ArticlePDF Available

The Beothukis/Culmofrons problem and its bearing on Ediacaran macrofossil taxonomy: Evidence from an exceptional new fossil locality

Authors:

Abstract

The late Ediacaran siliciclastic successions of eastern Newfoundland, Canada, are renowned for their fossils of soft-bodied macro-organisms, which may include some of the earliest animals. Despite the potential importance of such fossils for evolutionary understanding, the taxonomic framework within which Ediacaran macrofossils are described is not clearly defined. Rangeomorphs from a newly discovered fossil surface on the Bonavista Peninsula, Newfoundland, require us to reconsider contemporary use of morphological characters to distinguish between genera and species within Ediacaran taxa. The new surface exhibits remarkable preservational fidelity, resolving features smaller than 0.1 mm in dimension in both frondose and non-frondose taxa. Such preservation permits the recognition of rarely observed fourth-and fifth-order rangeo-morph branching, offering unparalleled opportunities to investigate the fine-scale construction of rangeomorph taxa including Culmofrons plumosa Laflamme et al., 2012. Our observations enable resolution of taxonomic issues relating to rangeomorphs, specifically overlap between the diagnoses of the frondose genera Beothukis Brasier and Ant-cliffe, 2009 and Culmofrons. We propose a taxonomic framework for all Ediacaran macrofossils whereby gross architecture, the presence/absence of discrete morphological characters and consideration of growth programme are used to distinguish genera, whereas morphometric or continuous characters define taxa at the species level. On the basis of its morphological characters, Culmofrons plumosa is herein synonymized to a species (Beothukis plumosa comb. nov.) within the genus Beothukis. This discussion emphasizes the need to standardize the taxonomic approach used to describe Ediacaran macrofossil taxa at both the genus and species levels, and raises important considerations for future formulation of higher-level taxo-nomic groups.
THE BEOTHUKIS/CULMOFRONS PROBLEM AND ITS
BEARING ON EDIACARAN MACROFOSSIL
TAXONOMY: EVIDENCE FROM AN EXCEPTIONAL
NEW FOSSIL LOCALITY
by ALEXANDER G. LIU
1
,JACKJ.MATTHEWS
2
and DUNCAN MCILROY
3
1
School of Earth Sciences, University of Bristol, Life Sciences Building, 24 Tyndall Avenue, Bristol, BS8 1TQ, UK; e-mail: alex.liu@bristol.ac.uk
2
Department of Earth Sciences, University of Oxford, South Parks Road, Oxford, OX1 3AN, UK; e-mail: jack.matthews@univ.ox.ac.uk
3
Department of Earth Sciences, Prince Philip Drive, Memorial University of Newfoundland, St. John’s, NL Canada, A1B 3X5; e-mail: dmcilroy@mun. ca
Typescript received 15 April 2015; accepted in revised form 4 September 2015
Abstract: The late Ediacaran siliciclastic successions of
eastern Newfoundland, Canada, are renowned for their
fossils of soft-bodied macro-organisms, which may include
some of the earliest animals. Despite the potential impor-
tance of such fossils for evolutionary understanding, the
taxonomic framework within which Ediacaran macrofossils
are described is not clearly defined. Rangeomorphs from a
newly discovered fossil surface on the Bonavista Peninsula,
Newfoundland, require us to reconsider contemporary use
of morphological characters to distinguish between genera
and species within Ediacaran taxa. The new surface exhi-
bits remarkable preservational fidelity, resolving features
smaller than 0.1 mm in dimension in both frondose and
non-frondose taxa. Such preservation permits the recogni-
tion of rarely observed fourth- and fifth-order rangeo-
morph branching, offering unparalleled opportunities to
investigate the fine-scale construction of rangeomorph taxa
including Culmofrons plumosa Laflamme et al., 2012. Our
observations enable resolution of taxonomic issues relating
to rangeomorphs, specifically overlap between the diag-
noses of the frondose genera Beothukis Brasier and Ant-
cliffe, 2009 and Culmofrons. We propose a taxonomic
framework for all Ediacaran macrofossils whereby gross
architecture, the presence/absence of discrete morphological
characters and consideration of growth programme are
used to distinguish genera, whereas morphometric or con-
tinuous characters define taxa at the species level. On the
basis of its morphological characters, Culmofrons plumosa
is herein synonymized to a species (Beothukis plumosa
comb. nov.) within the genus Beothukis. This discussion
emphasizes the need to standardize the taxonomic
approach used to describe Ediacaran macrofossil taxa at
both the genus and species levels, and raises important
considerations for future formulation of higher-level taxo-
nomic groups.
Key words: Rangeomorph, Ediacaran, Newfoundland, sys-
tematics.
THE late Ediacaran strata of Newfoundland record some
of the oldest fossil evidence for large, soft-bodied organ-
isms. Fossils are typically preserved beneath volcanic tuffs
or volcaniclastic sediments, deposited in deep-marine
turbiditic environments (Wood et al. 2003; Ichaso et al.
2007; Brasier et al. 2013; Mason et al. 2013) ~580 to
560 Ma (Benus 1988; Van Kranendonk et al. 2008). The
Newfoundland biota includes many taxa whose biological
affinities have been widely debated (summarized in Liu
et al. 2015a). In recent years, researchers have tended to
interpret Ediacaran macrofossils on a case-by-case basis,
proposing that a range of biological groups, including
bacteria (Callow and Brasier 2009a; Laflamme et al.
2011), and potentially algae (Hofmann et al. 2008), fungi
(Callow and Brasier 2009b) and metazoans (Liu et al.
2014a,b), co-existed in the Ediacaran marine ecosystems
of Newfoundland.
The most abundant and diverse macroscopic group
within the Newfoundland successions is the Rangeomor-
pha (Liu et al. 2015a), members of which are characterized
by the possession of self-similar branching architectures
(Narbonne 2004; Brasier et al. 2012). The preservation of
rangeomorphs in the manner typical of Avalonian (eastern
Newfoundland and the southern UK) Ediacaran
macrofossils, on siliciclastic bedding planes as external
moulds and casts (cf. Kenchington and Wilby 2014),
contributes to a dearth of diagnostic morphological char-
acters with which to constrain their phylogenetic affinities.
©The Palaeontological Association doi: 10.1111/pala.12206 45
[Palaeontology, Vol. 59, Part 1, 2016, pp. 45–58]
There is no clear and consistent taxonomic framework
within which to describe Ediacaran macrofossils, many of
which are not currently attributable to living clades.
Efforts have been made to formulate high-level schemes
by grouping genera within higher-order Linnaean ranks
(e.g. Sepkoski, in Schopf and Klein 1992) or non-
Linnaean groupings (Laflamme et al. 2013; Grazhdankin
2014). A robust taxonomic framework is important, as
the genera and species that form the basis of existing Edi-
acaran taxonomy have been employed as the foundation
both for efforts to determine relationships between taxa,
and for discussions of palaeoecological attributes such as
diversity, disparity and population structure (Clapham
et al. 2003; Darroch et al. 2013). Despite this, there has
been little agreement or even discussion of the characters
or features that might usefully define Ediacaran genera
with respect to species. If we are to make effective pro-
gress in Ediacaran palaeobiology, the formulation of gen-
erally accepted taxonomic protocols is imperative. This
attractive proposition is impeded by the need to consider
non-uniformitarian and often abiological interpretations
for many preserved structures (Brasier et al. 2013), and
by preconceptions of the possible physiology of Ediacaran
organisms (necessarily guided primarily by extant taxa)
which may introduce artificial biases to taxonomic
schemes. However, the greatest limitation to our under-
standing of the morphological construction and ontogeny
of the Ediacaran macrobiota is often the quality of the
available fossil record.
Much of the progress made in Avalonian Ediacaran
palaeobiology stems from data collected at a handful of
localities exhibiting high-quality macrofossil preservation
(e.g. the ‘D’ and ‘E’ surfaces of the Mistaken Point Eco-
logical Reserve, or the bedding plane at Spaniard’s Bay;
Clapham et al. 2003; Narbonne 2004). Recent discoveries
have significantly increased the number of known fossil-
bearing sites, most notably on the Bonavista Peninsula
(O’Brien and King 2004; Hofmann et al. 2008), and in
Charnwood Forest in the UK (Wilby et al. 2011). A
bedding plane on the east coast of Burnt Point, near the
town of Port Union, Bonavista Peninsula, formally docu-
mented here for the first time (Fig. 1), is remarkable for
its high-quality preservation of Ediacaran macrofossils.
The horizon exhibits high taxonomic diversity, and a
wide size-range of specimens including some of the
largest and smallest rangeomorphs yet discovered. The
surface is named the MUN Surface to reflect the long his-
tory of research on Newfoundland’s Ediacaran successions
by researchers from the Memorial University of New-
foundland (MUN). Importantly, preservational fidelity on
the MUN Surface permits assessment of the architectural
scheme of Brasier et al. (2012) in defining rangeomorph
taxa. Here, we use MUN Surface specimens to assess the
role morphological characters (including those not
relating to branching architecture) can play in rangeo-
morph systematics; resolve a taxonomic conundrum relat-
ing to the genera Beothukis and Culmofrons; and propose
an extension of the classification approach of Brasier
et al. (2012) for rangeomorphs to other Ediacaran macro-
fossil groups. We also consider some of the broader issues
surrounding incorporation of Ediacaran taxa into higher-
level taxonomic groups, which may aid future develop-
ment of a consistent, global, higher-rank taxonomic
scheme for these organisms.
THE MUN SURFACE FOSSIL
ASSEMBLAGE
The MUN Surface lies near the base of the Port Union
Member of the Trepassey Formation (cf. O’Brien and
King 2005; Fig. 1), within an interval dominated by med-
ium- to thick-bedded buff-grey turbiditic sandstones with
soft-sediment deformation, rounded intraclasts and cen-
timetre- to decimetre-scale carbonate concretions (Liu
et al. 2015b, appendix S1, figs 12, text 1). Fossils are pre-
served as positive and negative epirelief impressions on a
siltstone surface, beneath a ~6-mm-thick fine-grained tuff
layer (Liu et al. 2015b, appendix S1, fig. 3), which is itself
overlain by 3- to 100-cm-thick beds of grey to grey-green
siltstone and coarse sandstones (Liu et al. 2015b,
appendix S1, figs 12).
Fossils on the MUN Surface occur in densities of up
to 45 well-preserved individuals/m
2
, with over 250
exceptionally preserved identifiable specimens on a total
exposed surface (at low tide) of ~120 m
2
(Figs 2, 3;
Liu et al. 2015b, appendix S1, fig. 4). Around 500
small, indistinct impressions not included in this biotic
density count represent additional poorly preserved
juvenile specimens (Liu et al. 2015b, appendix S1, fig.
4). The most striking fossils are members of the Range-
omorpha (Narbonne 2004; Laflamme et al. 2013) and
include at least 40 specimens of a unipolar form (i.e.
possessing a single apical generative zone) comparable
to Culmofrons plumosa (Fig. 2A; Liu et al. 2015b,
appendix S1, fig. 6; see later discussion); seven large
specimens of Bradgatia aff. linfordensis Boynton and
Ford, 1995 (Fig. 2BC); 43 Primocandelabrum sp.Hof-
mann et al., 2008 (Fig. 3A); rare Fractofusus andersoni
Gehling and Narbonne, 2007 (Fig. 3C); and Charnia
masoni Ford, 1958 (Fig. 3D). Other taxa include
numerous Charniodiscus spp. including C.procerus
Laflamme et al., 2004 (Fig. 3B); isolated holdfast discs;
a single partial specimen of the possible cnidarian
Haootia quadriformis (Liu et al. 2014a, fig. 1f); Thec-
tardis avalonensis Clapham et al., 2004; four Hadrynis-
cala avalonica Hofmann et al., 2008; and several poorly
defined ivesheadiomorphs (cf. Liu et al. 2011). Speci-
46 PALAEONTOLOGY, VOLUME 59
mens range in size from 10 mm to 0.7 m in maximum
dimension, and can preserve features ~0.1 mm in
dimension (Fig. 2C), a quality of morphological resolu-
tion comparable to that of the younger Ediacara Hills
(South Australia) and White Sea (Russia) assemblages
(Gehling and Droser 2013). Some rangeomorph speci-
mens exhibit up to five orders of self-similar branching
(e.g. Fig. 2C; Liu et al. 2015b, appendix S1, fig. 5),
whereas other notable fossil surfaces in Newfoundland
typically preserve only two or three. In addition to taxa
assignable to known late Ediacaran genera, a dense fab-
ric of hundreds of thin filamentous impressions is also
present (Fig. 3E). Filamentous structures are arranged
in multiple orientations on the bedding plane and can
both drape and underlie macrofossil taxa. Individual fil-
aments are 0.20.6 mm in width, but can reach up to
400 mm in length. They exhibit no internal or external
ornament and show no clear branching.
A tectonic overprint is evident on the surface, both as
two sets of cleavage fractures oriented at 016°and 350°,
respectively, and as deformation of holdfast discs (assumed
from study of multiple specimens to have originally been
circular) to oval shapes. The holdfast discs have an aspect
ratio (width/length) of 0.73 (mean value taken from 20
specimens to 2 d.p., relative standard deviation 6.35%),
with long axes oriented along a NNESSW axis (i.e. broadly
along strike), consistent with regional compressive tecton-
ism associated with the formation of the Catalina Dome.
THE TAXONOMY OF EDIACARAN
RANGEOMORPHS
Taxonomy forms the framework for many current and pre-
vious attempts to understand Ediacaran organisms and
their palaeoecology. However, there are no clear definitions
of what constitutes a species-level characteristic as opposed
to a generic attribute within Ediacaran macrofossil taxon-
omy, let alone agreement on a framework for the higher-
order classification of these taxa. Until we have a better
handle on their biological affinities, it is difficult to be sure
about how fixed the phenotype of these organisms was,
and the extent to which they conform to the morphological
and biological species concepts. We consider that as under-
standing of Ediacaran palaeobiology and palaeoecology
improves and attention turns to global questions, a stan-
dardized, consistently applied scheme for the diagnosis of
Ediacaran macrofossil taxa would be advantageous. Here,
we address the characters used to define Ediacaran rangeo-
morphs in light of the new MUN Surface specimens.
The broad concepts of what constitutes a genus, and
how genera should be objectively defined, have long been
FIG. 1. Map and stratigraphic column (latter not to scale) showing the location of the MUN Surface. A, Newfoundland, Eastern
Canada. B, the Avalon and Bonavista Peninsulas (see box in A), showing the major Ediacaran fossil localities of Mistaken Point, Spa-
niard’s Bay, Ferryland and the Catalina Dome. C, geological map of the Catalina Dome (after Hofmann et al. 2008), showing major
settlements, and the location of the MUN Surface (yellow star). See the column for key to the geological units. Dates in the column
are taken from Benus (1988) and Van Kranendonk et al. (2008, after Bowring et al. 2003), from correlative units on the Avalon Penin-
sula. Note that these dates have only been published in abstract form. Stratigraphy follows O’Brien and King (2005).
LIU ET AL.: EDIACARAN MACROFOSSIL TAXONOMY 47
debated (Calman 1949; Mayr 1963; Melville 1995). Cal-
man (1949, p. 17) noted that the genus ‘has no objective
existence as a group but is merely a convenient device to
make easier the cataloguing and handling of numbers of
species’. Calman further proposed that in establishing
new genera, distinguishing characters ‘should be such as
we may reasonably suppose to be longer established in
phylogeny than those distinctive of species... [remaining]
FIG. 2. Ediacaran macrofossils from the MUN Surface, Bonavista Peninsula, Newfoundland. A, unipolar rangeomorph Beothukis plu-
mosa comb. nov., with frond preserved in negative epirelief. Note the clearly displayed second-order rangeomorph branches along the
length of the frond. B, silicon rubber cast (i.e. positive hyporelief) of a partial Bradgatia aff. linfordensis specimen. Primary branches
are distally inflated and can reach over 22 cm in length. C, close-up image of Bradgatia from the MUN Surface preserved in negative
epirelief, showing sub-millimetre resolution of preserved features. All scale bars represent 10 mm.
FIG. 3. Ediacaran macrofossils on the MUN Surface, Bonavista Peninsula, Newfoundland. A, juvenile Primocandelabrum sp., exhibit-
ing clear first and second-order branching. B, Charniodiscus procerus.C,Fractofusus sp.; this specimen is too small to confidently iden-
tify to species level. D, Charnia masoni. E, multiple superimposed macroscopic filaments, oriented in a variety of directions; filaments
vary from broadly straight to gently curving; note the unweathered tuff still covering the surface at top left. A, B, D and E, preserved
in positive epirelief; C, preserved in negative epirelief. All scale bars represent 10 mm except C which is 1 mm.
48 PALAEONTOLOGY, VOLUME 59
LIU ET AL.: EDIACARAN MACROFOSSIL TAXONOMY 49
unchanged amid variations in other characters’ when
assessing a group of related forms (Calman 1949, p. 17).
Late Ediacaran fossil assemblages commonly offer only
small populations of specimens, of variable preservational
quality. Assessment of the variability of characters within
populations can therefore be difficult. In recent years,
increased awareness and understanding of wider global
sections and sedimentological and taphonomic processes
have resulted in considerable taxonomic synonymization
(Fedonkin et al. 2007). Amongst Avalonian taxa, discoidal
forms have largely been synonymized (Gehling et al.
2000), but there has also been much work to formally
describe new non-discoidal taxa. Most existing Avalonian
Ediacaran genera are monospecific (see Liu et al. 2015a,
table 1). As rangeomorphs are common in Newfound-
land, and are one of the few groups with a widely recog-
nized shared morphological attribute (the possession of
self-similar branching within centimetre-scale specimens;
Narbonne 2004; Brasier et al. 2012; Hoyal Cuthill and
Conway Morris 2014), they offer a useful case study for
Ediacaran taxonomic questions.
Many Ediacaran macro-organisms are preserved only
as mould or cast impressions on rock surfaces, so when
discussing their taxonomy, we are exclusively dealing
with morphotypes. Macrofossil taxa in Newfoundland
were originally distinguished by the presence or absence
of characters such as central rods and stems/stalks, and
by variations in gross shape, branch shape, branching
angle, number of branches and polarity (Hofmann et al.
2008; Bamforth and Narbonne 2009; Narbonne et al.
2009; Laflamme et al. 2012). However, some of these
parameters, such as branch angle and number of
branches, have since been recognized to vary within
populations of certain species, for example through
ontogeny (Liu et al. 2012), and have therefore been sug-
gested to be unsuitable for use in taxonomic diagnosis
(Wilby et al. 2015). There has also been little consistency
in whether formal diagnoses have been assigned to the
genus (as with Beothukis,Pectinifrons,Primocandelabrum,
Parviscopa,Hapsidophyllas,Frondophyllas, Vinlandia), or
the species (e.g. Avalofractus,Culmofrons), in monospecific
taxa. Diagnosing the species within a monospecific genus
would prevent creation of further species within that
genus, so we urge future workers to only diagnose species
when multiple species exist within a genus. More enlight-
ening in terms of addressing taxonomic methodologies are
taxa with multiple species, such as Fractofusus or the
arboreomorph Charniodiscus, in which gross morphology
defines the genus, and variations in frond shape, number
of branches, lengthwidth ratios and the presence/absence
of subsidiary branches have been considered to be species-
level traits (cf. Laflamme et al. 2004; Gehling and Nar-
bonne 2007). More recently, the structural architecture of
rangeomorph branches has been considered a valid char-
acteristic with which to discriminate between rangeo-
morph taxa, leading to the formalization of a taxonomic
scheme based on branching architecture (Narbonne et al.
2009; Brasier et al. 2012). This scheme is largely consistent
with existing rangeomorph taxonomy, and it proposes
that the presence or absence of structural features such as
holdfast discs, and branching architecture (the way in
which branches are arranged within a frond) are genus-
level traits, whereas morphometric or continuous charac-
ters (such as number of branches and shape of the frond)
would either be better suited for the diagnosis of species,
or used only with caution as diagnostic criteria, as they
may have been subject to ecological or ontogenetic influ-
ences (Brasier et al. 2012).
The MUN Surface specimens conform to the concepts
of branch furling, display and inflation proposed as
suitable characters for taxonomic identification within
the rangeomorphs (Brasier et al. 2012). They may also
permit future extension of such concepts to consider
variation in branching morphologies within third- to
fifth-order sub-units. Exceptionally preserved MUN Sur-
face specimens of unipolar forms similar to Culmofrons
plumosa warrant further attention, as they lie at the heart
of a conundrum that has a bearing on how rangeo-
morphs are diagnosed.
The genera Beothukis and Culmofrons
Culmofrons plumosa was formally described in March
2012 (Laflamme et al. 2012), at a similar time to the pub-
lication of Brasier et al. (September 2012) in which mor-
phological architecture was suggested as a means of
characterizing rangeomorph taxonomic diagnoses. The
genus Culmofrons was erected using material from the
Mistaken Point region of the Avalon Peninsula
(Laflamme et al. 2012). Culmofrons plumosa, the type
species, was diagnosed as follows (note that the generic
diagnosis of Culmofrons is ‘as for species’):
Rangeomorph frond with a spatulate to ovate peta-
loid composed of few (less than five on each side)
alternating primary branches forming a zigzagging
central axis. Basal primary branches attach directly
to a long cylindrical stem and circular unorna-
mented holdfast. Primary branches composed of
several (between 8 and 12) sub-rectangular to
trapezoidal secondary modular units oriented per-
pendicularly to the primary branches. Secondary
modular units composed of cm-scale rangeomorph
frondlets. (Laflamme et al. 2012, p. 195)
It is clear from this description, and from images of
the type material (Fig. 4C; Laflamme et al. 2012), that
some of the best-preserved rangeomorph fossils on the
50 PALAEONTOLOGY, VOLUME 59
MUN Surface are encapsulated by the diagnosis of
C. plumosa (e.g. Figs 2A, 4B). However, some of these
MUN Surface specimens were figured as part of the
emended diagnosis of the rangeomorph genus Beothukis
(Brasier et al. 2012, fig. 8B), and provide the basis for the
following emended generic diagnosis of Beothukis:
Frond unipolar, comprising two rows of primary
branches arranged in irregularly spaced alternations
along a furled central axis, forming a linear suture.
Inflation of first- and second-order branches is moder-
ate to medial. Mature first- and second-order branches
typically have furled margins, with alignments that are
arranged in radiating to subparallel series. Rangeo-
morph elements of the first-order branches are usually
undisplayed, whereas those of second-order branches
are clearly displayed. A basal disc and stem is some-
times preserved. (Brasier et al. 2012, p. 1114)
As with the diagnosis of Culmofrons, it is clear that the
MUN Surface material could equally be assigned to the
genus Beothukis as defined by Brasier et al. (2012),
although our new material differs in possessing a zigzag
rather than a linear central axis. We are thus faced with a
taxonomic conundrum that requires resolution.
Following the Brasier et al. (2012) scheme for classify-
ing rangeomorph genera, focusing solely on branching
architecture, requires that the MUN Surface specimens be
assigned to Beothukis, which has taxonomic priority.
However, as stated by Laflamme et al. (2012), there are
several clear differences between the type material of
Beothukis mistakensis Brasier and Antcliffe, 2009 (cur-
rently the sole species of Beothukis), and Culmofrons;
namely the presence in the latter of a long stem, fewer
than five primary branches, and a zigzag central axis
(Table 1; Fig. 4D; Laflamme et al. 2012, pp. 197198),
and these differences are shared by our specimens (Fig. 4;
Liu et al. 2015b, appendix S1, fig. 6). It is therefore
imperative to consider whether the morphological details
in which these taxa differ reflect species- or genus-level
characters. Throughout this discussion, we suggest that
the taxonomic importance ultimately ascribed to a char-
acter is dependent on whether its morphological appear-
ance in an organism is considered to be the result of
inherent genetically based programming, or extrinsic
factors in the palaeoenvironment.
The course of the midline (straight or zigzag) is a weak
taxonomic character in rangeomorphs as it may be prone
to taphonomic variability (Laflamme et al. 2007; Brasier
et al. 2012). In contrast, the presence or absence of a
FIG. 4. A comparison of MUN Surface specimens with the type specimens of Beothukis and Culmofrons. A, holotype specimen of
Beothukis mistakensis, ‘E’ Surface, Mistaken Point Ecological Reserve, Newfoundland. B, specimen from the MUN Surface, Burnt Point,
Bonavista Peninsula, Newfoundland. C, holotype specimen of Culmofrons plumosa from the Lower Mistaken Point surface (cf. Clap-
ham and Narbonne 2002), Mistaken Point Ecological Reserve, Newfoundland. D, schematic diagram of a Beothukis frond, annotated
to show the morphological features discussed herein. Images have not been retrodeformed, and all fossil specimens remain in the field.
Rangeomorph branches preserved as negative epirelief impressions. All scale bars represent 50 mm, except B which is 10 mm.
LIU ET AL.: EDIACARAN MACROFOSSIL TAXONOMY 51
basal disc is considered a key character in the Brasier
et al. framework, and this line of reasoning could be
extended to the presence or absence of a stem. The origi-
nal diagnosis and description of Beothukis (Brasier and
Antcliffe 2009) do not mention a stem or a holdfast disc,
but the diagnosis was emended by Brasier et al. (2012) to
state that a ‘stem is sometimes preserved’ (2012, p. 1114).
Importantly, the emended specific diagnosis of B. mistak-
ensis by Narbonne et al. (2009) noted that although a
stem is typically not present, it can be observed, and is
<5% of the length of the frond when present. It therefore
appears that, although a stem is not frequently seen in
Beothukis mistakensis (perhaps due to taphonomic rea-
sons), it is present in some specimens. Specimens previ-
ously assigned to the genus Culmofrons (cf. Laflamme
et al. 2012) possess a stem that comprises 2942% of the
length of the organism (values calculated using data pre-
sented in Laflamme et al. 2012, table 1), demonstrating
that considerable variability in stem length is observed
within what has previously been considered to be a single
taxon. Given these discussions, it appears that Culmofrons
and Beothukis specimens as previously defined differ not
in the presence of a stem, but in its length. The presence/
absence of a stem is a key morphological attribute of
undoubted value to the organism, but we consider that
taxonomic diagnoses should not be based on the inferred
ecological function of a morphological characteristic (cf.
Laflamme et al. 2012) until we can be assured that the
character in question represents genetic (as opposed to
ontogenetic, taphonomic or ecophenotypic) variability. As
stem length in Culmofrons does show variation within
populations and is seemingly unlinked to other morpho-
logical differences, until further evidence can be presented
to the contrary, we consider it appropriate to suggest that
stem length in that taxon is likely to be an ecophenotypic
trait rather than a genetic one. The length of a stem
(relative to total length or frond length) could be
regarded as a continuous character. We therefore do not
consider stem length as a means to distinguish between
closely related taxa at the level of genus.
The number of primary branches in rangeomorphs has
previously been suggested to be a character that should
only be used with caution in their taxonomy (Brasier
et al. 2012; Wilby et al. 2015), as in some taxa it has
been demonstrated to vary during ontogeny (e.g. Charnia
in Antcliffe and Brasier 2007; Liu et al. 2012). The
number of branches is arguably a continuous character,
and therefore, we argue that it should only be used as a
species- or population-level trait. The suggestion that
some taxa have an upper limit on the number of
TABLE 1. Summary of the differences in characters used to diagnose Beothukis Brasier et al., 2012, and Culmofrons plumosa
Laflamme et al., 2012.
Character Beothukis Culmofrons Character type
Frond shape Not specified Spatulate to ovate Morphometric
Polarity Unipolar (Unipolar)
Number of rows 2 (2)
Number of first-order branches per row (Variable) 5 Continuous, potentially
ontogenetic
Discrete upper limit to
number of first-order branches per row
Seemingly no Yes Discrete
Central axis Furled, linear suture Zigzagging Potentially taphonomic
Number of second-order branches per first order Not specified 812 Continuous
Branch inflation Moderate to medial Not specified
First- and second-order branch furling Furled Not specified
First-order branch display Undisplayed (Undisplayed)
Second-order branch display Displayed (Displayed)
Alignment of first-order branches Radiating to subparallel Not specified
Alignment of second-order branches Not specified
(subparallel to radiating)
Subparallel
(to radiating)
Presence of basal disc Sometimes preserved* Circular, unornamented Discrete
Presence of stem Sometimes preserved* Preserved Discrete
Length of stem <5% frond length Long, cylindrical Continuous
Characters in parentheses are added by us to assist comparison, and have been interpreted from the original diagnoses and descrip-
tions. For those characters that are specified to differ between the two taxa (in italics), we state whether they are discrete or continu-
ous, as discussed in the text.
*Note that in the original description of Beothukis (mistakensis, Brasier and Antcliffe 2009), neither a stem nor a holdfast disc was
mentioned, though a holdfast was later recognized by Narbonne et al. (2009).
52 PALAEONTOLOGY, VOLUME 59
primary branches they possess (e.g. Culmofrons), while
others seemingly appear to add branches indefinitely (e.g.
Charnia, Liu et al. 2012; Wilby et al. 2015), may indicate
a substantially different growth programme that arguably
transcends species-level distinction. Capping the number
of primary branches produces a growth plan in Cul-
mofrons (Fig. 4C) that (in mature specimens) is entirely
dependent on inflation of primary branches and addition
only of higher-order (e.g. secondary) branches for growth.
In contrast, B. mistakensis can have as many as ten primary
branches per row (Laflamme et al. 2012). Both of these
growth plans are compatible with an overall indeterminate
mode of growth whereby the organisms continue to grow
indefinitely (as seen in other rangeomorphs such as Char-
nia; Wilby et al. 2015). Current data suggest that Cul-
mofrons had a finite limit on the number of primary
branches, whereas B. mistakensis did not. This strongly
supports distinction of these taxa at a higher taxonomic
level than the species level. However, we note that few large
specimens of B. mistakensis have been described, which
importantly means we cannot yet be sure that B. mistaken-
sis did not also cap its branch addition, but at a later stage
in its developmental programme. As it is currently not pos-
sible to refute this possibility, we suggest that it would be
unwise to separate these taxa on the basis of assumed
differences in growth strategy until further evidence is
available.
In summary, our assessment of figured material of both
Culmofrons and Beothukis specimens, and material from
the MUN Surface, leads us to conclude that differences
between these taxa in the linearity of the frond midline
and the length of the stem are either continuous vari-
ables, or subject to taphonomic influence. Although we
also consider the total number of primary branches to be
a continuous variable, we recognize that the apparent
presence of a discrete cap to the number of primary
branches in Culmofrons may reflect a significantly differ-
ent growth programme to that seen in Beothukis, where
branches appear to be added continuously throughout
growth. If this is demonstrated to be the case in the
future, we would consider such a difference in growth
programme to be a character of taxonomic significance
above the species level. However, given the paucity of
large (>25 cm) specimens of B. mistakensis, we cannot
currently refute the possibility that B. mistakensis also
limits branch addition at a later stage in its growth cycle.
We therefore suggest that these organisms can currently
be shown to differ only in characters we consider to
reflect variation at a species level. We recommend that
these taxa are grouped within the same genus, and as
Beothukis has taxonomic priority, we suggest inclusion of
Culmofrons plumosa within the genus Beothukis to create
Beothukis plumosa comb. nov. The relevant specimens we
describe from the MUN Surface (e.g. Figs 2A, 4B; Brasier
et al. 2012, fig. 8B; Liu et al. 2015b, appendix S1, fig. 6)
are hereby assigned to the newly described B. plumosa
comb. nov. The emended diagnosis of Brasier et al.
(2012) for the genus Beothukis requires minor changes to
the discussion of stems (see below). We also provide
emended diagnoses for the species B. mistakensis and
B. plumosa comb. nov. to consider their branch architec-
ture, and the importance of the morphological differences
between them.
SYSTEMATIC PALAEONTOLOGY
Genus BEOTHUKIS Brasier and Antcliffe, 2009
Type species. Beothukis mistakensis Brasier and Antcliffe, 2009,
from the late Ediacaran of the Mistaken Point Ecological
Reserve, Newfoundland, Canada.
Emended diagnosis. Frond unipolar, comprising two rows
of primary branches arranged in irregularly spaced alter-
nations along a furled central axis. Inflation of first- and
second-order branches is moderate to medial. Mature
first- and second-order branches typically have furled
margins, with alignments that are arranged in radiating to
sub-parallel series. Rangeomorph elements of the first-
order branches are undisplayed, whereas those of second-
order branches are clearly displayed. A basal disc and
stem can be present.
Beothukis mistakensis Brasier and Antcliffe, 2009
Figure 4A
1991 Rangea sp.; Gehling, pl. 3.1.
1992 ‘Flat recliner’; Seilacher, pp. 608609, fig. 1 (partim,
bottom row, third from right), fig. 2 (partim, upper
left).
1992 ‘Folding over’; Seilacher, p. 609, fig. 3 (partim, top
right).
1999 ‘other form’; Seilacher, p. 98, fig. 3 (partim, lower
right of fossil block sketches).
2001 ‘Small tree-like form’ and ‘Charnia composite
morph’; Narbonne, Dalrymple and Gehling, p. 26,
pl. 1E and H (partim).
non 2003 ‘small, unnamed frond-shaped fossil’; Wood et al.,
p. 1383, fig. 9.
2004 Unnamed frond; Laflamme et al., p. 830, fig. 3.1
(partim).
2004 ‘short-stemmed rangeomorph frond’; Narbonne, p.
1143, fig. 3BC.
2004 ‘Bush-like form’; O’Brien and King, pp. 207210,
fig. 3f, pl. 5a.
2005 ‘Spatulate rangid’ and ‘short stem rangid’; Nar-
bonne et al., p. 28, pl. 1K and 1N.
LIU ET AL.: EDIACARAN MACROFOSSIL TAXONOMY 53
2007 ‘Rangeomorph fronds’; Ichaso et al., p. 28, fig.
3CD.
2008 ‘Charnia antecedens’; Hofmann et al., p. 17, fig.
13.7 (pars), (non figs 13.813.10, 15.115.5).
2008a‘Rangeomorph frond’; Laflamme and Narbonne, p.
184, fig. 2.5.
2008b‘Spatulate rangeomorph’; Laflamme and Narbonne,
p. 170, figs 4.4, 4.6, 4.7.
2009 Beothukis mistakensis; Brasier and Antcliffe, pp.
382383, figs 17ab, 18ab.
2009 Beothukis mistakensis; Narbonne et al., pp. 508514,
figs 3.3 (partim), 3.6 (partim), 5.15.2, 6.16.7, 7,
8.18.6.
2012 Beothukis mistakensis; Dornbos et al., p. 58, fig. 5.2c.
2012 Beothukis mistakensis; Brasier et al., p. 1116, fig.
5CD.
2013 Beothukis sp.; Brasier et al., p. 130, figs 9D, 11BD.
2013 Beothukis; Darroch et al., p. 596, fig. 2B.
2013 Beothukis mistakensis; Laflamme et al., p. 562, fig.
2.12.4.
2013 Beothukis; Macdonald et al., p. 257, fig. 6C.
2014 Beothukis mistakensis; Xiao, p. R121, fig. 1b. [cop.
Narbonne et al. 2009, fig. 7].
2014 Beothukis mistakensis; Hoyal Cuthill and Conway
Morris, p. 13123, fig. 1.
2014 Beothukis; Ghisalberti et al., p. 2, fig. 1e (partim,
lower right).
2014 Beothukis cf. Beothukis mistakensis; Narbonne et al.,
p. 215, fig. 6.16.7.
2014 Beothukis; Zalasiewicz and Williams, p. 144, fig. 13.
2015a Beothukis mistakensis; Liu et al., p. 1361, fig. 2B.
2015 Beothukis; Burzynski and Narbonne, p. 37, figs 4a
(partim, upper left), 5B(b).
Emended diagnosis. Frond unipolar and spatulate to ovate
in shape, comprising two rows of five or more primary
branches (in specimens of >2 cm in length; juveniles may
have fewer) arranged alternately along a furled, broadly
linear central axis. First- and second-order branches typi-
cally exhibit furled margins, and moderate to medial
inflation. Second-order branches are arranged in a radial
to sub-parallel arrangement. A circular basal holdfast disc
and a short stem are sometimes present, but the length of
the stem is typically <5% of the length of the frond when
observed.
Beothukis plumosa comb. nov.
Figures 2A, 4BC; Liu et al. 2015b, appendix S1, fig. 6AJ, L
(partim), NP
2007 ‘Frond’; Laflamme et al., p. 249, fig. 6de.
v* 2012 Culmofrons plumosa; Laflamme et al., p. 196, figs
2.12.4, 2.7 (non figs 2.52.6).
v. 2012 Beothukis sp.’; Brasier et al., p. 1120, fig. 8b.
2014 Culmofrons; Kenchington and Wilby, p. 105, fig. 2a
[cop. Brasier et al. 2012].
2015a Culmofrons plumosa; Liu et al., p. 1361, fig. 2e.
Diagnosis. Frond unipolar and spatulate to ovate in
shape, comprising two rows of primary branches (with
fewer than five branches in each row) arranged alternately
along a furled, often zigzagging central axis. Second-order
branches (typically 812 per first-order branch, but
reducing in number distally) are arranged in a broadly
sub-parallel to radiating arrangement. A circular basal
holdfast disc and a long cylindrical stem (comprising ~30
to 40% of the length of the organism) are present. A
broad, smooth region is often present at the junction
between frond and stem.
Remarks. We note that for the specific case of distin-
guishing juvenile B. mistakensis with relatively few
branches from juvenile B. plumosa, if a clear stem is not
evident or the specimen is poorly preserved, it may be
advisable to use open nomenclature (Beothukis sp.).
EXTENSION OF AN ARCHITECTURAL
APPROACH TO WIDER EDIACARAN
TAXONOMY
The discussion above builds upon the use of branching
architecture and the presence or absence of key discrete
morphological features as generic characters in rangeo-
morphs, with morphometric or continuous variables used
to discriminate between species. Extension of a similar
approach to other Ediacaran taxa would be valuable, as
would a search for further potential synapomorphies with
which to group possible higher-level clades. The most
comparable group to consider is the arboreomorph/fron-
domorph clade (cf. Laflamme et al. 2013; Grazhdankin
2014), many members of which bear superficial morpho-
logical similarities to the rangeomorphs. A lack of rangeo-
morph elements in arboreomorphs precludes use of the
exact terms used by Brasier et al. (2012), but detailed
study of their architecture may yet reveal comparable
variability in branch construction (cf. Laflamme and
Narbonne 2008b). The use of branch architecture and
presence/absence of stems and holdfasts (i.e. discrete
characters) as genus-level characters in the Arboreomor-
pha, with morphometric differences (i.e. continuous char-
acters) being species-level discriminating characters,
should be straightforward. This approach is broadly con-
sistent with existing Charniodiscus taxonomy based on
material from Newfoundland (Laflamme et al. 2004;
Laflamme and Narbonne 2008b).
54 PALAEONTOLOGY, VOLUME 59
Extension of an architectural approach to non-frondose
taxa requires consideration of alternative independent
characters that can undergo morphometric or continuous
variation. Symmetry has been suggested as a basis for
higher-order taxonomy (Fedonkin 1985; Erwin et al.
2011; Laflamme et al. 2013), but more subtle characters
are required for fine-tuning at the generic and species
levels. The concepts of polarity, rows and inflation (cf.
Brasier et al. 2012) could equally be applied to the
‘pneus’ or ‘segments’ of taxa such as Dickinsonia or Pteri-
dinium, but other groups such as tubular body fossils
(e.g. Wutubus; Chen et al. 2014) may require inspection
of additional features, such as branching style, consistency
of width or segment cross-sectional profile.
Differences in growth programme between taxa solely
growing by inflation and those growing or by both addi-
tion and inflation of branches or segments can potentially
be of use in distinguishing taxa at the generic or higher
levels. However, such distinction should only be accepted
if the perceived difference in growth programme can be
clearly demonstrated, and is not simply a later change in
a growth plan common to both taxa (e.g. the limiting of
primary branches at eight branches rather than four).
Morphological characters used for taxonomic purposes
must be clearly independent of ecological, taphonomic
and ontogenetic variation.
The grouping of Ediacaran macrofossils into higher
clades, a topic that has long attracted discussion
(Fedonkin 1985), also requires consideration. Two
recent studies have resulted in broadly similar schemes
that jointly recognize some groups (e.g. the Rangeomor-
pha), but differ in suggesting distinct names and charac-
teristic features for some groups (e.g. Arboreomorpha vs
Frondomorpha), or including different taxa within simi-
lar overall groups (e.g. the Dickinsoniomorpha;
Laflamme et al. 2013; Grazhdankin 2014). Both studies
take a phenetic approach to classification, with
Laflamme et al. (2013) explicitly using unique synapo-
morphies to recognize clades whenever possible. How-
ever, both studies also propose groupings diagnosed by
non-unique characters or character combinations. For
example, the Frondomorpha of Grazhdankin (2014, pp.
271272) are described as taxa ‘composed of a large,
relatively flattened foliate section, a central stem and a
holdfast or rooting anchor’; a description that could be
used to describe several rangeomorphs (e.g. Trepassia or
Beothukis). Similarly, the suggestion that members of
the Arboreomorpha possess primary branches that ‘end
at an outer margin’ (Laflamme et al. 2013, following
Erwin et al. 2011, supplementary information) would
lead to the inclusion of rangeomorph specimens such as
Beothukis plumosa from the MUN Surface (Fig. 4B)
within that group. We should perhaps not expect all
authors to converge on the same higher-order groupings
for Ediacaran taxa, and the debate promoted by the dif-
ferences their schemes reveal is welcomed. However, the
characters by which those groups are distinguished
should, where possible, be chosen such that they are
unique to the proposed clade, potentially paving the
way for future application of phylogenetic approaches to
these fossils. Adoption of a ‘bottom-up’ approach to
Ediacaran macrofossil taxonomy (first reaching agree-
ment on methods to distinguish genera and species
before progressing to higher levels) seems an appropri-
ate course of action.
CONCLUSIONS
It is hoped that refinement of Ediacaran morphological
and taxonomic studies will lead to wider appreciation of
the similarities and differences between Ediacaran macro-
fossil taxa, although we must remember that ‘classifica-
tion can only reflect existing knowledge and must be
open to modification in the light of further discovery’
(Calman 1949, p. 21). The newly discovered high-defini-
tion fossil assemblage of the MUN Surface reveals impor-
tant morphological and constructional data that guide
our interpretation of the biology and palaeoecology of
rangeomorphs and associated organisms. Our observa-
tions of specimens from the MUN Surface conform to
the existing descriptive taxonomic scheme for rangeo-
morphs (Brasier et al. 2012), allowing resolution of taxo-
nomic problems within the Rangeomorpha, namely the
Beothukis/Culmofrons dilemma. We suggest that the archi-
tectural approach used herein, whereby species are distin-
guished on the basis of continuous characters whereas
genera differ in gross architecture or developmental
programme, can be extended with modification to aid
taxonomic understanding of the Arboreomorpha/Frondo-
morpha. We also note that certain variables related to size
and shape of organisms could be subject to the influences
of taphonomy, ontogeny or ecology, and should therefore
only be used in taxonomic diagnoses with caution.
Determining the extent of ecophenotypic variation in
Ediacaran fossil assemblages would be a worthwhile
avenue for future research. Other Ediacaran macrofossil
taxa could similarly benefit from this consistent approach
to taxonomy, allowing for the fact that higher-level classi-
fication of Ediacaran macrofossils is itself in a current
state of flux. Consideration of the points outlined herein,
and critical appraisal of existing taxonomic schemes,
will lead to an increasingly robust framework within
which to reconstruct systematic relationships amongst the
Ediacaran macrobiota.
Acknowledgements. Aspects of this research were supported by
the Natural Environment Research Council (grant numbers NE/
LIU ET AL.: EDIACARAN MACROFOSSIL TAXONOMY 55
J5000045/1 to JJM, and NE/L011409/1 to AGL); a Henslow
Junior Research Fellowship from the Cambridge Philosophical
Society to AGL; the National Geographic Global Exploration
Fund (grant number GEFNE22-11 to AGL); and the support of
NSERC (to DM). The field assistance of J. Stewart is gratefully
acknowledged. This manuscript has benefitted from useful
discussions with M. Brasier, L. Parry and C. Kenchington, and
reviews by D. C. Garc
ıa-Bellido, M. Laflamme and two other
anonymous referees. Readers are advised that access to the
MUN Surface for palaeontological research requires a permit
from the Government of Newfoundland and Labrador, in accor-
dance with Regulation 67/11 (Palaeontological Resource Regula-
tions) of the Historic Resources Act 2011.
Author contributions. AGL and JJM jointly discovered the site
and conducted the field and laboratory studies. All authors were
involved in data analysis and writing the manuscript.
DATA ARCHIVING STATEMENT
Data for this study, including information on the sedimentology and
stratigraphic setting of the MUN Surface, and six additional figures,
are available in the Dryad Digital Repository: http://dx.doi.org/
10.5061/dryad.6r4j8
Editor. Javier
Alvaro
REFERENCES
AN T C LI F F E, J. B. and BR A S I E R , M. D. 2007. Charnia and
sea pens are poles apart. Journal of the Geological Society,164,
4951.
B A M F O RT H , E. L. and NA R B O N N E , G. M. 2009. New
Ediacaran rangeomorphs from Mistaken Point, Newfound-
land, Canada. Journal of Paleontology,83, 897913.
doi:10.1666/09-047.1
BE N U S, A. P. 1988. Sedimentological context of a deep-water
Ediacaran fauna (Mistaken Point, Avalon Zone, eastern New-
foundland). 89. In LANDING, E., NARBONNE, G. M.
and MY R O W, P. M. (eds). Trace fossils, Small Shelly Fossils
and the PrecambrianCambrian boundary. New York State
Museum and Geological Survey Bulletin, 43.
BO W R IN G , S. A., MY R OW , P. M., L A N D I N G , E. and
RA M E NZ A N I , J. 2003. Geochronological constraints on
terminal Neoproterozoic events and the rise of metazoans.
NASA Astrobiology Institute (NAI General Meeting) Conference
Abstracts, 113114.
BO Y N TO N , H. E. and F O R D , T. D. 1995. Ediacaran fossils
from the Precambrian (Charnian Supergroup) of Charnwood
Forest, Leicestershire, England. Mercian Geologist,13, 165182.
BR A S IE R , M. D. and AN T CL I F F E , J. B. 2009. Evolutionary
relationships within the Avalonian Ediacara biota: new insights
from laser analysis. Journal of the Geological Society, London,
166, 363384. doi:10.1144/0016-76492008-011
--
and LI U A. G. 2012. The architecture of Ediacaran
fronds. Palaeontology,55, 11051124. doi:10.1111/j.1475-
4983.2012.01164.x
-
LI U , A. G., M E N O N , L. R., M A T T H E W S , J. J.,
Mc I L RO Y , D. and W A C E Y , D. 2013. Explaining the
exceptional preservation of Ediacaran rangeomorphs from
Spaniard’s Bay, Newfoundland: a hydraulic model. Precam-
brian Research,231, 122135. doi:10.1016/j.precamres.2013.
03.013
B U R Z YN S K I , G. and N A R B O N N E , G. M. 2015. The discs
of Avalon: relating discoid fossils to frondose organisms in the
Ediacaran of Newfoundland, Canada. Palaeogeography, Palaeo-
climatology, Palaeoecology,434,3445. doi:10.1016/j.palaeo.
2015.01.014
CA L L OW , R. H. T. and B R A SI E R , M. D. 2009a. Remarkable
preservation of microbial mats in Neoproterozoic siliciclastic
settings: implications for Ediacaran taphonomic models. Earth
Science Reviews,96, 207219. doi:10.1016/j.earscirev.2009.07.
002
--
2009b. A solution to Darwin’s dilemma of 1859:
exceptional preservation in Salter’s material from the late
Ediacaran Longmyndian Supergroup, England. Journal of the
Geological Society, London,166,14. doi:10.1144/0016-
76492008-095
CA L M AN , W. T. 1949. The classification of animals. An intro-
duction to zoological taxonomy. Methuen, London, 54 pp.
CH E N , Z., Z HO U , C., XI AO , S., WA N G , W., GUA N , C.,
HU A , H. and Y UA N , X. 2014. New Ediacara fossils pre-
served in marine limestone and their ecological implications.
Scientific Reports,4, 4180. doi:10.1038/srep04180
CL A P HA M , M. E. and N A R B O N N E , G. M. 2002. Ediacaran
epifaunal tiering. Geology,30, 627630. doi:10.1130/0091-7613
(2002)030<0627:EET>2.0.CO;2
--
and G E H L I N G J. G. 2003. Paleoecology of the old-
est known animal communities: Ediacaran assemblages at
Mistaken Point, Newfoundland. Paleobiology,29, 527544.
doi:10.1666/0094-8373(2003)029<0527:POTOKA>2.0.CO;2
---
G R E E NT R E E, C. and A N D E R S O N, M. M.
2004. Thectardis avalonensis: a New Ediacaran Fossil from the
Mistaken Point Biota, Newfoundland. Journal of Paleontology,
78, 10311036. doi:10.1666/0022-3360(2004)078<1031:TAA
NEF>2.0.CO;2
DA R R OC H , S. A., L A F L A M M E , M. and CL A P H A M , M.
E. 2013. Population structure of the oldest known macro-
scopic communities from Mistaken Point, Newfoundland.
Paleobiology,39, 591608. doi:10.1666/12051
DO R N BO S , S. Q., CL A P H A M , M. E., F R A I S E R , M. L.
and LA F L AM M E , M. 2012. Lessons from the fossil record:
the Ediacaran radiation, the Cambrian radiation, and the end-
Permian. 5272. In SO L A N, M., A S PD E N , R. and
PA T E RS O N , D. (eds). Marine biodiversity and ecosystem
functioning: frameworks, methodologies, and integration. Oxford
University Press, 254 pp.
ER W I N, D. H., L A F L A M M E , M., TW EE D T , S. M.,
SP E R LI N G , E. A., P I S AN I , D. and P ET E R SO N , K. J.
2011. The Cambrian Conundrum: early divergence and later
ecological success in the early history of animals. Science,334,
10911097. doi:10.1126/science.1206375
FE D O NK I N , M. A. 1985. Precambrian metazoans: the prob-
lems of preservation, systematics and evolution. Philosophical
Transactions of the Royal Society of London, Series B,311,27
44. doi:10.1098/rstb.1985.0136
56 PALAEONTOLOGY, VOLUME 59
-
GE H L IN G , J. G., G R E Y, K., NA R B O N N E , G. M. and
VI C K ER S - RI C H , P. 2007. The rise of animals: evolution
and diversification of the Kingdom Animalia. John Hopkins
University Press, Baltimore, 326 pp.
FO R D , T. D. 1958. Pre-Cambrian fossils from Charnwood For-
est. Proceedings of the Yorkshire Geological Society,31, 211217.
doi:10.1144/pygs.31.3.211
GE H L IN G , J. G. 1991. The case for Ediacaran fossil roots to
the metazoan tree. Memoir of the Geological Society of India,
20, 181224.
-
and DR O S ER , M. L. 2013. How well do fossil assem-
blages of the Ediacara Biota tell time? Geology,41, 447450.
doi:10.1130/G33881.1
-
and NA R B ON N E , G. M. 2007. Spindle-shaped Ediacara
fossils from the Mistaken Point assemblage, Avalon Zone,
Newfoundland. Canadian Journal of Earth Sciences,44, 367
387. doi:10.1139/e07-003
--
and AN D E R S O N M. M. 2000. The first named
Ediacaran body fossil, Aspidella terranovica.Palaeontology,43,
427456. doi:10.1111/j.0031-0239.2000.00134.x
GH I S AL B E R T I , M., GO LD , D. A., L A F L A M M E , M.,
CL A P HA M , M. E., N A R B O N N E , G. M., S U M M O N S ,
R. E., J O H N S T O N , D. T. and J A C O B S, D. K. 2014.
Canopy flow analysis reveals the advantage of size in the old-
est communities of multicellular eukaryotes. Current Biology,
24, 305309. doi:10.1016/j.cub.2013.12.017
GR A Z H D A N KI N , D. 2014. Patterns of evolution of the Edi-
acaran soft-bodied biota. Journal of Paleontology,88, 269283.
doi:10.1666/13-072
HO F M AN N , H. J., O B R I E N , S. J. and KI N G , A. F. 2008.
Ediacaran biota on Bonavista Peninsula, Newfoundland,
Canada. Journal of Paleontology,82,136. doi:10.1666/06-087.1
H O Y A L C U T H IL L , J. F. and C O N W A Y M O R R I S , S.
2014. Fractal branching organizations of Ediacaran rangeo-
morph fronds reveal a lost Proterozoic body plan. Proceedings
of the National Academy of Sciences,111, 1312213126.
doi:10.1073/pnas.1408542111
I C H A S O, A. A., D A L R Y M P L E , R. W. and N A R B O N N E ,
G. M. 2007. Paleoenvironmental and basin analysis of the
late Neoproterozoic (Ediacaran) upper Conception and St.
John’s groups, west Conception Bay, Newfoundland.
Canadian Journal of Earth Sciences,44,2541. doi:10.1139/
e06-098
KE N C HI N G TO N , C. G. and W IL B Y , P. R. 2014. Of time
and taphonomy: preservation in the Ediacaran. 101122. In
LA F L AM M E , M., S C H I F F B A U E R , J. D. and DAR -
R O C H , S. A. F. (eds). Reading and writing of the fossil record:
preservational pathways to exceptional preservation. The Paleon-
tological Society Papers, 20.
LA F L AM M E , M. and N AR B O NN E , G. M. 2008a. Competi-
tion in a Precambrian world: palaeoecology of Ediacaran
fronds. Geology Today,24, 182187. doi:10.1111/j.1365-2451.
2008.00685.x
--
2008b. Ediacaran fronds. Palaeogeography, Palaeocli-
matology, Palaeoecology,258, 162179. doi:10.1016/j.palaeo.
2007.05.020
--
and AN D E RS O N M. M. 2004. Morphometric anal-
ysis of the Ediacaran frond Charniodiscus from the Mistaken
Point Formation, Newfoundland. Journal of Paleontology,78,
827837. doi:10.1666/0022-3360(2004)078<0827:MAOTEF>2.0.
CO;2
--
GR E E NT R E E, C. and A N D E R S O N , M. M. 2007.
Morphology and taphonomy of an Ediacaran frond: Charnia
from the Avalon Peninsula of Newfoundland. 237257. In
VI C K ER S - RI C H , P. and K O M A R O W E R , P. (eds). The
rise and fall of the Ediacaran biota. Geological Society, London,
Special Publication, 286. doi:10.1144/SP286.17
-
SC H I FF B A UE R , J. D., N AR B O NN E , G. M. and
B R I G G S, D. E. G. 2011. Microbial biofilms and the preser-
vation of the Ediacara biota. Lethaia,44, 203213.
doi:10.1111/j.1502-3931.2010.00235.x
-
FL U D E, L. I. and N A R BO N N E, G. M. 2012. Ecological
tiering and the evolution of a stem: the oldest stemmed frond
from the Ediacaran of Newfoundland, Canada. Journal of Pale-
ontology,86, 193200. doi:10.1666/11-044.1
-
D A R R O CH , S. A. F., T W E E D T , S. M., P E T E R S O N,
K. J. and E R W I N , D. H. 2013. The end of the Ediacara
biota: extinction, biotic replacement, or Cheshire Cat? Gond-
wana Research,23, 558573. doi:10.1016/j.gr.2012.11.004
LI U , A. G., M CI LR O Y , D., A N T C L I FF E , J. B. and B R A -
SI E R , M. D. 2011. Effaced preservation in the Ediacaran
biota of Avalonia and its implications for the early macrofossil
record. Palaeontology,54, 607630. doi:10.1111/j.1475-4983.
2010.01024.x
--
MA T T HE W S , J. J. and B R A S I E R , M. D. 2012. A
new assemblage of juvenile Ediacaran fronds from the Drook
Formation, Newfoundland. Journal of the Geological Society,
169, 395403. doi:10.1144/0016-76492011-094
-
MA T T HE W S , J. J., M E N ON , L. R., M c IL R O Y , D.
and BR A S IE R , M. D. 2014a.Haootia quadriformis n. gen.,
n. sp., interpreted as a muscular cnidarian impression from
the Late Ediacaran period (approx. 560 Ma). Proceedings of
the Royal Society B,281, 20141202. doi:10.1098/rspb.
2014.1202
-
Mc I L RO Y , D., MA T T H E W S , J. J. and B R A S I E R , M.
D. 2014b. Confirming the metazoan character of a 565 Ma
trace-fossil assemblage from Mistaken Point, Newfoundland.
Palaios,29, 420430. doi:10.2110/palo.2014.011
-
K E N C H IN G T O N , C. G. and M I T C H E L L , E. G.
2015a. Remarkable insights into the paleoecology of the Aval-
onian Ediacaran macrobiota. Gondwana Research,27, 1355
1380. doi:10.1016/j.gr.2014.11.002
-
MA T T HE W S , J. J. and Mc I L RO Y , D. 2015b. Data
from: The Beothukis/Culmofrons problem and its bearing on
Ediacaran macrofossil taxonomy: evidence from an excep-
tional new fossil locality. Dryad Data Repository. doi:10.5061/
dryad.6r4j8
MA C D ON A L D , F. A., ST RA U S S, J. V., SP ER L I NG , E.
A., H A L V E R S O N , G. P., N A RB O N NE , G. M., J O H N -
ST O N , D. T., K U N Z M A N N , M., S C H RA G , D. P. and
HI G G IN S , J. A. 2013. The stratigraphic relationship between
the Shuram carbon isotope excursion, the oxygenation of
Neoproterozoic oceans, and the first appearance of the Edi-
acara biota and bilaterian trace fossils in northwestern Canada.
Chemical Geology,362, 250272. doi:10.1016/j.chemgeo.2013.
05.032
LIU ET AL.: EDIACARAN MACROFOSSIL TAXONOMY 57
MA S O N, S. J., N A R B O N N E , G. M., D A L R Y M P L E , R. W.
and O ’ BR IE N, S. J. 2013. Paleoenvironmental analysis of
Ediacaran strata in the Catalina Dome, Bonavista Peninsula,
Newfoundland. Canadian Journal of Earth Sciences,50, 197
212. doi:10.1139/cjes-2012-0099
MA Y R , E. 1963. Animal species and evolution. Belknap Press
(Harvard University Press), 797 pp.
ME L V IL L E , R. V. 1995. Towards stability in the names of ani-
mals: a history of the International Commission on Zoological
Nomenclature, 18951995. International Trust for Zoological
Nomenclature, 104 pp. doi:10.1017/S0007485300039432
N A R B O NN E , G. M. 2004. Modular construction in the Edi-
acaran biota. Science,305, 11411144. doi:10.1126/
science.1099727
-
DA L R YM P L E , R. W., G E H L I N G , J. G., W O O D , D.
A., C L A PH A M , M. E. and SA L A , R. A. 2001. Neoprotero-
zoic fossils and environments of the Avalon Peninsula, New-
foundland. Field Trip B5, Geological Association of Canada
Mineralogical Association of Canada Joint Annual Meeting, St
John’s, Newfoundland, 98 pp.
--
LA F L AM M E , M., G E H L I N G , J. G. and
BO Y C E, W. D. 2005. Life after snowball: mistaken Point
Biota and the Cambrian of the Avalon. In North American
Paleontological Convention Field Trip Guidebook, 98 pp.
-
LA F L AM M E , M., G R EE N T RE E , C. and TR U S L E R ,
P. 2009. Reconstructing a lost world: Ediacaran rangeomorphs
from Spaniard’s Bay, Newfoundland. Journal of Paleontology,
83, 503523. doi:10.1666/08-072R1.1
--
TR U S LE R , P. W., DAL R Y MP L E , R. W. and
GR E E NT R E E, C. 2014. Deep-water Ediacaran fossils from
northwestern Canada: taphonomy, ecology, and evolution.
Journal of Paleontology,88, 207223. doi:10.1666/13-053
O B RI E N , S. J. and K I N G , A. F. 2004. Ediacaran fossils from
the Bonavista Peninsula (Avalon zone), Newfoundland: pre-
liminary descriptions and implications for regional correlation.
Current Research, Newfoundland Department of Mines and
Energy Geological Survey,041, 203212.
--
2005. Late Neoproterozoic (Ediacaran) stratigraphy of
the Avalon zone sedimentary rocks, Bonavista Peninsula, New-
foundland. Current Research, Newfoundland and Labrador
Department of Natural Resources Geological Survey,051, 101
113.
SC H O PF , J. W. and K LE I N , C. 1992. The Proterozoic
biosphere. Cambridge University Press, 1374 pp.
SE I L AC H E R, A. 1992. Vendobionta and Psammocorallia: lost
constructions of Precambrian evolution. Journal of the Geologi-
cal Society, London,149, 607613. doi:10.1144/gsjgs.149.4.0607
-
1999. Biomat-related lifestyles in the Precambrian. Palaios,
14,8693. doi:10.2307/3515363
VA N K R AN E N DO N K , M. J., G E H L I N G , J. G. and
SH I E LD S , G. A. 2008. Precambrian. 2336. In O G G , J. G.,
O G G , G. and G R A D S T EI N , F. M. (eds). The concise geo-
logic time scale. Cambridge University Press, 228 pp.
WI L B Y, P. R., C A R N E Y , J. N. and HO W E , M. P. A. 2011.
A rich Ediacaran assemblage from eastern Avalonia: evidence
of early widespread diversity in the deep ocean. Geology,39,
655658. doi:10.1130/G31890.1
-
KE N C HI N G TO N , C. G. and WI L B Y , R. L. 2015. Role
of low intensity environmental disturbance in structuring the
earliest (Ediacaran) macrobenthic tiered communities. Palaeo-
geography, Palaeoclimatology, Palaeoecology,434,1427.
doi:10.1016/j.palaeo.2015.03.033
WO O D , D. A., DA L R Y M P L E , R. W., N A R B O N N E , G.
M., G E H L IN G , J. G. and C L A P H A M , M. E. 2003. Paleo-
environmental analysis of the late Neoproterozoic Mistaken
Point and Trepassey formations, southeastern Newfoundland.
Canadian Journal of Earth Sciences,40, 13751391.
doi:10.1139/e03-048
XI A O , S. 2014. Evolution: the making of Ediacaran giants.
Current Biology,24, R120R122. doi:10.1016/j.cub.2013.12.035
ZA L A SI E W IC Z , J. and WI L L IA M S , M. 2014. Ocean worlds:
the story of seas on Earth and other planets. Oxford University
Press, 320 pp.
58 PALAEONTOLOGY, VOLUME 59
1
The Beothukis/Culmofrons problem and its bearing on Ediacaran
1
macrofossil taxonomy: evidence from an exceptional new fossil
2
locality
3
ALEXANDER G. LIU, JACK J. MATTHEWS and DUNCAN MCILROY
4
5
SUPPORTING INFORMATION APPENDIX S1
6
7
8
9
SUPPLEMENTARY TEXT 1 Sedimentology and stratigraphic setting of the MUN Surface
10
11
SUPPLEMENTARY FIGURE 1 Sedimentary log through the MUN Surface cliff section
12
SUPPLEMENTARY FIGURE 2 Image of the MUN Surface bedding plane
13
SUPPLEMENTARY FIGURE 3 Sedimentary profile through the MUN Surface
14
SUPPLEMENTARY FIGURE 4 Density of fossils on the MUN Surface
15
SUPPLEMENTARY FIGURE 5 Bradgatia rangeomorph branching
16
SUPPLEMENTARY FIGURE 6 Beothukis plumosa specimens on the MUN Surface
17
2
SUPPLEMENTARY TEXT 1: Sedimentology and stratigraphic setting of the MUN
18
Surface
19
The Catalina Dome, north-eastern Bonavista Peninsula, Newfoundland, documents a ~650
20
m-thick siliciclastic sedimentary succession that has been lithostratigraphically correlated to
21
the late Ediacaran Conception and St. John’s Groups of the Avalon Peninsula (Hofmann et al.
22
2008). The upper horizons of the marine Drook Formation form the core of the anticlinal
23
Catalina Dome (O'Brien and King 2006; Fig. 1), with strata conformably younging and
24
shallowing upwards towards the shallow-marine-to-deltaic Renews Head Formation
25
(Hofmann et al. 2008).
26
The MUN Surface lies near the base of the Port Union Member of the Trepassey
27
Formation (cf. O'Brien and King 2005; Fig. 1), within an interval dominated by medium- to
28
thick-bedded buff-grey sandstones, often containing soft-sediment deformation, rounded
29
intraclasts, and zones of carbonate cementation (Supplementary Fig. 1). Beds are interpreted
30
to represent T(a)(b)cd units of turbidites (cf. Bouma 1962), with current ripples indicating a
31
downslope palaeoflow direction broadly towards the south-east. The Port Union Member has
32
most recently been interpreted to document slope deposition in a submarine channel/channel
33
lobe environment (Mason et al. 2013; see Supplementary Figs 12).
34
The main fossil-bearing surface dips at an angle of 24° to the ESE, and covers an area
35
of ~120 m2 at low tide (Supplementary Fig. 2). It sits above an upward-fining 13 cm-thick
36
bed of grey fine-sandstone to siltstone, with a noticeable ~1 cm-thick green-grey siltstone
37
horizon directly beneath the fossil surface, interpreted as a hemipelagite (Supplementary Fig.
38
3). Fossils are preserved as casts and moulds beneath a ~6 mm-thick fine-grained tuff layer
39
(Supplementary Fig. 3), which itself is overlain by a number of 3100 cm-thick grey to grey-
40
green siltstone beds and coarse sandstones (Supplementary Figs 12). The surface suffers
41
from modern wave erosion and freeze-thaw action, and possesses substantial areas of
42
3
unweathered tuff, tectonic cleavage affecting the sediment just beneath the fossil surface, and
43
natural variation in the colour of the surface (Fig. 2C; Supplementary Fig. 2). Despite these
44
obstacles to observation, a significant proportion of the surface is conducive to
45
palaeontological study.
46
The best-preserved fossils lie adjacent to the cliff-face at the up-dip edge of the
47
surface, where they have been protected from modern weathering and wave erosion to a
48
greater extent than specimens located closer to the tidal zone. This sheltered area of the
49
bedding surface has a golden hue (e.g. Fig. 2A). In contrast, much of the bedding plane
50
down-dip is stained red, considered to result from modern weathering and oxidation of pyrite
51
inferred to have been present at the substratetuff interface following burial (Gehling 1999;
52
Mapstone and McIlroy 2006; Supplementary Figs 23). Sedimentary grains directly beneath
53
the fossils are consistently <20 µm in diameter, while grains in the tuff overlying it are more
54
heterogeneous in grain size, ranging from 5110 µm. A second faulted outcrop of the same
55
surface 50 m to the southwest of the main site remains almost entirely covered by
56
unweathered tuff, and is thus currently unavailable for palaeontological study.
57
Lithostratigraphic correlation reveals that the surface is also visible in the Little Catalina
58
section ~2km to the north. However, that outcrop is relatively small, and modern wave action
59
has effaced the fossils there almost beyond recognition.
60
Almost all unipolar frondose taxa on the surface are aligned with their distal tips
61
oriented in a north-to-northwesterly direction, suggesting that they were tethered to the
62
seafloor at the time of burial (Supplementary Fig. 4). This orientation is significantly
63
different to the southeasterly direction of downslope flow indicated by ripple marks
64
immediately above this surface, or the easterly flow direction inferred from ripples in the
65
lower portion of this Member (Mason et al. 2013), suggesting other marine currents (e.g.
66
contourites) may be responsible for the observed alignment.
67
4
SUPPLEMENTARY FIGURE 1. Log
through the cliff section containing the
MUN Surface. Lower Port Union Member,
Trepassey Formation, Bonavista
Peninsula, Newfoundland. Vertical scale
of the log is in metres.
5
1
SUPPLEMENTARY FIGURE 2. A photograph of the MUN Surface bedding plane. Note
2
the variable coloration of the surface, and particularly the red staining inferred to be caused
3
by oxidation of pyrite at the substrate-tuff interface. J. Matthews at upper left for scale.
4
6
5
6
SUPPLEMENTARY FIGURE 3. The sedimentology of the MUN Surface. A) Bedding
7
plane viewed from an oblique angle. Grey siltstones underlie the fossil-bearing surface,
8
which is mantled by a thin, brown, fine-grained tuff still present upon much of the bedding
9
plane. At the interface between substrate and tuff, a thin veneer of orange iron oxyhydroxides
10
are visible these result from the modern oxidation of pyrite. Black coloration is due to
11
modern lichen cover on exposed surfaces. B) Plane-polarised light view of a thin section
12
through the same sedimentological interval as in (A).
13
7
14
15
SUPPLEMENTARY FIGURE 4. An oblique view of the MUN Surface bedding plane,
16
showing the high densities of diverse, well preserved Ediacaran macrofossils on the surface.
17
Note the strong current alignment of the frondose fossils, with distal ends of the fronds
18
broadly pointing in the same direction (arrow indicates the broad direction of perceived
19
flow). This current direction is towards the north, and differs significantly from the direction
20
of current flow obtained from turbidite beds 1 m stratigraphically above the surface (flowing
21
to the SE). Since the turbidites are assumed to indicate the palaeoslope direction, we infer the
22
frond alignment to result from the influence of contour currents (cf. Wood et al. 2003).
23
Bedding plane is naturally lit from the top. Scale bar = 50 mm.
24
8
25
26
SUPPLEMENTARY FIGURE 5. A close-up view of a partial Bradgatia specimen from the
27
MUN Surface, preserved as a negative epirelief mould of the specimen exterior. Progressive
28
orders of self-similar rangeomorph branching (cf. Narbonne 2004) are picked out by yellow
29
(2°), purple (3°), red (4°) and green (5°) shading. Scale bar = 10 mm.
30
9
31
32
SUPPLEMENTARY FIGURE 6. Beothukis plumosa specimens on the MUN Surface,
33
revealing the variation in taphonomic fidelity and style across the surface. AJ: Specimens of
34
B. plumosa. K: Specimen similar to B. plumosa, note more regular primary branch angles
35
than in other specimens, reminiscent of Charnia. L: B. plumosa (arrowed), next to a
36
Charniodiscus specimen. M: Effaced frond, with a stem and disc similar to those seen in B.
37
plumosa. NP: B. plumosa. All stems are positive epirelief features. Scale bars = 10 mm.
38
10
REFERENCES
39
40
BOUMA, A. H. 1962. Sedimentology of some Flysch deposits: a graphic approach to facies
41
interpretation. Elsevier, Amsterdam, 168 pp.
42
GEHLING, J. G. 1999. Microbial Mats in Terminal Proterozoic Siliciclastics: Ediacaran
43
Death Masks. Palaios, 14, 40-57. doi: 10.2307/3515360
44
HOFMANN, H. J., O'BRIEN, S. J. and KING, A. F. 2008. Ediacaran biota on Bonavista
45
Peninsula, Newfoundland, Canada. Journal of Paleontology, 82, 1-36.
46
doi:10.1666/06-087.1
47
MAPSTONE, N. B. and MCILROY, D. 2006. Ediacaran fossil preservation: Taphonomy and
48
diagenesis of a discoid biota from the Amadeus Basin, central Australia. Precambrian
49
Research, 149, 126-148. doi:10.1016/j.precamres.2006.05.007
50
MASON, S. J., NARBONNE, G. M., DALRYMPLE, R. W. and O'BRIEN, S. J. 2013.
51
Paleoenvironmental analysis of Ediacaran strata in the Catalina Dome, Bonavista
52
Peninsula, Newfoundland. Canadian Journal of Earth Sciences, 50, 197-212.
53
doi:10.1139/cjes-2012-0099
54
NARBONNE, G. M. 2004. Modular construction in the Ediacaran biota. Science, 305, 1141-
55
1144. doi:10.1126/science.1099727
56
O'BRIEN, S. J. and KING, A. F. 2005. Late Neoproterozoic (Ediacaran) stratigraphy of the
57
Avalon zone sedimentary rocks, Bonavista Peninsula, Newfoundland. Current
58
Research, Newfoundland and Labrador Department of Natural Resources Geological
59
Survey, 05-1, 101-113.
60
--- 2006. Lithostratigraphic and biostratigraphic studies on the eastern Bonavista Peninsula:
61
an update. Current Research, Newfoundland and Labrador Department of Natural
62
Resources Geological Survey, 06-1, 257-263.
63
WOOD, D. A., DALRYMPLE, R. W., NARBONNE, G. M., GEHLING, J. G. and
64
CLAPHAM, M. E. 2003. Paleoenvironmental analysis of the late Neoproterozoic
65
Mistaken Point and Trepassey formations, southeastern Newfoundland. Canadian
66
Journal of Earth Sciences, 40, 1375-1391. doi:10.1139/e03-048
67
... The Rangeomorpha is a group of Ediacaran organisms that is characterized by Rangea-like fronds (Pflug 1972) arranged asymmetrically in a number of orientations across a central axis or with glide-plane symmetry (Jenkins 1985;Brasier and Antcliffe 2009). The branching units have a fractal-like architecture in which the frond-shaped elements are composed of multiple smaller, similarly shaped branching units; there may be up to five orders of self-similar units recognized within a single rangeomorph frond (Narbonne 2004;Brasier et al. 2012;Liu et al. 2016). Rangeomorphs created a variety of body designs through different arrangements of these fronds (Jenkins 1985), including but not limited to: (1) Fractofusus, a biterminal spindle-shaped solitary rangeomorph that lay flat on the seafloor Dufour and McIlroy 2017); (2) short-stemmed fronds that are inferred to have been erect in the water column, such as Charnia (Laflamme et al. 2007;Antcliffe and Brasier 2008;Dunn et al. 2018a, b) and Culmofrons (Laflamme et al. 2012;Liu et al. 2016); (3) erect multifoliate fronds such as the cabbage-like Bradgatia (Flude and Narbonne 2008), Primocandelabrum (Hofmann et al. 2008;Kenchington and Wilby 2017), and possibly Charniodiscus s.s. ...
... The branching units have a fractal-like architecture in which the frond-shaped elements are composed of multiple smaller, similarly shaped branching units; there may be up to five orders of self-similar units recognized within a single rangeomorph frond (Narbonne 2004;Brasier et al. 2012;Liu et al. 2016). Rangeomorphs created a variety of body designs through different arrangements of these fronds (Jenkins 1985), including but not limited to: (1) Fractofusus, a biterminal spindle-shaped solitary rangeomorph that lay flat on the seafloor Dufour and McIlroy 2017); (2) short-stemmed fronds that are inferred to have been erect in the water column, such as Charnia (Laflamme et al. 2007;Antcliffe and Brasier 2008;Dunn et al. 2018a, b) and Culmofrons (Laflamme et al. 2012;Liu et al. 2016); (3) erect multifoliate fronds such as the cabbage-like Bradgatia (Flude and Narbonne 2008), Primocandelabrum (Hofmann et al. 2008;Kenchington and Wilby 2017), and possibly Charniodiscus s.s. (Brasier and Antcliffe 2009); and also (4) Pectinifrons, in which the paired fronds were located in parallel rows joined by a cylindrical central rod (Bamforth et al. 2008). ...
... Fossil-bearing bedding planes of the Mistaken Point Formation do not show evidence for soft-sediment deformation or loading/injection. Most authors agree that there was a microbial matground covering the seafloors that became the fossiliferous surfaces (cf. the death-mask model of Gehling 1999;Liu et al. 2016;Fig. 5c). ...
Article
Newfoundland’s Mistaken Point is home to some of the world’s oldest known complex body fossils. Detailed observation of newly discovered specimens has led to a reconsideration of Hapsidophyllas flexibilis, which is diagnosed as being a complex epifaunal multifoliate rangeomorph with a basal stolon. This study has revealed that published material of H. flexibilis includes a cryptic epifaunal recliner that grew on, and slightly into, the sealoor in a manner similar to the common Ediacaran epifaunal organisms Fractofusus and Beothukis. The new genus and species erected herein to accommodate these epifaunal organisms is Gigarimaneta samsoni, an organism that is broadly round in outline and composed of rows of allantoid units that are further sub-divided into smaller spherocylindrical units. There is no evidence of rangeomorph branching, instead the divisions seen in Gigarimaneta are considered to represent invaginations of the lower epithelium that increased the surface-area-to-volume ratio of these organisms without the creation of true branches. The resulting ‘pneu’-like divisions may have allowed this quasi-infaunal taxon to have gained nutriment from the substrate via the culturing of sulfur-oxidizing bacteria, either as endo- or epi-symbionts, perhaps coupled with the absorption of dissolved organic material from pore/seawater.
... These fossils are usually investigated using palynological maceration or thin sections of cherts and phosphatic rocks. However, in rare circumstances, filamentous organisms can be preserved as carbonaceous compressions [10,11] or even as casts and moulds in siliciclastic rocks [12][13][14], the latter making them difficult to differentiate from trace fossils [15][16][17][18]. ...
... 567-556 Ma, England) [13,71], the deep-marine strata of the Trepassey (approx. 565-560 Ma) and Briscal Formations of Newfoundland, Canada [12,14], and the deep-marine facies of the Charnwood Forest [12,72]. Callow & Brasier [13] and Liu et al. [73] also illustrated the occurrence of large filaments in the Drook Formation, which could represent the oldest occurrence of such structures (ca 574 Ma [74]). ...
... The 'typical' filament size reported varied between 100 and 1000 µm wide and 2 and 40 cm long. Liu et al. [14] also reported filament sizes from these beds ranging from 200 to 600 µm in width. These values are not far from the dimensions of LFSOB, especially those living in bundles inside large sheaths (see the electronic supplementary material, table S1; e.g. ...
Article
Full-text available
Precambrian filamentous microfossils are common and diverse. Nevertheless, their taxonomic assignment can be difficult owing to their overall simple shapes typically lacking in diagnostic features. Here, we report in situ communities of well-preserved, large filamentous impressions from the Ediacaran Itajaí Basin (ca 563 Ma) of Brazil. The filaments are uniserial (unbranched) and can reach up to 200 µm in width and up to 44 mm in length. They occur as both densely packed or sparsely populated surfaces, and typically show a consistent orientation. Although simple in shape, their preferred orientation suggests they were tethered to the seafloor, and their overall flexibility (e.g. bent, folded and twisted) supports a biological (rather than sedimentary) affinity. Biometric comparisons with modern filamentous groups further support their biological affinity, suggesting links with either large sulfide-oxidizing bacteria (SOB) or eukaryotes. Other morphological and palaeoecological characteristics further corroborates their similarities with modern large filamentous SOB. Their widespread occurrence and association with complex Ediacaran macrobiota (e.g. frondose organisms, Palaeopascichnus) suggest that they probably played an important role in the ecological dynamics of these early benthic communities by providing firm substrates for metazoans to inhabit. It is further hypothesized that the dynamic redox condition in the latest Ediacaran, with the non-continuous rise in oxygen concentration and periods of hypoxia, may have created ideal conditions for SOB to thrive.
... The Ediacaran successions of eastern Newfoundland comprise thick volcano-sedimentary successions of the Conception and St. John's Groups (Williams and King 1979;Narbonne 2005;Liu et al. 2016; Fig. 1). The most famous Ediacaran site is the Mistaken Point Ecological Reserve (MPER), a UNESCO World Heritage Site (https://whc.unesco.org/en/list/1 ...
... Each branch may be composed of multiple orders of rangeomorph units (Brasier et al. 2012), with up to four or five orders being recognized in some taxa (Narbonne 2004;Brasier et al. 2012;Kenchington and Wilby 2017). It is considered that rangeomorph units of all scales may grow in a number of different orientations, leading to a diverse expressions of the rangeomorph branch elements, creating a range of branching architecture that is used to classify the Rangeomorpha (Jenkins 1985;Brasier and Antcliffe 2009;Narbonne et al. 2009;Brasier et al. 2012;Laflamme et al. 2012;Liu et al. 2016). Primary order rangeomorph branches are the largest units and comprise rows on either side of a central axis in most rangeomorphs. ...
... In some cases, while the preservation is sufficient to recognize gross morphology, confidence in correctly identifying the details of the branching architecture necessary for taxonomic identification requires comparatively high-quality preservation. At present, there is still no clear consensus on whether: (1) continuous characters such as shape metrics and overall morphology or (2) discrete characters, mainly branching architecture and the number of growth poles, should hold more taxonomic weight (Brasier and Antcliffe 2009;Narbonne et al. 2009;Brasier et al. 2012;Laflamme et al. 2012;Liu et al. 2016;Dunn et al. 2017). ...
Article
The Avalon assemblage of Newfoundland, Canada contains abundant fossils of enigmatic soft-bodied Ediacaran organisms, many with remarkable preservation. One of the most numerically dominant groups of organisms in the assemblage is the Rangeomorpha, a frondose clade characterized by self-similar, repeating branching architecture known worldwide from rocks of Ediacaran age. Variations in branching characters and gross morphology have historically been used to divide this group, but there has been little consistency in taxonomic approach to the Rangeomorpha, concomitantly there are conflicting opinions that have resulted in some overlapping taxonomic diagnoses. Here we investigate one such taxonomic dispute, the Beothukis/Culmofrons problem. The two genera were recently synonymized into Beothukis based on the assertion that some characters were of different taxonomic rank than others. Subsequent debate has focused on which taxonomic characters displayed by the Rangeomorpha should be used for genus- and species-level subdivision. To test the validity of using continuous versus discrete characters in rangeomorph taxonomy we use a combination of morphometrics and statistical analysis to identify natural clusters within our specimen dataset which was collected from Beothukis sensu lato including material that was, until recently, attributed to Culmofrons. The results of the cluster assignment validates the differentiation between Beothukis mistakensis and Beothukis (Culmofrons) plumosa, but cannot—in isolation—be used to determine at what taxonomic rank that distinction should be made. We demonstrate a considerable degree of variation within Beothukis and Culmofrons, which has not yet been recorded for unifoliate rangeomorph taxa.
... The late Ediacaran fossil assemblages of Newfoundland, Canada preserve abundant, large, morphologically complex macro-organisms (Anderson & Misra, 1968;Narbonne, 2011;Liu et al. 2016) known as the Avalonian Assemblage (Waggoner, 2003), predominantly found on coastal exposures of the Avalon and Bonavista peninsulas (e.g. Narbonne, 2005;Liu et al. 2015). ...
... Jenkins, 1985;Brasier & Antcliffe, 2004, 2009Narbonne, 2004;Brasier et al. 2012). Some rangeomorphs have stems and/or basal discs and a range of growth programmes in the frondose portion that have created a diverse range of morphotaxa (Narbonne, 2004;Narbonne et al. 2009;Brasier et al. 2012;Laflamme et al. 2013;Liu et al. 2016). ...
... However, there is currently a lack of consensus concerning which morphological characters should be of significance at the genus and species level (e.g. continuous versus discrete branching characters; Liu et al. 2016;Kenchington & Wilby, 2017). Our re-examination of the type material of Beothukis has the potential to inform that debate by assessing the validity of the emended diagnosis of Beothukis that was broadened and inadvertently overlapped with the diagnosis of the simultaneously introduced genus Culmofrons (Laflamme et al. 2012). ...
Article
Beothukis mistakensis from the Ediacaran System of Newfoundland, Canada demonstrates complex fractal-like morphology through the development of primary-, secondary-and tertiary-order Rangea-like units. The primary-order rangeomorph units observed in B. mistakensis are tightly juxtaposed, show no evidence of being independent of one another and are made up of chamber-like secondary-order-probably mesoglea-filled-units. The growth of these rangeomorph units demonstrates that the frond developed from the tip towards the basal region through ontogeny. The tertiary-order units of Beothukis are considered to represent surface morphology on the secondary-order units. This is in contrast to palaeobiological reconstructions of Beothukis that invoke three-dimensional fractal-like branches with independent units, which has been used to infer an osmotrophic mode of life. It is considered here that the fractal-like morphology of the lower surface of B. mistakensis was an adaptation to increase surface area to volume ratio. The quilted morphology of Beothukis proposed here is consistent with a sessile, reclining, phagocytotic and/or chemosymbiotic mode of life similar to that invoked for the reclining rangeomorph Fractofusus.
... One of the most distinctive aspects of the earliest Ediacaran soft-bodied macrobiotas is that-with few rare exceptions-they were immotile, and in many cases grew to very large sizes on matgrounds [29,47,48] (Figure 1B). Being immotile on a porous organic-rich seafloor potentially results in serious biogeochemical challenges in the form of hydrogen sulfide buildup below the body tissues [49][50][51]. ...
... The earliest examples of Ediacaran fossils include the epibenthic Rangeomorpha, some of which had fractal-like lower surfaces and lived reclined on the seafloor [52,60,61] ( Figure 1D). Some rangeomorphs actively displaced sediment during growth such that they grew slightly below the ambient sediment-water interface [62] and as such were likely adapted to exploit sedimentary biogeochemical gradients, especially the very large reclining organisms (e.g., Bradgatia [47] and Gigarimaneta [48]). Fractal-like morphologies in reclining organisms are most consistent with sedimentary nutrient exploitation via symbioses with lithoautotrophic bacteria, based around the metabolism of methane, hydrogen, and hydrogen sulfide in particular. ...
Article
Full-text available
This review asks some hard questions about what the enigmatic graphoglyptid trace fossils are, documents some of their early fossil record from the Ediacaran–Cambrian transition and explores the idea that they may not have been fossils at all. Most researchers have considered the Graphoglyptida to have had a microbial-farming mode of life similar to that proposed for the fractal Ediacaran Rangeomorpha. This begs the question “What are the Graphoglyptida if not the Rangeomorpha persevering” and if so then “What if . . . ?”. This provocative idea has at its roots some fundamental questions about how to distinguish burrows sensu-stricto from the external molds of endobenthic sediment displacive organisms.
... One of the most distinctive aspects of the earliest Ediacaran soft-bodied macrobiotas is that-with few rare exceptions-they were immotile, and in many cases grew to very large sizes on matgrounds (Narbonne & Gehling 2003;Liu et al. 2016;Taylor et al. 2021;Fig. 1b). ...
... Some rangeomorphs actively displaced sediment during growth such that they grew slightly below the ambient sediment-water interface (Droser et al. 2014) and as such were likely adapted to exploit sedimentary biogeochemical gradients, especially the very large reclining organisms (e.g. Bradgatia (see Liu et al. 2016) and Gigarimaneta (Taylor et al. 2021)). Fractal-like morphologies in reclining organisms are most consistent with sedimentary nutrient exploitation via symbioses with lithoautotrophic bacteria, based around the metabolism of methane, hydrogen, and hydrogen sulfide in particular. ...
Preprint
Full-text available
This review asks some hard questions about what the enigmatic graphoglyptid trace fossils are, documents some of their early fossil record from the Ediacaran-Cambrian transition and explores the idea that they may not have been fossils at all. Most researchers have considered the Graphoglyptida to have had a microbial-farming mode of life similar to that proposed for the fractal Ediacaran Rangeomorpha. This begs the question “What are the Graphoglyptida if not the Rangeomorpha persevering” and if so then “What if…?”. This provocative idea has at its roots some fundamental questions about how to distinguish burrows sensu-stricto from the external molds of endobenthic sediment displacive organisms.
... Thus, Trepassia has a gridded appearance, but Beothukis is like a crazy quilt. Both Beothukis and Trepassia have a narrow stalk, and often lack a rounded holdfast at one end (Brasier and Antcliffe 2009;Laflamme et al. 2012Laflamme et al. , 2018Liu et al. 2016). Trepassia has a simple two-dimensional quilting pattern and lacks the prominent holdfast and wide tapering stalk of other Ediacaran fronds such as Charniodiscus and Rangea. ...
Article
Problematic fossils are described from Late Ediacaran to Early Cambrian red sandstones of the Arumbera Sandstone, Grant Bluff, and Central Mount Stuart Formations in central Australia, within a new systematic classification of Vendobionta. Arumberia banksi has been one of the most problematic of Ediacaran fossils, at first considered a fossil and then a sedimentary or organo-sedimentary structure. Our re-examination of the type material and collection of new material reveals misconceptions about its topology: it was a recessive fossil on the bed top, protruding down from the counterpart overlying slab. The concave-up body was 3 mm thick and chambered above a diffuse lower surface, so not a sedimentary structure. Also re-evaluated is the discoid fossil Hallidaya brueri, here including “Skinnera brooksi’ as its upper surface. A new species (Noffkarkys storaaslii gen. et sp. nov.) is a multilobed frond with regular, fine, trapezoidal quilts. Three new records of Trepassia wardae, Dickinsonia costata, and Ernietta plateauensis are reported from the Arumbera and Grant Bluff Formations. Reevaluation of palaeomagnetic and biostratigraphic data suggest an hiatus of 26 million years at the Ediacaran–Cambrian boundary within the Arumbera Formation, but some of this missing time is filled by the Grant Bluff and Central Mount Stuart Formations.
Article
Full-text available
The earliest record of animal life come s from the Ediacaran of Newfoundland , including dm scale fossil organisms , most of which are inferred to have be en epibenthic immotile eumetazoans. This work introduces the palaeobiology of the major fossil groups in the Newfoundland assemblages including strange fractal-like taxa and addresses some of biogeochemical challenges such as sulfide buildup that could most easily have been overcome by symbiogenesis. Specifically, the epibenthic reclining nature of some of the Ediacaran biota with their fractal-like high surface area lower surfaces-are considered to have been well designed for gaining nutriment from chemosynthetic, sulfur-oxidizing bacteria. This view constitutes a shift away from the view that most of the biota were anomalously large osmotrophs.
Article
Ediacaran frondose taxa are among the oldest large, morphologically complex organisms in the fossil record. Liu and Dunn describe fossilized filamentous structures that appear to represent physical connections between frondose individuals, suggesting that such organisms possessed a clonal reproductive strategy.
Article
Fossils of the Ediacaran macrobiota (∼571–539 mya) record phylogenetically diverse marine palaeocommunities, including early animals, which pre-date the “Cambrian Explosion” [1, 2, 3, 4]. Benthic forms with a frondose gross morphology, assigned to the morphogroups Rangeomorpha [5] and Frondomorpha (see also Arboreomorpha) [6, 7, 8], are among the most temporally wide-ranging and environmentally tolerant members of the Ediacaran macrobiota [6] and dominated deep-marine ecosystems ∼571–560 mya [9, 10, 11]. Investigations into the morphology [12, 13, 14], palaeoecology [10, 15, 16], reproductive strategies [17, 18], feeding methods [9, 19], and morphogenesis of frondose taxa together constrain their phylogenetic position to the metazoan (for Rangeomorpha) or eumetazoan (e.g., Arborea) total groups [14, 20], but tighter constraint is currently lacking. Here, we describe fossils of abundant filamentous organic structures preserved among frond-dominated fossil assemblages in Newfoundland (Canada). The filaments constitute a prominent component of the ecosystems, and exhibit clear physical associations with at least seven frondose taxa. Individual specimens of one uniterminal rangeomorph taxon appear to be directly connected by filaments across distances of centimeters to meters. Such physical linkages are interpreted to reflect evidence for stolonic connections: a conclusion with potential implications for the phylogenetic placement and palaeoecology of frondose organisms. Consideration of extant stoloniferous organisms suggests that Ediacaran frondose taxa were likely clonal and resurrects the possibility that they may have been colonial (e.g., [21, 22]). Video Abstract Download : Download video (91MB)
Article
Full-text available
The first challenge to the traditional interpretation of the Late Proterozoic Ediacara fossils came with a paper by A. Seilacher (1984, 1989) which not only proposed that Ediacaran organisms became extinct before the Cambrian, but that they represented a previously unrecognized kingdom of structurally unique multicellular organisms: the Vendozoa. This new model is based on a number of uncontested generalizations about size, shape, lifestyle and preservation, that have persisted in the literature. Many of these assumptions are now shown to be misconceptions, as a consequence of newly discovered material in Australia, Canada and the USSR, revealing a more diverse fossil assemblage and suggesting that the organisms were dominantly benthic. The interpretation of this biota in phylogenetic terms, is vindicated by the realization of strong links between some Ediacaran and Cambrian organisms. -from Author
Article
Full-text available
Rangeomorphs were important components of Ediacaran macrobenthic ecosystems, yet their biology and ecology remain poorly constrained. They formed high-density, tiered communities that were subjected to intermittent burial events, the largest of which killed entire communities. Abundant thin event beds in the Ediacaran succession of Charnwood Forest indicate the additional, frequent impact of minor obrution events. The type surface of Charnia masoni is immediately underlain by one such lamina (a tuff) and preserves a distinctly bimodal population. It is dominated by Charnia fronds that are of smaller or comparable length to the holotype (19.4 cm), but also includes notably larger specimens (> 45 cm) that would traditionally have been assigned to C. grandis. Multiple morphological- and morphometric parameters (length, width, spacing of primary branches) demonstrate that these are indistinguishable from the holotype of C. masoni, affirming the synonymy of the two taxa. Nevertheless, these outsized individuals are distinguished by their proportionally fewer primary branches per unit length. Taphonomic evidence indicates that they were survivors of an incumbent population, the rest of which was culled by a minor ashfall. We suggest that this temporary reduction in competition from neighbors allowed the survivors to grow larger and thereby gain access to a greater proportion of the water column. As the community recovered, their large size would have continued to provide them with an advantage, divorcing them from the density-dependent competition seen in the new understory. The interlude between cohorts implies that new recruits were substrate-sensitive, presumably awaiting re-establishment of the biomat. Sub-lethal disturbance events thus played a significant role in structuring Ediacaran communities, and help explain the observed bed-by-bed variability. Taken as a whole, the growth trajectory of C. masoni resembles that of extant organisms with indeterminate growth programmes and no genetically-controlled upper size limit.
Article
The Ediacaran fauna of Charnwood Forest is reviewed and several new forms are formally named and described, including a complex colonial form Bradgatia linfordensis and three new medusoid genera and species, Ivesia lobata, Shepshedia palmata and Blackbrookia oaksi. A new medusoid species Cyclomedusa cliffi is described. The frondose fossil Charnia grandis is recorded from Charnwood Forest for the first time. Three trails are also noted. -Authors
Article
Patterns of origination, evolution, and extinction of early animal life on this planet are largely interpreted from the fossils of the Precambrian soft-bodied Ediacara Biota, spanning nearly 40 m.y. of the terminal Ediacaran period. Localities containing these fossils are loosely considered as part of either the Avalon, White Sea, or Nama Associations. These associations have been interpreted to have temporal, paleobiogeographic, preservational, and/or paleoenvironmental signifi -cance. Surprisingly, elements of all three associations occur within the Ediacara Member of the Rawnsley Quartzite of South Australia. An analysis of over 5000 specimens demonstrates that fossil distribution is strongly controlled by facies and taphonomy rather than time or biogeography and that individual taxa vary considerably in their environmental tolerance and taphonomic integrity. The recognition that these taxa represent organisms living in various distinct environments, both juxtaposed and shared, holds strong implications for our interpretation of the record of early animal life on this planet and questions the biostratigraphic utility of the three associations. Furthermore, although in situ soft-bodied preservation provides a unique perspective on composition of benthic fossil assemblages, the record should not be interpreted as a simple "snapshot". Fossil beds represent a range of preservational modifi cations varying from current winnowed census samples of benthic communities at different depths and ecological maturity, to entirely transported assemblages. Unless the appropriate environments and taphonomic conditions are present for certain taxa, the absence of a particular taxon may or may not indicate its extinction in space or time.
Article
The strange biota of Neoproterozoic sea bottoms become more understandable if we assume that otherwise soft sediments were sealed by firm and erosion-resistant biomats. This allowed 'mat encrusters' (vendobionts; trilobozoan and other sponges) to get attached to sandy bottoms, and molluscan 'mat scratchers' to scrape off an algal film, as if they were living on rocks. Minute conical 'mat stickers' (Cloudina) probably required a sticky substrate to become stabilized in upright position. Horizontal burrows are interpreted as the works of worm-like 'undermat miners.' Only the latter lifestyle appears to go back to the Mesoproterozoic; the others emerged in Vendian times and virtually disappeared when matgrounds became restricted to hostile environments in the wake of the Cambrian ecological revolution.
Article
A variety of sedimentary structures and patterns in Proterozoic siliciclastic sedimentary rocks cannot be explained by known inorganic processes. In particular, certain bed-surface textures, and domed and disrupted sand lamination, are demonstrably the mechanical products of microbially bound sediment and microbial mats. In all but the most wave and current active marine environments of the terminal Proterozoic, the absence of effective grazing and burrowing allowed mat-communities of cyanobacteria to colonize sedimentary surfaces. The resultant microbial mats inhibited sediment erosion, formed partings when buried between sand beds, and restricted vertical migration of pore fluid and gas in both exposed and subaqueous environments. Distinctive 'petee' laminations, known from modern mat-bound, tidal-flat sediments, are recorded for the first time in the rock record from the terminal Proterozoic Rawnsley Quartzite of South Australia. The preservation of external molds of soft-bodied Ediacaran organisms is interpreted as a function of the early diagenesis of a sole veneer. A form of 'death mask' resulted from bacterial precipitation of iron minerals in the sand that smothered decaying microbial mats and megascopic benthic organisms. The appearance of three-dimensional trace fossils in Early Cambrian strata signaled a behavioral revolution; the evolution of efficient grazing reduced the development of benthic mat communities in all but the most extreme environments, while bioturbation disrupted buried mats and closed a taphonomic window of preservation for soft-bodied organisms.
Article
Impressions of soft-bodied Ediacaran megafossils are common in deep-water slope deposits of the June beds at Sekwi Brook in the Mackenzie Mountains of NW Canada. Two taphonomic assemblages can be recognized. Soles of turbidite beds contain numerous impressions of simple (Aspidella) and tentaculate (Hiemalora, Eoporpita) discs. A specimen of the frond Primocandelabrum is attached to an Aspidella-like holdfast, but most holdfast discs lack any impressions of the leafy fronds to which they were attached, reflecting Fermeuse-style preservation of the basal level of the community Epifaunal fronds (Beothukis, Charnia, Charniodiscus) and benthic recliners (Fractofusus) were most commonly preserved intrastratally on horizontal parting surfaces within turbidite and contourite beds, reflecting a deep-water example of Nama-style preservation of higher levels in the community. A well-preserved specimen of Namalia significantly extends the known age and environmental range of erniettomorphs into deep-water aphotic settings. Infaunal bilaterian burrows are absent from the June beds despite favorable beds for their preservation. The June beds assemblage is broadly similar in age and environment to deep-water Avalonian assemblages in Newfoundland and England, and like them contains mainly rangeomorph and arboreomorph fossils and apparently lacks dickinsoniomorphs and other clades typical of younger and shallower Ediacaran assemblages. Fossil data presently available imply that the classically deep- and shallow-water taxa of the Ediacara biota had different evolutionary origins and histories, with sessile rangeomorphs and arboreomorphs appearing in deep-water settings approximately 580 million years ago and spreading into shallow-water settings by 555 Ma but dickinsoniomorphs and other iconic Glades restricted to shallow-water settings from their first known appearance at 555 Ma until their disappearance prior to the end of the Ediacaran.
Article
When each of the Avalon-, Ediacara-, and Nama-type fossil assemblages are tracked through geological time, there appear to be changes in species composition and diversity, almost synchronized between different sedimentary environments, allowing a subdivision of the late Ediacaran into the Redkinian, Belomorian and Kotlinian geological time Intervals. The Redkinian (580-559 Ma) is characterized by first appearance of both eumetazoan traces and macroscopic organisms (frondomorphs and vendobionts) in a form of Avalon-type communities in the inner shelf environment, whereas coeval Ediacara-type communities remained depauperate. The Belomorian (559-550 Ma) is marked by the advent of eumetazoan burrowing activity in the inner shelf, diversification of frondomorphs, migration of vendobionts from the inner shelf into higher energy environments, and appearance of tribrachiomorphs and bilateralomorphs. Ediacaran organisms formed distinctive ecological associations that coexisted in the low-energy inner shelf (Avalon-type communities), in the wave-and current-agitated shoreface (Ediacara-type communities), and in the high-energy distributary systems (Nama-type communities). The Kotlinian (550-540 Ma) witnessed an expansion of the burrowing activity into wave-and current-agitated shoreface, disappearance of vendobionts, tribrachiomorphs and bilateralomorphs in wave-and current-agitated shoreface, together with a drop in frondomorph diversity. High-energy distributary channel systems of prodeltas served as refugia for Nama-type communities that survived until the end of the Ediacaran and disappeared when the burrowing activity reached high-energy environments. This pattern is interpreted as an expression of ecosystem engineering by eumetazoans, with the Ediacaran organisms being progressively outcompeted by bilaterians.