EVOLUTION & DEVELOPMENT 13:5, 408–414 (2011) DOI: 10.1111/j.1525-142X.2011.00499.x Eoandromeda and the origin of Ctenophora Feng Tang,a,b,∗ Stefan Bengtson,c,∗ Yue Wang,d Xun-lian Wang,e and Chong-yu Yina,b a Institute of Geology, Chinese Academy of Geological Sciences, Beijing, 100037, China b Key Laboratory of Stratigraphy and Paleontology, Chinese Academy of Geological Sciences, Beijing, 100037, China c Department of Palaeozoology and Nordic Center for Earth Evolution, Swedish Museum of Natural History, Box 50007, SE-104 05, Stockholm, Sweden d School of Resources and Environment, Guizhou University, Guiyang, 550003, China e School of Earth Sciences and Resources, China University of Geosciences, Beijing, 100083, China ∗Authors for correspondence (email: [email protected] and [email protected]) SUMMARY The Ediacaran fossil Eoandromeda octo- lacking crown-group synapomorphies such as tentacles, sta- brachiata had a high conical body with eight arms in he- toliths, polar fields, and biradial symmetry. It probably had a licospiral arrangement along the flanks. The arms carried pelagic mode of life. The early appearance in the fossil record transverse bands proposed to be homologous to ctenophore of octoradial ctenophores is most consistent with the Planulo- ctenes (comb plates). Eoandromeda is interpreted as an zoa hypothesis (Ctenophora is the sister group of Cnidaria + early stem-group ctenophore, characterized by the synapo- Bilateria) of metazoan phylogeny. morphies ctenes, comb rows, and octoradial symmetry but INTRODUCTION The new data on Eoandromeda suggest that it represents a stem-group ctenophore, implying that ctenophores were Megascopic biotas of the 635–541 Ma (million years ago) among the first eumetazoan lineages to appear, and that Ediacaran Period represent the foreplay to the “Cambrian planktic predators were present in the ecosystems already Explosion” of animals, but Ediacaran fossils are contentious during the heyday of the Ediacaran biota. as to their affinity (e.g., Dzik 2003; Seilacher et al. 2003; Narbonne 2005; Fedonkin et al. 2007; Budd 2008). The oc- toradial fossil Eoandromeda octobrachiata Tang et al. 2008 was described in two-dimensional preservation from 580– RESULTS 551 Ma black shales of the Doushantuo Formation in southern China (Tang et al. 2008; Zhu et al. 2008) and in Twelve specimens are figured in Fig. 1; three (Fig. 1B–E) semi-relief on bed soles in the Ediacara Member of the were previously figured by Wang et al. (2008) and are re- Rawnsley Quartzite in South Australia (Zhu et al. 2008). figured herein to illustrate some of the anatomical features. Both Tang et al. and Zhu et al. noted possible relationships Like the previous Doushantuo specimens (Tang et al. 2008; to the Ctenophora (comb jellies) and the Cnidaria, but left Zhu et al. 2008), the new ones are preserved as carbonaceous the question of systematic assignment open. films in a silty shale. Zhu et al. (2008) noted that of 38 Aus- Recently collected material from the type locality of E. tralian specimens, where the up versus down orientation was octobrachiata elucidates the anatomy of Eoandromeda and known, all had dextrally coiling arms (i.e., tips pointing in reveals new features that help a direct comparison with Re- the clockwise direction) when viewed from above. Assuming cent and fossil ctenophores. Cambrian ctenophore-like fossils that this symmetry holds true also for the Chinese specimens have been found in the ca. 505 Ma Burgess Shale (Conway (where the stratigraphic orientation is mostly unknown), Morris and Collins 1996) and the ca. 520 Ma Chengjiang Fig.1E,F,K,L,andMshowthefossilfromabove;Fig.1A– (Chen and Zhou 1997; Hou et al. 2004; Hu et al. 2007) de- D, G, and H from below. posits, and the report of a ctenophore embryo from the ca. Specimens hitherto reported from China and Australia 535 Ma Kuanchuanpu Formation extended the record to are symmetrically compressed disks. The specimens in near the beginning of the Cambrian (Chen et al. 2007). Edi- Fig. 1H–J, however, seem to represent obliquely to later- acaran records are problematic, however, notwithstanding ally compressed specimens. The one in Fig. 1H clearly be- Dzik’s (2002, 2003) ingenious attempts to relate sedentary longs to E. octobrachiata, but the center of radiation of the frond-like organisms (“petalonomans”) to the clade. arms is offset toward the periphery, suggesting an oblique 408 C 2011 Wiley Periodicals, Inc. Tang et al. The origin of Ctenophora 409 Fig. 1. Carbonaceous compressions of Eoandromeda octobrachiata. Scale bars 5 mm. Specimens in Institute of Geology, Chinese Academy of Geological Sciences, Beijing (prefix JK); School of Resources and Environment, Guizhou University, Guiyang (prefix JK-A); and School of Earth Sciences and Resource, China University of Geosciences, Beijing (prefix WH-A). (A) JK10909. (B and C) WH-A-04178. C—close-up of B; arrows point to transverse bands. (D) WH-A-04023. (E) WH-A-04209. (F) WH-A-04719. (G) JK-A-55-0025. (H) JK08014. (I) JK-A-54-0089. (J) JK08016. (K) JK-A-40-0013. (L) JK10903. (M) JK10329. 410 EVOLUTION & DEVELOPMENT Vol. 13, No. 5, September–October 2011 Fig. 2. Reconstruction of Eoandromeda octobrachiata. Artwork Javier Herbozo. compression of a high dome. The specimen in Fig. 1I has a agenetic halo), shows that the spiral morphology was fully rounded triangular shape. Although the anatomical features developed already at this growth stage. are poorly preserved, there is clear evidence of spiral arms The central part inside the arms is mostly indistinctly in the lower part. The elliptical specimen in Fig. 1J shows preserved, but two specimens (Fig. 1L and M) show a ring- even less anatomical details, but the longitudinal streaks and like structure, ca. 3 mm in diameter, connected to the inner the clear area near the top indicate that this specimen too is ends of the arms through a narrow neck. In Fig. 1L, the ring, a preservational variety of E. octobrachiata. We thus inter- neck, and central parts of the arms are made up of a lighter pret these three specimens as oblique to lateral compressions (thinner?) material than the rest of the arms. of an originally conical body, about as high as wide, with a The specimens figured herein represent different modes of bluntly rounded apex, and with arms curving down along its taphonomic degradation. The most conspicuous difference flanks (Fig. 1I). regards the expression of the transverse bands of the spiral Zhu et al. (2008) noted that both Chinese and Australian arms. They vary from distinct (Fig. 1A, C, E, F, H, and specimens show evidence of partitions in the arms and con- M), through more indistinct and frayed (Fig. 1D and G), cluded that “the live organism consisted of eight spiral tubu- to largely absent (Fig. 1I–L and parts of the specimens in lar arms with possible transverse structures.” Our new mate- Fig. 1B and H). The fact that different modes of expression rial confirms the existence of transverse structures, showing can be seen in a single specimen (e.g., Fig. 1B) indicates that them to be band-like elements that are regularly distributed this is a taphonomic rather than taxonomic difference. The along the arms (Fig. 1A, C–F, H, and M). The bands are transverse bands, when present, tend to be expressed in a 0.3–2 mm apart and roughly perpendicular to the arms; in darker material than the arm itself, which in turn is darker one specimen, they are particularly distinct and can be seen than the interband area, which in turn is darker than the to keep their perpendicular orientation even when an arm matrix surrounding the specimen (see, e.g., Fig. 1A and H). forms an angular kink (Fig. 1A, upper left). In two speci- They sometimes extend beyond the visible boundaries of mens (Fig. 1D and G), the bands are crescent-shaped, convex the arms, giving the latter a frayed edge (Fig. 1M). These side facing centripetally (Fig. 1D and G, left) or centrifugally observations suggest that the transverse bands represent a (Fig. 1G, right). tissue distinct from that of the arms proper. The tissue between the spiraling arms is usually expressed Other preservational differences are more subtle and by a dark stain, as in the holotype (Tang et al. 2008, Fig. 2A– mostly concern the distinctness of the arms and the expres- G). The arms are consequently not free tentacle-like struc- sion of the tissues within and between the arms. The specimen tures but are incorporated into the body. The specimen in in Fig. 1L has lighter central regions of the arms, suggesting Fig. 1K, the smallest one found (surrounded by a light di- that the center of the arms was thinner or less solid than the Tang et al. The origin of Ctenophora 411 periphery. Some specimens show an aggregation of irregu- Fig. 1L, on the contrary, shows arms with the interior lighter larly sized black grains up to 300 μm in size between the than the rim, suggesting that the interior was less dense. Zhu spiraling arms (Wang et al. 2008, Fig. 3H). These are ab- et al. (2008) interpreted the Australian specimens to show sent from the adjacent shale and may represent taphonomic tubular arms, possibly with transverse structures, and some- degradation of the interarm tissue. what deformed by compaction. A tubular structure is not The two laterally preserved specimens (Fig. 1I–J) have at odds with the preservation in the Doushantuo specimens, poorly expressed arms. This is probably at least partly due to and a homology with the eight meridional canals underlying the fact that they represent two arm-bearing flanks adpressed the eight comb rows in modern ctenophores (Ruppert et al. to one another.