Edrathina obditchia , an Ediacaran stem-group ctenophore from the Fermeuse Formation of Newfoundland

1 Introduction

The earliest well-constrained macroscopic Eumetazoa (Dunn et al., 2021) are known from deep-marine facies of the 575–550 Ma (million years/mega-annum) Avalon Assemblage (Matthews et al., 2021; McIlroy et al., 2026). These complex macro-organisms primarily belong to two enigmatic, entirely soft-bodied, clades: the Rangeomorpha and Arboreomorpha (Narbonne, 2005). Preservation within these earliest fossil assemblages is exceptional and likely reflects a confluence of taphonomic conditions (see Liu, 2016), including: 1) post-mortem microbial growth (death masks of Gehling, 1999); 2) casting by volcanic ash (Narbonne, 2005; Mason et al., 2013; Matthews et al., 2021); 3) an absence of macroscopic bioturbation (McIlroy and Logan, 1999; Seilacher, 1999); and 4) early permineralization (McKean et al., 2023; Tarhan et al., 2025). The Avalon Assemblage (sensu Waggoner, 2003) includes the oldest known evidence for Cnidaria, coming from c. 565 Ma trace fossils (Liu et al., 2010; Menon et al., 2013; Liu et al., 2014b; Liu and McIlroy, 2015), and body fossils including crown-group Cnidaria) staurozoan cnidarians (Dunn et al., 2022; McIlroy et al., 2024). Among these, Haootia quadriformis was first reported from c. 560 Ma strata in the Catalina Dome on the Bonavista Peninsula of Newfoundland (Liu et al., 2014a, 2015), with Mamsetia manunis more recently being documented from 150 m below the Haootia holotype (McIlroy et al., 2024). Other rare animal fossils from the Avalon Assemblage include candidate poriferans Thectardis Clapham et al., 2004 (Sperling et al., 2011; Suarez and Leys, 2022) and Lydonia Pasinetti et al., 2025. Ediacaran biotas from the contemporaneous (but shallow-marine) White Sea Assemblage (sensu Waggoner, 2003) include records of probable molluscs (Fedonkin and Waggoner, 1997), sponges (Serezhnikova and Ivantsov, 2007), and potential ecdysozoans (Howard et al., 2022; Hughes et al., 2024), indicating significant ecological and morphological metazoan diversity in the Ediacaran (Droser et al., 2017; Hughes et al., 2024).

The Kotlin Crisis, which defines the base of the Nama Assemblage and occurred at c. 550 Ma, is widely considered to be a global extinction event, greatly affecting the Rangeomorpha and Arboreomorpha (Darroch et al., 2018; McIlroy et al., 2026). The post-Kotlin Crisis biotas of the Nama Assemblage (Wood et al., 2023; Bowyer et al., 2024) are characterized by a soft-bodied biota dominated by discs, with Erniettomorpha and rare Rangeomorpha, alongside trace fossils (Darroch et al., 2021), mineralized tubicolous fossils (Schiffbauer et al., 2016), Conulariida (Leme et al., 2022) and sponges (Wang et al., 2024). These latest Ediacaran metazoans are considered to be survivors of the Kotlin Crisis and harbingers of the Cambrian biotas that flourished after the end-Ediacaran extinction c. 539 Ma (Linnemann et al., 2019; Kouchinsky et al., 2012; Gougeon et al., 2018).

Ctenophora and Cnidaria were historically grouped together as the Coelenterata (Hatschek, 1888; Scholtz, 2004). While the clade Coelenterata has since been largely rejected (see Dunn et al., 2015; Whelan et al., 2015), some molecular analyses have recovered a clade encompassing the Cnidaria and Ctenophora (Philippe et al., 2009; Nosenko et al., 2013; Pett et al., 2019). Two competing hypotheses currently dominate discussions of early animal evolution: 1) the Porifera-sister hypothesis, which proposes that sponges represent the earliest-diverging animal lineage, with all other animals forming a clade known as Eumetazoa (Pisani et al., 2015; Nielsen, 2019; Redmond and McLysaght, 2021); and 2) the ctenophore-sister hypothesis, which places the Ctenophora as the sister group to all other animals, implying that they diverged before all other animal phyla (Whelan et al., 2017; Schultz et al., 2023).

The early stem-group ctenophores were epibenthic and nektonic organisms that are increasingly well-known from Cambrian Lagerstätten, including the Burgess Shale and Chengjiang biotas (e.g. O’Brien and Caron, 2012; Ou et al., 2015; Zhao et al., 2019). Among these are the epibenthic dinomischids, which possess a body plan consisting of a subconical calyx surrounded at the apical margin by pinnulate tentacles (Zhao et al., 2019; Parry et al., 2025). Evidence for Ediacaran ctenophores has also recently emerged from the Jiangchuan Biota (Li et al., 2026), and a form comparable to the dinomischid genus Xianguangia Chen and Erdtmann, 1991 has been figured from the Tongshan Lagerstätte as an “undescribed rangeomorph” (Hou et al., 2025). While members of Dinomischidae Conway Morris, 1977 have occasionally been resolved within Cnidaria (Ou et al., 2017, 2022), their placement is now generally regarded as being within Ctenophora (Zhao et al., 2019, 2024; Parry et al., 2025). Although the precise evolutionary relationship between Ctenophora and Cnidaria remains unresolved, the preservation of the entirely soft-bodied Haootia and Mamsetia (McIlroy et al., 2024) demonstrates the potential for documenting other unmineralized metazoans within the Ediacaran (pre-Kotlin Crisis) strata of Newfoundland.

2 Geologic and taphonomic setting

The Ediacaran sedimentary successions of eastern Newfoundland and the Charnian Supergroup of the UK form part of the peri-Gondwanan terrane of Avalonia, and have similar lithostratigraphic and tectonostratigraphic histories (McIlroy et al., 1998). Both regions are characterized by volcano-sedimentary successions associated with volcanic arc activity (Mason et al., 2013; Pharaoh et al., 2023). In Conception Bay North, Newfoundland (Figures 1A, B), there is evidence of Ediacaran-aged strike-slip deformation between c. 560 and 540 Ma (Ichaso et al., 2007).

The basin fill consists of deep-marine volcaniclastic turbiditic deposits of the Conception Group (Ichaso et al., 2007; King, 1988; Khatrine and Lowe, 2025). These are overlain by the interbedded sandstones and siltstones, with abundant slump deposits that characterise the slope-to-shelf successions of the Trepassey and Fermeuse formations. The latter two formations are considered to have been deposited on a southwest-facing paleoslope, as part of a net-upward shallowing marine succession that culminates in the paralic Renews Head Formation of the St. John’s Group (King, 1988; Ichaso et al., 2007; Figure 1C). The Fermeuse Formation is composed primarily of fine-grained upward-fining sandstones deposited from density currents and mudstones, as well as rare volcanic ash beds especially near its base (King, 1988). One such ash bed overlies the Edrathina-bearing EM Coombs Surface at Inner Meadow, Upper Island Cove, and is dated at c. 551 Ma (McIlroy et al., 2026). Early Palaeozoic deformation of the Avalon Peninsula has resulted in regional folding and faulting that has compressed the fossils parallel to cleavage, necessitating the retrodeformation of the fossils at Inner Meadow to restore original morphologic proportions (Figures 2A–F; Rosse-Guillevic et al., 2026).

The Inner Meadow Lagerstätte is one of the stratigraphically-highest occurrences of Conception-type preservation (i.e. the preservation of organisms below volcanic ash beds; Narbonne, 2005). The fossil assemblage from Inner Meadow includes abundant rangeomorph and arboreomorph taxa including: Aninoides coombsorum Rosse-Guillevic et al., 2025; Arborea spinosa Laflamme et al., 2004; Charnia brasieri McIlroy et al., 2025; Charnia gracilis Wu et al., 2022; Hiemalora pleiomorpha Vodanjuk et al., 1989; Pectinifrons sp. Bamforth et al., 2008; Primocandelabrum sp. Hofmann et al., 2008; and Trepassia sp. Narbonne et al., 2009 (Figure 2; McIlroy et al., 2025). The preservation of fine details of the fossils on the EM Coombs Surface is exceptional. However, it differs from the preservational style of Cambrian “Burgess Shale type” Konservat-Lagerstätten in being of coarser fidelity, which obscures some features while still preserving lightly sclerotized body tissues. Differences in taphonomic pathways between the Ediacaran Avalon Assemblage and Phanerozoic Burgess Shale type faunas may account for some variations in the preserved morphology of candidate Metazoa common to both periods; consequently, taphonomic biases are considered when comparing candidate Ediacaran Ctenophora with Cambrian taxa.

3 Materials and methods

3.1 Material

Fossils from the Inner Meadow Lagerstätte are preserved on the upper bedding plane of a grey siltstone (EM Coombs Surface), which is overlain by a thin tuffite. The material figured herein remains uncollected in the field, in accordance with Newfoundland and Labrador provincial law. Jesmonite replicas of the type material are housed at The Rooms Provincial Museum of Newfoundland and Labrador (catalogue numbers NFM F-4037 through NFM F-4039).

3.2 Field and lab methods

A replica of the holotype was made by pouring Mold Star 20T silicone over the in situ fossil to create a mould. In the lab, a Jesmonite AC100 mixture was made and pigmented with brown dye, which was subsequently poured into the silicone mould to create a replica cast of the field specimen (plastotype).

3.3 Retrodeformation

Retrodeformation techniques are employed herein to determine the original morphology of fossils. Such retrodeformation of Ediacaran macrofossils generally involves using basal discs as strain markers, based on an assumption that the discs were all originally circular (cf. Jenkins and Gehling, 1978; Wood et al., 2003). That assumption is not always valid since basal discs may vary in ellipticity on a single surface (Liu et al., 2011). While retrodeformation does not achieve perfect precision, failure to correct for deformation introduces a greater source of error than using measurements based on carefully considered retrodeformation.

When an elongation of 156% perpendicular to the cleavage axis—or a strain factor of 0.801 parallel to cleavage and 1.249 perpendicular to cleavage for conservation of area—is applied to the fossils of the EM Coombs Surface at Inner Meadow, frondose body fossils become more geometrically coherent and many (though not all) discs approximate circular forms (see Figure 3).

3.4 Phylogenetic methods

The assessment of taxonomic affinity presented herein is primarily based on comparative morphology and taphonomic observations. Phylogenetic analyses were undertaken to integrate these data into existing matrices and to resolve the phylogenetic position of Edrathina obditchia using our morphological coding. The phylogenetic matrix used for this study is adapted from Zhao et al. (2024), and focuses on the Ctenophora and Cnidaria. The character “sclerotized pinnules” was modified to become “pectinate arrangement of pinnules” in order to better differentiate pinnules from branched limbs in other taxa. Two characters—propagation through lateral budding and number of tentacles in the polyp—were re-scored for the staurozoan genera Craterolophus James-Clark, 1863 and Haliclystus James-Clark, 1863. The titles of five characters containing the word “polyp” were revised to “polyp/polypoid” to avoid confusion, as these characters apply to both ctenophores and cnidarians, and the term “polyp” is typically restricted to cnidarians. The character “bulbous base at the terminate end of stalk” was modified to include bases that attach directly to a calyx (becoming “bulbous base at the terminate end of stalk/calyx”). A total of 95 taxa and 291 characters were used in this analysis. We also included the Ediacaran taxa Mamsetia and Haootia, building on preliminary phylogenetic analyses of McIlroy et al. (2024).

Bayesian inference was conducted using MrBayes 3.2.7 (Ronquist et al., 2012), under the mkv + gamma model (Lewis, 2001) with 4 gamma categories. The “variable” ascertainment bias correction was chosen because the matrix contains autapomorphies, excluding 7 constant characters from the analysis. The scripts were set up to run 30 million generations, with a temperature of 0.04 to slightly improve swap acceptance rates, sampled every 1000 generations and with a stopping rule at 0.008 for the average deviation of split frequencies. Convergence was assessed with the help of the Estimated Sampled Size (>200) and Potential Scale Reduction Factor (~1).

4 Systematic paleontology

Phylum Ctenophora Eschscholtz, 1829

Genus †Edrathina nov.

(Figures 3–5)

LSID. https://zoobank.org/nomenclaturalActs/AAD76CAE-ABB6-4ED6-81B2-33EB579C68DE.

Diagnosis. At least 6 limbs arise approximately equidistantly around the opening of a subconical calyx. Calyx tapers towards a basal disc. Calyx longer than wide. Limbs longer than calyx height. Limbs bear two rows of approximately 16 pinnules, alternately arranged and evenly spaced along their length. Pinnules predominantly arise ~60–70° towards the distal end of the limb.

Etymology. The genus name is derived from the Beothuk word edrathin meaning “comb”, referencing the fine pinnae on the limbs, which resemble the teeth of a two-rowed comb. The Beothuk were the inhabitants of the island of Newfoundland at the time of European colonization. Since they are now culturally extinct, their language is no longer spoken. However, word lists collected from the captured Beothuk women Oubee, Demasduit and Shanawdithit have been preserved. The use of these words serves to honour the Beothuk people and acknowledge that the land where the fossils were found is their unceded traditional territory. The genus name is to be treated as a noun in the nominative singular case. Gender feminine.

†Edrathina obditchia sp. nov.

(Figures 3-5)

Diagnosis. As for genus.

Etymology. The species name originates from the Beothuk word obditch meaning “beast”, chosen to reflect the bestial appearance of the organism, particularly the outstretched limbs bearing numerous small, spike-like outgrowths. It is to be treated as a feminine adjective in the nominative singular to agree with the genus name.

Material. All three specimens remain in situ per Newfoundland and Labrador law. Three replicas were created for study and are accessioned at The Rooms Provincial Museum (St. John’s, Newfoundland, Canada).

• Holotype: NFM F-4037 (Figures 3, 4)
• Paratypes: NFM F-4038 (Figure 5A), NFM F-4039 (Figure 5B)

Occurrence. EM Coombs Surface, Fermeuse Formation, at Inner Meadow near the community of Upper Island Cove, Newfoundland, Canada; late Ediacaran.

Description. The largest specimens are the holotype (Figures 3, 4; NFM F-4037) and paratype NFM F-4039 (Figure 5B), with paratype NFM F-4038 (Figure 5A) being smaller. Type material is preserved in positive epirelief. Measurements are primarily of retrodeformed material; uncorrected values of the deformed specimens are given in parentheses.

Disc. The sub-circular basal disc is attached to the base of the calyx and was restored to a roughly circular shape following retrodeformation (Figure 3). The basal disc is preserved with three to four inner rings in positive relief. The outer margin of the holotype disc has a 2.8 cm minor axis and 3.5 cm major axis (uncorrected 2.3 cm minor axis and 4.1 cm major axis). The negative depression within the inner-most ring of the holotype has a minor axis of 0.9 cm and a major axis of 1.2 cm (uncorrected minor axis 0.7 cm and major axis 1.5 cm).

Calyx. The species has a triangular calyx that tapers towards the basal disc and is inferred to result from collapse of an originally conical structure. Calyx height comprises ~35–45% of the total organism height. The calyx sides are slightly concave, and its upper margin is scalloped. The upper rim of the calyx shows the greatest positive relief. No stalk is present in the holotype, and the calyces of the paratypes are only partially preserved.

Crown. The limbs form a crown with an approximate spread angle of 75° as measured from the holotype. The crown is incompletely preserved in the paratypes, but if symmetry is assumed the spread angle is similar to that of the holotype. In the holotype, the crown is ~7.5 cm tall and ~9.9 cm wide (uncorrected ~6.3 cm tall and ~12 cm wide). We do not observe any convincing evidence for an oral cone, or other crown features except pinnulate tentacles.

Limbs. The six limbs emanating from the rim of the calyx are equidistant, and essentially straight to slightly more curved depending on their orientation (cf. NFM F-4039; Figure 5B). Most limbs broaden proximally, while some maintain a constant width (most evident in paratype NFM F-4038; Figure 5A). In the holotype, the best-preserved limb tapers in width from 0.6 cm proximally to 0.1 cm distally (uncorrected 0.6 cm proximally and 0.2 cm distally).

Pinnules. Limbs bear two rows of equally-spaced pinnules that taper distally and emerge at acute angles relative to the axis of the limb (~60–70°, measured from tip of limb to tip of pinnule). The proximal portion of the pinnules meet alternately on the outer surface of the limbs, forming raised structures where they overlap. Pinnules in the same row are sub-parallel and separated at their bases (~1 mm apart in the holotype). Preserved pinnule length ranges from 5–22 mm in the holotype, while their proximal widths remain relatively consistent (most are 2 mm, with those on the distal limb ends closer to 1 mm). A maximum of sixteen pinnules are documented from a single row (Figure 4B). The pinnules are predominantly straight, narrow structures with little to no overlap, suggesting a potential degree of stiffening.

Remarks. The paratype NFM F-4039 preserves only part of the calyx, with the bottom portion obscured by an underlying frond (Figure 5B). The strongly positive preservation at the calyx base is taken to suggest that the specimen may have twisted post-mortem. The paratype NFM F-4038 is underlain by an unidentified organism that obscures the top portion of the calyx (Figure 5A), which is inferred to have been conical, as in the holotype. This data is consistent with the calyx and crown of E. obditchia having been held upright above the seafloor (cf. McIlroy, 2026).

E. obditchia is interpreted to have at least six limbs (Figure 4A). Two structures located just outside and to the right of the crown that resemble limbs extending from the calyx lip are cleavage traces cutting the bedding plane (see Figure 3). Glacial polish in the centre of the crown of the holotype obscures the central portions of two limbs. While it bears geometric similarity to the oral dome of some Cambrian Ctenophora, it is considered a post-fossilization feature.

Given its preservation as a cast of the upper surface of the collapsed organism, it is possible that E. obditchia possessed more limbs than are preserved. While previous studies on Cambrian stem-group Ctenophora have estimated total limb/tentacle counts either by doubling the number preserved—assuming only one side of the tentacular whorl is preserved—or by extrapolating from limb spacing and calyx circumference (Zhao et al., 2019, 2024), these approaches were not applied herein. The holotype of E. obditchia preserves the calyx and limbs as low positive epireliefs, with no evidence of original biomineralization. As such, only the uppermost surface of the organism is preserved. In contrast, Cambrian ctenophoran fossils preserve body elements on adjacent sedimentary laminae. In such specimens, extrapolation of tentacle numbers can be used to account for elements on adjacent planes. Following a conservative approach, E. obditchia is reconstructed herein as having six limbs (Figure 6).

5 Discussion

5.1 Phylogenetic scoring

The phylogenetic scoring of Edrathina obditchia reflects the limitations imposed by unmineralized preservation, which limits the number of characters that can be confidently assessed. Given the inherent uncertainties associated with interpreting soft-bodied Ediacaran fossils, a conservative scoring approach was adopted, with character states scored only if they were reasonably supported by the available material.

A through-gut was scored as absent, as the calyx tapers proximally and shows no evidence of a secondary opening beyond the apical aperture. Symmetry was scored as present to reflect at least radial organization; however, the specific symmetry type was coded as uncertain (?) to account for potential unpreserved features or alternative tentacular arrangements that could modify the observed symmetry. As the character matrix distinguishes symmetry types (bilateral, biradial, triradial, pentaradial, etc.) but lacks a general “radial” category, we coded this character as unknown to remain consistent with the original coding framework proposed by Zhao et al. (2024). “Polypoid phase” was scored as present, with solitary polyp organization, consistent with the overall morphology of E. obditchia as a single, large polypoid organism.

The holotype shows no evidence of a stalk (instead exhibiting a tapered calyx). However, calyces of both paratypes are partially obscured making an unknown scoring for this character most appropriate. The character “circular depression on basal disc” is not scored as present, as basal disc sensu Zhao et al. (2024) refers to the discoid lower surface of the calyx of Xianguangia, which is not considered homologous to the anchoring structure of E. obditchia. “Bulbous base at the terminate end of stalk” was scored as present, as it is interpreted as a homologous anatomical structure, albeit one adapted for sticking to the firm matground present at Inner Meadow (McIlroy et al., 2026; Chida et al., 2026) rather than burial within soft substrates, which is the mode of life of Cambrian Ctenophora (cf. holdfast diversity in the Pennatulacea; Williams and Alderslade, 2011). No evidence of an oral dome with paired bracts was observed; pinnules are the only features preserved between tentacles.

E. obditchia was scored as possessing six equally-spaced tentacular outgrowths, arranged radially in a single cycle and emanating from the oral–aboral interface at the rim of the originally subconical calyx. The outer sheaths present on the external surfaces of tentacular outgrowths in dinomischids are not preserved in our material and were therefore scored as uncertain.

Pinnules are clearly preserved in all type material, emerging from the tentacular outgrowths, which was scored accordingly. The presence of cilia could not be confirmed owing to taphonomic style, and the presence of ciliary rows borne on pinnules was accordingly scored as uncertain.

The preservation of the calyx as a compressed but still three-dimensional form implies that E. obditchia retained some volumetric integrity post-mortem. This is in contrast to the post-mortem collapse of the hydrostatic skeleton of Cnidaria (cf. the flattened calyx and twisted arms of the Staurozoa Haootia quadriformis and Mamsetia manunis; McIlroy et al., 2024). Additionally, the consistent shape of the tentacles and pinnules of E. obditchia (Figure 5A) suggests a degree of stiffening, since consistently linear forms are unlikely to result from passive collapse of entirely unstiffened structures. The preservation of the organisms in strong positive relief (Figure 7) suggests that they were somewhat resistive to overburden pressure and decay prior to early diagenesis of the overlying tuff (cf. Narbonne, 2005). Despite this, E. obditchia was not scored as having skeletal elements, given that the preservation of the material could not confirm sclerotization, and a conservative scoring was preferred. The presence of a periderm was scored as unknown, as no strong evidence for or against this structure can be confidently determined in our material.

5.2 Phylogenetic analysis

Following the original approach of Zhao et al. (2024), we conducted three Bayesian analyses under different constraint scenarios: an unconstrained analysis (Figure 8A), a Planulozoa-constrained analysis (enforcing the monophyly of Cnidaria, Placozoa, and Bilateria; Figure 8B), and a ctenophore-first constrained analysis (enforcing Ctenophora as the sister-group to all other metazoans; Figure 8C). These analyses were intended to explore the phylogenetic implications of our morphological interpretations within those three scenarios. Consequently, the recovered placements should be viewed as the most probable hypotheses contingent upon currently available data and the applied constraints, although alternative topologies remain plausible in light of the uncertainties inherent to Ediacaran fossilisation.

Relative to the unconstrained topology of Zhao et al. (2024), our analysis recovers the clade Dinomischidae in an equivalent position within the overall tree and with similarly high support. Relationships among taxa within Dinomischidae are also identical to those recovered by Zhao et al. (2024). In our analysis, E. obditchia is recovered as a sister taxon to the Dinomischidae (Figure 8A), with moderate support for this relationship (posterior probability = 73%).

The Planulozoa hypothesis scenario produced a topology largely consistent with the unconstrained tree (Figure 8B). No changes in relevant clade composition or relationships were recovered relative to unconstrained results, and only very minor shifts in posterior probabilities were observed—including for the Edrathina + Dinomischidae clade and within Dinomischidae itself. The most notable difference between the two topologies is significantly reduced support for the total group Ctenophora, a pattern also observed in the results of Zhao et al. (2024).

The ctenophore-first hypothesis scenario produced a topology in which relationships within Ctenophora remain unchanged (Figure 8C). The sister-group relationship between Dinomischidae and E. obditchia was likewise recovered, though support for this clade was slightly lower than in the other two scenarios (posterior probability = 68%). The ctenophore-first hypothesis has been challenged on phylogenomic grounds (Pisani et al., 2015; Redmond and McLysaght, 2021) and also on morphological grounds, since it involves evolutionary scenarios lacking clear synapomorphies among the Porifera, Planulozoa, and Placozoa that require either independent evolution or secondary loss of key characters (see Nielsen, 2019).

In all three scenarios, Bayesian phylogenetic analysis recovers E. obditchia as a sister taxon to the Dinomischidae, which includes early-Cambrian ctenophore genera Calathites, Daihua, Xianguangia, and Dinomischus (Figure 8). The analysis also consistently resolves Haootia and Mamsetia as staurozoans within Cnidaria (Figure 8; cf. McIlroy et al., 2024), sharing a polytomy with the extant Staurozoa Craterolophus (Amyostaurida) and Haliclystus (Myostaurida).

The ctenophoran affinity of E. obditchia recovered in our analyses was likely affected by the limited number of characters that could be confidently coded from the available material. Many potentially informative features—such as cilia, ctenes and internal structures of the calyx—are not preserved and therefore could not be scored. The absence of these characters restricts the data available to the matrix and magnifies the influence of those that could be coded. For instance, the presence of pinnules was confidently determined and input, but some other characters were not scorable, so this single trait may have exerted disproportionate influence in determining the placement of E. obditchia. The inclusion of E. obditchia within Ctenophora nevertheless represents the most probable placement that can be determined with phylogeny, given the currently available data and taphonomic limitations that informed the analysis.

5.3 Morphological comparisons with other ctenophores and ctenophore-like taxa

A notable difference between Edrathina obditchia and Cambrian ctenophores is that the pinnules of E. obditchia are alternate rather than opposite. While this offset might represent a compactional artefact, if genuine it would not be expected to affect the biomechanical functionality of the crown regarding food capture. Likewise, any articulation that might exist between the calyx and tentacles, which is clear in Cambrian body fossils preserved in Burgess Shale-type Lagerstätten, is obscured by the different taphonomic mode in the Ediacaran material. Comparisons with Cambrian dinomischids (Table 1) should be regarded in the context of this taphonomic difference, and also the ~10 Myr evolutionary history that separates E. obditchia from established dinomischid taxa.

Table 1. Summary of key morphological features of Edrathina obditchia and comparable taxa, highlighting similarities and differences between them.

The Cambrian stem-group ctenophore Xianguangia sinica Chen and Erdtmann, 1991 is morphologically similar to E. obditchia in that both possess pinnules, lack a stalk, and have a radial arrangement of tentacles arising from a central calyx that lacks oral structures (Zhao et al., 2024; Table 1). The calyx (or column) of X. sinica is rounded with a blunt base (Ou et al., 2017), in contrast to the subconical calyx of E. obditchia. Xianguangia sinica is also interpreted to have eighteen tentacles, whereas E. obditchia bears at least six tentacles. The basal pit of X. sinica is not comparable to the basal disc of E. obditchia since the latter is a distinct discoidal structure with internal rings, whereas that of X. sinica is a pore confined to the basal surface of the calyx.

Daihua sanqiong Zhao et al., 2019, like E. obditchia, has a tapered conical calyx, pinnulate tentacles, and no documented elongate stalk. It differs from E. obditchia in lacking a bulbous base and in having eighteen tentacles (Zhao et al., 2019) in contrast to the six of E. obditchia (Table 1). Additionally, D. sanqiong possesses three oral domes, each with associated paired bracts (Zhao et al., 2019) that are absent or not preserved in E. obditchia.

Dinomischus venustus Peng et al., 2006 has a basal bulb, long stalk, and conical calyx with eighteen tentacles. The calyx of D. venustus is most comparable to that of E. obditchia, both being elongate and tapered in form, though D. venustus has a clear stalk not known from E. obditchia (Table 1). The tentacles of Dinomischus are structurally complex and encased in a thin outer sheath (Zhao et al., 2019) absent or not preserved in E. obditchia.

Calathites spinalis Luo et al., 1999 is characterized by a strongly sclerotized, tapering stalk with ridges (or spokes), and a round calyx surrounded by eighteen pinnulate tentacles (Zhao et al., 2024). The calyx–stalk junction forms a distinctive right angle, whereas the calyx of E. obditchia is more gradually tapered and lacks a preserved stalk. E. obditchia also bears only six tentacles as opposed to the eighteen of C. spinalis (Table 1).

Siphusauctum gregarium O’Brien and Caron, 2012 from the Burgess Shale, a soft-bodied taxon, shares a six-tentacle interpretation with E. obditchia (Table 1). However, the tentacles of S. gregarium lack pinnules and arise from a calyx that is enclosed within a sheath borne atop a narrow stalk (O’Brien and Caron, 2012). The overall configuration of the crown in S. gregarium is globular, forming a tulip shape (O’Brien and Caron, 2012), whereas the tentacles of E. obditchia splay outward from the calyx and are neither globular nor sheathed in arrangement. S. gregarium has also been compared with Namacalathus hermanastes Grotzinger et al., 2000, given similarities of key features such as a calyx, stalk, and hexaradial symmetry (Zhao et al., 2019). While the calcareous N. hermanastes (Zhuravlev et al., 2015) is notable as an Ediacaran–aged fossil with broadly comparable morphology to a Cambrian ctenophore, direct comparisons with E. obditchia are precluded by the absence of a calcareous skeleton.

The problematic Ediacaran Eoandromeda octobrachiata Tang et al., 2008—initially interpreted as a trace fossil (Ding et al., 1996)—has been proposed as a possible stem-group metazoan or ctenophore (Tang et al., 2011). Its assignment to the Ctenophora was inferred based on the presence of transverse bands, interpreted as comb plates, and on the spiral structures interpreted as arms (Tang et al., 2011). This interpretation, along with the proposed pelagic lifestyle, has been questioned based on morphology, abundance, and taphonomy, which collectively suggest a benthic mode of life (Zhu et al., 2008; Wang et al., 2020; Botha et al., 2023). Eoandromeda has most recently been considered as a possible trilobozoan based on co-occurrence and a similar spiral body plan (Wang et al., 2008; Ivantsov and Zakrevskaya, 2021).

5.4 Comparison with the Cnidaria

The preservation of Edrathina as external casts of body tissues differs from the preservation of cnidarians Haootia and Mamsetia, which are tetraradial polypoid staurozoans broadly contemporaneous in age to E. obditchia. The stauromedusae of Haootia and Mamsetia are similarly large and polypoid but mainly preserve internal muscular tissues of the calyx and arms, both of which were flexible (McIlroy et al., 2024). In contrast, the arms of E. obditchia appear less flexible, being preserved as largely linear structures that suggest some degree of stiffening. Additionally, no internal muscular tissues comparable to those reported from Haootia or Mamsetia are evident in E. obditchia.

The evolutionary history of the Ctenophora includes a shift from early sessile, polypoid forms to a predominance of motile planktonic taxa (e.g. scleroctenophores; Zhao et al., 2019, 2024). It is possible that stem-group ctenophores lie near the divergence of the Ctenophora and Cnidaria (cf. Carlisle et al., 2024) if polypoid morphology is interpreted as a synapomorphy of the two phyla. Such a scenario would align with the Coelenterata hypothesis, which has largely been dismissed (Dunn et al., 2015; Whelan et al., 2015).

While E. obditchia is resolved as a stem-group ctenophore (Figure 8), it should be noted that phylogenetic reconstruction of Ediacaran taxa is influenced by incomplete preservation, which introduces unavoidable uncertainty and potential bias. Accordingly, comparisons with the Cnidaria remain relevant given the possibility that E. obditchia could alternatively be placed within that phylum. The possible hexaradial symmetry of E. obditchia is shared with the cnidarian Hexacorallia, though it is noted that total-group ctenophores exhibit greater morphological disparity than extant taxa, encompassing a wide range of body plans, including hexaradial symmetry (e.g. Siphusauctum; Zhao et al., 2019, 2024).

5.5 Comparison with the Rangeomorpha & other Ediacaran taxa

Despite some broad similarities in overall body architecture with rangeomorph taxa, Edrathina cannot be assigned to the Rangeomorpha, as none of the type material exhibits the diagnostic pseudo-fractal or self-similar branching of that clade (see McIlroy et al., 2025), even with its exceptional fidelity of preservation (cf. Figure 5). The gross-scale dendriform morphology of Edrathina superficially resembles that of Primocandelabrum (Figure 2A) and Hylaecullulus Kenchington et al., 2018 (Hofmann et al., 2008; Kenchington and Wilby, 2017)—including a basal disc and a crown composed of several primary branches or limbs—but the architecture of Edrathina differs markedly from these forms (Table 1) in that:

• Branches of Edrathina are separated from one another at their bases, arising instead from inside the margin of a conical, tapered calyx and exhibiting radial (possibly hexaradial) symmetry. The most proximal branches of Primocandelabrum and Hylaecullulus arise directly from a thick stem (Kenchington and Wilby, 2017; Figure 2A).
• Hexaradial symmetry is not a typical character of dendriform rangeomorphs, nor has a calyx been described from any rangeomorph taxon.
• The 3 or 4 proximal (non-rangeomorph type) branches/limbs of the rangeomorphs Primocandelabrum and Hylaecullulus bifurcate (Hofmann et al., 2008; Kenchington and Wilby, 2017; Figure 2A), a character not seen in Edrathina.
• It might be argued that the pinnules of Edrathina represent the unpreserved axes of rangeomorph elements, however, if this were the case such axes should be zig-zagged. The branches do not have other branch axes arising from them, and there is no precedent for rangeomorphs being preserved just as branch axes without the typical rangeomorph branching architecture (cf. McIlroy, 2025; McIlroy et al., 2025).

The only feature of Edrathina that is comparable to the associated frondose Rangeomorpha is the basal disc, but similar “Aspidella” discs are common across the Rangeomorpha, Arboreomorpha (McIlroy, 2026) and even macroalgae (Wang et al., 2017), demonstrating convergent evolution. The large holdfast disc is likely an adaptation to anchoring the organism to matground surfaces, as its large size would have experienced significant drag and required robust basal attachment. This is analogous to the burrowing organ of some pennatulacean octocorals, which is modified in some taxa to become a discoidal holdfast for attachment to hard substrates (Williams and Alderslade, 2011).

Of all described Ediacaran material, a fossil from the June beds of the Mackenzie Mountains identified as “Primocandelabrum sp.” (Narbonne et al., 2014) bears the closest morphological resemblance to Edrathina (Table 1). However, as preserved, this specimen does not conform to the generic diagnosis of Primocandelabrum, and differs from the type material in lacking a stem, instead possessing a calyx-like structure. In addition, although the crown is incomplete, the bases of six limbs are visible, and higher-order branching is not preserved. These features more closely resemble Edrathina, as both taxa possess a basal disc with internal rings broadly comparable to “Aspidella”, a tapered calyx that broadens towards the crown, and a six-limbed body plan. Given these morphological features, the specimen figured by Narbonne et al. (2014) is not herein considered referable to Primocandelabrum. As a result, Edrathina remains without a close counterpart among formally described Ediacaran taxa.

5.6 Interpreted life attitude

Edrathina is interpreted to be an immotile sessile epibenthic marine organism with an upright life habit and a basal disc attaching it to the associated microbial matground. The preservation of E. obditchia in positive relief, rather than the deep negative reliefs of associated procumbent rangeomorphs such as Aninoides (Rosse-Guillevic et al., 2025), together with its occurrences overlying other procumbent organisms (Figure 5), strongly supports an upright growth habit. This interpretation is consistent with criteria recently proposed for determining the mode of life of Ediacaran organisms (McIlroy, 2026).

The shallow imprint of the disc suggests that it was a mat-sticker sensu Seilacher (1999), rather than having a buried bulbous anchor which is typical of the softground-dwelling Cambrian Dinomischidae. By analogy with Cambrian ctenophores and other pinnulate clades, it most likely filter-fed from the water column with tentacles outstretched in the direction of the current. The pinnules were likely held in a manner that optimized the capture of particulate food on their inner surfaces, which may have borne ctenes (cf. Zhao et al., 2024; Figure 6) that are not visible in Conception-type preservation.

6 Conclusion

Edrathina obditchia has a bauplan consisting of a discoidal basal disc attached to a subconical calyx with several biserially-pinnulate tentacles arising from and equally spaced around its margin. This body plan is unique among described Ediacaran taxa and is considered here to be most comparable to that of the stem-group ctenophores. The prominent basal attachment disc of E. obditchia is considered to be an adaptation to life on microbial matgrounds, which constitute a stable substrate that favours a large attachment surface.

Edrathina is similar to the much younger Cambrian Dinomischidae in possessing a calyx that bears several pinnulate limbs along its upper margin, an arrangement comparable to the tentacles of the stem-group ctenophores. Consistently, Bayesian phylogenetic analysis recovers E. obditchia as a basal stem-group ctenophore, with this placement retrieved under unconstrained and variously constrained phylogenetic analyses. This Ediacaran discovery adds to recent reports of possible ctenophores of similar age from China in Burgess Shale-type preservation (Hou et al., 2025; Li et al., 2026). Although this material awaits complete documentation, it may ultimately provide additional support for the interpretation presented herein.

The presence of pre-Kotlin Crisis Ediacaran metazoans in Avalonia is thus further evidenced by the documentation of stem-ctenophores alongside previous reports of crown-group Staurozoa (McIlroy et al., 2024) and other crown-group cnidarians (Dunn et al., 2022). The presence of Edrathina corroborates the emergence of diverse body plans at least 10 Ma before the Cambrian radiation event. This reinforces the view that key metazoan lineages had already begun to diversify in Ediacaran ecosystems prior to the Cambrian (Conway Morris, 2000) and before the Kotlin Crisis (McIlroy et al., 2026).

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. During the preparation of this work the authors used ChatGPT-5.1 to improve readability of this text. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fevo.2026.1882499/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Allison Daley, Université de Lausanne, Switzerland

Reviewed by: Zhe Chen, Chinese Academy of Sciences (CAS), China

Baptiste Coutret, University of Alberta, Canada