Abstract
Extant cnidarians are a disparate phylum of non-bilaterians and their diploblastic body plan represents a key step in animal evolution. Anthozoans (anemones, corals) are benthic polyps, while adult medusozoans (jellyfishes) are dominantly pelagic medusae. A sessile polyp is present in both groups and is widely conceived as the ancestral form of their last common ancestor. However, the nature and anatomy of this ancestral polyp, particularly of medusozoans, are controversial, owing to the divergent body plans of both groups in the extant lineages and the rarity of medusozoan soft tissues in the fossil record. Here we redescribe the enigmatic Conicula striata Luo et Hu from the early Cambrian Chengjiang biota, south China, which has previously been interpreted as a polyp, lophophorate or deuterostome. We show that C. striata possessed features of both anthozoans and medusozoans. Its stalked polyp and fully encasing conical, annulated organic skeleton (periderm) are features of medusozoans. However, the gut is partitioned by ∼28 mesenteries, and has a tubular pharynx, resembling anthozoans. Our phylogenetic analysis recovers C. striata as a stem medusozoan, indicating that the enormously diverse medusozoans were derived from an anemone-like ancestor, with the pharynx lost and number of mesenteries reduced prior to the origin of crown group Medusozoa.
Introduction
Among non-bilaterian animals, Cnidaria is the phylum with the most species richness as well as the most disparate range of morphologies (Daly et al., 2007). Members of Cnidaria share the presence of tentacles with stinging cells (cnidocytes) used in prey capture, a blind gastric cavity that is often partitioned by mesenteries/septa and a sessile polypoid phase in at least part of their lifecycle in most groups (Hyman, 1940). The phylum mainly includes two monophyletic subgroups: Anthozoa and Medusozoa (Marques and Collins, 2004; Schuchert, 1993). Anthozoans are primarily benthic, polypoid animals encompassing widely known groups such as sea anemones and corals (Brusca et al., 2016), whereas medusozoans usually have a biphasic lifestyle, with sessile polyps giving rise to swimming medusae (jellyfishes) via asexual reproduction (Collins, 2002). Interpretations of cnidarian phylogeny support a scenario in which the common ancestor of the crown group was a sessile polyp, and the swimming medusa represents a synapomorphy of Medusozoa (Collins et al., 2006; Kayal et al., 2018; Marques and Collins, 2004; McFadden et al., 2021). However, the anatomy of the polyps diverges in living anthozoans and medusozoans in several respects (Daly et al., 2007; Technau and Steele, 2011), rendering the inference of their ancestral forms uncertain.
Anthozoan polyps possess a tubular pharynx (actinopharynx) that extends from the mouth to a gastric cavity that is partitioned by well-developed mesenteries (Daly et al., 2007). The pharynx contains either one or two ciliated siphonoglyphs that impart a bilateral/biradial symmetry (Malakhov, 2016). In contrast, medusozoan polyps are relatively small and possess an exoskeleton called periderm, most of which are made of chitin (Mendoza-Becerril et al., 2016). This organic skeleton can encase the entire body of the polyp in some lineages such as coronate Scyphozoa, or be reduced to just the basal portion of the polyp (Mendoza-Becerril et al., 2016). In medusozoan polyps, the mouth is usually extended away from the body on a protuberance referred to as the scyphopharynx or hypostome in different subgroups. The gastric cavity is not always portioned by septa, as they are absent in the main subclades of hydropolyps (Bouillon and Boero, 2000) and cubopolyps (Chapman, 1978).
Molecular clock estimates suggest that the cnidarian crown groups radiated in the Ediacaran or the Cryogenian (dos Reis et al., 2015; Park et al., 2012), but cnidarian fossils before the Cambrian are rare and/or controversial (Liu et al., 2014; Van Iten et al., 2014; Van Iten et al., 2016). Cambrian deposits, however, yield a wealth of cnidarian fossils that exceptionally preserved delicate soft tissues (Cartwright et al., 2007; Conway Morris, 1993; Han et al., 2016; Hou et al., 2005) and even their early developmental stages (Dong et al., 2016), and are therefore crucial to understand the origin and early evolution of cnidarian clades. Here, we redescribe the enigmatic Conicula striata Luo et Hu (Luo et al., 1999) from the Cambrian (Epoch 2, Age 3) Chengjiang biota from Yunnan Province, south China. C. striata has previously been interpreted as a lophophorate (spiralian) (Luo et al., 1999), an actinarian (cnidarian) (Hu, 2005) or a phlogitid (presumed deuterostome) (Caron et al., 2010). However, with only one specimen reported, C. striata has remained as one of the most poorly understood early Cambrian fossils. In this study, we depict the detailed morphology of C. striata based 51 exquisitely preserved specimens, revealing mosaic morphological characteristics as seen in both extant anthozoans and medusozoans. These features bring C. striata in the stem of Medusozoa, unveiling the earliest medusozoan polyp as an anemone-like form encased in an extensive periderm.
Results
Systematic Palaeontology
Phylum Cnidaria Verrill, 1865
Stem group Medusozoa Petersen, 1979
Conicula striata Luo et Hu, 1999
(Figures 1-3, S1-S7)
Occurrence
Conicula striata occurs in Cambrian Series 2, Stage 3, the local lithostratigraphic unit is the Yu’anshan Member of the Chiungchussu Formation, corresponding to the Eoredlichia-Wutingaspis trilobite biozone (Babcock and Zhang, 2001).
Amended diagnosis (amended from Luo et al., 1999, p. 87). Solitary polypoid cnidarian encased in a conical, annulated periderm. The polyp possesses unbranched, flexible circumoral tentacles that can protrude outwards from the distal end of the periderm, and a blind digestive tract consisting of an elongate, tubular pharynx and a mesentery-partitioned gastric cavity.
Description
Specimens of Conicula striata usually exhibit a conical shape in lateral view (Figures 1, S1, and S2) and occasionally a circular cross section in oblique-lateral view (Figure S3). The body is 10-38 mm long with maximum width ranging between 4-17 mm (width/length ratio among complete specimens is 0.38-0.45). Some specimens preserve evidence of both a rigid, external skeleton (periderm) (Figure 2), circumoral tentacles (Figure 3A-C) and a columnar body with internal anatomical features (Figure 3D-E, G-H, J-L).
The external skeleton appears to have been originally inflexible and robust, and fully encloses the aboral portion of the body (“Pd”, Figures 2A, C, D and S2D-F). The skeleton forms a conical structure in the proximal portion, which during growth expanded and then narrowed to form a globular chamber, resulting in a shape resembling a classic, well-loaded ice cream cone. It is ornamented by parallel annulations (“Al”, Figure 2A-C), with slight relief under low angle light (Figures 2C and S1D) which are most visible in the aboral section. The space between adjacent annulations is around 0.2-0.4 mm. The appearance of the exoskeleton is consistent with the periderm of medusozoan cnidarians. Dark, spine-like structures project into the interior of the periderm. They are arranged in one whorl and surround internal features, particularly the region housing the digestive tract, and are termed here as peridermal teeth (“Pt”, Figures 2D-I, S2A, and S6E-G).
Soft tentacles are exhibited at the distal part of the body (“Te”, Figures 1F, H, J, 3A-C, and S1E, F), either approximately straight (e.g. Figure 1H) or curved (e.g. Figure 1F, J) and vary in aspect ratio, suggesting that they were capable of protruding and retracting. The tentacles are smooth, with no identifiable branches or pinnules (Figures 3A-C and S1E, F). When present, the tentacles are buckled and intertwined as a dark organic-rich mass within the globular chamber, making it difficult to trace each tentacle (Figures 1A-E, S2C, E, F, S5B, and S6A-D). It is challenging to surmise the total number of tentacles due to twisting and/or superposition of multiple tentacles as well as incomplete exposure and varying degrees of decay.
The tentacle base surrounds a disc with a central dark, tongue-shaped structure, likely representing the mouth opening (“Mo”, Figures 1D, E and S5C, D). An elongate, narrow tubular structure (about 6.5 mm and 7.7 mm long in YKLP 13484a and 13485a, respectively) connects the opening to the lower part of the body (“Ph”, Figures 1F-K, 3D, E, and S5E-H), in which an oval-shaped structure with a dark outline expands to near the body margin (“Gc”, Figures 1, 3D, E, and S5G, H). The tapering band and oval-shaped structure are distinct from other regions by their preservation in a red-brown or dark colour when viewed with cross-polarised light or fluorescence microscopy, respectively (Figures 3D, E, S3A, D, S4D, E, and S5G, H), and their higher amounts of carbon and iron in elemental maps (Figures S4F, G, S6E, G, and S7F, G). The tubular structure and oval-shaped structure are interpreted as a pharynx and a gastric cavity, respectively, based on their shape, size and position within the animal body (Figure 3F, I).
The gastric cavity varies in size and shape in different specimens, but frequently preserves evidence that was partitioned by dark, longitudinal lines (“Me”, Figures 3G, H, S2C, S3A-C, S4A-F, and S6A-D). In YKLP 13212a, these dark lines are nearly parallel with each other and extend from the base of the oval-shaped structure to the distal region of the columnar body (Figures 3G, H and S4D-F). The spacing between two adjacent lines in the central region is ∼0.39 mm, suggesting a total of 28 mesenteries or so (calculated by dividing this interval by the body circumference≈11.3 mm). However, dark lines are sometimes inconspicuous, with some superposition from either side of the body, making their precise number difficult to determine across specimens. These dark lines are carbonaceous in preservation (Figure 3H) and interpreted as mesenteries/gastric septa according to their appearance and position (Figure 3I).
A further dark patch, with a circular, crescent or triangular outline, attaches at the base of the gastric cavity (“Pc”, Figures 1 and S2). It contains a high abundance of carbon as revealed by elemental maps (Figure S6H, I). A slightly curved, narrower ribbon-like structure (“Pe”, Figures 1A, D, F, 3J, L, and S4A-C), sometimes with modest relief (Figure 3L), protrudes from the bottom of the dark patch to the proximal end of the periderm (e.g. YKLP 13212 and 13288). In YKLP 13212, it is roughly parallel sided (∼0.34 mm wide), and contains abundant weathered pyrite granules (Figure 3K). The dark patch and ribbon-like structure are herein interpreted as a peduncle chamber and a peduncle, respectively.
Phylogenetic position
Our phylogenetic analysis incorporating Conicula striata includes 99 taxa and 304 characters scored for a diversity of living and fossil cnidarians and outgroups. Analysis of this dataset without any topological constraints recovers C. striata in the medusozoan stem group and all other tubular fossils in a clade that is in a polytomy with extant medusozoan taxa (Figures 4C and S8A). When removing these tubular fossils, C. striata is still recovered in the medusozoan stem group while extant medusozoan taxa are monophyletic (Figure S8B). This analysis also recovers the paraphyly of Scyphozoa, a result that is not found in recent molecular phylogenies but has been found in many recent morphological phylogenies(Duan et al., 2017; Zhao et al., 2019). Constraining the in-group relationships of cnidarians still recovers C. striata in the stem group of Medusozoa (Figure S8C).
Discussion
C. striata is a stem medusozoan with a mixed anthozoan-medusozoan body plan
Conicula striata was erected in 1999 based on one incomplete specimen preserving a tentacular, annulated conical body (Luo et al., 1999). It has since been variously interpreted as a lophophorate (Luo et al., 1999), a sea anemone (Hu, 2005) or a deuterostome (Caron et al., 2010). Our 51 new specimens corroborate the validity of this genus and further provide new insights into its morphology, phylogenetic position and evolutionary significance. C. striata possesses an expansive, partitioned gastric cavity with a single opening (Figure 3D-E, G-H), unbranched, circumoral tentacles (Figure 1F, H, J) and a radially symmetrical body (Figure S3). These characters are incompatible with the previous interpretations as either a lophophorate or a deuterostome. Lophophorates possess a U-shaped gut (with a mouth and an anus) that is often conspicuously preserved in Chengjiang fossils (Zhang et al., 2013). Although blind guts are present in articulate brachiopods, C. striata shares no derived characters with brachiopods in any aspect of body construction. Cambroernids (presumed deuterostomes), a fossil group including Herpetogaster, phlogitids and eldoniids (Caron et al., 2010), also differ markedly as they have bilaterally arranged, paired, branched, even dendritic tentacles and a discrete, through gut (Caron et al., 2010; Hou et al., 2006).
The regular, longitudinal dark lines preserved in association to the columnar body wall and the gastric cavity (Figure 3G, H) are interpreted as mesenteries. Similar structures have been identified in Cambrian exceptionally preserved fossils before, such as Archisaccophyllia (Hou et al., 2005) and some dinomischiids (Zhao et al., 2019). Partitioning of the gastric cavity by mesenteries/septa is a widespread character in extant cnidarian polyps (Figure 3I) (Daly et al., 2007), with a higher number (at least 8) of well-developed mesenteries occurring in anthozoans while only 4 present in scyphopolyps and stauropolyps. In hydropolyps (Bouillon and Boero, 2000) and cubopolyps (Chapman, 1978) gastric septa are absent. The number of mesenteries observed in C. striata (∼28) is consistent with that in anthozoans. The elongate tubular structure of C. striata that extends from the gastric cavity to the oral disc and mouth corresponds in size and topological position to the actinopharynx of anthozoans (Figure 3D-F) (Daly et al., 2003; Daly et al., 2007), but is different from the scyphopharynx/hypostome of medusozoans which is an oral extension at the distal end of the body. The digestive system of C. striata therefore closely resembles the condition seen in extant anthozoans.
The exoskeleton preserves regularly arranged annulations in the proximal portion in some specimens (Figures 2A-C and S1A-D). While compaction could have enhanced their relief, they are unlikely to be artefactual due to the common occurrence across specimens. Such annulations resemble the growth lines caused by the marginal accretion of an exoskeleton, which are widely present in extant cnidarians (e.g. medusozoan periderm) and spiralians (such as molluscs, brachiopods and tubular polychaetes). Given that the polypoid body and internal structures of C. striata deviate from the body plans of bilaterians, we limit comparisons of the skeleton to those of cnidarians. Anthozoans also produce skeletons, which are found in living antipatharians, ceriantharians, scleractinians and octocorallians, and extinct rugose and tabulate corals, but presumably have multiple independent origins. Among these, C. striata only superficially resembles the ceriantharians, which produce tubes made of mucus and ptychocysts (Stampar et al., 2015). The accretionary exoskeleton of C. striata is directly comparable to the periderm of medusozoans and we infer that these features are homologous.
The dark, spine-like structures (Figure 2D-I) that encircle the basal portion of the columnar body are interpreted as the remains of peridermal teeth. Extant coronate polyps possess well-developed whorls of complex peridermal teeth, which are protrusions of the inner layer of the periderm towards the central chamber, to anchor the polypoid body (Jarms, 1991). Similar peridermal teeth also appear in tubular fossils as sheet-like ridges (cusps) in Sphenothallus (Dzik et al., 2017) or paired projections in Olivooides (Dong et al., 2016). These peridermal teeth are often repeated along the length of the tube wall (Jarms, 1991), but in C. striata they occur in a single whorl towards the aboral end of the tube, with the remainder of the lumen of the skeleton appearing smooth. In addition, the interpreted peduncle might also function for anchoring the polypoid body inside the theca, along with the peridermal teeth. In colonial hydropolyps, peduncle-like structures connect individual polyps together (Cartwright, 2004) and are therefore not readily comparable with that seen in C. striata.
C. striata is inferred here to have been a solitary organism with an annulated periderm fully encasing the polyp, evidencing a benthic and sessile lifestyle commonly seen among coronate and hydrozoan polyps. However, all specimens lack a holdfast at the aboral end and instead appear to taper naturally, with no evidence of attachment to other organisms or substrates, suggesting C. striata may have embedded the apex into the seafloor for anchoring, similar to some conulariids (Van Iten et al., 2013). Alternatively, C. striata may have been recumbent, but the conical skeleton does not curve to facilitate such a mode of life like that of horn corals (Scrutton, 1998). An alternative less plausible scenario is that C. striata was planktonic, with buoyancy provided by the inflated distal chamber, possibly having an intermediate lifestyle between the benthic polyps and pelagic medusae.
Evolutionary significance
Historically, many tubular fossils from the Ediacaran-Cambrian have been interpreted as cnidarian polyps (Table S1)(Van Iten et al., 2014), such as the microfossils Olivooidae, Carinachitidae and Hexangulaconulariidae(Guo et al., 2020) as well as macrofossils Corumbella and conulariids(Van Iten et al., 2016). The features revealed in these tubular fossils, such as radial symmetry, transverse ribs/crests, peridermal teeth and a single opening, are the primary lines of evidence for a cnidarian interpretation, with particularly close comparisons made to the peridermal tubes of medusozoan polyps(Conway Morris and Chen, 1992; Dong et al., 2016; Van Iten et al., 2006; Zhu et al., 2000). Soft tissues are extremely scarce among these fossil tubes, and their cnidarian affinities and interpretations are accordingly not without previous controversy(Steiner et al., 2014; Walde et al., 2019). They are recovered as a paraphyletic grade of total group medusozoans in our Bayesian analyses (Figures 4C, S8). C. striata not only shares with those tubular fossils similar exoskeletal features, but also provides unique new evidence for the soft tissues of early medusozoans, such as mesenteries, the digestive tract and tentacles, characters that are not available from the overwhelming majority of tubular fossil taxa, shedding light on character state changes that occurred early in medusozoan evolutionary history.
C. striata shows a tubular pharynx, which is similar to the actinopharynx of anthozoans in topological location, inferred function and architecture (Figure 3D-F). The presence of an anthozoan-like pharynx in the medusozoan stem group is also supported in our ancestral state reconstruction (Figure 4C). It is recovered as a plesiomorphic trait of cnidarians (Figure 4C), indicating that the tubular pharynx in the stem group of Medusozoa is homologous with the actinopharynx in anthozoans, but it was subsequently lost prior to the origin of crown group medusozoans. Whether other tubular fossil taxa have an anthozoan-like pharynx is not known, which would depend on further findings of soft tissues in these groups. Moreover, C. striata possesses about twenty-eight mesenteries lining the gastric cavity, a feature commonly seen in extant hexacorallians, which is also recovered as being plesiomorphic for medusozoans and cnidarians in our analysis (Figure 4C).
We infer that the periderm is a true medusozoan synapomorphy(Mendoza-Becerril et al., 2016), but is absent in the common ancestor of Anthozoa, and probably of Cnidaria (Figure 4C). This inference is in congruence with earlier cnidarian fossils from the Ediacaran-Cambrian, in which all potential medusozoan polyps share a well-developed, annulated exoskeleton (periderm). In contrast, all known anthozoan fossil taxa lack a comparable exoskeleton/periderm (Han et al., 2010; Hou et al., 2005). A polyp encased fully by a periderm is recovered as a plesiomorphic trait of the medusozoan total group (Figure 4C) and this is only retained in living coronates and some members of Hydrozoa.
In light of this, we reconstruct the ancestral medusozoan as an anemone-like polyp, which possessed a tubular pharynx (actinopharynx) connecting the mouth and gastrovascular cavity that is partitioned by ∼28 mesenteries and unbranched tentacles, with the body encased fully by an annulated exoskeleton (periderm) (Figure 4A, B). The body plan of C. striata bridges the long-known morphological gap between living anthozoan and medusozoan polyps (Figure 4D), suggesting that several features previously regarded as anthozoan apomorphies (e.g. an actinopharyx) might have a deeper origin and are shared by stem medusozoans.
Given the anatomical simplicity in crown medusozoan polyps(Ruppert et al., 2004), we infer that several characters, including well-developed mesenteries and a tubular pharynx, experienced subsequent reduction and even total loss in some or all extant medusozoan lineages (Figure 4D). Only a few living medusozoan polyps (e.g. coronate scyphozoans) have a well-developed periderm and it is completely restricted to the lower body or podocyst in some lineages (e.g. staurozoans)(Mendoza-Becerril et al., 2016). Our analyses indicate that this is also the result of secondary reduction and a component of a broader trend of polyp simplification in Medusozoa where the lifecycle is now dominated by the medusa stage.
Materials and Methods
Materials
51 specimens were collected in fieldwork from 2014 to 2019 in Haikou area, Kunming, eastern Yunnan province, south China. They were checked and prepared under a Leica M205C stereomicroscope. A fine needle was used to remove the matrix and expose the fossils. The specimen size was measured with ImageJ 1.51j8. All specimens are housed in the Yunnan Key Laboratory for Palaeobiology (YKLP), Yunnan University, China. The holotype (He-f-6-5-112/113) is deposited in the Yunnan Institute of Geological Survey, Kunming, China.
Photography
16 out of 51 specimens were figured. Photographs were taken either by a Canon EOS 5DS R digital camera mounted with Canon MP-E 100mm or 65mm macro lens (1-5X), using high/low angle cross-polarised light, or by a Leica DFC 5000 camera mounted on Leica M205C microscope (to obtain morphological details). Interpretative drawings were combined with camera lucida drawings done under Nikon SMZ1000 stereomicroscope and digital photographs. Fluorescence images were obtained using a Leica DFC7000 T digital camera linked to a Leica M205 FA fluorescence microscope. All figures were processed in Adobe Photoshop CC 2019 to adjust the levels, brightness and contrast. The reconstruction of the body plan was drafted in Adobe Illustrator CC 2019.
Scanning electron microscopy (SEM)
SEM images were collected by a FEI Quanta 650 FEG in low-vacuum mode using an accelerating voltage of 15kV (30kV in Figure S7F). Elemental composition analyses were carried out with an EDAX Pegasus using accelerating voltages of 15kV (10kV in Figure S6I). All above analyses were performed in the Institute of Palaeontology, Yunnan University.
Phylogenetic analysis
Phylogenetic analyses were performed in MrBayes 3.2.7 (Ronquist et al., 2012) under the mkv + gamma model (Lewis, 2001). 20,000,000 generations were requested, with analyses stopping automatically once the average deviation of split frequencies was <0.01. Ancestral character states for selected nodes were reconstructed in separate analyses using monophyly constraints also performed in MrBayes 3.2.7. Posterior probabilities for the character states at these nodes are plotted as pie charts at nodes shown in Figure 4C.
Supplementary information
Supplementary information includes eight figures, one table and phylogenetic information (character descriptions and codes).
Author contributions
Y.Z. and P.-Y.C. designed research; Y.Z., P.-Y.C., X.-G.H., Y.-J.L. and F.W. collected fossil material; Y.Z., L.A.P. and J.V. performed research, analysed data and prepared all figures; Y.Z., F.S.D and L.A.P collated the morphological data and conducted phylogenetic analyses; Y.Z. and L.A.P. wrote the initial manuscript with significant input from J.V., P.-Y.C. and all other co-authors.
Declaration of interests
The authors declare no competing interests.
Supplementary Information
This PDF file includes
Figures S1 to S8
Tables S1
Phylogenetic information
SI References
Other supplementary materials for this manuscript include the following
Datasets S1 Conicula phylogenetic code
Phylogenetic information
Our character dataset was mainly adopted from Zhao et al. (Zhao et al., 2019), which includes the main taxa and characters of cnidarians and ctenophores. Our changes to previously existing characters were mainly deleting duplicate or meaningless characters, correcting some mistakes and reordering most of characters. We deleted some redundant taxa (mainly octocorals and siphonophores) as well as Eolympia and Namacalathus that are useless in current studies. Except for Conicula, we also added 16 newly sampled taxa, including extant cnidarian Edwardsia, Zoanthus, Craterolophus, Carybdea, Chiropsalmus, Carukia, Atorella, Linuche, Crateritheca and Halammohydra, and tubular fossils Corumbella, Conularia, Eoconularia, Hexaconularia and Carinachites, and cup-shaped cnidarian fossil Cambroctoconus.
The phylogeny analyses were conducted using Bayesian inference under the mkv + gamma model in MrBayes 3.2.7 (Ronquist et al., 2012). We set the number of generations to be 20,000,000 and allowed the stop rule when the average deviation of split frequencies dropped below 0.01, with convergence checked for all parameters (ESS scores >200) using the command ‘sump’. We performed two sets of analyses without any topological constraints (convergence was achieved after <6,000,000 generations in both analyses), one included all sampled taxa and the other removed the newly sampled tubular fossils and Cambroctoconus, of which all lack the preservation of reliable soft tissues and therefore have large amounts of missing data (over 80%). However, some recovered clades within the cnidarian have in-group relationships that are inconsistent with the results of recent phylogenomic analyses (Kayal et al., 2018; e.g. McFadden et al., 2021; Zapata et al., 2015). To get further results of the position of Conicula in phylogenetic trees, we conducted additional analyses using the command of ‘constraint partial’. All fossil taxa were left unconstrained, so they can wander to any clades in the tree. The first topological constraint was combined with the recent common results of phylogenomic studies in cnidarian clades (Collins et al., 2006; e.g. McFadden et al., 2021). While the second focused on the ctenophore-first (ctenophores are the sister group to all other metazoans including sponges) (e.g. Dunn et al., 2008; Whelan et al., 2017) and Planulozoa (for this paper, we only considered cnidarians as the sister group to bilaterians, with the exclusion of Placozoa) (e.g. Pisani et al., 2015; Simion et al., 2017). Our commands using for implementing topological constraints in MrBayes 3.2.7 are listed below.
Character descriptions
The new characters are in bold type and marked with an asterisk.
Collar complex
0 – absent 1 – present
Multicellularity with extracellular matrix
0 – absent 1 – present
Extracellular digestion
0 – absent 1 – present
Ostia with porocytes
0 – absent 1 – present
Septate junctions (SJs)
0 – absent 1 – present
Tight junctions (TJs)
0 – absent 1 – present
Gap junctions (GJs)
0 – absent 1 – present
Adherens junctions (AJs)
0 – absent 1 – present
Hemidesmosomes
0 – absent 1 – present
Epithelia
0 – absent 1 – present
Basal laminae
0 – absent 1 – present
Collagen
0 – absent 1 – present
Nerve cells
0 – absent 1 – present
Acetylcholine used as a neurotransmitter
0 – absent 1 – present
Diffuse nervous system
0 – absent 1 – present
Hox/ParaHox gene
0 – absent 1 – present
Epidermis with pulsatite bodies
0 – absent 1 – present
Xenacoelomorph cilia
0 – absent 1 – present
Striated ciliary rootlets
0 – absent 1 – present
Diploblasts made of 2 cell layers
0 – absent 1 – present
Triploblasts made of 3 cell layers
0 – absent 1 – present
Spiral cleavage with 4d mesoderm
0 – absent 1 – present
Through-gut
Conicula possesses a partitioned, blind gut with only one opening, the type of gut is incompatible with the through-gut bearing two separated openings. Therefore, the character of through-gut is scored as absent in Conicula.
0 – absent 1 – present
U-shaped gut
0 – absent 1 – present
Protonephridia (or homologues)
0 – absent 1 – present
Fate of blastopore
0 – protostomy 1 – deuterostomy 2 – amphistomy
Body cuticle with chitin
0 – absent 1 – present
Body cuticle with α-chitin
0 – absent 1 – present
Body cuticle moulted
0 – absent 1 – present
Segmented body with jointed limbs
0 – absent 1 – present
Lobopods
0 – absent 1 – present
Slime papillae
0 – absent 1 – present
Telescoping mouth cone with protrudable stylets
0 – absent 1 – present
Respiration via metameric tracheae and spiracles
0 – absent 1 – present
Teloblastic segmentation
0 – absent 1 – present
Longitudinal ventral nerve cord(s)
0 – absent 1 – present
Circum-pharyngeal, collar-shaped brain with anterior and posterior rings of perikarya separated by a ring-shaped neuropil
0 – absent 1 – present
Introvert with scalid rings
0 – absent 1 – present
Flosculi
0 – absent 1 – present
Immunoreactivity of horseradish peroxidase (HRP)
0 – absent 1 – present
Lophophore
0 – absent 1 – present
Trochophore
0 – absent 1 – present
Radula
0 – absent 1 – present
Segmental metanephridia sacculus
0 – absent 1 – present
Chitinous microvillar appendages (chaetae)
0 – absent 1 – present
Parapodia with dorsal and ventral branches terminated by β-chitinous chaetae
0 – absent 1 – present
Eversible proboscis surrounded by rhynchocoel
0 – absent 1 – present
Complex jaw apparatus in pharynx
0 – absent 1 – present
Mesoderm
0 – absent 1 – present
Origin of mesoderm
0 – from the blastopore lips and as ectomesoderm
1 – from the walls of the archenteron or neural crest
Mixocoel (haemocoel) surrounded by segmented mesoderm 0 – absent 1 – present
Radial cleavage
0 – absent 1 – present
Coelom formation
0 –schizocoely 1 –enterocoely
Trimeric coelom
0 – absent 1 – present
Pharyngeal slits
0 – absent 1 – present
Endostyle (or homologues)
0 – absent 1 – present
Notochord
0 – absent 1 – present
Stomochord
0 – absent 1 – present
Haemal system with axial complex
0 – absent 1 – present
Calcareous endoskeleton composed of separate ossicles
0 – absent 1 – present
Tornaria type larva
0 – absent 1 – present
Longitudinal dorsal nerve cord
0 – absent 1 – present
Zig zag myomeres
0 – absent 1 – present
Endothelium that lines the inner wall of blood vessels
0 – absent 1 – present
Neural crest
0 – absent 1 – present
Neurogenic placodes
0 – absent 1 – present
Dorsoventral axis
0 – absent 1 – present
Anterior posterior axis
0 – absent 1 – present
Symmetry*
The single character of symmetry in Zhao et al. (character 69)(Zhao et al., 2019) is now split into two characters (69 and 70) to establish character polarity, as scoring the different types of symmetry would necessitate scoring the character of symmetry as absent for Choanoflagellata and Placozoa. The character is here scored as polymorphic for Porifera because the symmetry is present only in some subgroups of sponges, such as calcareous sponges (Manuel, 2009) or Palaeozoic sponges (Botting et al., 2014). The material of Conicula exhibits a circular cross-section, along with the conical gross morphology (with oral-aboral axis), suggesting the presence of symmetric nature.
0 – absent 1 – present
Type of symmetry
We are careful to score this character as uncertain in Conicula because the total number of tentacles or mesenteries in current material is yet to be determined.
0 – bilateral 1 – biradial 2 – triradial
3 – tetraradial 4 – pentaradial 5 – hexaradial
Compression in pharyngeal plane
0 – absent 1 – present
Compression in oral aboral axis
0 – absent 1 – present
Compression in tentacular plane
0 – absent 1 – present
Cydippid larvae
0 – absent 1 – present
Ciliary rosettes
0 – absent 1 – present
Radially-arranged outgrowths from the interface between the oral and aboral regions
Conicula possesses soft outgrowths (tentacles) extending from the oral disc of the polyp, so this character is scored as present in Conicula.
0 – absent 1 – present
Radial outgrowths fixed in globular configuration
0 – absent 1 – present
Radial outgrowths tentacular
0 – absent 1 – present
Outgrowths with pinnules
The tentacles in Conicula are unbranched with no signs of the existence of pinnules. We also code this character as inapplicable in those taxa without tentacular outgrowth because the possession of pinnules is contingent on the presence of tentacles.
0 – absent 1 – present
Outer sheaths on external surface of radial outgrowths
This character is coded as absent in Conicula because no evidence of outer sheaths is present in the external surface of the outgrowths in Conicula.
0 – absent 1 – present
Outgrowths with ciliary rows
0 – absent 1 – present
Ciliary rows paired
0 – paired 1 – unpaired
Orientation of ciliary rows relative to oral-aboral axis
0 – adaxial 1 – abaxial
Uniformity of ciliary rows
0 – uniform 1 – non uniform
Number of ciliary rows
0 – eight 1 – eighteen 2 – six
3 – twenty four 4 – more than 24 5 – sixteen
Cushion rings/plates or polster cells
0 – absent 1 – present
Cushion rings paired
0 – paired 1 – unpaired
Large compound cilia
0 – absent 1 – present
Large cilia fused to form locomotory plate
0 – absent 1 – present
Extension of the oral surface to form oral cone
0 – absent 1 – present
Aboral region represented only by apical organ
0 – absent 1 – present
Apical organ forming narrow pointed extension
0 – absent 1 – present
Oral macrocilia
0 – absent 1 – present
Oral lobes
0 – absent 1 – present
Morphology of tip of oral extension
0 – narrow 1 – voluminous 2 – manubrium-like
Mouth as margin of creeping sole
0 – absent 1 – present
Pharyngeal ridges
0 – absent 1 – present
Sealing ridges of pharynx
0 – absent 1 – present
Macrocilia on pharynx lining inside of mouth
0 – absent 1 – present
Ciliary dome
0 – absent 1 – present
Statolith
0 – absent 1 – present
Balancers
0 – absent 1 – present
Pole plates
0 – absent 1 – present
Aboral papillae
0 – absent 1 – present
Ciliary grooves
0 – absent 1 – present
Interplate ciliary groove (ICG)
0 – absent 1 – present
Pharyngeal canals
0 – absent 1 – present
Tentacular canals
0 – absent 1 – present
Meridional canals
0 – absent 1 – present
Diverticula of meridional canals
0 – absent 1 – present
Circumoral ring canal
0 – absent 1 – present
Termination of meridional and pharyngeal canals
0 – both terminate blindly
1 – branch to form a complex network
2 – united with the circumoral ring canal
Interradial canals
0 – absent 1 – present
Number of interradial canals*
This character ‘interradial canals’ in the previous matrix combined the presence and number of interradial canals (Zhao et al., 2019), and is now split into two separate characters.
0 – two 1 – four
Adradial canals
0 – directly branch from the infundibulum
1 – branch from interradial canals
Aboral canal
0 – absent 1 – present
Anal canals
0 – absent 1 – present
Anal pores
0 – absent 1 – present
Paired ctenophore Tentacles
0 – absent 1 – present
Colloblasts
0 – absent 1 – present
Tentilla
0 – absent 1 – present
Disposition of tentilla
0 – fringing along tentacles 1 – fringing along elongate oral margin
Tentacle sheaths
0 – absent 1 – present
Opening position of tentacles
0 – orally 1 – aborally
Auricles
0 – absent 1 – present
Brood chambers
0 – absent 1 – present
Sclerotised arms and calyx
0 – absent 1 – present
Tentacles extending beyond skeletal rods and outer sheaths
0 – absent 1 – present
Medial structures of skeletal elements
0 – absent 1 – present
Kinked spokes
0 – absent 1 – present
Spinose spokes
0 – absent 1 – present
Radiating flaps/lobes
0 – absent 1 – present
Oral structure separated by circumferential constriction
0 – absent 1 – present
Constriction type
0 – skirt or bell 1 – lappets
Feeding strategy in tentaculate metazoans
Metazoans who rely on pinnules/cilia to feeding are defined as the suspension-feeding groups, of which feeding strategies are mainly the granules suspended in the water current. Those metazoans bearing flexible, non-cilia tentacles are usually preying on macro-organisms. Conicula has flexible, free bending tentacles, with no clear signs of pinnules/cilia, it is coded as predominantly macro for feeding strategy.
0 – predominantly micro 1 – predominantly macro
Mesoglea
0 – absent 1 – present
Mesoglea cellular
0 – absent 1 – present
Embryonic development
0 – direct 1 – indirect
Structure of mitochondrial DNA
0 – circular 1 – linear
Cnidae
0 – absent 1 – present
Cnidae in gastrodermis
0 – absent 1 – present
Cnidocil
0 – absent 1 – present
Cnidae apical structure*
The typical feature of anthozoan cnidae is the absence of operculum (Brusca et al., 2016), and we score the sampled taxa mainly following Reft & Daly (Reft and Daly, 2012).
0 – flaps/cap 1 – operculum 2 – no flaps/operculum
Spirocyst
0 – absent 1 – present
Ptychocyst
0 – absent 1 – present
Stenoteles
0 – absent 1 – present
Euryteles
0 – absent 1 – present
Birhopaloids
0 – absent 1 – present
Rhopalonemes
0 – absent 1 – present
Desmonemes
0 – absent 1 – present
Mastigophores
0 – absent 1 – present
Isorhizas
0 – absent 1 – present
Basitrichous isorhizas
0 – absent 1 – present
Apotrichous isorhizas
0 – absent 1 – present
Heterotrichous anisorhizas 0 – absent 1 – present
Nemathybomes on scapus*
Edwardsia is characterised by prominent nemathybomes that is nematocyst-dense pockets in the epidermis, which is absent in Nematostella (Daly, 2002). It is inapplicable in other cnidarians as their body column is not composed of scapus.
0 – absent 1 – present
Zooxanthellae
0 – absent 1 – present
Mesogleal skeleton
0 – absent 1 – present
Ectodermal skeleton
0 – absent 1 – present
Composition of ectodermal skeleton
0 – proteinaceous 1 – calcitic
Corallum*
Corallum is a spiny, proteinaceous skeleton type widely present in Antipatharia (Daly et al., 2007).
0 – absent 1 – present
Columella
0 – absent 1 – present
Costae
0 – absent 1 – present
Octocorallian spicules
0 – absent 1 – present
Spicules in tentacle
0 – absent 1 – present
Gorgonin
0 – absent 1 – present
Periderm
Considering the chemical composition varies in different types of periderm based on Mendoza-Becerril et al. (Mendoza-Becerril et al., 2016), we also code the periderm as present in Hydra. Cornularia is a unique genus among Octocorallians in having a polyp covered by a theca-like chitinous outer sheath (López-González et al., 1995; Weinberg, 1978), we score Cornularia as present for a periderm as well. The conical external skeleton of Conicula is a robust structure ornamented by parallel annulations, encasing fully the internal polyp. These features are consistent with the tubular fossils with an external periderm, such as conulariids. We therefore code periderm as present in Conicula.
0 – absent 1 – present
Periderm type
Based on Mendoza-Becerril et al., the periderm is assigned into two types of corneous (chitin-protein) and coriaceous (calcium carbonate/phosphate), the former is widely present in extant organisms, while the latter is thought to be present in a few of Anthoathecata, such as Millepora, and most of fossil groups, like conulariids and Corumbella (Mendoza-Becerril et al., 2016). Cornularia is coded as having the corneous type of periderm because of the possession of a chitinous envelope (López-González et al., 1995; Weinberg, 1978). It is unknown the peridermal type of Conicula based on current evidence.
0 – corneous 1 – coriaceous 2 – fibrous
Cuticle layers in periderm
Conulariids had found evidence of two-layer cuticle present in the periderm (Jerre, 1994; Van Iten, 1992a). Corumbella may have one-layer cuticle in the periderm (Mendoza-Becerril et al., 2016). While the number of cuticle layers in Conicula is uncertain.
0 – one 1 – two 2 – five
Regions with periderm*
A new character describes the body region encased by the periderm and is scored mainly following Mendoza-Becerril et al. (Mendoza-Becerril et al., 2016). Basing on fossil material, Conicula and other tubular fossils have a periderm encasing the entire polyp.
0 – entire polyp
1 – hydrocaulus or hydrorhiza
2 – basal region or podocyst
3 – stolon and anthostele
Tubular periderm*
A new character defines the shape of periderm. The tubular periderm is present in most of extant taxa that have a periderm encasing the entire polyp or hydrocaulus. It also appears in Conicula and other tubular fossils, including conulariids, olivooids, Hexaconularia, Carinachites, Corumbella and Sphenothallus.
0 – absent 1 – present
Shape of tubular periderm*
Coronates, Conicula and tubular fossils have a cone-shaped periderm, while hydrozoans have two types of tubular periderm shape that are contingent on the encased region of the periderm.
0 – cone 1 – tube 2 – goblet-like
Cone slender and elongate*
Coronates possess a slender and elongate cone that differs from the pyramid-shaped or conical fossil tubes (such as conulariids and Conicula), except for the tubes of Sphenothallus (Dzik et al., 2017) and Corumbella (Pacheco et al., 2015) that appear to slender and elongate.
0 – absent 1 – present
Tapering end of cone*
There is a disc-like attachment structure present in the tapering end of the cone of coronates, similar structure could be also seen in the cone of Sphenothallus (Dzik et al., 2017). Conicula, olivooids (Dong et al., 2016; Liu, Y. et al., 2014a), Hexaconularia (Van Iten et al., 2010) and Carinachites (Han et al., 2018) lack the structure of attachment disc but instead exhibit a blunt tapering end. It is uncertain that conulariids have an attachment disc, because it is usually broken off at the tapering end of fossil specimens.
0 – blunt 1 – with an attachment disc
Periderm forming a globular expansion*
This new character defines the periderm shape of Conicula. All fossil and living taxa with a tubular periderm do not expand the tube to form a globular chamber in the distal region, and this character is scored as absent in them.
0 – absent 1 – present
Periderm with face*
Unlike the periderm of extant taxa, the periderm is divided into various numbers of faces in conulariids, olivooids, Hexaconularia and Carinachites. We follow the most common opinion to code the peridermal face as present in Corumbella (Babcock et al., 2005; Pacheco et al., 2011; Pacheco et al., 2015; Van Iten et al., 2016). This character is absent in Conicula and Sphenothallus since there are no evident faces present in fossil material.
0 – absent 1 – present
Face divided by corner groove*
This character is generally present in conulariids (Leme et al., 2008b), Hexaconularia (Van Iten et al., 2010) and Carinachites (Conway Morris and Chen, 1992), but is absent in Olivooides (divided by longitudinal ridges) and Corumbella.
0 – absent 1 – present
Corner groove wide and deep*
Carinachites possesses wide and deep grooves with triangular shape in cross-section (Han et al., 2018), which are distinct from the narrow and shallow grooves that appear in conulariids and Hexaconularia.
0 – absent 1 – present
Midline*
The character of midline is widely present in the middle of the face of conulariids (Van Iten, 1992b), but is absent in olivooids, Hexaconularia and Carinachites. We score this character as present in Corumbella following the common opinion (Babcock et al., 2005; Pacheco et al., 2011; Pacheco et al., 2015; Van Iten et al., 2016).
0 – absent 1 – present
Carinae*
This character is widely present in the internal periderm of conulariids (Van Iten, 1992b), and is absent in Corumbella following the recent study (Pacheco et al., 2015).
0 – absent 1 – present
Bipartite periderm stage*
This character is adopted from character 203 ‘embryonic stage retained in the tube morphology’ in previous matrix (Zhao et al., 2019). The tubular periderms of olivooids and Hexaconularia are consisting of two distinct body parts, the embryonic tissue and the post-embryonic tissue (Steiner et al., 2014; Van Iten et al., 2010). Extant cnidarians experience a planula phase before forming a polypoid or medusoid body, without a bipartite periderm stage at early development phase. We score this character as absent in extant cnidarians with a tubicolous periderm and as unknown in all other tubular fossils.
0 – absent 1 – present
Apical region with longitudinal ridge*
Olivooids have longitudinal ridges in the apical cone (embryonic tissue) that play a crucial role to determine the symmetric type (Steiner et al., 2014). Hexaconularia appears to with no evident ridges in the apical region (Duan et al., 2017).
0 – absent 1 – present
Number of ridges*
Olivooides has five ridges in the apical cone (Dong et al., 2016), while Quadrapyrgites exhibits four ridges (Liu, Y. et al., 2014a).
0 – four 1 – five
Periderm apertural end*
This new character describes differences of the apertural end of periderm across sampled taxa. The distal portion of periderm tube protrudes towards the central lumen to form apertural lobes surrounding the terminal opening, as seen in conulariids (Moore, 1956), olivooids (Dong et al., 2016; Liu, Y. et al., 2014a) and Carinachites (Han et al., 2018). The operculum appeared in the periderm tube of hydrozoan (Bouillon et al., 2006) and coronates (Werner, 1973). Sphenothallus (Chang et al., 2018; Dzik et al., 2017) and Corumbella (Babcock et al., 2005; Pacheco et al., 2015) do not develop an apertural operculum or oral lobes. Conicula and Hexaconularia are scored as unknown as it is uncertain based on current fossil material.
0 – open 1 – with lobes 2 – with an operculum
Cross-section of the terminal end of periderm*
This character is adopted from character 1 of Leme et al. (Leme et al., 2008a). The circle-shaped cross-section of the terminal end of periderm is present in Conicula, Sphenothallus and extant hydrozoans and coronates. In contrast, conulariids, olivooids, Hexaconularia, Carinachites and Corumbella have various types of polygonal cross-sections.
0 – circular 1 – polygon
Peridermal tooth
This character is modified based on the code in Zhao et al. (Zhao et al., 2019). Peridermal teeth refer to the inward protrusion of the inner layer periderm to form ridge-like or tooth-like structures towards the lumen. We now score the peridermal tooth as inapplicable in the taxa without a periderm. Several hydrozoans possess intrathecal septa or ridges within the periderm that is presumably homologous to the teeth or ridges of the periderm tube. We hereby add a new taxa Crateritheca with prominent intrathecal septa (Bouillon et al., 2006; Millard, 1975) into the sampled taxa.
0 – absent 1 – present
Tooth morphology*
Olivooides has peridermal teeth in the form of paired projections (Dong et al., 2016). Sphenothallus possesses simple sheet-like cusps (ridges) with smooth rim (Dzik et al., 2017), which may present in Eoconularia (Jerre, 1994) and Conicula as well. We also code Crateritheca as having sheet-like cusps in the perisarc (Bouillon et al., 2006; Millard, 1975).Coronates bear complex peridermal teeth with various shapes (Jarms, 1991; Jarms et al., 2002).
0 – paired projections
1 – sheet-like cusps (ridges) 2 – teeth-like cusps
Tooth disposition*
Sphenothallus (Dzik et al., 2017), Eoconularia (Jerre, 1994) and extant Crateritheca (Millard, 1975) have irregular arrangements of internal ridges. Conicula, Olivooides (Dong et al., 2016) and extant coronates (Jarms, 1991; Jarms et al., 2002) have peridermal teeth arranged in the whorls.
0 – irregular arrangement 1 – whorl
Periderm with annulation*
The external annulation in the periderm is widely present in the tubular fossil groups, including conulariids, olivooids, Corumbella, Carinachites and Conicula, as well as in extant coronates and most hydrozoans. Most tubes of Sphenothallus do not have obvious annulations, but some of them exhibit faint, fine striae (Chang et al., 2018; Muscente and Xiao, 2015; Zhu et al., 2000), we also code annulation as present in the tube of Sphenothallus.
0 – absent 1 – present
Annulation distribution* Most external annulations in the periderms of hydrozoans are confined to the basal portion, and this phenomenon appears also in Conicula. We score the external annulations in coronates, conulariids, olivooids, Sphenothallus, Corumbella and Carinachites as widespread in the periderm tube.
0 – confined to basal part 1 – widespread in periderm
Annulation continuous
The annulations in extant coronates and hydrozoans are continuous around the entire periderm. We score the continuous annulation as present also in Conicula, Sphenothallus and olivooids since they do not have external midlines or corner grooves.
0 – absent 1 – present
Location of annulation offset*
Conulariids are coded following characters 5 and 6 of Leme et al. (Leme et al., 2008a). The annulations of Corumbella (Pacheco et al., 2015) and Hexaconularia (Van Iten et al., 2010) offset in the midlines, and the annulations are interrupted in the corner sulcus of Carinachites (Conway Morris and Chen, 1992).
0 – interradii (midlines) 1 – perradii (corners)
Annulation type*
Based on the morphology of external annulations preserved in the fossil material, we score olivooids as the special crest-like annulations (Dong et al., 2016; Liu, Y. et al., 2014a), while living hydrozoans and coronates as well as Conicula and other tubicolous periderms (Babcock, 1991) exhibit the rib-like annulations, except for Sphenothallus, which has faint, fine striae that distinctly differ from the rib-like annulations (Chang et al., 2018; Zhu et al., 2000).
0 – rib-like 1 – crest-like 2 – striae
Ornament in rib*
The external annulations in extant coronates and hydrozoans are smooth without ornaments, which are also applicable in Conicula, Sphenothallus, Carinachites and Corumbella. The ornaments in ribs of conulariids are variable across different taxa (Leme et al., 2008a). Eoconularia does not have ornaments in the transverse ribs, while Conularia has nodes in the transverse ribs (Babcock, 1991; Leme et al., 2008a).
0 – absent 1 – present
Propagation through lateral budding
0 – absent 1 – present
Oocyte development
0 – oocytes develop without accessory cells
1 – oocytes develop with accessory cells
2 – oocytes develop within follicles
3 – oocytes develop from uptake of somatic or other germ line cells
Nectosome
0 – absent 1 – present
Pneumatophore
0 – absent 1 – present
Planula
0 – absent 1 – present
Planula ciliation
0 – absent 1 – present
Number of endodermal cells of the planula
0 – variable 1 – constant, n=16
Glandular cells in the planula
0 – absent 1 – present
Nervous cells in the planula
0 – absent 1 – present
Relationship between axes of planula and adult
0 – oral-aboral axis in the adult derived from the longitudinal axis of the planula
1 – oral-aboral axis in the adult derived from the transverse axis of the planula
Polypoid phase
In previous matrix Aegina and Halitrephes had been incorrectly coded as the presence of a polypoid phase (Zhao et al., 2019), we now rescored it as absent in these two taxa and newly sampled Halammohydra, and accordingly change other characters contingent on the presence of a polypoid phase to inapplicable in these three taxa.
0 – absent 1 – present
Polyp life mode
In all specimens, Conicula is living in solitary, neither with no signs of attaching to/with one another, or no stolon or other connected structures present.
0 – solitary 1 – colonial
Polymorphic polyps
0 – absent 1 – present
Polyp dominant*
Anthozoans only have a polyp phase. Although most medusozoans have polyp and medusa phases, the two phases are not equally important during the life cycles. The medusa phase is dominant in scyphozoans, cubozoans and staurozoans, while the polyp phase is generally dominant in hydrozoans.
0 – absent 1 – present
Stalk/peduncle in polyp
This character has changed inapplicable in the bilaterians which are here regarded as no polyps present. Conicula is scored as present for a peduncle based on a narrower ribbon-like structure extending from the aboral end in the fossil material.
0 – absent 1 – present
Pedal disc
0 – absent 1 – present
Coelenteron
Several lines of evidence indicate that Conicula possesses coelenteron-like features, including a blind cavity lined by mesenteries and an anthozoan-like pharynx. We score the coelenteron is present in Conicula
0 – absent 1 – present
Actinopharynx
Actinopharynx refers to a short, muscular tubular passageway located between the mouth and gastric cavity in a polyp, which is formed by the invagination of the epidermis (Daly et al., 2007). Conicula has a distinct tubular structure similar to the actinopharynx in some aspects, including shape, topological location and dimension. We score actinopharynx as present in Conicula.
0 – absent 1 – present
Siphonoglyph
This character was repeated in the previous matrix (characters 145 and 199) (Zhao et al., 2019). We retain the ‘character 199’ as it was correctly coded siphonoglyph as absent in Corynactis, Montastraea and Porites based on Daly et al. (Daly et al., 2003).
0 – absent 1 – present
Number of siphonoglyphs
0 – one 1 – more than one
Gastric cavity partitioned by mesentery
The dark longitudinal lines preserved only in the region of gastric cavity indicate the presence of mesentery in Conicula.
0 – absent 1 – present
Mesenterial filament *
We split the character ‘mesenteric filament’ in the previous matrix (character 201) (Zhao et al., 2019) into two separate characters. One is for the presence of mesenterial filament and the other one is considering the number of mesenterial filament strips. Mesenterial filament refers to the thickened, cordlike margin armed with cnidae, cilia and gland cells, which occurs in the free inner edge of each mesentery below the pharynx (Brusca et al., 2016). The gastric septa of medusozoan polyp lack the mesenterial filaments.
0 – absent 1 – present
Number of mesenterial filament strips
0 – two strips 1 – three strips 2 – one strip
Ciliated tract on mesenterial filament
0 – absent 1 – present
Acontia
Acontia are long thread-like extensions of the lower ends of mesenterial filaments, which are armed with numerous stinging cells (Lam et al., 2017). We rescore it as inapplicable in these taxa without mesenterial filaments.
0 – absent 1 – present
Number of mesenteries*
The sampled hexacorallians have more than twelve mesenteries in total except Antipathes which has only ten mesenteries. The sampled octocorallians have eight mesenteries, which is also scored as present in Cambroctoconus (Park et al., 2011). The sampled scyphozoans, cubozoans and staurozoans have four gastric septa that are also appeared in Eoconularia (Jerre, 1994). The estimated number of mesenteries is up to 28 in Conicula based on fossil material, and therefore we code Conicula as present for more than twelve mesenteries.
0 – more than twelve 1 – ten
2 – eight 3 – four
Mesentery in polyp
Cubopolyps have gastrodermal folds instead of the true gastric septa (Marques and Collins, 2004), we therefore code the mesentery is absent in cubopolyps (Straehler-Pohl and Jarms, 2011). This character is rescored as inapplicable in Aegina, Halitrephes and Halammohydra since they develop directly to the medusae phase with no polyps present.
0 – absent 1 – present
Coupled mesenteries*
We follow Daly et al. (Daly et al., 2003) to split the character ‘Pairing of mesentery’ of the previous matrix (Zhao et al., 2019) into two separate characters (coupled mesenteries and paired mesenteries), and score them accordingly.
0 – absent 1 – present
Paired mesenteries*
0 – absent 1 – present
Mesentery pair morphology
0 – members the same size 1 – members differ in size
Paired secondary cycle
0 – absent 1 – present
Perfect mesenteries*
This character is added primarily to establish character polarity before scoring the independent character ‘number of perfect mesenteries’. We score the perfect mesenteries as present in Conicula, because the presence of perfect mesenteries is presumably contingent on the presence of actinopharynx (Daly et al., 2003), a feature appeared in Conicula.
0 – absent 1 – present
Perfect mesenteries only*
This character is adopted from the character ‘Types of mesentery’ of the previous matrix from Zhao et al. (Zhao et al., 2019) with only considering the state ‘only perfect mesenteries’.
0 – absent 1 – present
Number of perfect mesenteries
0 – eight 1 – six or multiple of six
Directive mesentery
In this matrix, we assume the gastric septa of medusozoans are homologous to mesenteries of anthozoans, and therefore rescore this character is absent in scyphozoans and staurozoans rather than inapplicable.
0 – absent 1 – present
Number of directive mesentery pairs
0 – one pair 1 – two pairs
Gonads on mesenteries of 1st cycle
0 – absent 1 – present
Gonads on mesenteries of 2nd and subsequent cycles
0 – absent 1 – present
Number of tentacles in polyp
It is challenged to determine the exact number of tentacles in Conicula. With the incompletely exposed tentacles can be counted as up to 9, the total number of tentacles in Conicula is estimated to more than twenty. The Chengjiang (Ou et al., 2015) and Burgess Shale (Conway Morris and Collins, 1996) ctenophores lack evidence of tentacles, we rescore the characters that are contingent on the possession of tentacles are all inapplicable in these fossil taxa.
0 – six 1 – eight 2 – twelve
3 – sixteen 4 – eighteen 5 – more than twenty
Structure of polyp tentacles
The polyp tentacle of Hydra is now coded as hollow following Ruppert et al. (Ruppert et al., 2004). It is unknown the structure of polyp tentacles in Conicula and other fossil groups.
0 – hollow 1 – solid
Arrangement of tentacles
0 – scattered 1 – one cycle 2 – more than one cycle
Two-tentacle polyp stage
0 – absent 1 – present
Tentacles retractile
0 – non retractile 1 – retractile
Tentacle/coelenteron relationship
0 – one tentacle per endocoel and per exocoel
1 – one tentacle per exocoel, multiple per endocoel
Catch tentacles
Catch tentacles is a special type of tentacle longer than ordinary tentacles, probably using for social behaviour, which is only present in Metridium (Williams, 1975) of our sampled taxa. We score Conicula and dinomischids (Zhao et al., 2019) as absent for catch tentacles based on its gross appearance of tentacles.
0 – absent 1 – present
Acrospheres
0 – absent 1 – present
Marginal spherules
0 – absent 1 – holotrichous
Acrorhagi
0 – absent 1 – present
Organisation of the nervous system
0 – nets 1 – with nerve rings
Canal system in polyp
0 – absent 1 – present
Gastrodermic musculature
0 – not in bunches
1 – organised in bunches of gastrodermal origin
2 – organised in bunches of ectodermal origin
Mesogleal sphincter
0 – absent 1 – present
Ectodermal longitudinal muscle location
0 – tentacles and oral disc only 1 –whole body
Basilar musculature
0 – absent 1 – present
Retractor muscle
0 – weak 1 – defined
Parietal muscle
0 – absent 1 – present
Mesogleal lacunae
0 – absent 1 – present
Actinula
This is an easily confused term as it widely uses to describe the free-moving larva stage of Trachymedusae and Anthomedusae, but the two are not homologous (Bouillon and Boero, 2000; Petersen, 1990). Here, we refer the character actinula only to the larva stage of Anthomedusae, and accordingly score it present in Candelabrum and Ectopleura (Petersen, 1990).
0 – absent 1 – present
Ephyrae
The ephyra is a distinguishing characteristic of living scyphozoans and is absent in staurozoans and cubozoans. Based on fossil material, Olivooides may also have an ephyra stage during the development process (Dong et al., 2013).
0 – absent 1 – present
Marginal lappet type*
The marginal lappet of ephyrae generally comprises two portions (lappet stem and rhopalial lappet) as seen in Aurelia and Rhizostoma, while the lappet stem is absent in Linuche, Atorella and Nausithoe (Straehler-Pohl and Jarms, 2010). It is unknown the type of marginal lappet of Olivooides based on the fossil material.
0 – without lappet stem 1 – with lappet stem
Number of marginal lappets*
Aurelia and Rhizostoma have eight marginal lappets. Coronate Linuche and Nausithoe have sixteen lappets whereas Atorella possesses twelve lappets (Daly et al., 2007). Olivooides probably has five lappets (Dong et al., 2013).
0 – five 1 – eight
2 – twelve 3 – sixteen
Medusoid phase
The medusoid phase is one of the biphasic life cycles of cnidarians, referring to the presence of free-living medusa. It is uncertain in Conicula as all specimens are in the polypoid phase and the medusoid form is still yet to be determined. Considering the presence of ephyra (Dong et al., 2013), it is therefore scored this character as present in Olivooides.
0 – absent 1 – present
Location of medusa formation
Because of the presence of ephyra (Dong et al., 2013), we score Olivooides as producing medusa from apical or oral location.
0 – lateral budding from an entocodon
1 – apical or oral
2 – direct development without polyp stage
Type of apical medusa formation
Unlike the traditional opinion, several cubozoans also have strobilation-like metamorphosis process, such as the newly sampled taxon Carukia (Courtney et al., 2016). The ephyra is one of the results of strobilation, we score Olivooides as strobilation because of the possible fossil record of ephyra (Dong et al., 2013).
0 – strobilation 1 – without transverse fission
Strobilation type
Cubozoan Carukia produces monodisc strobilation (Courtney et al., 2016). Aurelia, Linuche and Nausithoe have polydisc strobilation (Helm, 2018), while Rhizostoma has both polydisc and monodisc strobilation (Holst et al., 2007). Olivooides has two tightly fused ephyrae (Dong et al., 2013), probably indicating the presence of polydisc strobilation.
0 – polydisc 1 – monodisc
Adult medusoid shape
Although the state ‘3 – actinuloid’ existed in the previous matrix, none of taxa has been scored as ‘actinuloid’ (Zhao et al., 2019). Here a new taxon Halammohydra is added into the sampled taxa, and is scored as the presence of actinuloid (Clausen, 1967).
0 – bell 1 – pyramidal
2 – cubic 3 – actinuloid
Shape of horizontal cross-section of the medusa 0 – circular 1 – four-part symmetry
Development of the umbrella
The sampled taxa in previous matrix (Zhao et al., 2019) were all coded as present for a fully developed umbrella except the score of inapplicable and unknown. Here we add a new taxon Halammohydra, of which the medusa develops only the aboral cone (Clausen, 1967).
0 – fully developed 1 – aboral cone
Umbrellar size*
Scyphomedusa is named as the ‘true jellyfishes’ that have a well-developed umbrella up to 2 metres (e.g. Cyanea), while hydromedusa is mostly small. Cubomedusa and stauromedusa are also small compared with scyphomedusa, but is larger than hydromedusa (Brusca et al., 2016).
0 – mostly small (up to 10cm in diameter)
1 – small (up to 30cm in diameter)
2 – big (up to 2m in diameter)
Umbrellar margin
0 – smooth and continuous 0 – lobed
Rhopalia/rhopalioids
0 – absent 1 – present
Complex eyes in rhopalia
0 – absent 1 – present
Statocysts
0 – absent 1 – present
Statocyst origin
0 – endodermic 1 – ectodermic
Statolith composition
0 – MgCaPO4 1 – CaSO4
Ocelli in medusa
0 – absent 1 – present
Giant fibre nerve net (GFNN) in medusae
The character was duplicated in the previous matrix (characters 23, 147 and 258) (Zhao et al., 2019), and here we score it inapplicable in Hydra and Candelabrum because of the absence of the medusoid phase.
0 – absent 1 – present
Nerve ring in medusa
0 – absent 1 – present
Number of rings in nerve ring
0 – one 1 – two
Manubrium
0 – absent 1 – present
Gastric filaments
0 – absent 1 – present
Gastric saccule*
This new character is widely present in the chirodropids of cubozoans (Daly et al., 2007). It is coded as absent in Carybdea, Carukia and other medusozoans.
0 – absent 1 – present
Coronal muscle
0 – well-developed 1 – marginal and tiny
Longitudinal muscles in the peduncle*
This character is contingent on the presence of peduncle, a character generally present in staurozoans. Haliclystus has longitudinal muscles in the peduncle but it is absent in Craterolophus (Miranda et al., 2016).
0 – absent 1 – present
Pedalium of coronate type
0 – absent 1 – present
Pedalium of cubozoan type
0 – absent 1 – present
Pedalial branching*
The branched pedalia bearing numerous tentacles appear in the chirodropids of cubozoans (Daly et al., 2007). This new character is present in Chiropsalmus (Gershwin, 2006) of our sampled taxa.
0 – absent 1 – present
Velum
We rescore the velum as absent in Obelia (Bouillon and Boero, 2000).
0 – absent 1 – present
Claustrum*
The claustrum refers to a membrane constituted by layers of mesoglea and gastrodermis that divides the gastric cavity, which is exclusive to some of the staurozoans and is absent in other cnidarians, including the cubozoans (Miranda et al., 2017). In our sampled taxa, claustrum is present in Craterolophus and is absent in Haliclystus (Miranda et al., 2016).
0 – absent 1 – present
Tentacles in medusa
0 – absent 1 – present
Structure of medusa tentacle
Obelia is rescored as the presence of solid marginal tentacles (Bouillon and Boero, 2000).
0 – hollow 1 – solid
Shape of medusa tentacle
0 – filiform 1 – capitate
Tentacular bulbs
Following the recent studies (Holst et al., 2021; Miranda et al., 2013; Miranda et al., 2016), we now score the tentacular bulbs as absent in staurozoans.
0 – absent 1 – present
Tentacular insertion
0 – at umbrellar margin 1 – away from margin
Number of tentacular whorls
In the matrix of Zhao et al. (Zhao et al., 2019), most medusozoans were coded as one tentacular whorl, but none of the taxa was scored as present for two tentacular whorls. We now add a new taxon Halammohydra, which has two whorls of tentacles (Clausen, 1967).
0 – one whorl 1 – two whorls
Oral arms with suctorial mouths
0 – absent 1 – present
Mesentery in medusa
0 – absent 1 – present
Mesenteric shape in medusa
0 – straight 1 – Y-shaped
Perradial mesenteries
0 – absent 1 – present
Radial canals
0 – absent 1 – present
Circular canal
0 – absent 1 – present
Circular canal partial
0 – absent 1 – present
Peripheral canal system
0 – absent 1 – present
Velarium
0 – absent 1 – present
Velar canals
0 – absent 1 – present
Frenulae
0 – absent 1 – present
Coronal furrow
0 – absent 1 – present
Gonadal location
Unlike the score in the Zhao et al. (Zhao et al., 2019), we add a new state (2 – both sides of gastric septa) to define the gonadal location of staurozoans (Holst et al., 2021; Miranda et al., 2013), since the radial canals are absent in staurozoans.
0 – manubrium 1 – radial canals 2 – both sides of gastric septa
Urticant rings
0 – absent 1 – present
Peronia
0 – absent 1 – present
Acknowledgments
We thank Qiang Ou (China University of Geosciences, Beijing) and Shixue Hu (Chengdu Center, China Geological Survey) for their insightful comments and suggestions, and Karen E. Sanamyan (Kamchatka Branch of Pacific Geographical Institute) for sharing extant sea anemones images in Figure 3F, I, along with Mengying Yin for creating the 3D model in Figure 4B, and Shangnan Zhang for her help with SEM analyses. This work was supported by the National Natural Science Foundation of China (42072019 and 42062001 to P.-Y.C. and F.W.), the Strategic Priority Research Program of Chinese Academy of Sciences (XDB26000000 to P.-Y.C. and Y.-J.L.) and State Key Laboratory of Palaeobiology and Stratigraphy (Nanjing Institute of Geology and Palaeontology, CAS) (193128 and 213111 to F.W. and Y.-J.L.). Y.Z. is supported by a graduate grant from China Scholarship Council (201907030012). F.S.D. is supported by fellowships from Merton College and 1851. L.A.P. is supported by an early career fellowship from St Edmund Hall.