Non-technical Summary
Xandarellids are Cambrian arthropods belonging to Artiopoda, the diverse extinct group that includes trilobites. They are known from six species belonging to five genera from the early Cambrian of China, and possibly a sixth genus known from Siberia. A new species from Australia’s most informative site for exceptional fossil preservation in the Cambrian, the Emu Bay Shale in South Australia, is the first xandarellid discovered in the southern supercontinent Gondwana. It closely resembles the Chinese genus Xandarella but is distinguished by a longer tail piece. This similarity reinforces biogeographic connections between Australia and South China in the early Cambrian.
Introduction
Phylogenetic analyses and systematic arrangements of Paleozoic euarthropods over the past 25 years have identified a clade known as Artiopoda Hou and Bergström, Reference Hou and Bergström1997 that accommodates the trilobites and their relatives. They are dorsoventrally flattened, mostly benthic species, with relatively undifferentiated biramous post-antennal appendages. The exopods typically have bipartite flaps, the more proximal of which bears imbricate lamellar setae (see Stein and Selden, Reference Stein and Selden2012 for re-diagnosis, and Ortega-Hernández et al., Reference Ortega-Hernández, Legg and Braddy2013 for phylogenetic definition). Artiopodans are diverse and abundant in exceptionally preserved Cambrian biotas. Although the interrelationships of the major artiopodan clades continue to vary between different phylogenetic analyses (e.g., Edgecombe et al., Reference Edgecombe, Paterson and García-Bellido2017; Lerosey-Aubril et al., Reference Lerosey-Aubril, Zhu and Ortega-Hernández2017; Chen et al., Reference Chen, Ortega-Hernández, Wolfe, Zhai, Hou, Chen, Mai and Liu2019; Du et al., Reference Du, Ortega-Hernández, Yang and Zhang2019, Reference Du, Guo, Losso, Pates, Li and Chen2024; Hou et al., Reference Hou, Williams, Sansom, Siveter, Siveter, Gabbott, Harvey, Cong and Liu2019; Schmidt et al., Reference Schmidt, Hou, Zhai, Mai, Belojević, Chen, Melzer, Ortega-Hernández and Liu2022; Zhu et al., Reference Zhu, Zeng, Liu and Zhao2023; Liu et al., Reference Liu, Fu, Wu and Zhang2024; McCoy et al., Reference McCoy, Herrera, Wittry, Mayer and Lamsdell2025 for recent examples), a subset of major groups, such as nektaspidids, helmetiids, and xandarellids, has consistently been identified, and a close alliance of these clades to trilobites is a recurring theme.
Within the trilobitomorph clade of artiopodans, Xandarellida currently comprises five genera from the early Cambrian Chengjiang Konservat–Lagerstätte of South China. These are the nominate genus, the monotypic Xandarella Hou, Ramsköld, and Bergström, Reference Hou, Ramsköld and Bergström1991, the monotypic Cindarella Chen, Ramsköld, Edgecombe, and Zhou in Chen et al., Reference Chen, Zhou, Zhu and Yeh1996, the likewise monotypic Sinoburius Hou, Ramsköld, and Bergström, Reference Hou, Ramsköld and Bergström1991 (following revisions by Chen et al., Reference Chen, Ortega-Hernández, Wolfe, Zhai, Hou, Chen, Mai and Liu2019 and Schmidt et al., Reference Schmidt, Liu, Hou, Haug, Haug, Mai and Melzer2021), two species of Luohulinella Zhang, Fu, and Dai, Reference Zhang, Fu and Dai2012, and the monotypic Zhugeia Zhu et al., Reference Zhu, Zeng, Liu and Zhao2023. Phytophilaspis Ivantsov, Reference Ivantsov1999, based on P. pergamena Ivantsov, Reference Ivantsov1999, from the early Cambrian Sinsk Biota of Russia, has been grouped with xandarellids in phylogenetic analyses (Paterson et al., Reference Paterson, Edgecombe, García-Bellido, Jago and Gehling2010, Reference Paterson, García-Bellido and Edgecombe2012), variably included or excluded from the group using different tree reconstruction methods (Lerosey-Aubril et al., Reference Lerosey-Aubril, Zhu and Ortega-Hernández2017; Jiao et al., Reference Jiao, Du, Zhang, Yang and Eggink2022; Schmidt et al., Reference Schmidt, Hou, Zhai, Mai, Belojević, Chen, Melzer, Ortega-Hernández and Liu2022; Zhu et al., Reference Zhu, Zeng, Liu and Zhao2023; Liu et al., Reference Liu, Fu, Wu and Zhang2024), and when nested within Xandarellida is closest to either Luohuilinella (Du et al., Reference Du, Ortega-Hernández, Yang and Zhang2019) or Xandarella (Chen et al., Reference Chen, Ortega-Hernández, Wolfe, Zhai, Hou, Chen, Mai and Liu2019). The only other species assigned to Xandarellida, Xandarella mauretanica Ortega-Hernández et al., Reference Ortega-Hernández, Abdelfattah, Hearing, Harvey, Edgecombe, Hafid and El Hariri2017, from the lower Cambrian Tatelt Formation of Morocco, was later identified as an ellipsocephaloid trilobite belonging to the palaeolenid genus Gigoutella Hupé, Reference Hupé1953 (El Albani et al., Reference El Albani, Mazurier, Edgecombe, Azizi and El Bakhouch2024) and accordingly reclassified within Trilobita as Gigoutella mauretanica (Ortega-Hernández et al., Reference Ortega-Hernández, Abdelfattah, Hearing, Harvey, Edgecombe, Hafid and El Hariri2017).
As in other Cambrian Burgess Shale-type biotas, non-biomineralized artiopodans are represented by several genera in the Emu Bay Shale Konservat–Lagerstätte on the north coast of Kangaroo Island, South Australia (Paterson et al., Reference Paterson, Edgecombe, García-Bellido, Jago and Gehling2010, Reference Paterson, García-Bellido and Edgecombe2012; Edgecombe et al., Reference Edgecombe, Paterson and García-Bellido2017). A new xandarellid described herein is a rare component of the Emu Bay Shale biota, with only two specimens having been found in 23 field seasons of excavating at Buck Quarry between 2007 and 2019. Before our field program, two additional specimens were collected at a coastal exposure of the Emu Bay Shale on the wave cut platform immediately east of the mouth of Big Gully; one of them was identified as Xandarella sp. in an unpublished Ph.D. thesis (Nedin, Reference Nedin1995). The stratigraphy, sedimentology, paleoenvironment and biota of the Emu Bay Shale at Buck Quarry are documented by Gehling et al. (Reference Gehling, Jago, Paterson, García-Bellido and Edgecombe2011), Paterson et al. (Reference Paterson, García-Bellido, Jago, Gehling, Lee and Edgecombe2016), and Gaines et al. (Reference Gaines, García-Bellido, Jago, Myrow and Paterson2024).
Material and methods
Four specimens from the Emu Bay Shale serve as the basis of a new species.
Photography used a Canon EOS 50D digital SLR camera with a Canon Compact-Macro EF 50 mm and Canon MP-E 65 mm macro lens through Canon EOS Utility 2 capture software. Camera lucida drawings were made with an Olympus SZX10 stereomicroscope. Images were added and figure plates assembled with Photoshop version 2024.
The phylogenetic position of the new Australian xandarellid was tested by coding it in a recent, densely sampled character matrix for Artiopoda (Zhu et al., Reference Zhu, Zeng, Liu and Zhao2023). One new character was added to that 93-character matrix, the revised matrix being available in nexus format as Morphobank Project 5963 (morphobank.org). Character 94 is as follows: length of posterior three thoracic tergites: similar to anterior thoracic tergites (0); progressively lengthening posteriorly (1). Applicability is restricted to trilobitomorph taxa with free thoracic segments and a pygidium (see “Terminology” below).
The matrix was analyzed using parsimony as an optimality criterion with TNT version 1.5 (Goloboff and Catalano, Reference Goloboff and Catalano2016) under implied character weights. Concavity constants k = 2 to 16 were explored, using basic weights. A traditional heuristic tree search with Tree Bisection–Reconnection (TBR) branch swapping was employed in each of those 15 analyses, using 2,000 stepwise addition sequences, saving 10 trees per iteration. Node support was evaluated with symmetric resampling (group present/contradicted, G/C ratio) with 1,000 replicates, each with 33% change probability.
Terminology
Nomenclature for the xandarellid trunk has varied according to different interpretations of tagmosis, such as whether the term “pygidium” accurately describes the multi-somitic terminal region (Ramsköld et al., Reference Ramsköld, Chen, Edgecombe and Zhou1997; Cotton and Braddy, Reference Cotton and Braddy2004). Recent work has used the terms “thorax” and “pygidium” to describe the trunk tergites (Chen et al., Reference Chen, Ortega-Hernández, Wolfe, Zhai, Hou, Chen, Mai and Liu2019; Hou et al., Reference Hou, Williams, Sansom, Siveter, Siveter, Gabbott, Harvey, Cong and Liu2019; Schmidt et al., Reference Schmidt, Liu, Hou, Haug, Haug, Mai and Melzer2021) rather than “thorax” and “abdomen” (Hou and Bergström, Reference Hou and Bergström1997), and we follow the former usage here. Thoracic tergites are abbreviated with the prefix “T”.
Repository and institution abbreviation
The four specimens serving as the basis of a new species are deposited in the paleontology type collection of the South Australian Museum, Adelaide, South Australia (prefixed SAMA P).
Systematic paleontology
Phylum Euarthropoda Lankester, Reference Lankester1904
Class Artiopoda Hou and Bergström, Reference Hou and Bergström1997
Order Xandarellida Chen, Ramsköld, Edgecombe, and Zhou in Chen et al., Reference Chen, Zhou, Zhu and Yeh1996
Genus Austroxandarella new genus
Type species
Austroxandarella poikar n. gen. n. sp.
Diagnosis
Cephalic shield broad, semi-elliptical; anterolateral margins evenly curved, without notches. Eye slits and dorsal bulges absent. Trunk with 10 pleura-bearing thoracic tergites (T1–T10), anterior seven of subequal length (sag., exsag.), with coincident spacing of transverse ridges across axis and parts of pleural field; T8 slightly longer; T9 and T10 and coincident transverse ridges two or three times longer (sag.)/more widely spaced than anterior seven tergites/ridges, respectively. Gut tubular, without paired midgut diverticula. Pygidium elongate, triangular; tapering axial keel rising as a sharp carina in its terminal part, not observed to extend as a median spine.
Etymology
Compounding Austro-, for the Australian provenance of the type species, and -xandarella, for inferred affinity to Xandarella. Gender feminine.
Remarks
The phylogenetic analysis (Fig. 1) supports membership of Austroxandarella in Xandarellida, the new genus resolving within a xandarellid clade under all explored concavity constants for implied character weights. Under aggressive concavity constants (k = 2 and 3), Xandarellida includes two taxa that are not usually classified within the group, Zhiwenia coronata Du et al., Reference Du, Ortega-Hernández, Yang and Zhang2019 and Tremaglaspis vanroyi Lerosey-Aubril et al., Reference Lerosey-Aubril, Ortega-Hernández, Kier and Bonino2013 (Fig. 1.1). These two species also grouped with and within Xandarellida, respectively, under k = 3 in the study by Zhu et al. (Reference Zhu, Zeng, Liu and Zhao2023). Strong but less aggressive concavity constants (k = 4 to 9) recover a scope of Xandarellida more consistent with previous classifications (Chen et al., Reference Chen, Ortega-Hernández, Wolfe, Zhai, Hou, Chen, Mai and Liu2019); the group is divided into a clade with eye notches (Cindarella and Luohuilinella) and a clade that unites Sinoburius and Zhugeia on the one hand and Austroxandarella, Xandarella, and Phytophilaspis on the other (Fig. 1.2). As homoplastic characters are less strongly downweighted (k = 10 to 16), the same two main clades are recovered within Xandarellida, but interrelationships of the five genera in the clade that includes Xandarella and Austroxandarella differ (Fig. 1.3). Node support is weak across all concavity constants, the only clade receiving a G/C ratio of 50% or greater being Sinoburius + Zhugeia.
Xandarellid interrelationships based on 94 morphological characters under implied weights. (1) Strict consensus of 10 trees for concavity constants k = 2 and 3. (2) Tree for concavity constants k = 4, 5, 6, 7, 8, and 9. (3) Tree for concavity constants k = 10, 11, 12, 13, 14, 15, and 16.
A thorax composed of 10 tergites with pleurae is shared by Austroxandarella and Xandarella, and in both cases the anterior seven are of uniform length, with a progressive elongation of the posterior three tergites. The two genera also resemble each other in having a semi-elliptical cephalic shield (albeit relatively wider in Austroxandarella), without lateral notches to accommodate the eyes as are present in Cindarella and especially Luohuilinella. The new genus is distinguished from Xandarella by having a longer, triangular pygidium, without a single strong pair of marginal spines seen in the latter genus. Austroxandarella has a carina arising on the posterior part of a median keel on the pygidium, but the posteromedian margin of the pygidium appears to be blunt, the keel not extending as a spine as is the case in Xandarella.
No specimen of Austroxandarella poikar clearly preserves the boundary between the cephalic shield and thorax in the sagittal region, so it is presently not possible to determine whether a reduced axial tergite is present at the front of the trunk as is the case in other xandarellids (including Xandarella); character 48, describing this character, is coded as uncertain in A. poikar. The anteriormost of a series of transverse rods in SAMA P50364a is much narrower than those associated with pleura-bearing thoracic segments and, although appearing to lie within the posterior part of the cephalon (Fig. 2.1, 2.2), may be associated with the homologue of the axial tergite in other xandarellids.
Austroxandarella poikar n. gen. n. sp., lower Cambrian Emu Bay Shale, Kangaroo Island, South Australia; holotype SAMA P50364a,b. (1, 4) SAMA P50364a, incomplete exoskeleton and close-up of transverse rods coincident with T2–T6. (2) Composite camera lucida drawing of part (SAMA P50364a) and counterpart (SAMA P50364b). (3) SAMA P50364b, close-up of pygidium. CS = cephalic shield; G = gut; Pyg = pygidium; Tn = thoracic tergites. (1–3) Scale bars = 5 mm; (4) scale bar = 2 mm.
Austroxandarella poikar new species
Holotype
SAMA P50364a,b (Fig. 2), from level 10.7 in Buck Quarry (34°35′25″S, 137°34′36″E), Emu Bay Shale, Big Gully, north coast of Kangaroo Island, South Australia; Cambrian Series 2, Stage 4, Pararaia janeae Zone.
Diagnosis
As for the genus, by monotypy.
Description
Holotype SAMA P50364a, b (Fig. 2).—Length of dorsal shield from anterior margin of cephalic shield to preserved posterior end of pygidium at least 57 mm.
Cephalic shield estimated (from largely complete right side) to be ~29 mm wide and at least 2.1 times wider than long; outline of cephalic shield broadly semi-elliptical with evenly curved anterior and lateral margins. Surface of cephalic shield smooth, without a visible dorsal bulge or eye slit. Transverse rod as described below in thorax confined to axial region in posteriormost part of cephalon.
Posterior margins of thoracic tergites spaced equally to that between raised transverse rods (in part, P50364a) or grooves (in counterpart, P50364b) across the thoracic axis and a variable extent of the pleural fields. Rods/grooves extend a short distance onto the pleural field in anterior half of trunk (at the boundary of T6 and T7), lengthening to span much of the pleural field in posterior half of trunk; rods coincident with T7 and T8 transverse across their width, gently reflexed forward distally in more anterior tergites; rods coincident with T9 and T10 gently arched forward medially, paralleling and situated close to posterior margin of their respective tergite. Spacing between rods coincident with T1 and T7, ranging from 2.0 to 2.6 mm sagittally; rods coincident with T8 separated by 3.0 mm; rods coincident with T9 and T10 separated by 5.2 and 6.6 mm (sag.). Rods subdivided into small, roughly parallel blocks across their width, a few of them expanded at distal tip (Fig. 2.4). Gut is a continuous tube between at least T3 and T8 (“G” in Fig. 2.2), preserved as irregular surface where phosphatized fill is locally present but has mostly split out.
Pygidium broadly triangular in outline, at least 11.6 mm long, length ~80% maximum width; lateral margins convex, without marginal spines. Axial keel along length of pygidium, tapering posteriorly from midlength, of low relief except near posterior termination, where it forms a sharp, narrow carina that terminates at pygidial tip.
Paratype SAMA P49130a, b (Fig. 3).—Length of dorsal shield between break across posteriormost part of cephalic shield and preserved posterior end of pygidium is 62 mm.
Austroxandarella poikar n. gen. n. sp., lower Cambrian Emu Bay Shale, Kangaroo Island, South Australia; paratype SAMA P49130a,b. (1, 3, 4) SAMA P49130a, incomplete exoskeleton and close-ups of exopods and endopods (3, endopods indicated by white arrowheads) and gut (4). (2) Composite camera lucida drawing of part (SAMA P49130a) and counterpart (SAMA P49130b). (5) SAMA P49130b, incomplete exoskeleton showing transverse rods coincident with tergites. CS = cephalic shield; DES = distal exopod shaft; En = endopods; LS = lamellar setae on exopods; G = gut; PES = proximal exopod shaft; Pyg = pygidium; Tn = thoracic tergites. (1, 2) Scale bars = 10 mm; (3, 5) scale bars = 5 mm; (4) scale bar = 2 mm.
Thoracic tergites 1–7 and spacings of transverse rods coincident with them range between 2.8 and 3.1 mm long sagittally; rods coincident with T8, T9, and T10 separated by 5.0, 7.0, and 10.2 mm (sag.), respectively. Posterior margins of tergites mostly straight, curving backward distally, terminating as sharp, falcate tips from at least T4. Lateral margin of tergites with a thin marginal rim (see tergite 4 in Fig. 3.2, 3.5). In anterior part of trunk (T1–T6), transverse rods nearly straight across their width on right side of body in part (SAMA P49130a), abruptly terminating near where pleural tips flex backward; rods with a sinuous extent on left side. Transverse rods coincident with T8 and T9 bowed forward medially, situated in posterior part of their respective tergite. Gut is a continuous tube with irregularly pinched and swollen margins along entire thorax and at least anterior half of pygidium, gently tapering from T8, with closely spaced transverse wrinkles of diagenetically altered fill (Fig. 3.4).
Closely spaced, parallel, longitudinally aligned lamellae beneath T4, T5, and T6 in part (P49130a) and counterpart (P49130b) interpreted as exopod lamellar setae (“LS” in Fig. 3.2, 3.3); in P49130a, slender ridges between and oblique to lamellar setae interpreted as proximal shafts of exopods. Three slender, weakly curved, tapering ridges on outer part of right pleural field (crossing T1 to T3) of P49130a possibly represent distal shafts of exopods (“DES” in Fig. 3.2, and in counterpart Fig. 3.5). Diagonal ridges impressed on outer part of pleural field of T9, T10, and pygidium in P49130a are likely to represent slender endopods, with each tergite exhibiting multiple ridge impressions (Fig. 3.2, 3.3) and thus multipodous.
Length from posteriormost transverse rod to preserved posterior end of pygidium 18 mm; pygidium ~16.8 mm long. Pygidium elongate triangular, apparently blunt posteromedially.
Paratype SAMA P52874 (Fig. 4).—Length of dorsal shield from anterior margin of cephalic shield to incomplete pygidium is 85 mm.
Austroxandarella poikar n. gen. n. sp., lower Cambrian Emu Bay Shale, Kangaroo Island, South Australia: paratype SAMA P52874nc. (1) Incomplete exoskeleton. (2) Camera lucida drawing. CS = cephalic shield; Pyg = pygidium; Tn = thoracic tergites. Scale bars = 10 mm.
Cephalic shield inferred (from complete left side) to be ~61 mm wide and 2.3 times wider than long; outline of cephalic shield broadly semi-elliptical with evenly curved anterior and lateral margins.
Thorax widest across T3, estimated (from complete left half) to be 66 mm wide. Thoracic tergites 3–7 and spacings of transverse rods coincident with them 2.7–3.5 mm long sagittally; T8 and its coincident transverse rod of similar length/spacing to anterior seven tergites; T9 and T10 lengthening progressively, T9 ~1.4 times length (exsag.) of T7 in mid-distal part of pleural field, T10 twice length of T7. Posterior margins of tergites transverse in axial region, straight across much of pleural field, transverse across pleural field in T4–T6, but weakly directed anteriorly in anterior part of thorax; distal part of pleurae curving backward, terminating as an angular falcate tip. Straight transverse rods situated near midlength of tergites in axial region from T2 to T8.
Paratype SAMA P40811a, b (Fig. 5).—Length of dorsal shield from anterior margin of cephalic shield to broken edge across pleural tips of T10 and pygidium is 62 mm.
Austroxandarella poikar n. gen. n. sp., lower Cambrian Emu Bay Shale, Kangaroo Island, South Australia: paratype SAMA P40811a,b. (1) SAMA P40811a, incomplete exoskeleton. (2) Composite camera lucida drawing of part (SAMA P40811a) and counterpart (SAMA P40811b). Anterior and posterior margins of tergites indicated by gray and black arrowheads, respectively. CS = cephalic shield; En? = probable endopods; Tn = thoracic tergites. Scale bars = 10 mm.
Reconstruction of the xandarellid euarthropod Austroxandarella poikar n. gen. n. sp. from the lower Cambrian Emu Bay Shale, Kangaroo Island, South Australia.
Cephalic shield inferred (from nearly complete outline on counterpart P40811b) to be ~51 mm wide and 2.2 times wider than long; outline of cephalic shield broadly semi-elliptical with evenly curved anterior and lateral margins; posterolateral corner not extended as a genal spine.
Thorax widest across T2, very slightly wider than cephalic shield. Thoracic tergites 1–6 measuring 2.1–2.9 mm long sagittally; T8 approximately twice as long as anterior tergites; T9 and T10 substantially longer. Posterior margins of tergites transverse in axial region, straight across much of pleural field, transverse across pleural field in T4–T6, but weakly directed anteriorly in anterior part of thorax; distal part of pleurae curving backward, terminating as an angular falcate tip. Anterior margins of thoracic tergites visible in recurved distal part of thorax show broad overlap by succeeding tergite.
Three slender, elongate, variably curved structures preserved in relief with distinctive gray coloration in P40811a originating at transition between axis and pleural field in posteriormost part of cephalic shield and first thoracic tergite, the posterior two extending across most of pleural field. Position and morphology consistent with endopods (“En?” in Fig. 5.2).
Etymology
From poikar, meaning ‘prawn’ in Ngarrindjeri Aboriginal language, the traditional custodians of the southern Fleurieu Peninsula who inhabited the adjacent Kangaroo Island.
Materials
Paratypes: SAMA P49130a,b (Fig. 3), from level 9.8 in Buck Quarry; SAMA P40811a,b (Fig. 4), SAMA P52874nc (no counterpart) (Fig. 5), from the wave-cut platform, immediately east of the mouth of Big Gully.
Remarks
Interpretation of the morphology of Austroxandarella poikar in the context of characters used in the systematics of Xandarellida is detailed in Discussion that follows.
Discussion
Interpretation of trunk structures
The four available specimens of Austroxandarella poikar n. gen. n. sp. complement each other, allowing for the identification of segmental structures in the entire trunk. Specimen P50364 most clearly displays segmentation of the trunk in the form of transverse rods, which are strongly expressed in all segments (Fig. 2.1, 2.2). The posterior margins of the thoracic tergites are faintly expressed but can be traced between all tergites, showing a one-to-one correspondence between tergites and transverse rods; however, the position of the rod varies with respect to the length (sag., exsag.) of each tergite. Specimen P49130b clearly depicts the posterior margins of most thoracic tergites across the entire pleural field (Fig. 3.1, 3.2) and demonstrates that the transverse rods are in segmental register with the tergites and are more consistently situated near the posterior margins of the tergites sagittally than in SAMA P50364. Three specimens show that the transverse rods span the axial region and a variable extent but not all of the pleural field (Figs. 2.4, 3. 5), providing another indication that the rods do not correspond to tergal margins. Specimen P40811 shows the anterior margin of some thoracic tergites covered by the preceding tergite in the distal part of the pleural field (Fig. 5.2), demonstrating the broad overlap between tergites seen in other xandarellids (Ramsköld et al., Reference Ramsköld, Chen, Edgecombe and Zhou1997, fig. 15), rather than edge-to-edge articulation.
On the basis of these relationships, we interpret the transverse rods as intra- rather than inter-segmental structures. Their preservational mode, in high relief (or as grooves in the counterpart: Fig. 3.3, 3.5, respectively) and with a distinctive blocky fabric, differs from that of the dorsal exoskeleton, and they are likely to be internal structures. An identity as endoskeleton is consistent with their position within the body, and their taphonomic distinction from the cuticular exoskeleton suggests that they may represent connective rather than cuticular endoskeleton (the latter including apodemes), both of which are widely present in euarthropods (Bitsch and Bitsch, Reference Bitsch and Bitsch2002; Gabbott et al., Reference Gabbott, Edgecombe, Theron and Aldridge2025). Their variably sinuous course across their widths or near their distal tips, where they are sometimes swollen, is consistent with an identity as tendinous rods.
An identity as ventral transverse tendons—a general character of euarthropods (Boudreaux, Reference Boudreaux1979) that is known in such fossil clades as naraoiids (Chen et al., Reference Chen, Edgecombe and Ramsköld1997) and trilobites (Whittington, Reference Whittington1993)—can be ruled out because the rods in Austroxandarella are affiliated with dorsal structures (i.e., tergites) rather than with ventral structures (sternites and appendage attachment sites). This follows from arguments elaborated below in which more than a single pair of biramous appendages are borne beneath the progressively elongated posterior thoracic tergites and pygidium. Among connective endoskeletal structures attached to the tergites in euarthropods are various endophragma, which often support the dorsal longitudinal muscles (Bitsch and Bitsch, Reference Bitsch and Bitsch2002). The extent of the transverse rods in Austroxandarella depicts a rather typical width with respect to sheets of dorsal longitudinal muscle.
Appendages and dorsoventral decoupling
Although the appendages are not well known in Austroxandarella poikar, available data are consistent with dorsoventral decoupling of ventral segmental structures (appendages and sternites) and tergites in the posterior half of the body that is characteristic of Xandarellida (Chen et al., Reference Chen, Ortega-Hernández, Wolfe, Zhai, Hou, Chen, Mai and Liu2019). A few possible endopods in SAMA P40811 would be associated with the cephalon and first pleura-bearing thoracic tergite (Fig. 5.2); the latter tergite has a one-to-one relationship with a biramous appendage pair in xandarellids. The exopod shafts and associated lamellar setae in SAMA P49130 underlying and in alignment with T4–T6 (Fig. 3.1–3.3) indicate a one-to-one relationship with the tergites, as in the corresponding region of the trunk in Xandarella (Hou and Bergström, Reference Hou and Bergström1997). Elongation of T8 and especially T9 and T10 relative to anterior tergites is suggestive of Austroxandarella having supernumerary appendages, by comparison with similar elongation of posterior tergites in Xandarella spectaculum. In the latter, the anterior seven tergites (not counting a reduced axial tergite at the head–trunk boundary) are associated with a single biramous appendage pair, with decoupling posterior to this, such that T10 covers as many as five pairs of appendages (Hou et al., Reference Hou, Siveter, Siveter, Aldridge, Cong, Gabbott, Ma, Purnell and Williams2017). In SAMA P49130a, the pleural fields of T9, T10, and the pygidium show multiple diagonal ridges identified as endopods, with the number on each tergite increasing posteriorly (Fig. 3.2, 3.3)—a pattern also recognized in Xandarella spectaculum (Hou and Bergström, Reference Hou and Bergström1997). Compared with consistent lengths of tergites along the trunk in other xandarellids, such as Cindarella and Luohuilinella—the typical plesiomorphic state for Artiopoda—the marked, progressive elongation of the last few posterior tergites in Xandarella and Austroxandarella provides a new character for phylogenetics, here added to the character matrix of Zhu et al. (Reference Zhu, Zeng, Liu and Zhao2023).
The ridges interpreted as proximal and distal exopod shafts in SAMA P49130 (Fig. 3.2) are slender, more like those of Xandarella spectaculum (Hou and Bergström, Reference Hou and Bergström1997, figs. 68C, 69), compared with the flap-like proximal and distal exopod lobes of Cindarella eucalla (Ramsköld et al., Reference Ramsköld, Chen, Edgecombe and Zhou1997). The slender shaft is scored as apomorphic under character 11 in the Zhu et al. (Reference Zhu, Zeng, Liu and Zhao2023) matrix.
Pygidial keel
A median keel along the posterior part of the pygidium in the holotype of Austroxandarella poikar (Fig. 2.3) is reminiscent of crests that give rise to axial spines on the posterior three thoracic tergites in Cindarella eucalla. An axial spine had likewise been interpreted as being present on the terminal thoracic tergite in Sinoburius lunaris (Edgecombe and Ramsköld, Reference Edgecombe and Ramsköld1999), but restudy of the latter taxon disputed its presence (Chen et al., Reference Chen, Ortega-Hernández, Wolfe, Zhai, Hou, Chen, Mai and Liu2019). Comparison needs to be made with a dorsal keel that gives rise to a terminal spine on the pygidium of Xandarella spectaculum (Hou and Bergström, Reference Hou and Bergström1997, fig. 68B), but as noted above, the posterior carina on the keel of Austroxandarella does not appear to extend beyond the posteromedial margin as a spine.
Absence of eyes
The lateral margin of the cephalic shield is sufficiently complete in SAMA P52874 (Fig. 4) and P40811 (Fig. 5) to indicate that an anterolateral notch to accommodate a stalked ventral eye, as in Luohuilinella and Cindarella, is absent in Austroxandarella; the anterolateral margin in both A. poikar specimens has a uniform curvature. The dorsal surface of the cephalic shield is imperfect in all Emu Bay Shale specimens but is adequately preserved and sufficiently complete to suggest that a dorsal bulge that accommodates ventral eyes, as in Xandarella and Sinoburius, is absent, and no specimen shows evidence of an eye slit, as in Xandarella. Given that anterolateral notches, eye slits, and dorsal bulges are all reasonably conspicuous structures in Chengjiang xandarellids, their absence in Austroxandarella poikar (coded as absent under characters 28, 23, and 22, respectively) might indicate that the species is blind, although we are reluctant to overweight negative evidence.
Gut morphology
Two of the four specimens of Austroxandarella poikar preserve the gut as a continuous narrow tube along much or all of the length of the thorax, and in SAMA P49130a, the gut extends into the pygidium (Fig. 3.1, 3.2). The gut of Austroxandarella appears to be lacking paired midgut diverticula (Fig. 3.4), which are best known in Cindarella eucalla (see Ramsköld et al., Reference Ramsköld, Chen, Edgecombe and Zhou1997, figs. 5a, 6f, 12a), in which reniform diverticula with an internal canal system are developed in an extensive series of thoracic segments. Similar reniform gut diverticula are also observed in Xandarella (Hou et al., Reference Hou, Siveter, Siveter, Aldridge, Cong, Gabbott, Ma, Purnell and Williams2017, fig. 20.39; Chen et al., Reference Chen, Ortega-Hernández, Wolfe, Zhai, Hou, Chen, Mai and Liu2019). By contrast, the gut diverticula of Luohuilinella deletres Hou et al., Reference Hou, Williams, Sansom, Siveter, Siveter, Gabbott, Harvey, Cong and Liu2019 are tubular and unbranched (Hou et al., Reference Hou, Williams, Sansom, Siveter, Siveter, Gabbott, Harvey, Cong and Liu2019). Accordingly, the absence of gut diverticula in Austroxandarella poikar may contain taxonomic signal, providing a distinction from the most similar genus, Xandarella.
Acknowledgments
We thank landowners P. and C. Buck for generously allowing access to the field site, and M.-A. Binnie for access to the SAMA collections. We also acknowledge our previous collaborators and volunteers in Emu Bay Shale research: R. Atkinson, M. Betts, R. Bicknell, M.-A. Binnie, G. Brock, A. Camens, A. Daley, H. Drage, E. Ellis, R. Gaines, J. Gehling, M. Gemmell, J. Holmes, T. and C. Ireland, J. Jago, K. Kenny, P. Kruse, J. Laurie, M. Lee, B. McHenry, M. Mills, J. Ortega-Hernández, L. Reid, D. Rice, and N. Schroeder. M. Koolmatrie, Ngarrindjeri Elder, kindly advised on the species name. This research was supported by grants from the Australian Research Council (LP0774959 and DP200102005 to J.R.P. and G.D.E.; FT120100770 to J.R.P.; FT130101329 to D.C.G.-B.) as well as financial and logistical assistance from the South Australian Museum and SeaLink. TNT was made available courtesy of the Willi Hennig Society. We thank the journal’s referees for advice.
Competing interests
The authors declare none.
Data availability statement
Data available from the Morphobank Digital Repository: https://morphobank.org/index.php/MyProjects/List/select/project_id/5963
