1. Introduction
The artiopods were a major component during the evolution of the Cambrian marine fauna (Hesselbo, Reference Hesselbo1992; Hou & Bergström, Reference Hou and Bergström1997; Ortega-Hernández et al. Reference Ortega-Hernández, Legg and Braddy2013; Lerosey-Aubril et al. Reference Lerosey-Aubril, Zhu and Ortega-Hernández2017; Du et al. Reference Du, Bruton, Yang and Zhang2023). Within the artiopods, the Vicissicaudata exhibited their highest diversity during the Cambrian and remained fairly widespread and diverse throughout the Ordovician (Lerosey-Aubril et al. Reference Lerosey-Aubril, Zhu and Ortega-Hernández2017). Post-Ordovician, the Vicissicaudata are represented by only a few taxa. Most are cheloniellids, a rare but important group of artiopods that originated in the Late Ordovician (Braddy & Dunlop, Reference Braddy and Dunlop2021; Van Roy et al. Reference Van Roy, Rak, Budil and Fatka2022). The post-Ordovician cheloniellids include Pseudarthron whittingtoni from the late Silurian Ludlow Series in Scotland (Selden & White, Reference Selden and White1983), Cheloniellon calmani from the Lower Devonian Hunsrück Shale in Germany (Stürmer & Bergström, Reference Stürmer and Bergström1978), and Paraduslia talimaae from the Lower Devonian Severnaya Zemlya Formation from Russia (Dunlop, Reference Dunlop2002). In addition to cheloniellids, the post-Ordovician vicissicaudatans include Carimersa neptuni from the Silurian Herefordshire Lagerstätte in the UK (Briggs et al. Reference Briggs, Siveter, Siveter, Sutton, Legg and Lamsdell2023). Here, we describe a new vicissicaudatan artiopod which resolves as the sister taxon to the cheloniellids, Tardisia broedeae gen. et sp. nov., from the Late Carboniferous Mazon Creek fossil site. This new record extends the stratigraphic range of the vicissicaudatan artiopods to the late Paleozoic in the paleotropics and provides new morphological data to reconcile the evolutionary history of crown vicissicaudatans.
2. Materials and methods
2.a. Specimens
Two specimens from the Invertebrate Paleontology collections of the Field Museum of Natural History in Chicago, Illinois, USA were included in this study: FMNH PE928 and FMNH PE88856. FMNH PE928 was collected and then donated to the Field Museum by George Langford and is one-half of a siderite concretion without its counterpart. The counterpart of this specimen was at one point in the possession of George Langford; his notes, including a photograph of the part and counterpart of this specimen and his line-drawing reconstruction, were reproduced and published by the Earth Science Club of Illinois (Langford, Reference Langford and Wittry2018). Langford donated his specimens to either the Field Museum or the Illinois State Museum (Langford, Reference Langford and Wittry2018); however, we could not find the counterpart to this specimen in either collection. FMNH PE88856 is similarly one-half of a siderite concretion without its counterpart. It was originally collected by Thomas V. Testa and was donated to the Field Museum by Jack Wittry.
2.b. Geologic background
The Late Carboniferous (∼308.6–308.4 Ma (Montañez et al. Reference Montañez, McElwain, Poulsen, White, DiMichele, Wilson, Griggs and Hren2016)) Mazon Creek fossil site is famous for its exceptionally diverse and abundant assemblage of fossils preserved inside siderite (FeCO3) concretions (Clements et al. Reference Clements, Purnell and Gabbott2019; Burke et al. Reference Burke, Mayer and McCoy2024). The concretions are found in a three- to eight-metre-thick layer located near the base of the Francis Creek Shale, a grey shale that is up to 25 metres thick (Baird, Reference Baird and Nitecki1979, Reference Baird, Shabica and Hay1997; Baird et al., Shabica & Hay, Reference Baird, Shabica, Hay, Shabica and Hay1997; Clements et al. Reference Clements, Purnell and Gabbott2019; Burke et al. Reference Burke, Mayer and McCoy2024). The Mazon Creek fossils are traditionally divided into two assemblages (Baird et al. Reference Baird, Shabica, Anderson and Richardson1985 a; Baird et al, Reference Baird, Sroka, Shabica and Kuecher1986; Burke et al. Reference Burke, Mayer and McCoy2024): the Braidwood assemblage, which includes terrestrial fauna and flora washed in from a nearshore coal swamp as well as a low-diversity brackish-freshwater fauna; and the Essex assemblage, which encompasses marine-to-brackish animals living in the nearshore delta where the concretions formed. However, some authors argue that the Braidwood/Essex biota split is not accurate, because it represents the endpoints of a continuum rather than two discrete biotas (Clements, Purnell & Gabbott Reference Clements, Purnell and Gabbott2019).
Of the two fossils included in this study, FMNH PE928 does not have very specific locality information. The general region given for where it was collected, between Wilmington, Braidwood, and Coal City, mostly encompasses localities known for the Braidwood assemblage, although also some Essex assemblage localities (Baird et al. Reference Baird, Shabica, Anderson and Richardson1985 a). FMNH PE88856 was found in Pit 11, one of the classic Essex assemblage localities (Johnson & Richardson, Reference Johnson and Richardson1966; Burke et al. Reference Burke, Mayer and McCoy2024). The specific environment in which T. broedeae lived is therefore not clearly constrained.
2.c. Photography
Specimens were photographed with a Canon EOS R5 with a 60mm Macro lens, aperture of f/11, exposure of 1/5 second, ISO of 12800, and no flash under polarized light. Measurements were taken from specimen photographs using ImageJ (Schneider et al. Reference Schneider, Rasband and Eliceiri2012).
2.d. Computed tomography
Both specimens were scanned together separated by a thin piece of wrapping foam using a GE dual tube μCT scanner at the PaleoCT lab in the Department of Organismal Biology and Anatomy at the University of Chicago. The scan used a 240 kV micro-focus tube at 120kV and 130 μA at 250 ms exposure timing, using filter of 0.3 mm Cu. Resolution of voxel x = y = z = 15.1080 μm with 2000 projection images was achieved. The μCT volumes were visualized using Avizo Software 3D 2023.1 (Thermo Fisher Scientific, Waltham, MA, United States). For features that were only visible in the CT scans, measurements were taken from CT scan images using imageJ (Schneider et al. Reference Schneider, Rasband and Eliceiri2012). The computed tomography data have been deposited at https://www.morphosource.org/concern/media/000624704?locale=en
2.e. Phylogenetic analysis
The new arthropod was coded into the dataset of Briggs et al. (Reference Briggs, Siveter, Siveter, Sutton, Legg and Lamsdell2023) in order to ascertain its phylogenetic affinities. The character matrix was unmodified except for the addition of the new species, resulting in a dataset consisting of 104 characters coded for 79 taxa (Tables S1, Table S2). This matrix was analyzed under maximum parsimony optimality criteria using TNT v. 1.5 (Goloboff & Catalano, Reference Goloboff and Catalano2016). The parsimony search strategy employed 10,000 random addition sequences with all characters unordered and of equal weight (Congreve & Lamsdell, Reference Congreve and Lamsdell2016), each followed by tree bisection-reconnection (TBR) branch swapping. Jackknife (Farris et al. Reference Farris, Albert, Källersjö, Lipscomb and Kluge1996), Bootstrap (Felsenstein, Reference Felsenstein1985) and Bremer (Bremer, Reference Bremer1994) support values were also calculated in TNT. Bootstrapping was performed with 50% resampling for 1,000 repetitions, while jackknifing was performed using simple addition sequence and tree bisection-reconnection branch swapping for 1,000 repetitions with 33% character deletion.
3. Systematic palaeontology
Arthropoda Siebold (Reference Siebold1848)
Artiopoda Hou and Bergström (Reference Hou and Bergström1997)
Vicissicaudata Ortega-Hernández et al. (Reference Ortega-Hernández, Legg and Braddy2013)
Remarks. The newly described species is assigned to the Vicissicaudata because it shows several characteristic vicissicaudatan features (Lerosey-Aubril et al. Reference Lerosey-Aubril, Zhu and Ortega-Hernández2017; Briggs et al. Reference Briggs, Siveter, Siveter, Sutton, Legg and Lamsdell2023). The Vicissicaudata have a trunk of 6–19 segments, all except for 1–4 of the posterior-most ones with wide pleura. The new species has seven thoracic tergites in the trunk with wide pleurae, and a one-segment postabdomen, which lacks wide pleurae. In addition, the postabdomen of the new species bears a pair of posteriorly directed caudal appendages, a feature commonly seen in Vicissicaudata. Finally, posterior to the postabdomen is a telson. This assignment is further supported by the results of our phylogenetic analysis.
Tardisia gen. nov.
Etymology. The genus name Tardisia (feminine) is inspired by the TARDIS time machine in the TV show Dr. Who, and refers to the large stratigraphic gap between this species and the next youngest members of the Vicissicaudata.
Type species. Tardisia brodeae gen. nov. sp. nov. by monotypy.
Localities and horizon. As for the type species, by monotypy.
Diagnosis. As for the type species, by monotypy.
T. broedeae sp. nov.
Etymology. The species name broedeae honours Irene Broede, a Mazon Creek collector who volunteers for the Field Museum and is an active member of the Earth Science Club of Northern Illinois (ESCONI).
Holotype. FMNH PE928 (Fig. 1a–c). Almost-complete specimen in dorsal view, preserving part of the head, thorax, postabdomen with caudal appendages and telson. Preserved in a siderite concretion, part only.
Tardisia broedeae gen. et sp. nov. specimens, line drawings, and CT scans. (a–c) Holotype, FMNH PE928. (a) Fossil specimen, photographed under cross-polarized light. (b) Line drawing of FMNH PE928, black dotted lines indicate overlap between the tergites, brown dotted lines show where the specimen is overlain by concretionary material, and the red dotted line at the anterior of the head indicates the anterior margin of the head as reconstructed by Langford. HS is for head shield, TT is for thoracic tergite, PA is for postabdomen, t is for telson, and CA is for the caudal appendages. (c) CT scan of FMNH PE928, reconstructed as if looking through the concretion, and showing the surface of the fossil currently obscured by the concretion. The arrows indicate the bulbous bases of the appendages. (d–f) Paratype FMNH PE 88856. (d) Fossil specimen, photographed under cross-polarized light. (e) Line drawing of FMNH PE 88856, grey dotted lines indicate where the tergite curvature changes, which likely delineates the axis. Brown dotted lines indicate the preserved edges of the postabdomen, which likely do not reflect the complete original shape. HS is for head shield, TT is for thoracic tergite, PA is for postabdomen, t is for telson, and CA is for the caudal appendages. (f) CT scan of FMNH PE88856, reconstructed as if looking through the concretion, and showing the surface of the fossil currently obscured by the concretion. Note the rounded anterior margin of the head. Inset is a digital slice of the CT scan, at the point in the fossil indicated by the black line, showing the appendages in cross section. All scale bars = 1 mm.
Additional material. Paratype FMNH PE88856 (Fig. 1d–f). Almost-complete specimen in dorsal view, preserving the head, thorax and postabdomen with caudal appendages, but missing the telson. Preserved in a siderite concretion, part only.
Localities and horizon. Francis Creek Shale Member of the Carbondale Formation. The holotype, FMNH PE928, was collected between Wilmington, Braidwood, and Coal City, in Will County, Amazon Creek Region, of Illinois, USA. FMNH PE88856, was collected from Pit 11, Peabody Coal Company Northern Mine Locality T31N R9E Sec 6 SE1/4, SE1/4, 41.189128, −88.226618, Mazon Creek Region, Kankakee County, Illinois, USA.
Diagnosis. Head shield smaller than the first thoracic tergite, no eyes. Seven thoracic tergites with wide pleural lobes ending in posteriorly directed points. Eighth trunk segment differentiated into a postabdomen without wide pleural lobes, bearing a pair of long, posteriorly directed, caudal appendages. Short, wide, shield-shaped telson overlaps the bulbous base of the caudal appendages where they insert into the postabdomen.
Description. The holotype, specimen FMNH PE928 (Fig. 1a–c) comprises one-half of a siderite concretion with the fossil preserved in negative relief (Fig. 1a). The overall shape of the body approximates an elongated oval, with its widest point at the third thoracic tergite (Fig. 1a–c). This specimen is missing the anterior-most portion of the head shield, but the preserved portion of the body is 10.6 mm long from the anterior-most preserved part of the head to the posterior-most tip of the telson, and 6.2 mm wide at its widest point in the third trunk tergite. According to Langford’s photograph and line drawing (Langford, Reference Langford and Wittry2018), the head shield was fully preserved in the half of the concretion that is now lost and was rounded and sub-semicircular (∼106° arc), smaller than the first thoracic tergite and lacking eyes (Fig. 1b).
The trunk comprises seven tergites, a postabdomen and a telson (Fig. 1a–c). Seven tergites comprise the thorax and terminate in posteriorly directed points. In four of these, the posterior margin of the tergite overlaps the anterior margin of the next tergite, suggesting that each tergite overlaps the posterior one (Fig. 1a, b). The eighth trunk segment is differentiated into a narrower postabdomen, lacking wide pleura, which bears a pair of caudal appendages. The seven trunk tergites and the postabdomen are all approximately 1 mm long. From the first to the seventh trunk tergite, their respective widths are 5.7 mm, 5.9 mm, 6.2 mm, 5.7 mm, 4.9 mm, 3.9 mm and 3.0 mm; the postabdomen is 2.5 mm wide. The telson is shield-shaped (Fig. 1a–c), 1.6 mm long and 2.1 mm wide at its widest point (its anterior end). The axis is not clearly differentiated in this specimen.
The caudal appendages on the postabdomen are the only appendages preserved in this specimen, either on the concretion half we have, or on the half that is now lost (Langford, Reference Langford and Wittry2018); no appendages preserved inside the concretion were observed in the CT scan. The caudal appendages on the postabdomen are not fully preserved, both lack their posterior-most end, and the region where the left appendage intersects the telson is obscured by some concretionary material (Fig. 1a–c). The right appendage is 0.5 mm wide where it intersects the telson, and the left appendage (which has a longer preserved length) extends 4.5 mm beyond the telson. In the fossil (Fig. 1a, b), both appendages narrow continuously from anterior to posterior. In the CT scan (Fig. 1c), it is apparent that these appendages have bulbous anterior ends where they meet the postabdomen, which is covered by the telson. Based on the CT scans (Fig. 1c), the appendages are about 1 mm wide where they meet the postabdomen, and the maximum preserved appendage length is 6.1 mm. These appendages are straight, and there is no evidence as to whether they were flexible or movable. Their preservation in these specimens when the trunk biramous appendages were not preserved suggests the caudal appendages were heavily sclerotized, with similarly robust cuticles as the dorsal exoskeleton.
The paratype specimen FMNH PE88856 (Fig. 1d–f) also comprises one-half of a siderite concretion with the fossil preserved in negative relief (Fig. 1d). The overall shape of the body is oval, with its widest point at the third trunk tergite (Fig. 1d–f). This specimen is missing the telson, and the postabdomen is poorly preserved (Fig. 1d–f), but the preserved portion of the body measures 8.6 mm long from the head to the posterior-most preserved part of the postabdomen, and 5.6 mm wide at its widest point in the third thoracic tergite. The head shield of this specimen is completely preserved. It is smaller than the first trunk tergite and bears no eyes. In the fossil, the head shield appears to have a straight anterior margin (Fig. 1d). However, in the CT scan, it is clear that the head is rounded and sub-semicircular (∼94° arc) (Fig. 1f).
The trunk comprises seven tergites, a poorly preserved postabdomen, and the telson is not preserved (Fig. 1d–f). The first seven tergites, which form the thorax, terminate in posteriorly directed points, although these are less pronounced than in FMNH PE928 (Fig. 1). The curvature of some of the trunk tergites changes abruptly, which likely delineates the axis of the body (Fig. 1d–f). The axis is ∼2 mm wide in all the segments where it can be measured. The first three tergites have tergopleurae that curve slightly towards the anterior of the body (Fig. 1d–f); this might just be due to flattening the three-dimensional body at an angle (Van Roy et al. Reference Van Roy, Rak, Budil and Fatka2022). The eighth trunk segment is too poorly preserved to confirm its differentiation into the postabdomen (Fig. 1d–f); however, it does bear a pair of caudal appendages. The first four trunk tergites are ∼1 mm long, the fifth through seventh trunk tergites are each ∼0.6 mm long and the postabdomen is too poorly preserved to measure. From the first to the seventh trunk tergites, their respective widths are 5.0 mm, 5.4 mm, 5.6 mm, 5.4 mm, 4.6 mm, 3.8 mm and 2.9 mm. The telson is not preserved.
As with FMNH PE928, the caudal appendages on the postabdomen are the only appendages preserved in this specimen and no appendages preserved inside the concretion were observed in the CT scan. In this specimen as well as in FMNH PE928, there is no evidence as to whether the caudal appendages were flexible or movable. The appendages are not fully preserved, and both lack their posterior-most end (Fig. 1d–f). The bulbous base of the appendages, where they meet the postabdomen, is clearly visible for both appendages; this was obscured by the telson in FMNH PE928 (Fig. 1a–c) and indicates that the telson is not preserved in FMNH PE88856. The appendages are ∼0.6 mm wide where they meet the postabdomen and extend 4.7 mm posterior to the postabdomen. These appendages have a striated appearance (Fig. 1d). These appendages are preserved with high relief and are semicircular in cross section (Fig. 1f inset); however, the appendages may have been compressed during fossilization, in which case they could have originally had more relief, potentially to the point of being circular in cross section. The cross section (Fig. 1f inset) also shows that the two appendages are separate. Based on the features revealed in the CT scan – two separate structures that are semicircular to circular in cross section – we can discard a tailspine interpretation for these appendages.
4. Discussion
The lack of preserved trunk biramous appendages in T. broedeae gen. et sp. nov. makes it difficult to directly evaluate it for the diagnostic artiopod features, which all refer to the appendages (Stein & Selden, Reference Stein and Selden2012). However, the results of our phylogenetic analysis are congruent with those of Briggs et al. (Reference Briggs, Siveter, Siveter, Sutton, Legg and Lamsdell2023) and place T. broedeae within the artiopods and specifically within the Vicissicaudata as the sister group to the Cheloniellida (Fig. 2a; Fig. S1). The cheloniellids, including the position of Tardisia as sister taxon, are one of the more well-supported groups in the phylogeny with high internal node support values. T. broedeae does exhibit the morphological characters which unite the Vicissicaudata (Lerosey-Aubril et al. Reference Lerosey-Aubril, Zhu and Ortega-Hernández2017; Briggs et al. Reference Briggs, Siveter, Siveter, Sutton, Legg and Lamsdell2023): trunk of 6–19 segments, all except for 1–4 of the posteriormost ones with wide pleura, the last trunk segment bearing a pair of caudal appendages. It is unusual in that it lacks eyes, contrary to most vicissicaudatans (Lerosey-Aubril et al. Reference Lerosey-Aubril, Zhu and Ortega-Hernández2017), although many cheloniellids, such as Neostrabops martini, Duslia insignis, an unnamed Duslia-like cheloniellid from Morocco, and Paradusila talimaae are reported to lack eyes (Dunlop, Reference Dunlop2002; Braddy & Dunlop, Reference Braddy and Dunlop2021).
Phylogenetic analysis, artist’s reconstruction of Tardisia broedeae gen. et sp. nov., and line-drawing comparisons to some cheloniellids. (a) Phylogenetic tree showing the placement of T. broedeae as the sister taxon to the cheloniellids. (b) Artist’s reconstruction of T. broedeae in the Mazon Creek ecosystem; appendages other than the caudal appendages are not reconstructed; art by Sofia Herrera. (c–f) Line drawings of T. broedeae and other cheloniellids (not to scale with each other) showing key morphological comparisons. Appendages other than the caudal appendages are omitted from the line drawings. The head shield is white, the thoracic tergites are grey, the postabdomen is red, the telson is yellow, and the caudal appendages, if present, are blue. (c) T. broedeae. (d) Cheloniellon calmani, redrawn and simplified from Stürmer and Bergström (Reference Stürmer and Bergström1978). (e) Triopus draboviensis, redrawn and simplified from Van Roy et al. (Reference Van Roy, Rak, Budil and Fatka2022). (f) Paraduslia talimaae, redrawn and simplified from Dunlop (Reference Dunlop2002).
Cheloniellids are united by a number of features, some of which can be evaluated in T. broedeae. Like cheloniellids, T. broedeae has a wide, oval body, with a head shield narrower than the first tergite (Dunlop, Reference Dunlop2002; Braddy & Dunlop, Reference Braddy and Dunlop2021). Many vicissicaudatans have a narrow, elongate tail spine (Lerosey-Aubril et al. Reference Lerosey-Aubril, Zhu and Ortega-Hernández2017; Briggs et al. Reference Briggs, Siveter, Siveter, Sutton, Legg and Lamsdell2023); in contrast, cheloniellids have a short pointed telson described as conical or cap-shaped (Stürmer & Bergström, Reference Stürmer and Bergström1978; Braddy & Dunlop, Reference Braddy and Dunlop2021; Van Roy et al. Reference Van Roy, Rak, Budil and Fatka2022), which, if dorsoventrally flattened, would likely appear very similar to the shield-shaped telson of T. broedeae. However, the cheloniellid telson, along with its postabdomen, are reduced in size, and are relatively much smaller than the postabdomen and telson of T. broedeae. The tergites of cheloniellids, with the exception of Neostrabops martini (Braddy & Dunlop, Reference Braddy and Dunlop2021), typically exhibit a pronounced, narrow axis (Dunlop, Reference Dunlop2002). The axis of T. broedeae is not particularly distinct, and the axis and tergopleurae each make up about one-third the width of the thoracic tergites. Moreover, the tergopleurae in cheloniellids exhibit a distinct radial arrangement (Braddy & Dunlop, Reference Braddy and Dunlop2021; Van Roy et al. Reference Van Roy, Rak, Budil and Fatka2022) which is not seen in T. broedeae. Van Roy et al. (Reference Van Roy, Rak, Budil and Fatka2022) identified a unique articulating device in Triopus draboviensis and suggested it may represent a cheloniellid apomorphy; T. broedeae does not appear to preserve a similar articulating device. Finally, the elongate pair of posteriorly directed caudal appendages on the postabdomen of T. broedeae seem generally very similar to the furca on the postabdomen of cheloniellids, and particularly to the elongate filament-like furca with expanded anterior ends in C. calmani (Stürmer & Bergström, Reference Stürmer and Bergström1978), although it is not clear from these specimens if these appendages of T. broedeae insert dorsally as do the furca in cheloniellids (Dunlop, Reference Dunlop2002; Lerosey-Aubril et al. Reference Lerosey-Aubril, Zhu and Ortega-Hernández2017; Braddy & Dunlop, Reference Braddy and Dunlop2021; Van Roy et al. Reference Van Roy, Rak, Budil and Fatka2022).
T. broedeae, from the Late Carboniferous Mazon Creek deposit, is the youngest vicissicaudatan yet described and extends the stratigraphic range of the group into the late Paleozoic. The next youngest, from the Devonian, are C. calmani (Stürmer & Bergström, Reference Stürmer and Bergström1978) and P. talimaae (Dunlop, Reference Dunlop2002). However, post-Ordovician vicissicaudatans are rare and typically represented by only a small number of near-complete, well-preserved specimens (Stürmer & Bergström, Reference Stürmer and Bergström1978; Dunlop, Reference Dunlop2002; Briggs et al. Reference Briggs, Siveter, Siveter, Sutton, Legg and Lamsdell2023). Further investigations of Late Paleozoic Lagerstätten may yield additional examples. In particular, other Mazon Creek-type fossil assemblages, which have similar environments, similar faunas, and similar preservation (Baird et al. Reference Baird, Sroka, Shabica and Beard1985 b; Briggs & Gall, Reference Briggs and Gall1990) could be a fruitful place to explore.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S001675682400044X
Acknowledgements
We thank the Field Museum of Natural History for facilitating access to these specimens, April Neander for aid with X-ray tomography, Mane Pritza for specimen photography, Tom McCoy for help with Latin declensions, Sofia Herrera for the fossil reconstruction, as well as Javier Ortega-Hernández and an anonymous reviewer for helpful comments. Funding for this work was provided by the Negaunee Integrative Research Center (Field Museum) and by A. Susman to FH.
Competing interests
The authors declare none.
Open access
