2).
- 1894 Allopus littoralis n. sp.; Marsh, p. 83, pl. 2 (4, 4a).
- 1894 Baropus lentus n. sp.; Marsh, p. 83, pl. 2 (5).
- 1903b Theranopus (?) mcnaughtoni n. sp. Matthew, p. 103, pl. 2, fig. 1.
- 1903e Thenaropus heterodactylus King; Matthew, p. 112, fig. 1 (1, 2).
- 1903e Limnopus vagans Marsh; Matthew, p. 112, fig. 3 (2).
- 1903e Baropus lentus Marsh; Matthew, p. 113, fig. 3 (6).
- 1910 unnamed; Hausse, p. 3–19, pls. 2–7.
- 1912 Permomegatherium zeilleri n. sp.; Delage. p. 241, 8 pls.
- 1926 unnamed; Tilton, p. 389, pl. 2 (A–E).
- 1931 Baropus waynesburgensis n. sp.; Tilton, p. 551, fig. 4.
- 1932 Allopus littoralis Marsh; Branson & Mehl, p. 390, text fig. 2, pl. 10, fig.2.
- 1952 Limnopus vagus Marsh; Baird, p. 835, text figs. 1–2, pl. 122.
- 1952 Limnopus littoralis Marsh; Baird, p. 836, text fig. 3, pls. 123, 124, figs 1–3.
- 1952 Limnopus waynesburgensis (Tilton); Baird, p. 837, text fig. 4, pl. 124, fig. 4.
- 1959 Limnopus vagus Marsh; Schmidt, p. 78, fig. 32e.
- 1959 Baropus haussei n. sp.; Schmidt, p. 81, fig. 33d.
- 1965 Limnopus cutlerensis n. sp.; Baird, p. 47, figs. 14 (B–C).
- 1970 Limnopus vagus Marsh; Haubold, p. 96, fig. 5B.
- 1970 Limnopus littoralis Marsh; Haubold, p. 96, fig. 5H.
- 1970 Limnopus waynesburgensis (Tilton); Haubold, p. 96, fig. 5G.
- 1970 Limnopus cutlerensis Baird; Haubold, p. 96, fig. 5D.
- 1970 Limnopus heterodactylus (King); Haubold, p. 96, fig. 5.
- 1970 Limnopus haussei (Schmidt); Haubold, p. 99, fig. 5E.
- 1971 Limnopus vagus Marsh; Haubold, p. 17, figs. 12 (5), 13 (2).
- 1971 Limnopus littoralis Marsh; Haubold, p. 17, fig. 13 (4).
- 1971 Limnopus waynesburgensis (Tilton); Haubold, p. 17, fig. 13 (3).
- 1971 Limnopus cutlerensis Baird; Haubold, p. 17, fig. 12 (4).
- 1971 Limnopus heterodactylus (King); Haubold, p. 17, fig. 13 (5).
- 1971 Limnopus haussei (Schmidt); Haubold, p. 17, fig. 13 (6).
- 1971 Theranopus? mcnaughtoni Matthew; Haubold, p. 17.
- 1983 Limnopus palatinus n. sp.; Fichter, p. 46, figs. 28–35.
- 1988 Limnopus zeilleri (Delage); Gand, p. 98, figs. 28, 29 (A–I), 30 (A–E), 31.
- 1991 Limnopus glenshawensis n. sp.; Martino, p. 960, figs. 4–13.
- 1995 Limnopus vagus Marsh; Hunt et al., p. 264, figs. 2F, 4 [non figs. 2E, 3].
- 1996 Limnopus vagus Marsh; Haubold, p. 48.
- 1996 Limnopus waynesburgensis (Tilton); Haubold, p. 48.
- 1996 Limnopus cutlerensis Baird; Haubold, p. 48.
- 1996 Limnopus zeilleri (Delage); Haubold, p. 48.
- 2000 Limnopus cutlerensis Baird; Haubold, p. 12.
- 2000 Limnopus zeilleri (Delage); Haubold, p. 12.
15 REMARKS: The synonymy list above for Limnopus heterodactylus is updated from Voigt’s (2005) list of material that he considered belonged to the genus Limnopus; here we include Lucas and Dalman’s (2013) work.
Limnopus variabilis (Matthew 1910)
(Figs. 2–3)- 1910 Hylopus (?) variabilis Matthew, p. 120, pl. 2, figs. 1–3.
- 1935 Hylopus (?) variabilis Matthew; Abel, p. 78.
- 1975 Hylopus? variabilis Matthew; Häntzschel, p. W185.
- 2007a (?)Hylopus variabilis Matthew; Falcon-Lang and Miller, p. 949.
Display large image of Figure 2
Display large image of Figure 3
17 MATERIAL: NBMG 3041, 3042, 3043, reposited at the New Brunswick Museum. ‘Lower Cordaite’ shales, ‘Fern Ledges’, Lancaster Formation, Saint John, New Brunswick (approximately 45°14’42.39”N; 66°04’56.83”W). As Matthew (1910) did not select a holotype, we here select specimen NBMG 3041 and its counterpart NBMG 3043 as the lectotype with specimen and designate NBMG 3042 as a paralectotype.
18 AGE: Late Carboniferous, Early Pennsylvanian (Langsettian), Lancaster Formation, Cumberland Group.
19 EMENDED DIAGNOSIS: Limnopus variabilis differs from the only other currently recognized ichnospecies of the genus, Limnopus heterodactylus (=Limnopus vagus), in having slender digits in both the manus and the pes, and digit distal tips that taper sharply. Both manus and pes digits are shorter in length than in Limnopus heterodactylus, but otherwise follow the emended ichnogenus concept described in this paper.
20 DESCRIPTION: Hylopus variabilis was described from a single manus and pes that are not clearly distributed into manus and pes sets (Matthew 1910). Tracks have low relief and may represent shallow underprints, but retain their footprint plantigrade morphology. The deepest impressions are the digit terminations. Digits are well spaced at regular intervals, with wide hipices between digits. The pes is plantigrade with short, broad sole impressions. The pes is only slightly wider than long, measuring 25 mm by 28 mm for length and width, respectively. The pes is pentadactyl with regularly spaced digits that increase in length from I to IV and are set forward. Digit V is shorter than digit IV and set to the side. Matthew described a callus or lump on the side of the palm of dubious validity.
21 The manus is wider than long, measuring 13 mm by 24 mm for length and width, respectively. The manus is tetradactyl with regularly spaced digits. Digits I to III are roughly equal in length, only increasing slightly from 4.0 mm to 5.3 mm to 6.2 mm. Digits I to II are directed inward towards the midline, whereas digit III is directed forward. Digits I to III are closely grouped with digit IV and directed away from the midline and slightly set back. Matthew (1910) described the stride, however, with only a single manus and pes with no repeated footfalls present on the specimen, measuring other morphometrics such as stride, pace, and inner and outer trackway widths is not possible.
DISCUSSION
Limnopus from ‘Fern Ledges’
22 Matthew (1910) described three specimens of Hylopus variabilis from ‘Fern Ledges’. Hylopus variabilis was considered a non-tetrapod track by Abel (1935) and Häntzschel (1975). Although the lectotype specimen (NBMG 3041) was illustrated in a New Brunswick Museum newsletter (Anonymous 1987), it has only rarely been mentioned elsewhere (e.g., Sundberg et al. 1990), and has undergone no systematic reassessments. Indeed, the comprehensive tetrapod footprint compendia of Kuhn (1963) and Haubold (1971) do not list the ichnotaxon. Falcon-Lang and Miller (2007a) tentatively vindicated Matthew’s ichnospecies, stating that it may represent “possible poorly preserved undertracks of a tetrapod emplaced in soft sediment”, but they provided no further detailed study of its ichnological features; they added that assignment to Hylopus is questionable. From our reassessment, we conclude that the specimens represent actual tetrapod tracks, which we reassign to Limnopus. One of the three specimens labeled as Hylopus variabilis by Matthew (NBMG 3043) in the NBM collection has been recognized as the counterpart to the Matthew’s figured specimen of Hylopus variabilis (NBMG 3041). Matthew assigned a third specimen (NBMG 3042) to Hylopus variabilis; we consider this specimen to be a deep underprint of Limnopus variabilis.
23 Subsequent explorations of the New Brunswick Museum collections have identified three additional, previously undescribed and unillustrated specimens with putative tetrapod tracks from the Lancaster Formation. Two of these specimens (NBMG 12926, Fig.4A; and NBMG 2191, Fig.4B) were collected and labeled by Matthew for associated plant remains but contain previously unrecognized tetrapod tracks, both tentatively interpretable as strongly extramorphologically distorted deep underprints of Limnopus variabilis; these specimens do not warrant further description.
Display large image of Figure 4
24 A third specimen with tetrapod tracks (NBMG 3494) was labeled as the type specimen for the plant Odontopteris crassa, but had associated tetrapod tracks that were labeled simply as ‘Batrachian tracks?’ (Fig. 4C). Matthew (presumably), using red circles, identified 13 individual but unconvincing footprints on the surface; they may represent the tracks of a small amphibian with a short stride. On the specimen label, Matthew indicated the toes are distinct in the last footfalls. Deep, elliptical impressions arranged in groups of up to three are present at the anterior of the ‘trackway’. Given the short blunt digits measuring a maximum of 4.7 mm in length, the tracks may represent deep, digitigrade underprints of a tetrapod trackway. The putative trackway is digitigrade at best, with no clear diagnostic features to assign it to an existing vertebrate ichnotaxon with confidence.
25 Limnopus variabilis differs from the other valid species of the ichnogenus, Limnopus heterodactylus (which incorporates the type ichnospecies originally named Limnopus vagus). As defined by Lucas and Dalman (2013), Limnopus heterodactylus has thick digits with blunt to rounded terminations (Fig. 5). These authors used this concept to synonymize all previously described Limnopus forms into a monospecific ichnogenus (Figs. 5A–D). In contrast, Limnopus variabilis has short, slender digits with sharp, tapering digit terminations that are well spaced with wide interdigit hipicies (Fig. 5E). The lectotype of Limnopus heterodactylus has a manus that is longer than wide, which by strict definition is applicable to Limnopus variabilis. The pes of Limnopus heterodactylus has a length-to-width ratio of roughly 2:1, whereas Limnopus variabilis has a length-towidth ratio that is nearly 1:1, being only slightly longer in length by 3 mm in the lectotype specimen.
Display large image of Figure 5
26 Given the low relief of the footprints, the type specimen of Limnopus variabilis is likely a shallow underprint, but it is plantigrade and displays the complete morphology of the manus and pes. If the digit impressions were preserved as underprint fallout, the sediment distortion around the digits would widen rather than be preserved as slender impressions. The digits do not exhibit extramorphological digit extensions or extractions, which would lengthen the digits; the digits preserved on the lectotype specimen are short. If digit extractions were present, the digit relief would shallow distally towards the anterior, as the foot was removed from the substrate. The digits are in fact deepest at their anterior, indicating they do not represent extramorphological toe drags.
27 Alternatively, the slender and short nature of the digits of Limnopus variabilis may be due to digit-impression wall collapse. Wall collapse is related to the cohesiveness of the sediments in which the footprint was emplaced. In poorly consolidated sediments that are not cohesive, sediment slumping occurs where the digit imprint walls are steeper than the angle of repose, allowing for sediment collapse into the imprint from either side. This would artificially narrow the base of the imprint. However, the loss of sediment from the trace wall would cause a widening of the digit imprint at the sediment surface. As the trace is probably a shallow underprint, it is unlikely that these surface features would be observed, and thus we are indeed seeing the true foot morphology within the shallow underprint fallout. If better examples were to be discovered, they might show that Limnopus variabilis is a minor morphological variant of Limnopus heterodactylus.
Limnopus versus Batrachichnus
28 Tucker and Smith (2004) recognized the close similarity between the ichnogenus of Limnopus and the variable ichnogenus Batrachichnus first described by Woodworth (1900). Using a multivariate taxonomic analysis of Limnopus and Batrachichnus specimens, Tucker and Smith concluded that size seems to be the only ichnotaxobase upon which to separate the two ichnogenera. They observed two separate but overlapping size distributions. Although size is not a major factor in distinguishing ichnotaxa, based on the possibility that Batrachichnus and Limnopus may be produced by separate biotaxa with separate stratigraphic ranges, as well as the fact the two names are well established in literature, Tucker and Smith retained both ichnotaxa. However, they reduced Batrachichnus to a subgenus of Limnopus, and the designation of Limnopus subgenus Batrachichnus automatically created a second subgenus, the autonym Limnopus subgenus Limnopus. Limnopus vagus was considered to be the only ichnospecies of Limnopus (Limnopus) at the time, while Limnopus (Batrachichnus) was composed of two ichnospecies, Limnopus (Batrachichnus) salamandroides and Limnopus (Batrachichnus) plainvillensis.
29 The taxonomic proposals of Tucker and Smith (2004) were rejected by Lucas et al. (2011) on the basis that a trinomial nomenclature is cumbersome and confuses the ichnotaxonomic assessment of Batrachichnus and Limnopus. Although technically valid, we concur that the use of subichnogenera and subichnospecies should be discouraged as confusing, cumbersome and ultimately unhelpful. Despite morphological similarities, Batrachichnus and Limnopus can be separated arbitrarily at the 20 mm pes-length size boundary. Batrachichnus is considered to extend from 2.38 mm (Stimson et al. 2012) to 20 mm (Lucas et al. 2011; Voigt 2005; Melchor and Sarjeant 2004) in pes length. Limnopus is considered to be morphologically similar but to occupy all size ranges above 20 mm in pes length up to 140 mm (DiMichele et al. 2012). With a pes length of 25 mm, Limnopus variabilis can thus be considered within the size range of Limnopus. If the two ichnogenera were to be considered synonymous, Batrachichnus would have priority; but a systematic review of that ichnogenus is beyond the scope of this manuscript. If the two ichnogenera are considered to be nearly identical and distinguished on size only, then much of the extramorphological variability included within the ichnospecies Batrachichnus salamandroides identified by Haubold et al. (1995) should also be included within the ichnogenus concept of Limnopus.
30 Both Batrachichnus and Limnopus are considered to have been produced by temnospondyls (Haubold 1971, 1984; Melchor and Sarjeant 2004; Van Allen et al. 2005; Stimson et al. 2012) or terrestrial microsaurs (Stimson et al. 2012) and may represent ontogenetic variations. Knowing the size ranges of Limnopus and Batrachichnus, Limnopus variabilis is a medium-sized footprint. It could be argued that Limnopus variabilis is an ontogenetic variation of either: an adult trackway to the smaller juvenile produced Batrachichnus; or a small example of Limnopus made by a juvenile of a larger animal that would produce larger footprints assigned to Limnopus heterodactylus (see Stimson et al. 2012 for detailed discussion of juvenile temnospondyl footprints). Limnopus vagus exhibits the typical morphology of Limnopus heterodactylus and is of similar size to some specimens of Limnopus heterodactylus. Given the wide range in size of Limnopus heterodactylus, we do not consider Limnopus variabilis to be an ontogenetic variation of this ichnospecies. If Limnopus and Batrachichnus are as similar as previous authors (e.g., Tucker and Smith 2004) would suggest, then the extensive research and treatment of Batrachichnus could stand as a model for the taxonomic treatment of Limnopus. The above-mentioned taxonomic work on Batrachichnus suggested a wide range of extramorphological variability that was included within the genus concept (Haubold et al. 1995). By comparison, the ichnotaxonomy for the ichnogenus concept for Limnopus should be handled in a similar manner and it would thus encompass a wide range of extramorphological variations. Limnopus variabilis redescribed here is at least a shallow underprint or a surface trace with low relief that preserves its original morphology that can be distinguished from Limnopus heterodactylus at the ichnospecies level. The morphological differences are minor enough that they should be considered variability under the genus concept by analogy to the treatment of Batrachichnus.
Nanopus (?) vetustus and Bipezia bilobata
31 Matthew (1910) described two other ichnotaxa; Nanopus vetustus (NBMG 3045) (Fig. 6) and Bipezia bilobata (NBMG 3037, 3038) (Fig. 6). Nanopus vetustus was considered by Abel (1935) and Häntzschel (1975) to be a nomen dubium; these authors thus discarded the ichnotaxon. Matthew described Nanopus vetustus (Figs. 6A, B) as a small footmark comparable to other examples of the genus, with maximum dimensions of 12 mm by 12 mm. He described digits as mere lobes, and none of the “footprints” exhibit more than three digits. Abel (1935) considered Nanopus vetustus to be an unrecognizable form and likely not footprints of a tetrapod. We agree that this form was incorrectly identified by Matthew as a tetrapod track and should be regarded as a nomen dubium.
Display large image of Figure 6
32 Bipezia bilobata was described from two specimens (Fig. 7A and Figs 7B–D), although only one of these was illustrated by Matthew (1910) (Fig. 7D). He described the ichnogenus as containing spindle-shaped footprints that were pointed at both ends, found usually in laterally coalescing pairs, one print opposite the other. Matthew noted that some imprints are short and round. Imprints are up to 10 mm in length and up to 3 mm in width. They were later considered by Glaessner (1957) to be a junior synonym of the invertebrate ichnotaxon Isopodichnus, and were included as such in Häntzschel (1975). Species of Isopodichnus are now considered junior synonyms of either Cruziana or Rusophycus (Keighley and Pickerill 1996).
Display large image of Figure 7
33 We have re-examined the type specimen of Bipezia bilobata (NBMG 3038). Indeed, we concur that the surface impression strongly resembles Rusophycus. However the three-dimensional nature of the fossil can be examined along the broken edge of the specimen (Fig. 7C). In cross section the ‘trace’ is a cylindrical tube with an organic rich rind, infilled by fine grey siltstone similar to the sediment encasing it. The surrounding sediment is deformed around the tube. The organic rind suggests that Bipezia bilobata is not a surface trace of either an invertebrate or vertebrate, and the deformed sediments are similar to those of Rygel et al.’s (2004) vegetation-induced sedimentary structures. These observations suggest that Bipezia bilobata represents a hollow tube that the sediments surrounded and encased, rather than any form of infaunal burrow. Bipezia bilobata likely represents the remains of a plant stem or root that has been cast in sediment rather than an ichnofossil; Matthew’s other, unillustrated, specimen of Bipezia bilobata (NBMG 3037) shows that the bilobate impressions do not represent an ichnofossil but are broken fragments of shale that superficially resemble the type specimen. We thus agree with Glaessner (1957) and Häntzschel (1975) that Bipezia bilobata does not represent a vertebrate trackway and regard it as a nomen dubium.
CONCLUSION
34 As Matthew (1910) concluded, specimens he identified as Hylopus (?) variabilis are poorly preserved tetrapod footprints; we consider that they can now be assigned to Limnopus. Matthew published his vertebrate tracks amongst descriptions of arthropod body fossils and invertebrate ichnofossils found in the Upper Carboniferous Lancaster Formation at ‘Fern Ledges’ and a second locality at Barrack Shore (Fig. 1A). Certainly, some of the vertebrate ichnotaxa he described are indeed dubious as discussed above, but the dismissal of the entire assemblage by Abel (1935) and the subsequent assignment of the ichnotaxa to ‘unrecognized or unrecognizable’ in the Treatise on Invertebrate Paleontology (Häntzschel 1975) essentially removed Hylopus (?) variabilis from further review by vertebrate ichnologists. Dismissal by previous authors was based in part on the poor quality of preservation of the tracks, and lack of access to the type material. The detailed re-examination of the original material presented here vindicates Matthew’s ichnotaxon Hylopus (?) variabilis as a vertebrate ichnotaxon and it is here reassigned to the ichnogenus Limnopus based on the similar ichnotaxobase of their ichnogeneric concepts. Although Lucas and Dalman (2013) considered Limnopus a monospecific ichnogenus, Limnopus variabilis can be distinguished from Limnopus heterodactylus (the other valid ichnospecies of Limnopus) in its ichnospecific diagnosis. Matthew’s type specimen does not closely match the genus concept of Hylopus, nor does it conform to the species concept of Limnopus heterodactylus with its short, slender digits with sharply tapered digit terminations. Until further data is available, the ichnospecies Limnopus variabilis should be retained.
The authors thank the City of Saint John for the use of the aerial photograph of Seaside Park and the ‘Fern Ledges’. RFM benefitted from on-site discussions of the ‘Fern Ledges’ geology with Howard Falcon-Lang and Robert Wagner. We thank John Calder, Robert Fensome, and Fabio Petti for their careful reviews of this manuscript.