The advent of animals: The view from the EdiacaranMary L. Drosera,1 and James G. Gehlingb,c

aDepartment of Earth Sciences, University of California, Riverside, CA 92521; bSouth Australia Museum, Adelaide, SA 5000, Australia; and cUniversity ofAdelaide, Adelaide, SA 5000, Australia

Edited by Neil H. Shubin, The University of Chicago, Chicago, IL, and approved December 9, 2014 (received for review April 15, 2014)

Patterns of origination and evolution of early complex life on thisplanet are largely interpreted from the fossils of the Precam-brian soft-bodied Ediacara Biota. These fossils occur globally andrepresent a diverse suite of organisms living in marine environments.Although these exceptionally preserved fossil assemblages are typi-cally difficult to reconcile with modern phyla, examination of themorphology, ecology, and taphonomy of these taxa provides keys totheir relationships with modern taxa. Within the more than 30 milliony range of the Ediacara Biota, fossils of these multicellular organismsdemonstrate the advent of mobility, heterotrophy by multicellularanimals, skeletonization, sexual reproduction, and the assembly ofcomplex ecosystems, all of which are attributes of modern animals.This approach to these fossils, without the constraint of attemptingphylogenetic reconstructions, provides a mechanism for comparingthese taxa with both living and extinct animals.

Ediacara | animals | Ediacaran | South Australia | fossils

Fossils of the Ediacara Biota consisting of macroscopic, mor-phologically diverse and generally soft-bodied organisms (1)

occur globally in strata spanning 575–541 Mya, marking the endof the Precambrian (2). The record of these organisms predatesthe well-known Cambrian Explosion by nearly 40 million y andprovides critical information concerning evolutionary innovationsin early complex multicellular life forms on Earth. However, theirphylogenetic affinities and their relationship to Cambrian shellyand nonbiomineralized biotas, and thus their overall place in an-imal evolution on this planet (1), remain poorly constrained. In-deed, most are classified only to the genus and species level. Theapparent discontinuity between the Precambrian and the Cam-brian fossil record is largely based on the absence of skeletal hardparts until the very end of the Ediacaran period and the lack ofCambrian-type constructional morphologies among the Ediacarabiota. With rare exceptions (3–6), fossils of the Ediacara biota arenot found in Cambrian strata, and those that are reported are nottypical Ediacara morphologies. For lack of a strong alternative,much of the biota is, thus, commonly interpreted to have goneextinct by the end of the Ediacaran period (2, 7). A few Ediacaranfossils have been interpreted as stem group metazoans, but theCambrian period marks the unequivocal appearance of mostmajor phyla.Historically, the majority of Ediacara fossils were interpreted

as members of modern animal phyla. Radially symmetrical andsea-pen-like forms have generally been assigned to the Cnidaria,and segmented forms with generally bilateral symmetry have beenassociated with annelids and arthropods (e.g., refs. 8 and 9). How-ever, the challenges of finding unequivocal morphologic characterslinking the majority of these fossils to modern phyla, as well as theirunusual style of preservation as casts and molds in medium-grainedsandstone, prompted suggestions of a wide range of alternativephylogenetic affinities, ranging from an extinct kingdom of “Ven-dobionts” (10) to prokaryotic colonies (11), protists (12), lichens(13), and fungi (14). The lack of forms with clear morphologic ties toeven Cambrian fauna further complicates this issue. Lacking clearrelationships to modern taxa, recently, a novel classification of thesetaxa based on morphologic similarities and differences among justthe Ediacara fossils has led to the interpretation of at least six, andpossibly nine, clades (1, 15). This provides testable hypotheses for

relationships within the Ediacara Biota and, importantly, reveals themorphologic disparity of these taxa.However, although fossils of the Ediacara Biota are not easily

classified with modern taxa, they nonetheless provide the recordof early animals. One of the primary issues is that they are soft-bodied and preserved in a manner that is, in many cases, unique tothe Ediacaran. An alternative venue for providing insight into theseorganisms and the manner in which they fit into early animal evo-lution is offered by examination of the paleoecology, morphology,and taphonomy of fossils of the Ediacara Biota. In this article, wereview the framework of the fossils of the Ediacara Biota and dis-cuss some of the new evidence that demonstrates how these taxa fitinto the record of early metazoan evolution without the constraintsof attempting phylogenetic reconstructions.

The Temporal and Spatial RecordFossils of the Ediacara biota occur at 40 localities worldwide,with four particularly good localities; namely, southeastern New-foundland, the Flinders Ranges of South Australia, the White SeaRegion of Russia, and Namibia (2, 16). The Ediacara fossil assem-blages have traditionally been demarcated as three successiveassemblages: the Avalon, the White Sea, and the Nama, whichhave been interpreted as a reflection of evolutionary controls, asevidenced by the radiometric ages of several key localities (2).However, there are also strong secondary preservational (17, 18)and paleoenvironmental controls (19, 20).What is apparent is that the older Avalon Assemblage contains

a distinct assemblage of taxa that exhibit only a few broad mor-phologic [e.g., rangeomorphs and arboreomorphs (cf. 21)] types.Particularly striking are the common fractal-like self-similarbranching rangeomorph forms with surface area/volume ratios,which are comparable to modern osmotrophic bacteria. Althoughthis feeding type is restricted to these bacteria in the modern era,these self-similar growths may have represented a strategy forovercoming physiologic constraints that typically make osmo-trophy prohibitive for macroscopic life forms (22). Further, these

Significance

Patterns of evolution, origination, and extinction of early ani-mal life on this planet are largely interpreted from fossils of thesoft-bodied Ediacara Biota, Earth’s earliest multicellular com-munities preserved globally. The record of these organismspredates the well-known Cambrian Explosion by nearly 40million years and provides critical information concerning earlyexperimentation with complex life-forms on Earth. Here weshow that, although in appearance, these organisms look verystrange and unfamiliar, many of them may have had a biologyand/or ecology similar to animals today, and some were mostcertainly bilaterians, cnidarians, and poriferans.

Author contributions: M.L.D. and J.G.G. designed research; M.L.D. and J.G.G. performedresearch; M.L.D. analyzed data; and M.L.D. and J.G.G. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.1To whom correspondence should be addressed. Email: Mary.Droser@UCR.EDU.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1403669112/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1403669112 PNAS | April 21, 2015 | vol. 112 | no. 16 | 4865–4870

EVOLU

TION

EART

H,A

TMOSP

HER

IC,

ANDPL

ANET

ARY

SCIENCE

SSP

ECIALFEATU

RE

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0

early unusual and distinct macroorganisms may have been im-portant in cycling dissolved organic carbon that may have beenabundant in the Ediacaran times (22).Elements of the Avalon Assemblage persisted (19, 21), but

there are major increases in biologic and ecologic complexityfollowing in the wake of the appearance of the older Avalon As-semblage that more strongly relate to animals as we understandthem. The younger, “second wave” of the Ediacara Biota com-prising the White Sea and Nama assemblages includes a widerange of new taxa (Fig. 1) (2, 21) that exhibit pronounced biologicand ecologic innovation, as evidenced by dramatic increases inbody plans and ecospace use; it is a radiation in its own right.Notable aspects of this radiation include dramatic increases inmobility (18); the appearance of undisputed bilaterians, such asburrowing organisms and stem-group mollusks (e.g., refs. 23 and24); the advent of sexual reproduction (25); the appearance of thefirst biomineralizers (26, 27); and the advent of active heterotro-phy by multicellular organisms (24, 28–30, 31).Moreover, this diversification in macrofauna body plans and

ecologies was accompanied by a diversification in organicallybound microbial substrates (microbial mats) recorded in the formof extensive sedimentary textures both physically and biologicallymediated. These textured organic surfaces (TOS) are diversepatterned assemblages of structures that partially or completelycover bedding surfaces. Lacking autonomous taxonomic traits,they nonetheless possess discrete, repeating characters (11, 32).The constituent organisms may have been members of a pro-karyotic, microbial community. However, it is likely that they wereeukaryotic forms (32) and that certain large and multicomponentTOS may actually consist of assemblages of multicellular bodyfossils (25, 32). Regardless of phylogenetic affinity, TOS played

a distinctive role in Ediacara seafloor ecology, both as members ofEdiacara communities and likely as preservational mediators (17).Although TOS are known from Avalon fossil assemblages (33),

they achieved unprecedented abundance, morphologic diversity,and complexity as part of the second wave radiation (32). Many ofthe new body and trace fossil taxa characteristic of second waveassemblages occur in intimate association with TOS and appear torepresent matground-based, heterotrophic lifestyles.

Environments of the Ediacara BiotaThere are abundant data and general consensus that all thedeposits that bear Ediacara fossils are marine in origin (19, 34).Although a terrestrial origin was recently proposed (35), thishypothesis has been rigorously tested and is not consistent withany of the sedimentologic data (31, 34, 36–40). Fossils of theEdiacara Biota are indeed preserved in a variety of marineenvironments (2), from outer shelf and slope settings in volcanicsediments of forearc basins of the Avalon Terrane (41, 42) toprograding carbonate platforms such as southern China (43, 44)and shallow marine prograding siliciclastic environments of theeast European Platform (19, 20, 45, 46). Burial events enabledsamples of benthic communities to be conserved that includedsessile body fossils and trace fossils. However, these snapshotsalso include elements of the history of each surface, beginningwith recruitment of benthic taxa (18, 25), ambient currents andwave events, ecologic succession (47), degradation of organisms,and evidence of current deformation and loss of sessile frondousforms in the final burial event (48, 49).Within each Ediacaran assemblage there are considerable

differences in the biotic composition, depending on wave energy,the sedimentary substrate, and the depth of benthic communities

Fig. 1. Common Ediacara biota fossils from the Ediacara Member, Rawnsley Quartzite, Flinders Ranges, South Australia. (A) Dickinsonia costata with pre-contraction outline. (B) Funisia dorothea preserved as body casts and external molds where casts have been lost. (C) Two specimens of Parvancorinaminchami. (D) Kimberella quadrata. (E) Multilayered sandstone case of Bradgatia sp. (F) Sandstone cast of scratch traces, Kimberichnus teruzzii, produced byK. quadrata. (G) Spriggina floundersi. (H) Internal cast of three walls of Pteridinium simplex. (I) Helminthoidichnites isp., groove and levee traces preserved ona bed base. Specimens D and G are from the Ediacara Conservation Park; other specimens are from the Nilpena Heritage Site. (Scale bar: 1 cm.)

4866 | www.pnas.org/cgi/doi/10.1073/pnas.1403669112 Droser and Gehling

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0

(19, 20). As a consequence, the most useful field exposures arethose in which whole bedding surfaces can be exposed andstudied from differing paleoenvironmental settings within a re-gion, particularly in field sites where multiple specimens andsurfaces can be excavated and studied, such as the coastal ex-posure in eastern Newfoundland, the White Sea region of north-western Russia, the Nilpena site in the Flinders Ranges of SouthAustralia, and the Aar Farm site in southern Namibia.

Heterogeneity of the Ediacaran Seafloor: An Example fromthe Ediacaran of South AustraliaOne of the defining characteristics of the Ediacara Biota fossils isthat they are typically preserved in situ. As a result, extensivebedding planes preserve individual communities as accurately asis possible in the fossil record. The Flinders Ranges region ofSouth Australia contains one of the best-exposed and most com-plete successions of Neoproterozoic to early Paleozoic rocks in theworld and includes the type section for the Ediacaran Period.Ediacara fossils occur within the Ediacara Member of the RawnsleyQuartzite, preserved within a succession of shallow marine anddeltaic facies and representing common components of the WhiteSea Assemblage.Well-known fossils of the Ediacara Member, including Dick-

insonia (Fig. 1A), Kimberella (Fig. 1D), Parvancorina (Fig. 1C),Spriggina (Fig. 1G), and other taxa, occur abundantly on the baseof thin-bedded rippled quartz sandstones representing deposition

in shallow marine settings between fair-weather and storm wavebase. The rich fossil assemblages of this wave-base sand faciesrepresent benthic communities typically smothered by sand de-posited by waning storm surges. Fossils are also preserved on thebase of the beds of the sheet-sand facies, where fluidized sandsmothered benthic communities in deeper water, well belowwave base.Excavation of 28 fossiliferous beds 4–25 m2 in area within

these two facies demonstrates marked heterogeneity in bothabundance and diversity between beds (Fig. 2). Within each fa-cies, successions of beds of nearly identical sedimentology sug-gest the possibility there was “community-scale” differentiationof taxa in the Ediacara ecosystem, where taxa were responding tosmall-scale changes in environmental conditions, resulting ina patchy distribution of taxa. These bedding plane data providean unusually good opportunity to examine abundance diversityrelationships, as time-averaging can be assumed to be as minimalas possible. Rarefied generic diversity curves indicate consider-able variation between beds, but overall, the wave-base facies hasa greater diversity overall than the sheet-sands facies (Fig. S1).Individual beds of the Ediacara Member exhibit a wide range ofevenness values (Fig. S2). Although beds with low evenness andhigh dominance are not rare, there are beds with evenness valueson par with those of the Phanerozoic. Furthermore, althoughthese data are not similarly collected, the results are consistentwith the suggestion of Powell and Kowalewski (50) that evenness

Fig. 2. The relative abundance of different fossils on the excavated beds from South Australia. The total number of fossils found on the bed is given inparentheses after the bed name. All the beds shown, with the exception of the beds represented by the four right columns, represent deposition betweenfair-weather and storm wave base [the wave-base sand facies of Gehling and Droser (19)]. The four beds on the right are within the sheet-flow sand facies.The body fossil Funisia is preserved on beds MM2, Matt, and STC-X in abundances that range in the thousands, and thus, these fossils actually dominate beds.However, their dense packing and typically poor preservation do not allow for accurate counts on these beds.

Droser and Gehling PNAS | April 21, 2015 | vol. 112 | no. 16 | 4867

EVOLU

TION

EART

H,A

TMOSP

HER

IC,

ANDPL

ANET

ARY

SCIENCE

SSP

ECIALFEATU

RE

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0

increases through time (Fig. S2). In most Ediacaran cases, thesame taxa are present on most of the beds, but it is the varyingabundances of each that characterize these assemblages. Forexample, Dickinsonia and the frond holdfast Aspidella both occuron nearly all of the beds, but in some cases, Dickinsonia domi-nates beds and in others, Aspidella dominates, indicating thepresence of abundant frondose organisms.Both the bed heterogeneity data and evenness values show

that even though these represent nearly the oldest animal com-munities (following on the heels of those of the Avalon As-semblage), there is a sophistication of community assembly onpar with that of the Phanerozoic. This is striking, considering thelack of evidence for predation, infaunalization, and widespreadskeletonization. Although there is little evidence of interactionsbetween taxa, there are abundant indications that several of thetaxa were interacting with the mat.

Population Structure and ReproductionPreservation of extensive bedding planes also permits analysis oflarge numbers of a single taxon in situ. Funisia dorothea (Fig. 1B)occurs in huge abundance on a number of bedding planes. It isup to 30 cm long and 12 mm in diameter and is divided longi-tudinally into serial units 6–8 mm in length throughout the lengthof the tube (Fig. 1B). Units within the tube decrease in widthprogressively toward the apex, suggestive of growth via terminaladdition. Individuals can occur within dense assemblages,sometimes greater than 1,000/m2. Funisia tubes are directlyconnected to attachment structures that range from 1 to 8 mm.Importantly, attachment structures of a similar size and de-velopmental stage are spatially clustered, even within individualbedding surfaces.The phylogenetic affinity of F. dorothea is problematic. The

lack of evidence for polypoid openings or pores in the body walllimits our understanding of its taxonomic affinities. However,although it is difficult to place these fossils within Metazoa, themorphology and ecology are suggestive of cnidarian or poriferangrade animals. The branching patterns and rarity of branching ofFunisia is consistent with metazoan asexual budding. The con-sistency of tube widths on individual bedding surfaces, thedensely packed nature of the attachment structures, and theclustering pattern of developmental stages of attachment struc-tures on individual bedding planes suggests the juveniles settledas aggregates in a series of limited cohorts.These solitary organisms thus exhibit growth by the addition of

serial units to tubes and by the division of tubes, and dispersedpropagation via the production of spats. Among living organisms,spat production is almost ubiquitously the result of sexual re-production but is known to occur rarely in association with asex-ual reproduction. Hence, despite its morphologic simplicity, theF. dorothea provides evidence of a variety of growth modes anda complex arrangement for the propagation of new individuals.Aggregation is not uncommon among some elements of the

Ediacara biota, being present in the frondose holdfast Aspidella.These typically occur in dense assemblages, but in contrast toF. dorothea, their right skewed size distribution is consistent withcontinuous recruitment, rather than being periodic (18). Re-cently, Darroch and colleagues (51) analyzed the populationstructures of three rangeomorph taxa and one nonrangeomorphfrom Mistaken Point, using the Bayesian Information Criterionlikelihood-based model selection. They found that although thepopulations of these taxa had wide distributions, each populationwas best described as a single distribution. The authors suggestthat, assuming these organisms reproduced sexually, their findingsmost strongly support a continuous or year-round reproductionstrategy for these taxa, as this would allow for unimodal populationswith large overall size ranges.Only a few studies of large numbers of individual taxa pre-

served on a single bedding plane have been done. However, with

continued excavation at various sites, these types of studies havethe promising potential to elucidate the biology of these organ-isms, if not the phylogeny, providing another step toward un-derstanding the link with Cambrian and younger animals.

Skeletonization and a Link to the CambrianThe apparent lack of taxonomic continuity between the Pre-cambrian and Cambrian fossil records has led to controversialand conflicting interpretations about the Ediacara biota andtheir place in the evolution of metazoan life on this planet. Thishas been further complicated by the absence of similar modes ofconstruction between these faunas and the rarity of Precambrianskeletonized fossils. A relatively new Ediacara fossil, Corona-collina acula, described by Clites and colleagues (27), is preservedin the Ediacara Member (Rawnsley Quartzite) of South Australiaand represents the oldest multielement organism (Fig. 3). Coro-nacollina consists of a triradial truncated cone associated withruler-straight spicules, up to 37 cm in length, diverging radiallyfrom the cone. The spicules most commonly disarticulated afterdeath and only rarely are found attached to the truncated cone.The morphologic consistency between articulated and dis-

articulated spicules suggests they were made of a rigid substance,such as opaline silica or calcium carbonate. In life, the spicules likelyprovided structural support in a manner similar to the Cambriandemosponge, Choia. Although generally preserved flattened, Choiaspecimens from the Cambrian Fezouata Formation of Moroccoexhibit raised central regions, perhaps representing soft tissuereplaced by pyrite. The presence of Ediacaran sponge-gradeorganisms has been suggested by both biomarkers (52) and otherbody fossils. A recent review questions the Poriferan origin of manyof these taxa, suggesting that some are, for example, holdfasts ormicrobial in origin (53). However, the sharply preserved spiculatestructures and the nature of preservation in some of these taxa suchas Paleophragmodictya cannot be easily reconciled with organic fibers,tentacles, or microbial structures or a holdfast origin. Constructed

Fig. 3. C. acula represents the oldest evidence of skeletonization. (A) Areconstruction of C. acula, after Clites and colleagues (27). This reconstruc-tion is based on known specimens that have only up to four spicules, but it islikely it had more. (B) The holotype SAM P43257 of C. acula. The centraldepression is the mold of the thimble-like main body of the specimen. Notethe rigid spicules that radiate from the main body.

4868 | www.pnas.org/cgi/doi/10.1073/pnas.1403669112 Droser and Gehling

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0

from a framework of rigid and brittle elements,Coronacollina revealsa constructional mode only recently recognized among members ofthe Ediacara biota. It provides a link in constructional mode acrossthe Cambrian boundary and sheds light on the development ofstructural support in early sponges. Although in many respectsmorphologically simple, Coronacollina is nonetheless one of the mostcomplicated of the Ediacara taxa because it was composed of at leasttwo different materials: the soft-bodied truncated cone, which wassolid enough to withstand compaction, and biomineralized spicules.

Mobility and the Presence of BilateriansThe majority of organisms comprising the Ediacara Biota areinterpreted to represent stationary or attached organisms.However, it now appears that at least four different Ediacaranorganisms were capable of movement. The association of certainbody fossil taxa with motility and feeding traces provides theearliest evidence that the Ediacara Biota included bilaterian-gradeanimals. External molds of Kimberella (Fig. 1D) have been docu-mented in close association with casts of arrays of fanned sets ofbifid scratch traces (Kimberichnus teruzzii; Fig. 1F) from theEdiacara Member of the Rawnsley Quartzite, South Australia,and the Verkhovka Formation, Zimnie Gory Formation, and inthe basal part of the Erga Formation of the White Sea region ofnorthwestern Russia. From their orientation and proximity, thesebody and trace fossils are interpreted as evidence of mat grazingactivities (31, 54, 55). The suggestions of molluscan affinities forKimberella (23, 24, 53) should not overlook some important dif-ferences from extant gastropod grazing behaviors (31, 56). Twogenera of the so-called dickinsoniomorphs, Dickinsonia (Fig. 1A)and Yorgia, are preserved as external molds at the end of serial setsof roughly oriented faint casts of what are interpreted as “resting” or“feeding” traces, where these organisms remained stationary forperiods of time, decomposing the microbial mat substrate, beforemoving to the next site, until the sediment-smothering events thatpreserved them and their “footprints” (57). Suggestions that theseare evidence of fungal “fairy-rings” (58), or random products of waveor current transport (59), are easily countered by the repeated ob-servation that the “footprints” were made in a definite order, asdetermined by identification of the body axis and anterior enlarge-ment of body divisions in these organisms. The implication that themat on the substrate decomposed sufficiently to enable a cast sug-gests that such a process could have been a source of nutrition fordickinsoniids. Alone among the Ediacara biota, dickinsoniids featurepreserved external body molds with apparent contraction marks andvariable patterns of body divisions that might be explained by mus-cular peristaltic contraction. There is no equivalent evidence of dif-ferential contraction of body elements among the array of taxa ofPetalonamae, such as Pteridinium (Fig. 1H) and Phyllozoon (9).The most abundant trace fossils in shallow marine environ-

ments associated with rich assemblages of the Ediacara Biota arevery common groove and levee trails of Helminthoidichnites (Fig.1I) that are found in both part and counterpart on numerous

bedding surfaces with or without body fossil impressions.Uniquely, Helminthoidichnites is preserved on bed soles asgroove and levee and shows alternation of relief, but only whenthe sandstone layer is less than 15 mm thick and not covering anentire bed. This association is interpreted as a product of matmining by an animal too small to be characterized from casts andmolds and limited by the redox state within partly buried matsubstrates. On the basis of fine-grained sandy layers, scallopedlevees enable the direction of burrowing to be determined (Fig.1I, Upper Left). On bed tops, Helminthoidichnites displays evi-dence of avoidance behavior exhibited by parallel spirals andchanging depth where crossings occurred. These traces arecommon in the younger Ediacaran formations from relativelyshallower-water environments such as the Ediacara Member ofthe Rawnsley Quartzite in South Australia (19), the BlueflowerFormation of the Mackenzie Mountains of northwest Canada(60), and the Shibantan Member of the Dengjing Formation insouth China (61, 62). It is unlikely these complicated but de-finitive trace fossils could have been made by anything other thana bilaterian.The serial arcuate looping forms, Paleopascichnus and Yelo-

vichnus, and chains hemispheres, referred to as Neonereites, arenow considered encrusting body fossils, reinterpreted as xen-ophyophores by Seilacher and colleagues (63), but disputed byAntcliffe and colleagues (64). The oldest described trace fossilsseem to be those of actinian-grade polyps that left simple straightor curved locomotive trails in ash-coated, deep-water substratesin the Mistaken Point Formation assemblage in southeasternNewfoundland (65).

Concluding RemarksTaken as a whole, the Ediacara Biota represents an enigmaticassemblage of fossils that are not easily related to modern taxa.However, examination of aspects of the ecology, such as tracefossils, taphonomy, and morphology, reveal that these fossilsshow characteristics of modern taxa. It is clear that bilaterians,cnidarians, and poriferans are represented among the EdiacaraBiota. Although we may never be able to reconcile the phylogenyof all, or even most, of the Ediacara taxa, it is likely that withthese approaches, we will be able to continue to better relatethese taxa with both modern and extinct animals.

ACKNOWLEDGMENTS. We are indebted to Jane and Ross Fargher for access totheir property. Fieldwork was facilitated by D. Rice, M. Dzaugis, M. E. Dzaugis,P. Dzaugis, D. A. Droser, R. Droser, V. Droser, C. Droser, N. Anderson, E. Gooch,C. Casey, M. Laflamme, J. Doggett, J. Paterson, C. Armstrong, J. Perry, D. Reid,J. McEntee, I. Smith, M. Fuller, P. Trusler, members of the South AustralianMuseum Waterhouse Club, and the Ediacaran Foundation. S. Finnegan,L. Tarhan, and C. Hall aided with this manuscript. The research was supported byNational Science Foundation Grant EAR-0074021, NASA Grant NNG04GJ42GNASA Exobiology Program (to M.L.D.), and Australian Research CouncilDiscovery Grant DP0453393 (to J.G.G.).

1. Xiao S, Laflamme M (2009) On the eve of animal radiation: Phylogeny, ecology andevolution of the Ediacara biota. Trends Ecol Evol 24(1):31–40.

2. Narbonne GM (2005) The Ediacara biota: Neoproterozoic origin of animals and theirecosystems. Annu Rev Earth Planet Sci 33:421–442.

3. Conway Morris S (1993) The fossil record and the early evolution of the Metazoa.Nature 361:219–225.

4. Hagadorn JW, Fedo CM, Waggoner BM (2000) Early Cambrian Ediacaran-type fossilsfrom California. J Paleontol 74(4):731–740.

5. Jensen SR, Gehling JG, Droser ML (1998) Ediacara-type fossils in Cambrian sediments.Nature 393:567–569.

6. Lin JP, et al. (2006) A Parvancorina-like arthropod from the Cambrian of South China.Hist Biol 18(1):33–45.

7. Vickers-Rich P, Komarower P, eds (2007) The Rise and Fall of the Ediacaran Biota(Geological Society of London, London).

8. Glaessner MF (1984) The Dawn of Animal Life: A Biohistorical Study (Cambridge Univ.Press, Cambridge).

9. Gehling JG (1991) The case for Ediacaran fossil roots to the metazoan tree, GeologicalSociety of India Memoirs (Geological Society of India, Bangalore) Vol 20, pp 181–224.

10. Seilacher A (1992) Vendobionta and Psammacorallia: Lost constructions of Precam-brian evolution. J Geol Soc London 149(4):607–613.

11. Steiner M, Reitner J (2001) Evidence of organic structures in Ediacara-type fossils andassociated microbial mats. Geology 29(12):1119–1122.

12. Zhuravlev AYu (1993) Were Ediacaran Vendobionta multicellulars? (with 3 figures inthe text). Neues Jahrb Geol Palaontol Abh 190(2):299–314.

13. Retallack GJ (1994) Were the Ediacaran fossils lichens? Paleobiology 20(4):523–544.14. Peterson KJ, Waggoner B, Hagadorn JW (2003) A fungal analog for newfoundland

ediacaran fossils? Integr Comp Biol 43(1):127–136.15. Erwin DH, et al. (2011) The Cambrian conundrum: Early divergence and later eco-

logical success in the early history of animals. Science 334(6059):1091–1097.16. Fedonkin MA, Gehling JG, Grey K, Narbonne GM, Vickers-Rich P (2008) The Rise of Animals:

Evolution and Diversification of the KingdomAnimalia (Johns Hopkins Univ Press, Baltimore).17. Gehling JG (1999) Microbial mats in terminal Proterozoic siliciclasitcs: Ediacaran death

masks. Palaios 14(1):40–57.18. Droser ML, Gehling JG, Jensen SR (2006) Assemblage palaeoecology of the Ediacara

biota: The unabridged edition? Palaeogeogr Palaeoclimatol Palaeoecol 232(2):131–147.

Droser and Gehling PNAS | April 21, 2015 | vol. 112 | no. 16 | 4869

EVOLU

TION

EART

H,A

TMOSP

HER

IC,

ANDPL

ANET

ARY

SCIENCE

SSP

ECIALFEATU

RE

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0

19. Gehling JG, Droser ML (2013) Howwell do fossil assemblages of the Ediacara Biota telltime? Geology 41(4):447–450.

20. Grazhdankin D (2004) Patterns of distribution in the Ediacaran biotas: Facies versusbiogeography and evolution. Paleobiology 30(2):203–221.

21. Laflamme M, Darroch SAF, Tweedt SM, Peterson KJ, Erwin DH (2013) The end of theEdiacara biota: Extinction, biotic replacement, or Cheshire Cat? Gondwana Res 23(2):558–573.

22. Laflamme M, Xiao S, Kowalewski M (2009) From the Cover: Osmotrophy in modularEdiacara organisms. Proc Natl Acad Sci USA 106(34):14438–14443.

23. Fedonkin MA, Waggoner BM (1997) The Late Precambrian fossil Kimberella isa mollusc-like bilaterian organism. Nature 388(6645):868–871.

24. Seilacher A (1999) Biomat-related lifestyles in the Precambrian. Palaios 14(1):86–93.25. Droser ML, Gehling JG (2008) Synchronous aggregate growth in an abundant new

Ediacaran tubular organism. Science 319(5870):1660–1662.26. Germs GJB (1972) New shelly fossils from Nama Group, South West Africa. Am J Sci

272:752–761.27. Clites EC, Droser ML, Gehling JG (2012) The advent of hard-part structural support

among the Ediacara biota: Ediacaran harbinger of a Cambrian mode of body con-struction. Geology 40(4):307–310.

28. Ivantsov AYu, Malakhovskaya YE (2002) Giant traces of Vendian animals. Dokl EarthSci 385:618–622.

29. Fedonkin MA (2003) The origin of the Metazoa in the light of the Proterozoic fossilrecord. Paleontological Research 7(1):9–41.

30. Seilacher A, Buatois LA, Mangáno MG (2005) Trace fossils in the Ediacaran-Cambriantransition: Behavioral diversification, ecological turnover and environmental shift.Palaeogeogr Palaeoclimatol Palaeoecol 227(4):323–356.

31. Gehling JG, Runnegar BN, Droser ML (2014) Scratch traces of large Ediacara bilateriananimals. J Paleontol 88(2):284–298.

32. Gehling JG, Droser ML (2009) Textured organic surfaces associated with the EdiacaraBiota in South Australia. Earth Sci Rev 96(3):196–206.

33. Laflamme M, Schiffbauer JD, Narbonne GM, Briggs DEG (2011) Microbial biofilms andthe preservation of the Ediacara biota. Lethaia 44(2):203–213.

34. Xiao S, et al. (2013) Affirming life aquatic for the Ediacara biota in China and Aus-tralia. Geology 41(10):1095–1098.

35. Retallack GJ (2013) Ediacaran life on land. Nature 493(7430):89–92.36. Antcliffe JB, Hancy AD (2013) Reply to Retallack (2013): Ediacaran characters. Evol Dev

15(6):389–392.37. Chen Z, et al. (2014) New Ediacaran fossils preserved in marine limestone and their

ecological implications. Sci Rep 4(4180).38. Gehling JG, Droser ML (2014) Howwell do fossil assemblages of the Ediacara Biota tell

time?: REPLY. Geology 42(4):e333.39. Menon LR, McIlroy D, Brasier MD (2013) Evidence for Cnidaria-like behavior in ca. 560

Ma Ediacaran Aspidella. Geology 41(8):895–898.40. Xiao S, et al. (2014) Affirming life aquatic for the Ediacara biota in China and Aus-

tralia: REPLY. Geology 42(3):e326.41. Mason SJ, Narbonne GM, Dalrymple RW, O’Brien SJ (2013) Paleoenvironmental

analysis of Ediacaran strata in the Catalina Dome, Bonavista Peninsula, Newfound-land. Can J Earth Sci 50(2):197–212.

42. Wood DA, Dalrymple RW, Narbonne GM, Gehling JG, Clapham ME (2003) Paleo-environmental analysis of the late Neoproterozoic Mistaken Point and Trepasseyformations, southeastern Newfoundland. Can J Earth Sci 40(10):1375–1391.

43. Zhu M, Gehling JG, Xiao S, Zhao Y, Droser ML (2008) Eight-armed Ediacara fossilpreserved in contrasting taphonomic windows from China and Australia. Geology36(11):867–870.

44. Jiang G, Shi X, Zhang S, Wang Y, Xiao S (2011) Stratigraphy and paleogeography ofthe Ediacaran Doushantuo Formation (ca. 635-551Ma) in South China. Gondwana Res19(4):831–849.

45. Gehling JG (2000) Environmental interpretation and a sequence stratigraphic frame-work for the terminal Proterozoic Ediacara Member within the Rawnsley Quartzite,South Australia. Precambrian Res 100(1):65–95.

46. Hall M, et al. (2013) Stratigraphy, palaeontology and geochemistry of the late Neo-proterozoic Aar Member, southwest Namibia: Reflecting environmental controls onEdiacara fossil preservation during the terminal Proterozoic in African Gondwana.Precambrian Res 238:214–232.

47. Liu AG, McIlroy D, Matthews JD, Brasier MD (2012) A new assemblage of juvenile,Ediacaran fronds from the Drook Formation, Newfoundland. J Geol Soc London169(4):395–403.

48. Tarhan LG, Droser ML, Gehling JG (2010) Taphonomic controls on Ediacaran diversity:Uncovering the holdfast origin of morphologically variable enigmatic structures.Palaios 25(12):823–830.

49. Liu AG, McIlroy D, Antcliffe JB, Brasier MD (2011) Effaced preservation in the Ediacarabiota of Avalonia and its implications for the early macrofossil record. Palaeontology54:607–630.

50. Powell MG, Kowalewski M (2002) Increase in evenness and sampled alpha diversitythrough the Phanerozoic: Comparison of early Paleozoic and Cenozoic marine fossilassemblages. Geology 30(4):331–334.

51. Darroch SAF, Laflamme M, Clapham ME (2013) Population structure of the oldestknown macroscopic communities from Mistaken Point, Newfoundland. Paleobiology39(4):591–608.

52. Love GD, et al. (2009) Fossil steroids record the appearance of Demospongiae duringthe Cryogenian period. Nature 457(7230):718–721.

53. Antcliffe JB, Callow RHT, Brasier MD (2014) Giving the early fossil record of spongesa squeeze. Biol Rev Camb Philos Soc 89(4):972–1004.

54. Fedonkin MA, Simonetta A, Ivantzov AYu (2007) New data on Kimberella, the Ven-dian mollusk-like organism (White Sea region, Russia): Palaeoecological and evolu-tionary implications. Geol Soc Lond Spec Publ 286(1):157–179.

55. Ivantsove AYu (2013) Trace fossils of Precambrian metazoans “Vendobionata” and“Mollusks”. Stratigr Geol Correl 21:252–264.

56. Ivantsov AYu (2010) Paleontological evidence for the supposed Precambrian occur-rence of mollusks. Paleontol J 44(12):1552–1559.

57. Gehling JG, et al. (2005) Ediacara organisms: Relating form to function. Evolving Formand Function: Fossils and Development, ed Briggs DEG (Yale Univ., New Haven, CT),pp 43–66.

58. Retallack GJ (2007) Growth, decay and burial compaction of Dickinsonia, an iconicEdiacaran fossil. Alcheringa 31(3):215–240.

59. McIlroy D, Brasier MD, Lang AS (2009) Smothering of microbial mats by macrobiota:Implications for the Ediacara biota. J Geol Soc London 166(6):1117–1121.

60. Carbonne C, Narbonne GM (2014) When life got smart: The evolution of behavioralcomplexity through the Ediacaran and Early Cambrian of NW Canada. J Paleontol88(2):309–330.

61. Chen Z, et al. (2013) Trace fossil evidence for Ediacaran bilaterian animals withcomplex behaviors. Precambrian Res 224:690–701.

62. Meyer M, et al. (2014) Interactions between Ediacaran animals and microbial mats:Insights from Lamonte trevallis, a new trace fossil from the Dengying Formation ofSouth China. Palaeogeogr Palaeoclimatol Palaeoecol 396:62–74.

63. Seilacher A, Grazhdankin D, Legouta A (2003) Ediacaran biota: The dawn of animallife in the shadow of giant protists. Paleontological Research 7(1):43–54.

64. Antcliffe JB, Gooday AJ, Brasier MD (2011) Testing the protozoan hypothesis forEdiacaran fossils: A developmental analysis of Palaeopasichnus. Palaeontology 54(5):1157–1175.

65. Liu AG, McIlroy D, Brasier MD (2010) First evidence for locomotion in the Ediacarabiota from the 656 Ma Mistaken Point Formation, Newfoundland. Geology 38(2):123–126.

4870 | www.pnas.org/cgi/doi/10.1073/pnas.1403669112 Droser and Gehling

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 18

, 202

0