, Volume 90, Issue 4, pp 649–657

Forbidden fruits in the Garden of Ediacara

Research Paper

DOI: 10.1007/s12542-016-0327-3

Cite this article as:
Grazhdankin, D. PalZ (2016) 90: 649. doi:10.1007/s12542-016-0327-3


The expulsion of jellyfish from the "Garden of Ediacara", as described by Adolf Seilacher, has challenged the “gelatinous ocean” stereotype; however, not all discoidal fossils can be interpreted as holdfast structures, sandy skeletons of benthic organisms, microbial colonies, fungal fairy rings, or erosional scratch circles. Here I describe a late Ediacaran (~550 Ma) medusiform organism, Bjarmia cycloplerusa gen. et sp. nov., preserved as a composite mould in a steep crescentic erosional scour cast in fine-laminated sandstone from the Erga Formation in the Southeast White Sea area. Biostratinomic features point to an allochthonous burial of a bowl-shaped body as it trapped mud pebbles when it was suspended in a sediment-laden flow. An unprecedented range of preserved characters, including moulds of a coronal and longitudinal muscles, suggests affinities with scyphomedusae. The organism is reconstructed as a coronate-like jellyfish, with numerous pedalia separated one from another by deep radiating slits, four deep subgenital pits in the floor of the subumbrella, and a skirt of poorly differentiated tentacle-like structures surrounding the large four-cornered mouth opening. Rhopalia and marginal lappets are not preserved in the specimen. Bjarmia cycloplerusa gen. et sp. nov., if borne out by future research, can be used as evidence for a substantial branching by the late Ediacaran within-stem cnidarian lineages—a largely cryptic component of the pre-Cambrian biota—and raises questions about the nature of late Ediacaran food webs.


Ediacaran Scyphozoa Taphonomy Palaeoecology 


Die “Vertreibung” von Quallen aus dem von Adolf Seilacher beschriebenen “Garten von Ediacara” lässt das Bild eines “gelatinösen Ozeans” fraglich erscheinen. Dennoch können nicht alle discoidalen Fossilien als Haftorgan-Strukturen, sandige Skelette benthischer Organismen, mikrobielle Kolonien, kreisförmig auftretende Pilz-Fruchtkörper (Hexenringe), oder Scharrkreise interpretiert werden. In vorliegender Arbeit wird der medusenförmige Organismus Bjarmia cycloplerusa gen. et sp. nov. aus dem späten Ediacarium (~550 Ma) beschrieben. Das Fossil ist als Abdruck in einer fein laminierten Sandsteinfazies der Erga-Formation aus der südöstlichen Weißmeer-Region erhalten. Biostratinomische Merkmale, wie beispielsweise im zentralen Bereich des Fossils befindliche Tonklasten, deuten auf eine allochthone Einbettung des ehemals schalenförmigen Körpers innerhalb eines Suspensionsstromes hin. Eine bisher nie dagewesen Fülle von erhalten gebliebenen Merkmalen, darunter z. B. Abdrücke koronaler und longitudinaler Muskeln, deuten auf eine Verwandtschaft mit Scyphomedusen hin. Der Organismus wird als Coronata-artige Qualle mit zahlreichen, jeweils durch tiefe radiale Fugen getrennten Pedalia, vier tiefen, an der Unterseite der Subumbrella gelegenen Subgenitalhöhlen, sowie schlecht differenzierten tentakelartige Strukturen, welche die große viereckige Mundöffnung saumartig umgeben, rekonstruiert. Sowohl Rhopalia als auch randliche Hautlappen sind in dem untersuchten Exemplar nicht erhalten. Sollten diese Interpretationen durch zukünftige Studien noch weiter bekräftigt werden können, dann dokumentiert Bjarmia cycloplerusa gen. et sp. nov. eine bereits im späten Ediacarium auftretende Verzweigung innerhalb der Stammlinie der Cnidaria—einem weitestgehend kryptischen Bestandteil der präkambrischen Lebewelt—wirft jedoch auch gleichzeitig Fragen bezüglich der Nahrungsnetze im späten Ediacarium auf.


Ediacarium Scyphozoa Taphonomie Paläoökologie 


From the perspective of molecular biology, most extant eumetazoan phyla have their stem lineages rooted in the Ediacaran Period; however, no diagnostic apomorphies have yet been identified in the fossil record (Erwin et al. 2011). In particular, the recognition of polypoid or medusoid cnidarians in the Ediacaran fossil biota has been a matter of controversy for over half a century, since Sprigg (1947, 1949) described fossil jellyfish from a locality near abandoned Ediacaran mines in South Australia. The current fascination with Ediacaran fossils derives largely from (Seilacher 1984, 1989, 1992, 1994, 2007; Buss and Seilacher 1994) iconoclastic hypothesis that most of them represent an extinct kingdom of macroscopic protistan-grade organisms, the Vendobionta. As for the discoidal impressions, Seilacher (1984) concluded that none of them could be confidently interpreted as a fossil jellyfish. There is currently consensus among palaeobiologists that Ediacaran composite discoidal moulds, with a centrally positioned slit, invagination, or a small sunken boss surrounded by a series of radiating cracks, and three-dimensional discoidal casts represent holdfast structures (Gehling 1988; Runnegar 1991; Grazhdankin 2000; Gehling et al. 2000; Serezhnikova 2005) or sandy skeletons of benthic organisms (Seilacher and Goldring 1996; Savazzi 2007). Positive hyporelief, slightly raised-, flat-, to low-relief discoidal moulds on microbial substrate can be interpreted as microbial colonies and fungal fairy rings (Grazhdankin and Gerdes 2007). As for the sharp concentric ridges preserved on erosional surfaces, these are interpreted as having formed by current- or wave-induced rotation of anchored tubular or branching organisms (Jensen et al. 2002).

Fossil record of the cnidarian medusa is very obscure and sporadic (Young and Hagadorn 2010). Even so, the oldest reliable fossil medusae described from the Cambrian Series 3 strata of Laurentia suggest a significant degree of diversification within the Cnidaria by the mid-Cambrian (Cartwright et al. 2007). Although controversial, phosphatized embryos (Dong et al. 2013; but see Han et al. 2013; Yasui et al. 2013) and conulariid-like fossils (Van Iten et al. 2014), as well as a problematic fossil Haootia quadriformis (Liu et al. 2014), point to a protracted and cryptic history of the Medusozoa. It is, however, widely accepted that none of the previously described Ediacaran discoidal fossils can be confidently identified as a pelagic medusa. In this study I describe Bjarmia cycloplerusa gen. et sp. nov., a possible scyphozoan free-swimming medusa occurring together with an Ediacaran assemblage in the Southeast White Sea area.

Geological Age

The taxon is described from a single specimen collected 12 m above the base of the Erga Formation of the Valdai Group cropping out in Zimnie Gory (Winter Mountains) coastal cliffs in the Southeast White Sea area of northwestern Russia (Grazhdankin 2004a) (Fig. 1). The Valdai Group comprises, from bottom to top, the Lamtsa, Verkhovka, Zimnegory, and Erga formations, and forms a 500-m-thick succession of undeformed unmetamorphosed shallow marine, fluviodeltaic, and alluvial tuffaceous-siliciclastic rocks. A U–Pb zircon date of 555.3 ± 0.3 Ma (recalculated as 552.85 ± 0.77 Ma; Schmitz 2012) for a volcanic tuff in the Zimnegory Formation, approximately 70 m below the fossiliferous bed with Bjarmia cycloplerusa gen. et sp. nov., is a maximum for the age of the described taxon (Martin et al. 2000).
Fig. 1

Fossil locality and geological age of Bjarmia cycloplerusa gen. et sp. nov. Stratigraphic position of the fossil horizon in the composite section of the Ediacaran strata for the Southeast White Sea area (left) and on a correlation scheme of White Sea Winter Coast sections (right)

The base of the Zimnegory Formation is an erosional unconformity that is revealed in drill-core sections to the northeast of Zimnie Gory (Grazhdankin 2003, 2004b; Maslov et al. 2008; Grazhdankin and Maslov 2009) (Fig. 1). The erosional relief reaches at least 100 m of incision into the underlying Verkhovka Formation. The valley fill begins with thin (0.1–0.2 m), laterally discontinuous conglomerate followed by quartz sandstone, with flaser laminations and channel casts. This is overlain by a thick unit of finely laminated shale with volcanic tuff interbeds; the lowermost of these interbeds yielded a U–Pb zircon date of 552.85 ± 0.77 Ma. The shale in turn progrades into laminated siltstone with channelised sandstone packages exhibiting planar and cross-bedding. Measured trough-cross-set azimuths are unimodal and southwest-directed, indicating strong current influence.

The Erga Formation can be divided into two members. The lower member fills an incised valley, with erosional relief up to 29 m, cut into the Zimnegory Formation (Grazhdankin 2003, 2004b; Maslov et al. 2008; Grazhdankin and Maslov 2009) (Fig. 1). The incised valley fill is a coarsening-upward succession consisting of five fining-upward sequences well exposed in coastal cliffs of the Zimnie Gory. The first sequence commences with conglomerate (0.3–0.5 m), which is overlain by an interval (1.8–2.4 m) of graded siltstone–shale couplets, and then by a thick unit (reaching 8 m in thickness) of bluish grey, occasionally variegated laminated shale. Each of the remaining four sequences comprises packages (0.8–1.9 m up to 4.1 m) of grey and yellowish grey fine-grained sandstone overlain by an interval (1.1–2.2 m up to 3.0 m) of graded siltstone–shale couplets. Thinner sandstone beds tend to consist of fine horizontal laminations. However, thicker units exhibit planar and hummocky stratification, convoluted laminations, amalgamation surfaces, ball-and-pillow structure, isolated shale clasts, and wave ripple laminations. In addition, the intervals of graded siltstone–shale couplets host laterally discontinuous sandstone units (up to 15 cm thick). The specimen of Bjarmia cycloplerusa gen. et sp. nov. was collected from a sandstone package in the base of the second sequence (Fig. 1).

Where the incision has not occurred, such as in sections along the Torozhma and Zimnyaya Zolotitsa rivers further northeast (Fig. 1), the base of the Erga Formation is lined with packages of conglomerate and quartz sandstone exhibiting wave ripple laminations overlain by the upper member. Idiomorphic zircon crystals of primary magmatic origin from shales ~0.75 m above the base of the conglomerate in a section along the Zimnyaya Zolotitsa River (Fig. 1) have yielded a U–Pb date of 550.2 ± 4.6 Ma (Iglesia Llanos et al. 2005). This date constrains the age of Bjarmia cycloplerusa gen. et sp. nov. from above.

The lower member of the Erga Formation is characterised by a diverse Ediacaran fossil assemblage that includes (but is not limited to) Andiva ivantsovi, Anfesta stankovskii, Archaeaspinus fedonkini, Armillifera parva, Aspidella terranovica, Bomakellia kelleri, Brachina delicata, Charniodiscus concentricus, Cyanorus singularis, Dickinsonia costata, Dickinsonia tenuis, Dickinsonia lissa, Eoporpita medusa, Ivovicia rugulosa, Kimberella quadrata, Mawsonites spriggi, Paravendia janae, Parvancorina minchami, Protolyellia simplex, Rangea schneiderhoehni, Solza margarita, Temnoxa molliuscula, Tribrachidium heraldicum, and Yorgia waggoneri (Fedonkin et al. 2007; personal observations).

Systematic palaeontology

Class Scyphozoa Götte, 1887

Subclass Scyphomedusae Ray Lankester, 1877

Stem-group Coronatida Vanhoeffen, 1892

Family Incertae sedis

Genus Bjarmia gen. nov.

Derivation of name From the old Norse term Bjarmia, referring to the Southeast White Sea area.

Type species Bjarmia cycloplerusa gen. et sp. nov.

Diagnosis A coronate-like medusa with numerous pedalia separated one from another by deep radiating slits; four deep subgenital pits in the floor of the subumbrella; and a skirt of poorly differentiated tentacle-like structures surrounding the large four-cornered mouth opening.

Bjarmia cycloplerusa sp. nov.

Figures 2 and 3
Fig. 2

Holotype of Bjarmia cycloplerusa gen. et sp. nov. (positive-hyporelief composite mould) showing four subgenital pits (sp), a coronal muscle (mc), radial musculature (mr), pedalia (p), a marginal ring canal (mrc), short contractile marginal projections (mp), and long cylindrical tentacle-like structures (t). The central area is obscured by a cluster of flattened ovoidal mud pebbles (pb)

Fig. 3

Preservation of radial musculature (mr), pedalia (p), marginal contractile projections (mp), and marginal ring canal (mrc) seen from different angles and lit from different directions

Etymology From the Greek Open image in new window (closing the circle).

Holotype Paleontological Institute (Moscow), specimen no. 3993-7038.

Type locality Approximately 4.5 km south of Zimnie Gory Lighthouse, Winter Mountains, Winter Coast of the White Sea, Arkhangelsk District, northwest Russia.

Preservation The only discovered specimen (Fig. 2) is preserved as a positive-hyporelief composite mould surrounded by crescentic erosional scour cast in fine-laminated sandstone with gently sloping cross-bedding. Scouring next to the specimen suggests that in the process of burial, the flow was deformed around the body, leading to a localised erosion by relatively strong and persistent vortices. The scour also designates the stoss side of the body. The central area of the specimen bears shallow casts of flattened ovoidal mud pebbles, which appear to be stacked along the stoss side and spaced in the down-current direction. There are at least 17 pebbles varying in length from 5 to 17 mm. Confinement of the pebbles to the central area points to an originally concave bowl- or saucer-like shape of the body that was capable of trapping the sediment and mud clasts during burial. Furthermore, a certain regularity in the arrangement of the pebbles, specifically the rhomboid outline of their cluster (Fig. 2), suggests that the mud clasts infilled a four-cornered depression in the body.

Sediment erosion from the stoss side produced no evidence that would suggest an extension of the body deep into the underlying sediment (e.g., Grazhdankin 2000), which makes it unlikely that the organism was a mud-dweller or a holdfast structure. It is also important that the adjacent bedding surface is devoid of mud clasts. All this implies allochthonous burial of the body that was engulfed by a benthic sediment flow and trapped the pebbles when it was suspended. Loaded with pebbles, the body then sank, convex-down, on the muddy sea floor and was smothered with sand. The positive relief of the fossil, therefore, most likely corresponds to the original concave side of the body, while composite mould features could be interpreted as overprinted internal structures. The counterpart of the specimen consisted of poorly consolidated mudstone and is therefore not preserved.

The central area of the specimen (reaching 45 mm in diameter) is surrounded by a ring consisting of irregular sharp concentric folds and four symmetrically placed crescent-shaped tubercles (see Fig. 2). The tubercles are inflated, with smooth surface, along the lee side of the body, but are flattened, occasionally overprinted by concentric folds along the stoss side. The taphonomy suggests that the tubercles correspond to sediment-filled cavities reaching 14 mm in length. In addition, a patch of radiating shallow grooves is seen along the lee side of the specimen. The grooves have various widths (0.1–1.0 mm) and are equidistantly spaced 0.3 mm apart. Although the folds, tubercles, and grooves constitute a ring, it is discontinuous along the upstream side, partially hidden by pebbles. Unlike pebbles, the ring is offset downstream relative to the centre of the circular mould. Considering the inferred bowl-like original shape of the body, its central concave part was filled with pebbles and served as an anchor at burial, while the downstream offset of its marginal soft parts could result from deformation due to current stress.

The circular rim of the bowl-shaped mould is quilted with radiating slits (Fig. 3), which continue also along the upstream side, indicating that their preservation is unaffected by current stress. The quilts are convex, although occasionally flattened or concave, suggesting collapsed preservation of soft, non-resistant parts, or deflated internal cavities. The slits are most likely to correspond to internal rigid radiating septa overprinted during collapse and composite moulding. They terminate 3–4 mm before the margin, which is preserved as a load structure, suggesting collapse of a marginal ring canal.

The lee side of the margin is decorated with radiating projections (see Fig. 3). Their preservation in positive relief indicates collapse of tentacle-like marginal extensions. Some of them are uniformly wrinkled, resembling shrinkage features of cylindrical bodies. The circular margin is also partially discontinuous along the lee side, where it is overprinted by a 45-mm-wide patch of fine, up to 12-mm-long ridges. In addition, small patches of fine ridges can be seen along the margin from the right-hand side (Fig. 2), as well as overprinting the left-hand side of the mould. The ridges are interpreted as collapsed cylindrical tentacle-like structures, but in contrast to the marginal projections, these are never wrinkled, have blurred outlines, and lack composite mould features.

Environmental context Fossil horizon with Bjarmia cycloplerusa gen. et sp. nov. is part of the interstratified sandstone and shale lithofacies interpreted as wave- and current-agitated prodelta deposits (Grazhdankin 2003, 2004a, 2004b; Maslov et al. 2008; Grazhdankin and Maslov 2009). The sandstone packages could be a result of periodic rapid progradation of sandy shoreface, whereas the amalgamated units with soft-sediment deformation features suggest episodes of high sediment supply rate causing local destabilisation of slopes and promoting slumping.

Reconstruction The specimen is reconstructed as a bowl-shaped organism with tentacles differentiated into contractile short marginal projections and a skirt of simple long tentacles springing from its concave side. Structures of the central area, if any, are hidden by the pebbles and could not be identified in the specimen. However, the regular arrangement of the pebbles suggests a four-cornered depression in the concave side lacking any large structures projecting from the central area. The overprinted internal anatomy may indicate that the organism had a four-ray rotational symmetry along the axis of the bowl. The reconstructed body plan resembles a medusiform organism, and is further described in terms of a jellyfish morphology and anatomy.

The fossil lacks composite mould features that could be immediately attributed to an internal gastrovascular cavity. Stomach impression is expected in the central area, which instead shows pebbles only. Four-cornered depression is interpreted as a mouth opening and pharynx. Also, the four sediment-filled cavities indented into the subumbrella resemble the so-called subgenital pits in the recent scyphomedusae, where they arch slightly into the gastric septa (Figs. 2, 3). Similar structures are known in fossil representatives (Brandt 1871; Haeckel 1874; von Ammon 1906). If this interpretation is correct, then the cavities suggest the presence of four septa that divided the coelenteron into a central stomach and four gastric pouches. Furthermore, the radiating slits could be reconstructed as thickened radial septa separating gastric pouches from a marginal ring-sinus, which in turn sends out a radiating vessel in each marginal projection. The reconstructed quilts resemble gelatinous thickenings of pedalia (Figs. 2, 3).

Concentric folds in the specimen of Bjarmia cycloplerusa gen. et sp. nov. are interpreted as a taphonomic feature of muscular contraction (Fig. 2). The swimming muscle in recent scyphozoans consists of a highly folded subepithelial stratum of striated fibrils arranged on larger mesogleal folds (Gladfelter 1972). The thin elastic nature of the mesoglea and the firm muscle anchorage leads to a deformation of the body as the muscle contracts. Muscular contraction folds are an obvious feature in the fossil scyphomedusae from Solnhofen Plattenkalk (Haeckel 1866, 1874; Maas 1902; von Ammon 1906; Kieslinger 1939; Leich 1995) and in taphonomic experiments involving modern scyphozoans (Rozhnov 1998), but are not readily discernible in the Cambrian fossil scyphomedusae preserved stranded on a tidal flat (Hagadorn et al. 2002; Hagadorn and Belt 2008).

Fine grooves extending along the lee side of the studied specimen are interpreted as impression of the muscular tissue, specifically bundles of the longitudinal (radial) muscles (Figs. 2, 3). An alternative interpretation, that the fine grooves are stretch wrinkles, is not supported by their restricted and confined occurrence in the area within the ring of concentric folds. Nor could they be interpreted as tool marks, which are always preserved in hyporelief as ridges.


A rotational (radial) symmetry in the arrangement of body parts, combined with the putative evidence for (1) the central cavity with branches and pockets, (2) the differentiated tentacles surrounding the mouth opening, and (3) the putative musculature, suggest cnidarian affinity of Bjarmia cycloplerusa gen. et sp. nov. Possible scyphomedusoid affinities within the Cnidaria are indicated by (1) a body plan consistent with an acraspedote medusa with a free-swimming lifestyle, as inferred from the preservation; (2) a subumbrella indented by genital cavities; and (3) a seemingly highly developed muscular system consisting of circular and radial muscle bands. There are also a few possible coronate features, such as (1) subdivision of the aboral surface into a central dome-like apex and a peripheral zone with marginal tentacles (coronal groove in the exumbrella is not preserved); (2) four crescent-shaped fusions that form gastric septa partially separating the stomach from the peripheral pouches; and (3) radiating septa between the pedalia that further subdivide the stomach into peripheral pouches. Assuming the features are correctly interpreted, the set of characters in the studied specimen fits the group of coronate scyphomedusae, but precludes reference to any of the families without their drastic and unwarranted redefinition.

Interpretation of Bjarmia cycloplerusa gen. et sp. nov. as a scyphozoan also raises questions about the nature of late Ediacaran food webs. From the perspective of ecosystem functioning, scyphomedusae represent an important group that can act as planktonic top predators capable of causing direct and indirect changes in lower trophic levels through cascading effects (Pitt et al. 2007; Pauly et al. 2009). However, the ecological effects of a top predator are variable, and depend on the complexity of the aquatic community (Strong 1992; Polis and Strong 1996). In simple, species-poor late Ediacaran communities, the jellyfish’s functional role could be completely different. Furthermore, as suggested by the fossil record, the origin of meso-zooplankton, primary prey of modern scyphozoans and an important component of the food chain, seems to postdate the Erga Formation by at least 20 million years (Butterfield 2007; Harvey and Butterfield 2008). Prior to the Cambrian “explosion” and the introduction of pancrustaceans, cnidarians may have preyed on benthic micro-metazoans, and the correlative innovation of true endomesoderm in bilaterians and the cnidocyte in their immediate sister group, the cnidarians, has been suggested as a coevolutionary response between these two lineages early in animal evolution (Erwin et al. 2011). In addition, the ability to assimilate dissolved organic matter allows scyphozoan medusae to feed at very different trophic levels (Skikne et al. 2009), and may help explain their success in relatively oligotrophic and eutrophic environments.

The advent of gelatinous zooplankton, with rapidly sinking jelly biomass, could have increased the efficiency of the biological pump, further oxygenating the water column and triggering a large-scale ecosystem change (Lebrato et al. 2013; Lenton et al. 2014). Interestingly, Bjarmia cycloplerusa gen. et sp. nov. was also coeval with the oldest ichnofabric consisting of burrows with meniscate backfill structure Nenoxites isp., up to 6.5 mm in width (Rogov et al. 2012). The advent of bioturbation was followed by a major wave of Ediacaran extinction, including the complete disappearance of several clades of soft-bodied organisms (Laflamme et al. 2013; Grazhdankin 2014). The concurrent evolution of biologically controlled mineralization and reef-building habits in metazoans could not be coincidental (Penny et al. 2014). Taken together, the discovery of a possible scyphozoan medusa reinforces a picture of early origination in Cnidaria, and provides further evidence of the ecological turnover in the late Ediacaran.


This study was supported by the Russian Science Foundation grant 14-17-00409. I thank the reviewers (Alex G. Liu, James W. Hagadorn, and M. Reich) for valuable comments on the manuscript.

Copyright information

© Paläontologische Gesellschaft 2016

Authors and Affiliations

  1. 1.Trofimuk Institute of Petroleum Geology and Geophysics, Prospekt Akademika Koptyuga 3Novosibirsk State UniversityNovosibirskRussia