Content uploaded by Colin Scrutton
Author content
All content in this area was uploaded by Colin Scrutton on Nov 14, 2021
Content may be subject to copyright.
PROCEEDINGS OFTHE YORKSHIRE GEOLOGICAL SOCIETY,
VOL.51,
PART 3,
PP.
177-208,
1997
The Palaeozoic corals, I: origins
and
relationships
COLIN
T.
SCRUTTON
(Presidential Address delivered
at
York,
2nd
December, 1995)
Department
of
Geological Sciences, University of Durham, South
Road,
Durham
DH1 3LE, UK
SUMMARY:
The
status, origins
and
relationships
of
the various groups
of
Palaeozoic corals
are
reviewed. Five orders
are
currently
recognized: Rugosa, Tabulata, Heterocorallia, Cothoniida
and
Kilbuchophyllida,
to
which
I add the
Tabulaconida
and
Numidia-
phyllida.
The
Rugosa
and
Tabulata
are
considered
to be
broadly monophyletic clades,
and the
Tabulata
are
confirmed
as
zoan-
tharian corals. Morphological features, particularly aspects
of
septal insertion
in
both groups,
are
discussed
as
clues
to
their likely
origins
and
relationships. They
are not
considered
to
have
had a
skeletonized common ancestor,
but
they
may
have arisen
as
sep-
arate skeletonization events from
the
same broad group
of
anemones, represented
by the
living Zoanthiniaria.
The
Rugosa
are not
considered
to be
ancestral
to the
Scleractinia.
The
latter, together with
the
Permian Numidiaphyllida,
are
considered
to
have
evolved through skeletonization events among
a
group
of
anemones derived from
the
Actiniaria/Corallimorpharia,
a
member
of
which also gave rise
to the
Kilbuchophyllida
in the
Ordovician.
The
pattern
of
septal insertion
in the
Heterocorallia
is
controversial
and
the
relationship
of
these corals
to
contemporary coral groups remains uncertain.
The
increasingly important record
of
Cam-
brian coralomorphs
is
assessed,
and
considered
to
include several genera
of
zoantharian corals. However, although similarities
are
apparent, none
is
regarded
as
directly ancestral
to the
post-Cambrian coral clades.
The
history
of
diversification
and
extinction
of
corals through
the
Palaeozoic
is
briefly reviewed.
? Class
Hydroconozoa
Class
Scyphozoa
Class Hydrozoa
Class Anthozoa
Subclass Ceriantipatharia
(serial insertion
of
unpaired
mesenteries; black coral, burrowing
anemones)
Subclass
Oetoeorallia
(8 unpaired mesenteries; sea pens,
sea fans, soft coral, precious coral)
Subclass Zoantharia
(paired mesenteries;
6
primary couples,
others inserted
at 2,4 or 6
positions;
"true"
corals and sea anemones)
Order Zoanthiniaria (?Prec/Camb-Rec)
Actiniaria/Corallimorpharia (?Prec/Camb-Rec)
(2 orders cross-cutting conventional divisions)
unclassified Cambrian corals
Order Tabulaconida (early Camb)
Order Cothonida (mid Camb)
Order Tabulata (early Ord-Perm)
Order Rugosa (mid Ord-Perm)
Order Kilbuchophyllida (mid Ord)
Order Heterocorallia (?mid/late Dev-mid Carb)
Order Numidiaphyllida (Perm)
Order Scleractinia (mid Trias-Rec)
Classification
of
the Cnidaria. Ordinal divisions
are
given
for the
Subclass Zoantharia only.
For
relationships among
the
Zoantharians, see
Figure
2 and
discussion
in
text. Based
on
various sources, principally Scrutton (1979), Hill (1981), Oliver
&
Coates (1987)
and
Scrutton
&
Clarkson (1991).
The phylum Cnidaria
is
represented
in the
geological record
principally
by the
corals. This informal grouping
is
usually
taken
to
mean
the
skeletonized members
of the
Subclass
Zoantharia, Class Anthozoa
(Fig. 1).
Other anthozoans,
and
members
of the
classes Hydrozoa
and
Scyphozoa, contain
few
skeletonized representatives
and
generally have
a
restricted
fossil record (Scrutton 1979). Despite this, cnidarians
are
thought
to be a
major component
of
late Precambrian
Edi-
acaran faunas (Glaessner
&
Wade
1966;
Glaessner
1984;
Jenkins 1992; Fedonkin 1992)
and
thus have
one of the
longest
fossil records among
the
Metazoa.
All
three classes have been
claimed
to be
represented
in the
Ediacaran, although
the
paleobiology
of
many
of
these forms
has
been disputed
(Seilacher 1989). However, these faunas throw
no
light
on the
origin
of
the skeletonized corals
and the
Zoantharia, except
to
suggest that
the
anthozoans differentiated
in the
late Precam-
brian. They will
not be
discussed further here.
© Yorkshire Geological Society 1997 0044-0604/97 $10.00
The earliest putative corals occur
in the
Early Cambrian,
from
the
Tommotian onwards.
In
recent years
the
diversity
of these records
has
rapidly increased,
in
large part
the
result
of
a
growth
of
interest
in
archaeocyathan build-ups.
Although many
of
these generally coralline forms,
the
Coralomorpha
of
Jell (1984),
are of
uncertain affinities,
among them
are
undoubted corals. Their significance
has
been clouded
by a
strong desire
to
assign them
to one or
other
of the
principal Palaeozoic coral groups, despite
the
lack
of any
evidence suggesting continuity
of
descent. With
the demise
of the
archaeocyathan build-ups
at the end of the
Early Cambrian,
the
coralomorphs
all but
disappear, with
only
a
very
few
Middle
and
Upper Cambrian records, most
of which
are in
need
of
revision.
A
significant
gap
separates
the most convincing Cambrian corals from
the
earliest
accepted representatives
of the two
main Palaeozoic coral
groups (Fig.
2).
The great increase
in
diversity among calcified skeletal
organisms
of the
early
to
mid-Ordovician included
the
first
178 C. T. SCRUTTON
Prec Palaeozoic post-Palaeozoic
Vend Camb Ord Sil Dev Carb Perm Trias post-Trias
Coralomorpha
—
--
inc.
Cothoniida,
Tabulaconida ?•« 1 Heterocora Ilia [ 50 genera
i Numidiaphyllida
Fig.
2
Generalized phylogeny
of the
Zoantharia Anthozoa. Possible alternatives
to the
relationships shown here
are
discussed
in the
text.
The
scale
in
numbers
of
genera applies
to the
coral groups only, based
on
data from Hill (1981) and Wells (1956); horizontal scale proportional
to time. Position
of
the Heterocorallia arbitrary. Minor groups
of
uncertain affinities, such
as
the Tetradiida, which may
not
belong here, are
omitted. Intrites
and
Bergaueria
are
considered
to be
actinian trace fossils
and
Mackenzia
is a
possible fossil actinian anemone. Other
anemone trace fossils are known later in the Phanerozoic but not shown here
(for
example Dolopichnus Alpert & Moore (1975)). Anemone
stocks indicated
by
dashed lines; recent work suggests that
the
Actiniaria
and
Corallimorpharia
are
complexly interrelated. Polyphyletic
origins
of the
Scleractinia shown schematically.
The
Zoanthiniaria
and
their descendants
are
designated Group 1 Zoantharians
and the
Actiniaria/ Corallimorpharia and their descendants are designated Group 2 Zoantharians
by
Oliver (1996). Modified from Oliver & Coates
(1987,fig.ll.38A)
and
Scrutton
&
Clarkson
(1991,
fig.
5).
appearances
of the
Tabulata
and the
Rugosa. Together these
groups dominated Palaeozoic coral faunas until their extinc-
tion
at the end of the
Permian
(Fig. 2). At
least three other
post-Cambrian Palaeozoic coral clades
are
known.
The
scler-
actiniamorph Kilbuchophyllida appeared
in the
mid-Caradoc
(Ordovician)
but
survived only briefly
in a
very local area
(Scrutton
&
Clarkson 1991; Scrutton
et al. in
press),
and the
short lived
but
rather more successful Heterocorallia became
established after
the
near total extinction
of
tabulate
and
rugose corals
in the
late Devonian (Schindewolf 1941; Hill
1956,1981;
but see
Tourneur
&
Herrmann
1995 for a
possible
earlier record). Another small scleractiniamorph clade,
the
Numidiaphyllida, differentiated
in the
Permian (Ezaki
1997;
Scrutton
et al. in
press).
The
third major group
of
fossil corals
is
the
post-Palaeozoic Scleractinia, which appeared
in the mid
Triassic following
the end
Permian extinctions (Wells 1956).
They have survived
to the
present
day and
include both deep
water forms
and the
major contributors
to
living reefs.
The
Scleractinia, which
in
terms
of
skeleton acquisition
was
prob-
ably polyphyletic
in
origin (Roniewicz
&
Morycowa
1993;
Veron 1995),
is not
considered
to be
directly descended from
any skeletonized Palaeozoic corals (Oliver 1980; Scrutton
&
Clarkson 1991; Oliver
1996;
Scrutton
et al. in
press). Their
origins, together with
the
Permian numidiaphyllids,
are
thought
to lie
with
a
branch
of the
same anemone stock that
gave rise
to the
Ordovician kilbuchophyllids
(Fig. 2). The
Actiniaria/Corallimorpharia plus
the
Scleractinia have been
grouped
as
suborders within
the
order Hexactiniaria (Wells
&
Hill
1956) on the
basis
of
close similarity
in the
structure
and
development
of
their soft parts;
the
Kilbuchophyllida
and
Numidiaphyllida,
on
indirect evidence, could also
be
placed
in
such
a
division, designated
as
Zoantharia Group
2 by
Oliver
(1996).
The purpose
of
this review
is to
consider
the
defining
characteristics, relationships
and
history
of the
Palaeozoic
coral groups.
It is not
intended
to
investigate
the
classifi-
cation within these groups
to any
great extent,
but
major div-
isions
of the
Rugosa
and
Tabulata
are
shown
in
Figure
3.
Although more recent work
has
added many
new
genera,
the
most comprehensive
and
balanced source
of
information
on
Palaeozoic corals
to
generic level
is
still that
by
Hill (1981).
The Rugosa, Tabulata
and
their contemporaries
are
con-
sidered first,
and the
possibilities
of any
linkages among
them,
and
between them
and
later corals
are
assessed. This
is
followed
by an
analysis
of the
Cambrian coral record
and its
significance,
and a
brief review
of the
geological history
of the
Palaeozoic corals.
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 179
O i- CO CD
CO CO 2
CO
.2 Q.
CP
_ CO C (C
CO C £= c
Q.
O
sz
O
x:
Q.
O
"5 _g5 a>
O Q. CO
o c 2
<
3 >* ~ ™ ^
CO
O Auloporina i CO
CO
o
Qu
>»
r:
Q.
Favositina
Rugosa Tabulata
Fig. 3 Ranges and generic diversity of suborders (and selected superfamilies) of Rugosa and Tabulata (modified from Scrutton 1985).
1.
RUGOSA
1.1. Scope and definition of group
The rugose corals have been relatively stable as a group since
the last century. The name Rugosa was introduced by Edwards
& Haime (1850) and takes priority over Pterocorallia (1890)
and Tetracorallia (1866), of which the latter is still favoured by
some authors. Since then the scope of the category has
expanded enormously, with a multitude of new genera and
species. However, although phylogenetic relationships within
the Rugosa often remain obscure, there are relatively few
corals whose inclusion within the order raises questions (see
below). The principal characteristics of rugose corals are listed
in Table 1. Many of these features are not unique to the
Rugosa among Palaeozoic corals, as Lafuste & Plusquellec
Table 1
Characteristic features of rugose corals.
1.
Septa inserted serially in four sectors (quadrants).
2.
Septa typically of two orders, major and minor, rarely with third
or higher order septa.
Bilateral symmetry of corallite usually apparent.
Earliest stages conical or conico-cylindrical, ?aseptate.
Epithecate, with rare exceptions.
Horizontal partitions present in most, usually differentiated into
axial tabulae (tabularium) and peripheral dissepiments
(dissepimentarium).
7.
Axial structure may be developed.
8. Solitary or modular (compound, colonial); corallites c. 4-140 mm
diameter.
9. Modules without mural pores, mural tunnels or open connecting
tubules, but some massive coralla may have dividing walls
incomplete or lacking.
10.
Skeleton calcitic (<8 mol% MgC03), fibrous and trabecular.
(1990) have shown, but differences in emphasis are significant
(compare Tables 1, 3, 4). The combination of predominantly
solitary form, normally well developed septa, horizontal ele-
ments within the corallite usually differentiated into dissepi-
ments and tabulae, tendency to form axial structures and lack
of mural pores constitutes some of the more important fea-
tures.
The protocorallite is almost exclusively conical, either
with a flattened tip (c. 1-2 mm across) or more commonly a
flattened side where it was attached to a grain, shell fragment
or hard substrate (Elias 1984; Neuman 1988), although in one
unusual group, the Diffingiina (Fedorowski 1985), the proto-
corallite appears to have had a small flat basal disc, c. 1 mm
diameter, not unlike that in a scleractinian coral. Among
Phanerozoic corals with well developed septa, the most sig-
nificant distinguishing characteristic of the Rugosa appears to
be the pattern of septal insertion (Oliver 1980), which is
usually regarded as reflecting in some way the pattern of inser-
tion of mesenteries in the polyp secreting the skeleton.
However, some tabulate corals, in which as a group septa are
normally absent to poorly developed, have been shown to
possess a pattern of septal insertion very similar to the Rugosa
(see below), and these two orders may be closely related.
The early stages of septal insertion (Fig. 4) are contained
within about a millimetre or less of the (predominantly)
conical tip of the solitary coral, or protocorallite in modular
(colonial) forms. This part of the insertion sequence has been
investigated in relatively few corals (for example Jull 1973;
references in Hill 1981; Scrutton 1983; Fedorowski 19916) and
when studied by serial sections is very dependent for its accu-
racy on the precise orientation of the sections parallel to
growth increments. In some cases an aseptal stage is pre-
served, and may have been general (Fedorowski 19916), but
usually the protosepta extend to the tip of the lumen, which
may contain a plug of sclerenchyme (see Table 2 for definitions
180 C. T. SCRUTTON
,— bilateral
I symmetry
minor
septa KLmK KL major
septa
protosepta
Fig. 4 Sequence of septal insertion in rugose corals. A-D. In the
earliest stage, the tip of the cone may be filled by
sclerenchyme, or some or all protosepta may extend down to
eliminate or obliterate an early aseptal stage. Sequence of
appearance of protosepta after Carruthers (1906); Duerden
(1906) and others have noted the simultaneous appearance of
the six protosepta. E, F. In most mature rugosans, the pinnate
symmetry resulting from the insertion pattern is lost and
septal arrangement becomes more or less radial. In mature
solitary corals, the alar (pseudo)fossulae are often suppressed
and the alar septa cannot be distinguished, although the
cardinal fossula is usually visible in corals with well developed
septa. In colonial corals, all trace of insertion points, cardinal
fossula/septum and bilateral symmetry is usually lost.
Compare with Figure 5. C, cardinal septum; A, alar septum;
KL,
counter lateral septum;
K,
counter septum;
M,
major septa
(metasepta); m, minor septa (see Table 2 for definitions of
septal types).
of septal types mentioned here). In sections of several rugose
coral species described by Duerden (1906), six protosepta
were reported as appearing simultaneously. The subsequent
insertion of metasepta in four loci adjacent to the cardinal and
alar septa (Fig. 4) can be much more widely demonstrated,
both through internal serial sections and external septal
grooves on the epitheca. In addition, this pattern of septal
insertion, serial in four quadrants, can result in a characteristic
arrangement of septa in the calice (Fig. 5), although great care
must be exercised in basing assumptions concerning sequence
and locus of septal insertion on mature septal disposition
alone. In most rugosans, major septa become increasingly
uniform in length, and more perfectly radially arranged, in
later growth stages. The term metasepta was introduced for
major septa only, inserted subsequently to the protosepta
(Duerden 1902), but it has been suggested that all later septa,
major and minor, should be included (Flower 1961; Wright
1969,
Weyer 1974). This has not been universally accepted
(Oliver 1980; Hill 1981) and (for reasons similar to those of
Duerden, set out below) is not favoured here.
Both Weyer (1972,1974) and Oliver (1980) regarded septal
grooves as offering a most important guide to septal insertion
(Fig. 6), as in cross-sections or calices, septal expression may
be unclear or misleading as noted above. Furthermore, a cross-
section combines features formed at different stages of
ontogeny, depending on calical shape. In the predominantly
bowl-shaped calices of Rugosa, axial areas of cross-sections
Fig. 5 Septal arrangement in various rugose corals. A. Amplexizaphrentis enniskilleni. septa pinnately arranged, with well marked cardinal fossula
surrounding the cardinal septum and alar pseudofossulae on the counter sides of the alar septa,
X3.
B. Palaeocyclus porpita. slight residual
pinnate symmetry, particularly in the counter quadrants adjacent to the alar septa,
X3.
C. Heliophyllum halli. radial septal arrangement
with short cardinal septum in weak cardinal fossula, but alar septa obscure, XI. D. Acervularia ananas, septal arrangement more or less
perfectly radial, x2. Cardinal septum oriented downwards in solitary
corals.
Arrows indicate cardinal and alar septa in A and
B,
and cardinal
septum in C. Compare with Figure 4.
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 181
Fig. 6 A. Idealized arrangement of septal grooves on the epitheca of a solitary rugose coral. The grooves may be equally spaced at origination, or
the minor (m) and next following major (M) septal groove may be more or less tightly clustered, as shown here. Three intercalated
hyposeptal grooves (h) are shown in the counter quadrant. B. Alternate origin of major and minor septal grooves according to Kunth's
rule.
C. Major septal groove originates a minor septum on its counter side, then a major septum on it's cardinal side according to Weyer
(1974).
D-F. Phaulactis sp., Silurian; Island of Gotland, Sweden, X3. D, Cardinal and E, F, alar views of epitheca showing origination of
major and minor septal grooves, with weak hyposeptal grooves faintly visible in places. G. Stereolasma rectum, Middle Devonian,
Ludlowville Formation, Wanakah Shale; Athol Springs, Erie Co., New York, U.S.A., X2. Alar view with major, minor and hyposeptal
grooves clearly visible. For further discussion see text. Specimen figured in G kindly provided by Bill Oliver.
represent much later ontogenetic stages than peripheral areas.
In Rugosa, new septal grooves usually diverge from
pre-existing grooves, suggesting that the locus from which the
new infold in the basal ectoderm controlling septal secretion
developed was seeded from a pre-existing septal fold (Fig. 6).
In epithecate Scleractinia, septal grooves are intercalated
(Weyer 1974). Fedorowski (1991b) pointed out that septal
grooves are normally missing from the conical tip of the coral-
lite (and are completely missing from some rugosan epithecae)
and that mismatches in the appearance of grooves and septa
may occur at this level. Grooves representing the early septal
apparatus may appear more or less simultaneously at a par-
ticular growth increment, despite other evidence for a sequen-
tial insertion of protosepta. This led Fedorowski (1991b) to
conclude that septal grooves were a secondary phenomenon.
Fedorowski's observations are important and septal grooves
may not help with the very earliest ontogenetic stages. In
addition, instances are known where external longitudinal
grooves do not have a direct one-to-one relationship with
septa internally (i.e. Ditoecholasma and relatives, Sutherland
1965).
However, with due regard for occasional irregularities
and exceptions, I agree with Weyer and Oliver that septal
grooves normally provide a reliable guide to what is essentially
the metaseptal and minor septal insertion sequence.
The pattern of insertion of minor septa has been disputed
and has appeared to be more variable than that of major septa.
In recent years, discussion has centred increasingly on the
interpretation of septal grooves on the epitheca (Hill 1956,
1981;
Weyer 1972,1974; Fedorowski 1991b). Weyer (1974) pro-
posed that the youngest metaseptum in each quadrant first
originated a minor septum on its counter side and then a major
septum on its cardinal side (Fig. 6C). Fedorowski (19916)
accepted this pattern as one variant, but considered that the
historical pattern proposed by Kunth (1869), in which
metasepta and minor septa appeared alternately towards the
cardinal side in each quadrant (Fig. 6B), was the most general.
However, the distinction between these two models hinges on
whether the new septal groove, when it appears, is the proxi-
mal part of the subsequent minor septum (Kunth) or the next
following major septum (Weyer) and in practice the distinc-
tion is often not easy to make. Either interpretation represents
a version of the Cyathaxonia-type, of insertion of Hill (1935).
However, Weyer's scheme more logically explains the distinc-
tion between major and minor septa. An alternative pattern,
in which each major septum originates a major septum on the
cardinal side and then a minor septum on the same side into
the loci just defined, proposed by Vollbrecht (1928), has not
been supported by subsequent observations (Weyer 1974; Hill
1981;
Fedorowski 19916). The Zaphrentis-type of insertion
(Hill 1935), where minor septa seem suddenly to appear
altogether at a late stage of ontogeny, is probably not distinct
but reflects the suppression or enclosure of the minor septa
within the corallite wall in earlier ontogeny.
However they may be inserted, the alternation of major and
minor septa is characteristic of rugose corals, and those
without overt minor septa may have them buried in the
peripheral stereozone and/or possess minor septal grooves. As
Weyer (1995c) has remarked, no rugose coral has been con-
firmed as lacking this alternation of major and minor septa.
Rare irregularities in the development of minor septa are
known (i.e. Strusz 1968) and rarely, both major and minor
septa may bifurcate towards the peripheral wall, forming the
diplosepta of Weyer (1980). Third or higher orders of septa
have been increasingly observed in various rugose coral
182 C. T. SCRUTTON
genera from the Ordovician through to the Permian (Oliver
1980;
Weyer 1980). Hyposepta are distinguished as third-order
septal grooves, arising by intercalation immediately following
origination of the flanking major and minor septal grooves, but
which are not always expressed as projections into the calice
(Weyer 1974,1980,1995c; Fig. 6). They are the predominant
type of third order septa, although the sequence of insertion
has yet to be established in some genera. These corals pos-
sessing third and higher order septa are otherwise typical
rugosans, even if their pattern of major septal insertion is
assumed rather than demonstrated.
In virtually all cases where a study has set out to demon-
strate septal insertion in rugose corals, the characteristic
pattern or a close variant of it has usually been detected.
However, in rare instances this has not been so. For example,
Ml (1973) was unable either to identify the protosepta or to
detect the typical pattern of septal insertion in either the pro-
tocorallite or offsets of the Devonian modular rugosan Hexag-
onaria anna. However, the characteristic alternation of major
and minor septa becomes clear in the more mature stages of
corallite development in this coral despite all septa remaining
very short.
Ultimately, most corals are assigned to the Rugosa because
they are closely similar to, or appear related to, corals in
which the characteristic pattern of septal insertion has been
demonstrated, or because they possess well developed septa
and a combination of the other characters listed in Table 1.
In corals with much reduced septa, the combination of a
conical early stage and solitary form still appears to be
uniquely rugosan, although these characters separately occur
in other groups of corals. However, there is sometimes a
problem in assigning certain modular corals to either the
Rugosa or the Tabulata, and this is discussed further in
section 3.5, after the characteristics of the tabulates have
been established.
One other rugosan whose affinities have been questioned is
the Devonian slipper-shaped, operculate coral Calceola.
Despite early uncertainty, the genus has been regarded as a
rugose coral since the middle of the last century (Kunth 1869).
Recently Stolarski (1993) has suggested that Calceola and all
the other Siluro-Devonian operculate corals should be placed
in a monophyletic grouping with the Cambrian coral Cotho-
nion (see below) and separated from the Rugosa. However,
Cambrian and Silurian operculae contrast significantly in
septal arrangement; this, together with the large stratigraphi-
cal gap, strongly suggest that the evolution of operculae was
convergent. Furthermore, whilst the pattern of septa insertion
is unknown for Cothonion, it is certainly rugosan in Gonio-
phyllum (Wright, pers. comm. 1996) and apparently so in Cal-
ceola (Termier & Termier 1948). These Siluro-Devonian
operculate corals, on the best available evidence, should be
retained as cystiphyllid rugosans.
Table 2
Definitions of septal terras used in this paper. See text for further
comments and discussion.
alar septa
cardinal septum
counter septum
counter-lateral septa
diplosepta
entosepta
exosepta
hyposepta
major septa
metasepta
minor septa
protosepta
pair of laterally placed major septa, with loci
of septal insertion on their counter sides,
considered by most, and herein, to be
protosepta q.v.; symbol A. (Fig. 4).
protoseptum q.v. in plane of bilateral
symmetry, flanked by loci of septal insertion;
symbol C. (Fig. 4).
protoseptum q.v. in plane of bilateral
symmetry opposite cardinal septum; symbol
K. (Fig. 4.).
pair of major septa flanking the counter
septum, considered by some, and herein, to
be protosepta
q.v.;
symbol KL. (Fig. 4).
major and minor septa in rugose corals that
bifurcate towards their peripheral ends; rare,
septa defined by their locus of development
corresponding to the space within the
mesenterial pairs of the polyp (entocoels)
(Fig. 7).
septa defined by their locus of development
corresponding to the space between the
mesenterial pairs of the polyp (exocoels)
(Fig- 7).
third order septa in rugose corals, usually
only visible as external septal grooves where
they appear serially by intercalcation
between major and minor septal grooves
(Fig. 6).
the longer of the two principal orders of
septa in rugose corals; represented by the
protosepta and the metasepta; rarely
apparently the only order of septa
developed, when minor septa confined to the
corallite wall (Figs 4,6,7).
the major septa in rugose corals inserted in
four loci after the insertion of the
protosepta.
the shorter of the two principal orders of
septa in rugose corals, serially inserted
between all major septa except in some
cases between the counter septum and the
counter-lateral septa; rarely apparently
missing, when they are confined to the
corallite wall and (usually*) represented by
septal grooves (Figs 4,6,7).
the first set of major septa inserted in the
corallite (or protocorallite in modular
forms),
considered to correspond to the first
cycle of mesenterial couples of the polyp and
defining the loci of metaseptal insertion;
interpreted here as six in number and
comprising the cadinal, counter, two alar and
two counter-lateral septa (Figs 4,7).
* rarely, septal grooves are not present on the epithecae of rugose
corals.
1.2. The rugosan polyp and the problem of protosepta
Clues to the nature of the rugosan polyp can be sought in the
general similarity of rugosan skeletal elements to those of
scleractinian corals, and observations on scleractinian polyps
and living zoantharian anemones (Duerden 1902, 1904;
Hyman 1940; Wells & Hill 1956; Fig. 7). In the first place, the
close agreement in septal structure and growth between
Rugosa and Scleractinia (Sorauf 1993) suggests that the septal
relationship to the secreting surface of the polyp is likely to
have been similar. This is normally extended to the supposi-
tion that the infolds of the basal ectoderm in which at least
the major septa develop corresponded to entocoels defined
by paired mesenteries in the polyp (see Table 2 for definitions
of septal types mentioned here). Such a relationship has been
denied by Birenheide (1965), who claimed that metasepta in
Rugosa originated by splitting off from septa formed earlier,
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 183
and that this process was comparable to the formation of
exosepta in Scleractinia. Scleractinian exosepta appear
between mesenterial pairs as a septal cycle preceding inser-
tion of the next cycle of paired mesenteries. When this next
cycle of paired mesenteries develop, the exosepta are split to
accommodate them and their accompanying entosepta. This
is the process known as septal substitution (Vaughan & Wells
1943;
Wells 1956). This process of septal splitting appears to
be necessarily associated with mesenterial development and
thus would not support Birenheide's contention. However,
septal splitting as such (forming the diplosepta of Weyer 1980)
occurs only rarely in the Rugosa (although septal grooves
suggest that the loci of new septal development were seeded
from the peripheral extremities of the invagination of the cal-
cioblast layer associated with the immediately pre-existing
series of septa), and the pattern of septal development shows
no comparable features to the characteristic septal layout
resulting from septal substitution (Schindewolf 1967;
Schouppe & Stacul 1968). To my mind, the most reasonable
supposition is that at least the major septa in Rugosa are
entosepta.
Among living zoantharian anemones, the initial insertion of
six primary mesenterial couples arranged in pairs (Vaughan &
Wells
1943;
Wells 1956) appears to be standard, although irreg-
ularities in insertion are not uncommon. Of these, one group,
the Zoanthiniaria (zoanthids) shows serial insertion of further
paired mesenteries in two primary exocoels only, following the
insertion of the first six pairs (Fig. 7B). The direction of inser-
tion of successive secondary mesenterial pairs is adjacent to
and towards the ventral pole (Duerden 1902, 1904; Hyman
1940;
Wells & Hill 1956). If an analogy is made with the Rugosa,
this suggests that the cardinal septum represents the ventral
pole and that secondary mesenterial insertion is taking place in
primary exocoels equivalent to the sectors between the cardi-
nal and alar septa in the Rugosa. In addition, Duerden (1905,
1906) equated the cardinal fossula of rugose corals with the
position of the ventral stomodaeal groove (siphonoglyph) of
living zoanthid polyps. The Rugosa only differ, as far as the
insertion of major septa are concerned, by showing insertion in
a second pair of sectors, between the alar septa and the counter-
lateral septa. A further distinction between the Zoanthiniaria
and the Rugosa is the insertion of minor septa in the latter.
Fig. 7 Arrangements of mesenterial pairs and patterns of septal insertion in some corals and anemones. A. Serial septal insertion in Rugosa,
together with conjectural mesenterial pattern, reflecting possible descent from a zoanthiniarian ancestor. Minor septa are interpreted as
exocoelic. B. Mesenterial pattern in the Zoanthiniaria, with serial insertion of new pairs in two primary exocoels only, adjacent to the ventral
directive couple. C. Cyclic septal insertion and mesenterial pattern in Scleractinia (lacking septal substitution). D. Mesenterial pattern in
the Corallimorpharia, the anemone group considered to be ancestral to the Scleractinia and the Kilbuchophyllida. Septa in black;
mesenteries with muscular pleats indicated in grey. D, dorsal,
V,
ventral. Rugosan notation as in Fig. 4. After Wells & Hill (1956); Duerden
(1902).
184 C.T. SCRUTTON
Duerden (1902) considered the rugosan major septa to be
entosepta and the minor septa to be exosepta in comparison
with living corals. Although the analogy may be flawed, in view
of the penecontemporaneous insertion of these two principal
septal types (see above), this interpretation has considerable
merit as a logical explanation for the developmental distinction
between them (Fig. 7A). It was for this reason that Duerden
restricted the term metasepta to the major septa in the Rugosa.
However, the important point is that minor septal insertion also
appears to be serial as with the major septa. The common
fundamental pattern of serial mesenterial/septal insertion
identifies the Zoanthiniaria as the living anemone group closest
to the Rugosa and suggests the possibility that an ancient zoan-
thiniarian could have been ancestral to the Rugosa, or that the
two orders could have had a common ancestor.
Hill (1956) had considered the Order Rugosa to be zoan-
tharians, although she later took a more agnostic position (Hill
1981),
elevating the group to a sub-class of equivalent rank to
the Zoantharia. The close general similarity of structure and
organisation of rugosans to living zoantharians convinces me
that her earlier view was fully justified. The identity of the pro-
tosepta in Rugosa has been much debated, with various
authors arguing for two (Hudson 1936; Wright 1969), four
(Kunth 1869; Flugel 1975) or six protosepta (Duerden 1902,
1906;
Hill 1935). However, one inference that follows from a
zoantharian classification is that the Rugosa should be con-
sidered to have six protosepta. The primary nature of the car-
dinal and counter septa are widely accepted but the
significance of the counter-lateral septa and to some extent the
alar septa has been more contentious. Wright (1969) con-
sidered the alar septa to be the first metasepta inserted in two
cardinal-lateral fossulae flanking the cardinal septum. Wright
used the term fossula for a site of septal insertion, although its
original and more normal use is for a gap, or discontinuity, in
the radial distribution of septa, with (true fossula) or without
(pseudofossula/septofossula) a corresponding depression in
the surface of the horizontal skeletal elements (see Neuman
1984).
Fossulae thus defined are often associated with sites of
septal insertion. However, as all agree, the alar septum define
two principal sectors on either side of the plane of symmetry
in which further septal insertion takes place. This would seem
to qualify the alar septa as protosepta beyond dispute. The role
of the counter-lateral septa is more difficult to resolve, as the
loci between them and the counter septum in most rugose
corals are not obviously different in character from most other
interseptal loci. Rather than protosepta, they could be the first
septa inserted in the alar quadrants (for example Fedorowski
19916). However, the locations between the counter septum
and the flanking counter-laterals could be analogous to the
two primary exocoels flanking the dorsal pole in zoanthiniar-
ians,
between which no further mesenterial (metaseptal) pairs
are inserted. From time to time, some metaseptal insertion has
been reported in these loci, although these claims have either
been shown to be misinterpretations or have never been prop-
erly substantiated (Oliver 1980; Hill 1981). However, several
studies by serial section have demonstrated a clear six septal
stage in early ontogeny (i.e. Duerden 1902,1906; Carruthers
1906;
Scrutton 1983). In addition, in many cystiphyllids, no
minor septa are inserted in the counter/counter-lateral loci
(e.g. Birenheide 1974), which suggest they are functionally not
the same as normal interseptal loci. Thus theoretical consider-
ations and, perhaps less persuasively direct observation,
suggests that the Rugosa should be regarded as possessing six
protosepta, although metaseptal insertion is only active in four
of these primary interseptal loculi.
Thus the Rugosa are considered to be a relatively homoge-
neous group, an order within the Subclass Zoantharia, pos-
sessing six protosepta/mesenterial pairs in line with other
zoantharians, and most closely related to the Zoanthiniaria
among known zoantharians.
1.3. Relationship to the Scleractinia
The relationship of the Rugosa to the broadly homoeomorphic
post-Palaeozoic Scleractinia has excited debate for many years.
The most prominent difference between the two groups has
been their contrasting modes of septal insertion, serial in four
quadrants in Rugosa and cyclic in six sextants in Scleractinia
(Fig. 7). Some workers had suggested direct descent of the
latter from the former, largely based on one or other of two
lines of argument: the interpretation of septal pattern in some
Permian corals as representing intermediate conditions
between the two (Schindewolf 1942; Il'ina 1965,1984); or by
emphasis on the bilateral ("polarity gradient") aspect of septal
insertion in scleractinians (Cuif 1977,1981). However, Oliver
(1980) published the definitive analysis of this relationship and
showed convincingly that no intermediates linking the two pat-
terns of septal insertion existed. That together with the less
compelling evidence of the lack of corals in the Lower Triassic
and the contrast in mineralogy between the calcitic Rugosa and
the aragonitic Scleractinia, led Oliver to conclude that there
was no direct relationship between the two groups. Reports of
calcitic scleractinians were regarded as doubtful by Sorauf
(1981) and the recently recorded aragonitic coral in the
Permian (Wendt 1990a) may not be a rugosan (see below),
although another late Permian coral may have aragonite in its
skeleton (Wendt 19906). Sorauf (1993) has concluded that the
rugosan (and tabulate) skeleton was constructed of calcite with
up to 8 mol% MgC03. Furthermore, the wide diversity of form
among the earliest scleractinians is now widely interpreted as
indicating a polyphyletic origin for these corals (Roniewicz &
Morycowa
1993;
Veron 1995), which would have required mul-
tiple lines of descent from among the last of the Rugosa.
Oliver's conclusion was greatly strengthened by the later dis-
covery of convincing scleractiniamorph corals, the Kilbu-
chophyllida, in the mid-Caradoc (Ordovician) of southern
Scotland (Scrutton & Clarkson 1991; Scrutton 1993,1996) and
now Northern Ireland (Scrutton et al. in press) (Fig. 8). All the
material so far recovered is mouldic, but the recognition in the
material from Northern Ireland of composite moulds of
kilbuchophyllids associated with composite moulds of
aragonitic molluscs, whilst calcitic fossils in the same beds occur
as 3-D moulds, is compelling evidence that the skeleton was
aragonitic (Scrutton et al. in press). Palaeozoic "scleractinian"
corals had been claimed previously, but without critical
supporting evidence, and none had been substantiated. The
kilbuchophyllids, however, show incontrovertible cyclic septal
insertion in six sextants and were concluded to represent an
earlier, ultimately unsuccessful skeletonized clade derived
from the same group of anemones which later gave rise to the
Scleractinia. This was the first reliable evidence for the
antiquity of a possible scleractinian ancestor (other than the
Rugosa), presumed to be the actinian/corallimorpharian
anemones. These anemones, particularly some corallimor-
pharians, are almost identical to the polyps of scleractinian
corals (Wells 1956) (Fig. 7C, D).
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 185
Fig. 8 Kilbuchophyllid corals. A. Kilbuchophyllia clarksoni, Ordovician, Caradoc Series, Bardahessiagh Formation; Pomeroy, Co. Tyrone,
Northern Ireland. Calical mould of mature corallum, X2. Dorsal (uppermost) - ventral axis denned by plate on crest of axial structure. B,
C.
Kilbuchophyllia discoidea, Ordovician, Caradoc Series, Kirkcolm Formation; Kilbucho, Southern Uplands, Scotland. Calical mould (B)
and interpretive sketch (C) of a juvenile corallum, xlO. Note the prominent, slightly wider troughs left by the six protosepta, with cyclic
insertion in each sextant. Protosepta marked by asterisks in C. Dorsal (uppermost) - ventral axis defined by axial plate, with sextants
flanking ventral pole retarded and the beginning of acceleration in the left lateral sextant. Ventral retardation is a feature of
kilbuchophyllids and also occurs in scleractinians: significant lateral acceleration is apparently restricted to K. discoidea. Compare C with
Fig. 7C.
Since then, another scleractiniamorph coral of Permian age
has been claimed embracing the enigmatic aragonitic Numidi-
aphyllum (Ezaki 1997, but see Wendt 1990a for a rugosan
interpretation). Ezaki's evidence is not wholly conclusive, but
nevertheless persuasive. He classifies Numidiaphyllum as a
member of the Scleractinia, but although it shows similarities
with some middle Triassic corals, there is no evidence for direct
descent. Recent studies analysing r-DNA sequence data from
living corals and anemones have suggested that the scleractin-
ian corals had a common ancestor more recent than their
divergence from the actinian/corallimorpharian anemones,
and that these anemones resolve into two mixed groups (Chen
et al. 1995; Veron et al. 1996; see discussion in Oliver 1996 and
Scrutton et al. in press). Because these results suggest that the
two anemone orders are not individually monophyletic, the
Actiniaria/Corallimorpharia are represented as a single
lineage in Figures 1 and 2. Similar r-DNA sequence analysis
by Romano & Palumbi (1966) and Romano (1995) suggest
that two major scleractinian lineages diverged probably more
that 250 Ma and at least 300 Ma ago, long before skeleton
acquisition. The Scleractinia, defined by the presence of a
skeleton, are thus polyphyletic, as had previously been sug-
gested (for example Roniewicz & Morycowa 1993; Veron
1995),
and it is likely that Numidiaphyllum represents another
small independent skeletonized clade, perhaps derived from
this same ancestral scleractinian stock, and which became
extinct in the end Permian extinction event. I therefore regard
the status of the numidiaphyllids at the moment to be more
properly reflected by classifying them as the order Numidia-
phyllida, of equivalent status to the Kilbuchophyllida and
some other small skeletonized clades in the Cambrian (see
section 4.5). Finally, some of the earlier claims for Palaeozoic
scleractiniamorphs, currently insufficiently well known, may
ultimately be substantiated (for example, Erina & Kim 1981).
Thus,
it is now well established that no direct link existed
between the Rugosa and the Scleractinia. The Palaeozoic
corals appear to have left no direct descendants, although
there can be no doubt that groups of anemones survived the
Permian, and other extinction events. Indeed early Triassic
anemone trace fossils have been identified (Alpert & Moore
1975;
Stanley 1988). With skeletonized corals now recorded
back to the early Cambrian, it seems likely that anemone
groups initially diversified, with other anthozoans, in the latest
Precambrian.
1.4. Origin of the Rugosa
Several genera of rugose corals, representing both solitary and
modular (dendroid and cerioid) forms, appeared more or less
simultaneously in the Blackriverian (early Mohawk, mid-
Ordovician), although there is one possible earlier Chazyan
(late Whiterock, mid-Ordovician) record of the very simple
solitary coral, Lambeophyllum (Welby 1961), which is little
more than a weakly septate conical cup (Fig. 9C). Possible
relationships between these early forms have been discussed
by Scrutton (1979) and Neuman (1984). They dismissed earlier
scenarios, either envisaging an intimate relationship between
early rugose and tabulate corals (Flower 1961), or the widely
held view of many Russian workers that an auloporid tabulate
coral was ancestral to a large subset of, or all rugose corals
(Ivanovskii 1972; Sytova 1977). Flower's scheme envisaged a
cerioid tabulate coral as ancestral to an early rugose coral of
similar modular structure such as Favistina, from which more
simple early rugosans descended by the loss of the ability in
the polyp to bud, loss of tabulae and reduction of the septa.
Such a series of degenerate changes is not otherwise recorded
among corals of any age and was thus considered unlikely. The
attraction of an auloporid ancestor was the great similarity
between the simple conical form of the auloporid corallite and
the rugosan (and tabulatan) protocorallite. However, early
records of auloporids have not been substantiated, and Scrut-
ton (1984, 1990) considered the first auloporids to have
evolved in the mid-Caradoc from the fasciculate tabulate coral
Eofletcheria, post-dating the earliest rugosans. At the moment,
there is no early tabulate coral lineage that shows evidence
suggesting likely ancestry to the rugose corals. Scrutton and
Neuman concluded that the Rugosa represented a mono-
phyletic clade, which most likely originated through the acqui-
sition of a skeleton by a solitary anemone resulting in a form
such as Lambeophyllum or Primitophyllum (Fig. 9). The possi-
bility, discussed in section 3.4, that septal insertion in some tab-
ulate corals shows a pattern similar to that in the Rugosa
suggests that a relationship between the two groups may be
186 C.T. SCRUTTON
B
Fig. 9 Some early rugose corals. A. Primitophyllum primum. B. Hillophyllum priscum. C. Lambeophyllum profundum. D. Streptelasma primum.
E.
Favistina stellata. A, B, D, E are cross-sections; C is a calical view showing pinnate arrangement of low septal ridges on floor of calice.
All from the middle Ordovician, Xl.5. All lack dissepiments, which did not appear until the upper Ordovician, and A and C lack any
horizontal partitions. Cardinal septum, where identified, oriented downwards. Modified from Scrutton (1979, fig.6) and sources quoted
therein.
closer than previously realized. However, this does not of itself
make common ancestry from a known coral more likely. This
is discussed further below.
2.
HETEROCORALLIA
2.1.
General characters and classification
Of the strongly septate corals coeval with but not here
included in the Rugosa, the most important group is the Order
Heterocorallia. The distinctive features of the heterocorals
were recognized by Yabe & Sugiyama (1940), and Schindewolf
(1941) gave them ordinal status equivalent to the Rugosa. Hill
(1981) considered the possibility that they were a division of
the Rugosa, although earlier (Hill 1956) she followed Schin-
dewolf,
as have most other workers (see Fedorowski 1991a for
review). The first heterocorals sensu stricto appeared in the
Famennian at a time when the two main groups, Rugosa and
Tabulata, were drastically reduced in diversity by late
Middle-early Upper Devonian extinctions (although a poss-
ible heterocoral has recently been reported from the Middle
Devonian (Eifelian) of Spain by Tourneur & Herrmann
(1995)). They underwent a limited diversification in various
shelf and basinal environments, particularly biohermal car-
bonates and dysphotic cephalopod-bearing limestones, before
disappearing in the mid-Namurian (Weyer 1995a; Weyer &
Polyakova 1995). Their characteristics are listed in Table 3.
They are distinguished by a unique and somewhat contro-
versial pattern of septal insertion and abaxial septal growth
(Wrzolek 1981) as well as the lack of an epitheca and no calice
such as that in rugosans. It is unclear as to how much of the
external surface may have been invested in soft tissue, and how
skeletons with a diameter of, for example, only 1 mm or less
and a length of up to 80 mm were stabilized in the environ-
ment. Cossey (1995) considered the corallite in Hexaphyllia to
be effectively endoskeletal, and although Fedorowski (1991a,
p.
45) thought the heterocoral polyp must have hung con-
siderably down the external wall, his reconstructions (Fig. 14)
do not show this. Schindewolf (1941) postulated a pseudo-
planktonic lifestyle for juvenile heterocorals, which Weyer
(1995a) suggested is supported by their wide distribution,
whilst Rozkowska (1981) regarded some Upper Famennian
heterocorals to have been suspended from (presumably ben-
thonic) algae. Some species have small, distally directed, more
or less hook-shaped spines arranged along the external septal
ridges which would help to anchor the delicate corallum to
algal fronds, although Cossey regarded these to have had a
protective function. However, in those species where the basal
parts are known, a talon or encrustation structure suggesting
benthonic attachment is present. The proximal tip of the coral-
lite has been reconstructed as a discoidal aseptate pad
(Wrzolek 1981; Fedorowski 1991a), on which the first septa
develop (Fig. 10A), although Weyer (1995a,b) has since
figured a basal talon plate, on which an initial aseptate hollow
tube is formed, as the initial stages of Oligophylloides sp. from
the late Famennian (Wocklumaria-Stufe) of Morocco.
Sugiyama (1984) distinguished two distinct ecological groups
of heterocorals, one with a benthonic holdfast, the other stabil-
ized by attachment to other, presumably soft bodied, organ-
isms,
but Cossey (1995) has since described a basal
encrustation as the initial stage of at least some members of
the latter group.
The fact that the vast majority of specimens are incomplete,
also hampers study of the earliest stages of septal insertion.
Most workers agree that some or all septa increase by the
Table 3
Characteristic features of heterocorals.
1.
Septal insertion by splitting of both ends of primary septum
(oblique septum); thereafter by repeated splitting of peripheral
septal ends (disputed).
2.
Septa well developed, of one type, with characteristic pattern
reflecting insertion.
3.
Bilateral symmetry usually expressed.
4.
Distal surface convex; no calice.
5.
Earliest stages discoidal, aseptate (based on limited evidence as
early stages rare).
6. Complex wall, non-epithecate.
7.
Tabulae well developed.
8. Predominantly solitary, some weakly modular (branching);
corallites slender c. 1-15 mm diameter.
9. Increase non-parricidal (lateral).
10.
Skeleton calcitic (low-Mg calcite), fibrous and trabecular.
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 187
Fig. 10 A-C. Heterocoral morphology illustrated by Oligophylloides pachythecus (after Wrzolek 1981). A. Reconstruction of basal disc, initially
aseptate but here with first formed septa. B. Solitary growth form with expanded base. C. Dendroid modular form with lateral branch. D.
Relationship of the Dividocorallia (Fedorowski
1991A)
to the Rugosa (modified from Wrzolek 1993). For explanation see text. Figures
reproduced by kind permission of Tomasz Wrzolek and the Editors of Acta palaeontologica polonica.
division of peripheral septal ends to produce a distinctive
pattern of dichotomous branching from the axial area towards
the periphery of the corallite (Schindewolf
1941;
Poty 1978a,6,
1981;
Sutherland & Forbes
1981;
Sugiyama 1984; see review by
Fedorowski 1991a). Schindewolf and others thought that the
initial stages possessed four septa but observations by Lafuste
(1979),
Poty (1981) and Fedorowski (1991a) have demon-
strated that the primary condition is a single septum (? anal-
ogous to the counter-cardinal septum in Rugosa), called by
Lafuste the oblique septum (Fig. 11B). One level in the series
of sections prepared by Sutherland & Forbes (1981), the first
serial section study of septal insertion in these corals, sug-
gested the possibility that some later septa might be inserted
with free axial ends. Fedorowski (1991a) argued, however, that
this appearance was the result of sections across the highly
domed tabular surfaces confusing two levels of development.
He considered that all septa originated by successive gener-
ations of dichotomous division of the peripheral ends of pre-
existing septa, often with the unequal development of
generations in different radial sequences (Fig. 11B). This has
not been universally accepted (Sugiyama 1995), and Weyer
(1995a,6) has described changing axial connections between
septa, and loss and replacement of septa, through sequences of
serial sections, and considers that some septa formed against
pre-existing septa, rather than by branching.
Recently, Fedorowski (1991a) erected a division of Palaeo-
zoic corals, the Dividocorallia, to include the Heterocorallia
and a new group, the Calyxocorallia, claimed to be inter-
mediate in structure between the heterocorals and the Rugosa
(Fig. 10D). Members of the Calyxocorallia, which contain both
solitary and modular forms and possess an epitheca and a
rugosan calyx, are distinguished by a pattern of septa like that
of the Heterocorallia, but with minor septa in addition. The
validity of this classification rests on the claimed mode of
septal insertion in the Calyxocorallia. It appears that
Fedorowski (1991a) had based his conclusions on the arrange-
ment of septa in juvenile and mature corallites, where major
septa meet and join in pairs and groups axially and periaxially
in a pattern with similarities to that in heterocorals (compare
Figs.
10D and 11A, B). Unfortunately this is not yet supported
by detailed serial sectioning of the earliest stages of these
corals.
Septal patterns in corallites can be misleading, and
similar arrangements to those illustrated in calyxocorals can
be found in other corals belonging to groups in which rugosan
septal insertion has been convincingly demonstrated (for
example Scrutton 1968, pi. 10, figs. 1, 2). Thus the validity of
the Calyxocorallia, and the necessity for the Dividocorallia as
a taxonomic unit, must remain doubtful or uncertain pending
detailed serial sectioning of calyxocorals (Wrzolek 1993).
Otherwise these corals would be considered to be normal
Fig. 11 Septal insertion in heterocorals, and patterns of septal bifurcation in scleractinian corals for comparison. A. Septal arrangement in cross-
section in Heterophyllia grandis, Carboniferous, Visean; Northumberland, England, x6. B. Uneven development of successive generations
of septal bifurcation in heterocorals as envisaged by Fedorowski (1991a) Initial oblique septum present in the centre. C-E. Process of
septal substitution in scleractinian corals (after Wells 1956). F. Process of repeated bifurcation of third order septa in the micrabacid
scleractinian coral Leptopenus discus. Traces of costae omitted for clarity (modified from Cairns 1982, fig. 1). For explanation see text.
188 C.T. SCRUTTON
rugosans, with which they agree in all other characteristics (e.g.
Berkowski 1995; Poty 1995).
Fedorowski (1991a) regarded his Dividocorallia to be phy-
logenetically descended from the Rugosa, either with a
common ancestor for the Calyxocorallia and the Heterocoral-
lia, or with the latter derived from the first supposed calyxo-
coral Pseudopetraia. Both Pseudopetraia and his earliest
heterocoral Tetraphyllia are recorded from the late Lower
Devonian. However, Weyer (1991,1995a) has interpreted the
septal insertion in a new species of Pseudopetraia, P. issa, to be
rugosan and rejected the poorly known Tetraphyllia as a het-
erocoral on the basis of its wall structure. The earliest member
of the Heterocorallia may now be the new record from the
Eifelian of Spain (Tourneur & Herrmann 1995), in which case
leaving a gap before the group becomes common in the
Famennian.
2.2.
Relationship to the Rugosa
Quite apart from the question of the calyxocorals, the hetero-
corals have been considered to be related to the Rugosa (Hill
1981).
This relationship seems to be based on the well devel-
oped septa of heterocorals, their predominantly solitary
nature and their contemporaneity. However, their lack of
epitheca and strongly everted distal surface distinguish them
sharply from rugosans, and what is currently agreed of the het-
erocoral pattern of septal insertion contrasts fundamentally
with than of the Rugosa. Neither does it show convincing simi-
larities with septal/mesenterial insertion in any other zoan-
tharians, although Fedorowski (1991a) has described possible
patterns of paired mesenteries in the heterocoral polyp that
might explain the sequence of septal division. The trabecular
nature of the septa in heterocorals certainly suggests that their
relationship to the secretory surface of the polyp is likely to
have been similar to that in the other corals. However, an
origin close to the Rugosa is not the only option. A better
model may be the way in which exocoelic septa split to
accommodate the delayed insertion of the next cycle of mesen-
terial pairs with their accompanying entosepta in scleractini-
ans showing septal substitution (Vaughan & Wells
1943;
Wells
1956;
Birenheide 1965) (Fig. 11C-E). Although exosepta,
where developed, usually appear in the second or third cycle
in scleractinians, the first cycle may be exocoelic in rare
instances. Alternatively, Cairns (1982) illustrated multiple
bifurcations of third-order septa unique to some rare, deep-
water, micrabaciid scleractinians which are very similar to the
pattern seen in heterocorals (Fig. 11F). The heterocoral inser-
tion pattern might result from the heterochronic extension of
septal substitution or bifurcation back to the protoseptal cycle,
with the almost total suppression of entosepta in early stages.
If later septa are not inserted by division, they may represent
true entosepta. Birenheide (1965) argued for a lack of mesen-
teries in both Rugosa and Heterocorallia on the basis of septal
splitting in comparison with septal substitution in scleractini-
ans,
and erected for these two orders the subclass Euanthozoa.
However, the sort of septal splitting envisaged in the hetero-
corals is extremely rare in the Rugosa, as noted above, and
there may be no close relationship between these two orders.
On the other hand, there may be some merit in Birenheide's
analogy as applied to the heterocorals, although this does not
mean that mesenteries were completely lacking in that group.
Although highly speculative, this scenario is no more far
fetched than derivation from the Rugosa, and has the merit
that the initial aseptate basal disc and lack of epitheca are fea-
tures of most scleractinians. We now know indirectly that the
anemone group ancestral to the scleractinians was present in
the Palaeozoic, so perhaps the origins of the heterocorals
should be sought there. However, all these septal arrange-
ments in scleractinians are hexamerally based in contrast to
the heterocoral pattern, and for the moment, it is best to con-
sider this group of uncertain origin. Indeed, they may not be
closely related to any of the other groups of corals and may
have arisen from a group of anemones now extinct.
3.
TABULATA
3.1.
Scope and definition of group
Unlike the Rugosa, the concept of the Tabulata has had a much
more chequered history (see for example Hill 1981). Although
the origin of the grouping goes back to the middle of the last
century and the influential work of Edwards & Haime (1850),
it was not stabilized in its present general form until the work
of Sokolov (1950,1962) and Hill & Stumm (1956). Even then,
major differences in the relationships and treatment of the
heliolitids and chaetetids distinguished these two seminal
contributions. These and other debates have continued to sur-
round the Tabulata, and some have regarded the concept as no
more than a convenient grouping of more or less weakly
related corals.
The problems stem from the very simple structure of the
corallites of many tabulates and the lack of any very distinc-
tive characteristics that clearly define the group. The principal
features of tabulate corals are listed in Table 4. They agree with
the Rugosa in their conical earliest stages, predominately
epithecate/holothecate corallum and similar modes of
increase, but septa are often poorly developed or absent and
until recently, little was known of septal insertion in these
corals. They are exclusively modular; the two, apparently soli-
tary, corals assigned here by Hill (1981) are poorly known and
not convincingly tabulates. Excluding these, it will be useful to
start with Hill's (1981) definition of the Tabulata and to con-
sider first those elements which more or less certainly should
be excluded from the group.
Table 4
Characteristic features of tabulate corals.
1.
Septa often 12, but otherwise variable; rarely with insertion
similar to Rugosa.
2.
Septa typically weakly developed, predominantly of one type,
often spinose, may be absent.
3.
Bilateral symmetry of corallite usually not expressed.
4.
Earliest stages conical or conico-cylindrical, ?aseptate.
5.
Predominantly epithecate.
6. Horizontal partitions present in most, usually undifferentiated
(tabularium).
7.
Axial structure rare.
8. Exclusively modular (compound, colonial); corallites slender c.
0.5-20 mm diameter.
9. Modules incommunicate or communicate (mural pores, tunnels
or connecting tubules), some with shared colonial tissue
(coenenchyme).
10.
Increase almost exclusively non-parricidal (lateral or
coenenchymal).
11.
Skeleton calcitic (<8 mol% MgC03), fibrous and trabecular.
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 189
3.2.
Non-tabulates and doubtful tabulates
There is very clear evidence that the chaetetids are sponges
rather than corals and should be removed from the Tabulata
(Fig. 12B, C). Sokolov (1955,1962) had previously excluded the
chaetetids from the Tabulata on the basis of their distinctive lon-
gitudinal fission mode of increase, although he placed them with
the Hydrozoa. With renewed interest in the living demosponges
with massive basal skeletons ('sclerosponges'), similarities
between ceratoporellids, and particularly the tabulosponges,
and the chaetetids supported their reassignment to the sponges
(Hartman & Goreau 1972,1975; Fischer 1977; West & Clark
1984),
and later some chaetetids yielded spicules (Deici et al.
1977;
Kazmierczak 1979; Gray 1981). More recent reviews have
demonstrated the chaetetid morphology to be a repetitive grade
of skeletal organisation in demosponges from the Cambrian
onwards (Wood 1990). Two other former tabulates, now more
reasonably regarded as sponges, are Nodulipora and Desmido-
pora (Hartman & Goreau 1975; Stel & Oekentorp 1981).
More problematical is the position of the Family Tetradi-
idae.
This exclusively Ordovician group possesses distinctive
tetrameral symmetry and a four-fold longitudinal fission mode
of increase (Bassler 1950; Hill 1981), unknown elsewhere
among the tabulates (Fig. 12D). In addition, the skeleton was
almost certainly aragonitic (Scrutton 1984, Wendt 1989). I am
increasingly convinced that they are not closely related to the
rest of the Tabulata and should be excluded from that group.
Similarities between tetradiid longitudinal fission and that of
chaetetids suggest a possible relationship with the sponges, but
the tetradiids include a spectrum of growth-forms from
massive to delicately branching (Bassler
1950;
Webby & Seme-
niuk
1971;
Webb 1995), of which the latter are unlike any other
calcified sponges. Their affinities are thus uncertain at the
moment, but their broad similarity to some of the Early Cam-
brian coralomorphs is discussed below.
3.3.
Should all the Tabulata be classified as sponges?
Increasing knowledge of sponges with massive basal skeletons
and the removal of some ex-tabulates to the sponges has
revived the suggestion, first made by Kirkpatrick (1912), that
other, if not all tabulate corals should be considered sponges
(Fliigel 1976; Stel & de Coo 1977; Kazmierczak 1984). I have
countered the general arguments elsewhere (Scrutton 1979,
1987) but one continuing persistent claim is the presence of
spicules in various tabulate corals (Kazmierczak 1984, 1989,
1991,1993). Many of these claims have already been disputed
(Oekentorp 1985; Finks 1986; Oliver 1986; Wood et al. 1990;
Elias & Lee 1993) and shown to be artefacts created by minute
microbial borings, but the latest (Kazmierczak 1994) concerns
structures in the walls of the Silurian tabulate coral Angopora
hisingeri which are not borings (Fig. 13). Kazmierczak
describes mid-wall tracts of vertical monaxons, interrupted by
zones in which meshworks of subhorizontally oriented irregu-
lar monaxons occur. However, this coral has a trabeculate wall
structure, and in this case, the 'spicular' structures are the dia-
genetically altered, very fine crystallites in the cores of the tra-
becule, which result in clear, vertical calcitic rods in the walls.
In places alteration has proceeded to the point where a more
or less continuous plate of clear neomorphic calcite defines the
centres of the corallite walls, and this provides the fabric inter-
preted as the "meshwork of subhorizontal spicules". The true
nature of these diagenetic effects is emphasized by the cores
of the septal spines in the corallites, which show the same clear
axial rods of neomorphosed calcite as seen in some corallite
walls.
Thus there are no convincing records of spicules
amongst these tabulate corals, although there are now many
fossil groups regarded as sponges which only rarely or never
contain spicules, such as the stromatoporoids, archaeocyathids
and chaetetids. The recognition of the bulk of the Tabulata as
corals is based, rather, on the positive evidence of anthozoan
A B C D
Fig. 12 Comparative patterns of increase and corallite/calical arrangement in tabulate corals and their homoeomorphs. A. Favosites, showing
lateral increase, and a median plate in the corallite walls representing the fused epithecae of adjacent corallites. B, C. Chaetetid grade
demosponge showing intramural increase (B) and longitudinal fission (C), and amalgamated wall structure. D. Tetradium, showing
characteristic quadripartite longitudinal fission effected by the four 'septa' fusing in the axis of the corallite. Wall structure unknown as
commonly neomorphosed. A-C after Scrutton (1987, figs 2A~F); D after Bassler (1950, pl.4, figs 11,12).
190 C. T. SCRUTTON
Fig. 13 The tabulate coral Angopora hisingeri, Silurian, Telychian, Uggool Limestone; Charlestown, Co. Mayo, Ireland. A. Cross-section, with
trabecular walls, x8. B. Longitudinal section, x8. C, D. Cross-section (C) and longitudinal section (D) detail, showing clear rod-like
structures in the coraliite walls, interpreted as spicules pseudomorphed in calcite by Kazmierczak (1994), X20. In fact the structures are
formed by the progressive neomorphism of trabecular cores in the walls, which in places has resulted in a discontinuous to continuous
sheet of clear calcite in the mid-wall area. Similar clear rods can also be seen in the axes of the trabecular septal
spines.
The small (<0.15 mm
diameter) vertical tubes in the coraliite walls are the result of infestation by a parasite or symbiont.
structures and similarity to other groups of corals (Oliver 1979,
1986;
Scrutton 1987).
It would be unwise to claim that there are no sponges yet
lurking among the Tabulata which may be revealed by future
detailed research. However, I am convinced that the bulk of
the group, excluding the forms mentioned in the previous
section, are corals and not sponges. One line of evidence, one
of the most spectacular recent discoveries among corals, is the
fossilized remains of polyps in a small population of favositid
tabulates from the Lower Silurian (Llandovery, Aeronian)
Jupiter Formation of Anticosti Island, Quebec (Copper 1985;
Copper & Plusquellec 1993) (Fig. 14). These structures are
formed of a set of 12 calcified, transversely wrinkled, radiating
segments in lateral contact, withdrawn from the coraliite walls
and tapering into the axial area. They rest on the uppermost
tabula of each coraliite across the colony surface. The regular-
ity of their development excludes the possibility that they rep-
resent the chance association of an unrelated epibiont.
Despite the misgivings expressed by some (Oekentorp & Stel
1985;
Kazmierczak 1991,1993), and difficulties in explaining
the process of fossilisation involved, these are very convincing
polyps. Early suggestions that they might form in a manner
similar to pseudoperculae (Dunbar 1927) can be discounted.
Pseudoperculae are incomplete or modified tabulae, whereas
these structures sit on top of the last formed tabulae and con-
trast with them in microstructure and surface morphology
(Copper & Plusquellec 1993). In living scleractinian polyps,
minute CaC03 crystals are seeded within calcioblastic ecto-
derm (Sorauf 1993), and proliferation and/or malfunction of
such calcioblast cells leading to death, perhaps as the result of
a genetic mutation in this small population, might provide a
Fig. 14 Preserved polyps in Favosites sp., Lower Silurian (Llandovery), Jupiter Formation; Anticosti Island, Canada. A. Hemispherical colony
showing polyps resting on last formed tabulae of many of the corallites across the surface, x4. B. Detail, X12. C. Concentrically and radially
striated polyp stalk, with twelve inwardly retracted tentacles of generally alternating length, x50. Photographs courtesy of Paul Copper.
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 191
possible mechanism for the preservation of these polyps.
Whatever their taphonomic history, I regard them as demon-
strating convincingly that the favositids and their relatives are
corals.
3.4.
Relationship of the Tabulata to other corals
Despite their chequered history, many now regard the Tabu-
lata to be a cohesive group (Hill 1981; Scrutton 1979, 1984;
Pandolfi 1989). It is possible to develop a phylogenetic scheme
for the group and to suggest realistic lines of descent connect-
ing the diverse patterns of sub-ordinal modular organisation
(Scrutton 1984). In particular, the heliolitids, which have been
regarded as a separate group, particularly by Russian workers
(i.e.
Sokolov 1955,1962), may have evolved from a nyctoporid
ancestor by the heterochronic development of a persistent
undifferentiated state in lateral buds. In one nyctoporid
species, the potential is demonstrated by extended juvenile
offsets that mimic tubular coenenchyme, before developing
into adult corallites (Scrutton 1984; Figs. 15 A, B). The retarded
offsets suggest an origin for the coenenchyme in the heliolitid
colony, and for the coenenchymal offsetting characteristic of
heliolitids. Furthermore, other coenenchymal colonial
organisations, such as those found in the syringophyllids and
the halysitids, exist among the Tabulata, and coenenchymal
tissues probably developed independently on several
occasions (Scrutton 1984).
Tabulates as a whole have been considered either to be of
uncertain relationship to other zoantharians (Hill 1981; Scrut-
ton 1979, 1984), or to be more or less closely related to the
Rugosa (Flower 1961). Many tabulates with septa consistently
have 12 equally developed radiating plates or sets of spines, for
example the heliolitids and some favositids. In this respect, it
is interesting that the favositid polyps consistently have 12 ten-
tacles,
although occasionally with a vague suggestion of alter-
nating size (Copper & Plusquellec 1993; Fig. 14C). Mistiaen
(1989) has suggested a fundamental dodecahedral symmetry
for tabulate corals on the basis of these observations, although
it is more likely to reflect an underlying hexameral symmetry
in common with other zoantharians. However, other tabulates
may have a range of septal numbers, both more and less than
12 and some, particularly those with short spines, appear to
have no regular number.
Kim (1974) claimed to show a rugosan pattern of septal
insertion in the rare Ordovician tabulate coral Agetolites, but
this was based on the serial sectioning of offsets rather than a
protocorallite. However, the well developed nature of the
septa in this genus suggests that further work could provide
more evidence. Kim considered that Agetolites should be
reclassified with the Rugosa, but the genus has well developed
mural pores. More recently, patterns of serial septal insertion
have been convincingly demonstrated in the Devonian pleu-
rodictyid Kerforneidictyum (Lafuste & Plusquellec 1976) and
in the rather bizarre Carboniferous tabulate coral Palaeacis
(Plusquellec et al. 1990; Fig. 16). The pattern is similar to that
in rugosans (and rugosan septal notation has been applied to
these corals), although the counter area is largely obscured by
mural pores in Kerforneidictyum and is not entirely clear in
Palaeacis. In each corallite, the cardinal septum is located
towards the periphery of the colony and a strongly pinnate
pattern of metasepta can be traced from the floor of the calice
in the cardinal sectors, and less clearly in the alar sectors.
Around the counter septum, the interpretation of the pattern
of major and minor septa in Palaeacis is confused because of
the contratingent nature of the weak and spiny septa (Fig.
16C).
The preferred interpretation of Plusquellec et al. is that
the minor septa lean against the major septa on their counter
side,
as in contratingent relationships in the Rugosa, which
means that no minor septa are present between the counter
septa and the counter-lateral septa. Although contrasting with
most rugose corals, such a pattern is known in the cystiphyl-
lide rugosans (Birenheide 1974).
Plusquellec et al. (1990) discuss the possibilities, that these
are tabulate corals, rugose corals or neither. Both genera
possess features peculiar to tabulates, such as well developed,
if complex, mural pores, and are classified in groups that in all
other respects are long established tabulate corals, so the first
option seems the most probable. Some other attempts to
investigate septal insertion in tabulates with well developed
septa have not demonstrated a rugosan pattern, although
these have usually examined hystero-ontogenies, which may
not provide such clear evidence (Jul! 1976). Furthermore, the
Fig. 15 Origin of coenenchyme in heliolitid tabulate corals. A,
B.
Nyctopora goldfussi, Upper Ordovician (Richmondian), Waynesville Limestone;
near Oxford, Ohio, U.S.A., X8. New offsets remain narrow, with close spaced tabulae, for the first ~2 mm of vertical growth before
broadening out to become normal mature corallites. The narrow parts of the corallites, in both cross- (A) and longitudinal (B) section,
have a structure and appearance similar to coenenchyme in heliolitids. C, D. Heliolites megastoma, Silurian (Wenlock), Much Wenlock
Limestone Formation; Pattin's Rock Quarry, Benthall Edge, near Ironbridge, Shropshire, x5. Cross- (C) and longitudinal (D) sections for
comparison with Nyctopora goldfussi.
192 C.T. SCRUTTON
Fig. 16 Septal insertion in Palaeacis. A. Colony of Palaeacis sp. with the positions of the probable cardinal septa in peripheral corallites indicated,
X3.
B. Cross-section of Palaeacis cuneiformis subsp. A Webb, showing positions of probable cardinal (to base), alar and counter septa,
with metasepta between pinnately arranged, x9. C. Sketch of septa, represented by lines, in Palaeacis sp. (after Plusquellec etal. 1990, fig.
9Cc).
Minor septa are somewhat irregularly developed. The probable cardinal (to base), alar and counter septa are indicated by asterisks,
and the pinnate arrangement of the other septa is clear. Of the septa flanking the counter septum, indicated by
arrows,
it is unclear whether
the inner or outer septum of each pair is a minor septum (see text for discussion). Photographs courtesy of Yves Plusquellec, reproduced
by kind permission of the Editor of Lethaia.
corallites in the majority of tabulate corals lack any marked
bilateral symmetry and septa, when developed, usually show
neither regular alternations of length, nor any differentiation
of what might be protosepta. The present indications, there-
fore,
are that the tabulates may well be derived from the same
group of anemones as the Rugosa, but they remain a distinct
group, and direct descent of the latter from the former is not
thought to be likely (see section 1.4 above).
The earliest tabulate coral appears still to have been the
simple cerioid modular genus Lichenaria in the early Ordovi-
cian (Scrutton 1984) (Fig. 17A). Apart from the phacelo-
cerioid Eofletcheria, all the early tabulate coral genera had
essentially the same structure, although mural pores soon
appeared (Fig. 17). More recently, convincing early Cambrian
corals have been described, some of which show some striking
similarities with tabulate corals. However, they are not
thought to be direct ancestors, and the Tabulata are regarded
as a clade originating with a skeletonisation event in the early
Ordovician among a group of modular anemones related to
the ancestors of the Rugosa, and probably also to the ancestors
of some of the Cambrian corals (see discussion under section
4.5 below).
3.5.
Distinguishing rugose and tabulate corals
Although the bulk of the members of the two orders, Rugosa
and Tabulata, present no problems with their assignment,
there are some genera whose relationships are not so clear cut.
Apart from those genera discussed in section 3.4 above, the
uncertainty is essentially restricted to a very small number of
modular corals, lacking any intercorallite communication, and
in which septa are weakly developed or absent. Fletcheria,
Fig. 17 Selected early tabulate corals. A. Lichenaria major. B. Eofletcheria orvikui. C. Foerstephyllum simplissimum. D. Nyctopora billingsi. E.
Saffordophyllum undulatum. F. Lyopora favosa. All from the middle Ordovician,
X3.
Mural pores are absent from Lichenaria (although
some later species may have very rare pores), Eofletcheria and Foerstephyllum but are present in Saffordophyllum. The gaps in the
trabecular wall, rare or disputed in Nyctopora but well developed in Lyopora, are analogous to mural pores but may not be homologous.
Modified from Scrutton (1979, fig. 4) and sources quoted therein, Bassler (1950, pis 11-14) and Sokolov (1962, figs 65a, b).
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 193
which is dendro-phaceloid, rarely cerioid, and cerioid Foer-
stephyllum, placed respectively in the Rugosa and the Tabulata
by Hill (1981), have both been considered as members of the
other main coral order. Neither have mural pores or connect-
ing tubules, but Fletcheria produces new corallites by a form of
axial or peripheral parricidal increase, which is not uncommon
among Rugosa but is unknown among the Tabulata. Some
species of Foerstephyllum have abundant short septa, which
are not differentiated into two orders and show no obvious
pattern of insertion (Ml 1976). Thus in both of these cases,
Hill's ordinal assignments are the most likely. The tabulate
corals (heliolitines) Karagemia and Rhaphidophyllum are
homoeomorphic with the Rugosa (Hill 1981), but require
further revision to clarify their affinities. These two genera
appear to have septa of alternating length and unusual
coenenchymal structures, and along with Cyrtophyllum they
possess more that the 12 septa normally characteristic of the
Heliolitina. Lafuste & Plusquellec (1990) have reviewed the
relationship between the rugosan Utaratuia, which at one time
was placed in synonymy with Tabellaephyllum, a genus subse-
quently shown to be a tabulate coral (Oliver & Sando 1977).
Utaratuia possesses scattered, irregular lacunae in the interco-
rallite walls, a feature found in some cerioid rugosans, but quite
distinct from the regularly developed mural pores of Tabellae-
phyllum, which are characteristic of many tabulates. Other-
wise,
the two genera are strongly homoeomorphic. Lafuste &
Plusquellec concluded that with respect to modular corals,
very few features are unique to one or other order. After
mural pores in tabulates, they list the dodecahedral septal
state,
also in tabulates and discrete axial structures in rugosans.
I would add multiple axial or peripheral parricidal offsets as
unique to Rugosa. Of course, none of these features occurs in
all members of either order. Despite this, in most cases as
stated above, members of the two orders can be readily sepa-
rated on combinations of the characters listed in Tables 1 and 4.
4.
THE CAMBRIAN CORALS AND
CORALOMORPHS
4.1.
The Cambrian records
There are now more than 40 proposed genera of Cambrian
cnidarians, of which about half are of broadly coralline char-
acter. The number has grown steadily, particularly in the last
20 years. Coralline cnidarians are very largely Early Cambrian
in age, principally from Australia, Canada and Siberia, but with
scattered records elsewhere and from later in the Cambrian.
They first appear in the Tommotian but become common in
the Botomian, the earliest forms solitary, whilst modular forms
predominate from the Botomian onwards. Most genera disap-
pear at the end of the Botomian or towards the end of the suc-
ceeding Toyonian. These extinctions relate to the two
successive events that almost eliminated the archaeocyaths
and led to the end of the period of early Cambrian archaeo-
cyath reef development (Zhuravlev & Wood 1996). A number
of these genera, including the majority of those most convinc-
ingly corals, occur associated with the archaeocyathan reefs.
The modular forms almost always contribute to frame build-
ing, although some, i.e. Labyrinthus (Kobluk 1979), may be
cryptic, whereas the solitary forms are almost invariably the
habitants of crypts within the fabric (Wood 1995). Zhuravlev
etal. (1993) have reviewed microstructure in some of the early
Cambrian forms. A few, mainly poorly known genera, occur in
the Middle Cambrian and possibly only one in the Upper
Cambrian, leaving a gap of some 30-40 Ma between the bulk
of these records and the first Ordovician tabulate corals.
Jell (1984) coined the term Coralomorpha for the Cambrian
coral and coral-like organisms. This is an informal grouping,
which I regard here as embracing those organisms discussed
in section 4.3 and listed in Table 5a, together with a few of the
forms mentioned in section 4.2 for which cnidarian coralline
affinities cannot be completely discounted. Those forms con-
sidered to be true corals (section
4.4;
Table 5b) require formal
classification.
4.2.
Non-corals and unlikely cnidarians
Of the claimed Cambrian cnidarians, some have been reas-
signed to other phyla. Bija (Vologdin 1932) is either a
chaetetid or a solenoporacean alga, Amgaella and Yakutina
(Korde 1959) are green algae, and Archaeotrypa (Fritz 1947,
1948) may be a bryozoan (see Scrutton 1979; Jell 1984;
Rozanov & Zhuravlev 1992; Zhuravlev et al. 1993). Cambro-
phyllum (Fritz & Howell 1955) from the early Upper Cam-
brian of Montana, which has septal-like processes within the
'corallites' is insufficiently known, but apparently has thick,
perforate walls (Scrutton 1979; Jell 1984; Kobluk 1984) and I
earlier considered it likely to be a chaetetid sponge. Pro-
toaulopora (Sokolov 1952), originally described as from the
late Cambrian of Kazakhstan, was considered of great signifi-
cance in Russian phylogenetic schemes which envisaged an
auloporid as ancestral to some or all of the Tabulata and
Rugosa. However, the original material was apparently a loose
specimen of probable Carboniferous age (Andrei Iwanowskii,
pers.
comm., 1987), and Zhuravlev et al. (1993) regarded it as
a solenoporacean alga. Heliomitra and Sillimorpha (Sedlak
1977) are mineralization artefacts (Scrutton 1979; Zhuravlev
etal. 1993).
Rozanov & Zhuravlev (1992) have recently discussed the
possibility that some of the late Vendian to early Cambrian
small shelly fossils, such as Anabarites and related forms, might
be cnidarians (Missarzhevsky 1974, Valkov 1982, Fedonkin
1983).
However, they regarded as unlikely Valkov's assign-
ment of a Subclass Angustiomedusa, embracing these forms,
to the Scyphozoa on the basis of supposed similarity to the
Conulata. These are a group of small (3-5 mm long), calcare-
ous,
tri-radiate, straight or twisted conical tubes and, to my
mind, are very doubtful cnidarians of any description. Tynan
(1983) assigned a new order of Early Cambrian, phosphatic
organisms from California, the Paiutiida, to the Cnidaria and
possibly to the Anthozoa. This is also a group of very small
(<1 mm diameter, <5 mm long), conico-cylindrical fossils, bilat-
erally symmetrical and with up to 7 continuous or discontinu-
ous internal longitudinal ridges which Tynan interpreted as
septa. He deduced a sequence of appearance for the ridges
that mimicked cyclic septal insertion in corals. However, I
regard this comparison as fortuitous and reject any relation-
ship to the Anthozoa. The mineralogy, size and morphology of
the paiutiids suggest that they are another of the many clades
of enigmatic tubular organisms that are common in the early
Cambrian and which are mainly of doubtful or unknown affin-
ity (Bengtson & Conway Morris 1992).
The Atdabanian to Botomian khasaktids (Sayutina 1980)
include both solitary and modular (sheet-like, branching and
catenulate) forms with various combinations of denticles, cysts
Table 5a
Cambrian coralomorpha
Name Form Size0 Structure Septa Tabulae Increase Age Remarks Reference
Cambrotrypa modular
c.
1
mm cerioid -
phaceloid
(Kobluk)
none none M. Cambrian,
North America. connecting processes
between separated modules
(Kobluk); not confirmed by
Bolton & Copeland
Fritz & Howell
1959;
Bolton &
Copeland 1963;
Kobluk 1984
Labyrinthus modular,
?rare
pores
< 5 mm incomplete
cerioid none ? rare
incomplete longitudinal
fission Lower Cambrian
(Toyonian),
Labrador
may be cryptic;
? aragonitic Kobluk 1979
Rosellatana modular ? oval,
<2.9 mm;
mean 1.1
x 1.5 mm
cerioid laminar, involved
in increase none longitudinal
fission Lower Cambrian
(Botomian),
Br Columbia
? aragonitic; similar to
Cambrotrypa, v. similar to
Paleoalveolites but for 4-
fold increase (Kobluk 1984)
Kobluk 1984
Flindersipora modular,
? small
pores
< 3 mm cerioid 6-16, long,
laminar + blunt
spines
sparse,
complete,
concave up
longitudinal
fission Lower Cambrian
(Botomian),
Australia
apparent pores may be
breaks in wall Lafuste et al,
1991;
Scrutton
1992
Hydroconus
(Hydroconozoa) solitary 6 mm conical no free septa;
? septal plates in
base
none none Lower Cambrian,
Atdabanian-
Botomian,
Siberia
solid base with central pit
and radiating structures of
uncertain significance.
Korde 1963
Rozanov &
Zhuravlev 1992
Dasyconus
(Hydroconozoa) solitary 4.5 mm conical none none none Lower Cambrian,
Atdabanian-
Botomian,
Siberia
Korde 1963
Zhuravlev et al.
1993
Aploconus
(? Hydroconozoa) solitary < 8 mm conical irregular short
spines or ridges none none Lower Cambrian,
lower Toyonian,
Nevada
similar to hydroconozoans Debrenne et al.
1990a; Zhuravlev
etal.
1993
Cysticyathus solitary -10 mm,
<30 mm conical,
irregular none few, complete none Lower Cambrian,
Tommotian,
Sibiria
perforate walls Rozanov &
Zhuravlev 1992;
Kmszetal.
1995
Khasaktia
(excluding other
khasaktids)
solitary -10 mm
(bowl
shaped)
bowl-
/trough
shaped
none few, complete none Lower Cambrian,
late Tommotian-
Botomian, Siberia
rather variable form Sayutina 1980
Table 5b
Cambrian zoantharian corals
Name Form Size0 Structure Septa Tabulae Increase Age Remarks Reference
Cothonion
(Cothoniida) solitary,
modular <10 mm ceratoid-
patellate,
operculate
prominent on
operculae, v. short
ridges in cup
none peripheral
parricidal earliest
M. Cambrian,
Australia
strongly bilateral Jell & Jell 1976
Tabulaconus
(Tabulaconida) solitary,
modular,
aporous
*<16 mm
(from fig.) fasciculate rudimentary or
absent complete and
incomplete,
vesicular
lateral Lower Cambrian,
Botomian /Lenian
western Canada,
Alaska, Russia
complex wall structure
diagenetic (Sorauf &
Savarese 1995)
Debrenne et al.
1987
Moorowipora
(Tabulaconida) modular,
aporous <5 mm cerioid -
fasciculate 0-20, short,
laminar mainly
complete, flat peripheral
parricidal Lower Cambrian,
Botomian,
Australia
conical colonies, fasciculate
at margins due to sediment
fouling
Fuller & Jenkins
1994;
Sorauf &
Savarese 1995
Arrowipora
(Tabulaconida) modular,
aporous <14 mm cerioid
0-35,
v. short,
spinous vesicular
tabellae lateral;
peripheral
parricidal
Lower Cambrian,
Botomian,
Australia
thick, irrregular walls; colony
in tabular sheets separated by
sediment layers
Fuller & Jenkins
1995
Lipopora modular <3 mm dendroid 8 single or paired
ridges none ?lateral earliest
M. Cambrian,
Australia
Jell & Jell 1976
Rackovskia modular 1-2 mm dendroid -
catenoid 7-8 ridges,
prominent none ? Lower Cambrian,
Atdabanian-
Botomian, Altay
Sayan, Mongolia
probably the only coralline
khasaktid Sayutina 1980
* - Diameter given as <7 mm for solitary cups and <27 mm for colonial forms (presumably the whole colony?). However Debrenne et al. (1987, fig. 10.2) shows a coraliite at least 16 mm
diameter if the scale is correct.
196 C. T. SCRUTTON
and laminae. Sayutina considered them to be ancestral to stro-
matoporoid-grade poriferans and whilst Webby (1986) con-
sidered only the genera Khasaktia and Vittia to be possible
direct ancestors, he recognized similarities between the group
as a whole and stromatoporoid- or sponge-grade organization.
However, Debrenne et al. (19905) pointed out that in Khasak-
tia vesiculosa, earlier descriptions of the wall structure had
been based on inaccurately oriented sections. They considered
this species to be strikingly similar to Cysticyathus (see below),
and that khasaktid microstructure supported the possibility of
a zoantharian assignment (Debrenne et al. 19905; Rozanov &
Zhuravlev 1992; Zhuravlev et al. 1993). Zhuravlev et al. recog-
nize that these microstructures are unlikely to be primary, but
claim consistent differences from primary and secondary
archaeocyathid and stromatoporoid microstructures. Never-
theless, I am sceptical of the value that can be placed on
microstructure in these early forms and I regard the khazak-
tids generally as very unlikely to be cnidarians. However the
group appears to be inhomogeneous. Khasaktia itself is some-
what difficult to assess and I will keep an open mind until
material referred to the genus is thoroughly revised. Apart
from that, one possible further exception is the faciculo-
catenulate genus Rackovskia Vologdin which, as noted by
Webby (1986), bears some similarity to Lipopora (Table 5b;
see below).
There are rare but generally accepted records of unminer-
alized cnidarians, cnidarian trace fossils and mineralized non-
zoantharian cnidarians from the Cambrian (Scrutton 1979;
papers in Lipps & Signor 1992), but these fall outside the scope
of this paper. Of the rest, some are too poorly known to
warrant further consideration at the moment, some are doubt-
ful corals and finally, there is a core of genera that are con-
vincingly corals and show striking morphological similarities
to later Palaeozoic corals. These latter two groups are dis-
cussed below.
4.3.
Doubtful corals
Several calcified modular skeletons have similarities with
coralline organization without being convincing corals (Table
5a).
Cambrotrypa (Fritz & Howell 1959; see also Bolton &
Copeland 1963; Scrutton 1979; Jell 1984), from the Middle
Cambrian of North America, could be a recumbent cerioid to
phaceloid coral but requires more study. Neither the type of
increase nor the composition of the skeleton is known. A new
cerioid species of Cambrotrypa from the ?Early Cambrian of
Poland appears to have irregular tabulae but no septa (Gunia
1967),
although Zhuravlev et al. (1993) regard this record to
be of doubtful affinity and age. Labyrinthus (Kobluk 1979) is
a cluster of polygonal tubes, < 5 mm across, lacking tabulae,
and with very incomplete walls giving cross-sections a
labyrinthine appearance (Fig. 18C). In Rosellatana (Kobluk
1984),
the tubes are better organized, with polygonal to
rounded lumena generally
1.1-1.5
mm diameter (Fig. 18D).
Both are preserved in neomorphic calcite, suggesting original
aragonitic mineralogy, and appear to have a form of increase
similar to longitudinal fission which suggests similarities to
chaetetids or tetradiids rather than mainstream corals (Scrut-
ton 1979; Jell 1984). Kobluk (1984), described the early stages
of the colony as recumbent in Rosellatana. He considered
Rosellatana to show similarities to Cambrotrypa (although
longitudinal fission has not been described from the latter),
but to be even closer to massive Paleoalveolites, a tetradiid
lacking tabulae. Work in progress may help to clarify the mor-
phology and relationships of these two genera (Brian Pratt,
pers.
comm., 1995, 1996). More recently Flindersipora
bowmani (Lafuste et al. 1991) was described as "the earliest
tabulate coral", but although the tubes are well organized
(<3 mm across) and contain well developed septa, septal
number is variable and tube cross-sectional shape is irregu-
larly polygonal to meandroid; again, increase is by longitudi-
nal fission (Fig. 18A, B). Wall microstructure is uncertain in
view of likely diagenetic alteration, but there appears to be no
mid-line as is present in corals. Claimed rare mural pores may
simply be incomplete walls, perhaps following division. I have
already expressed doubts about the relationships of Flinder-
sipora and rejected any direct relationship to the tabulates, or
indeed any other later corals (Scrutton 1992), a view subse-
quently supported by others (Zhuravlev et al. 1993; Sorauf &
Savarese 1995). Forms such as Flindersipora, Rosellatana and
Labyrinthus share characteristics, such as modular organiz-
ation and increase by longitudinal fission, with the Ordovician
tetradiids, former tabulate corals now best removed from that
group (see above), although they do not share the distinctive
quadripartite increase of tetradiids (Fig. 12D). These in turn
show some similarities with the chaetetid grade of organiz-
ation in the Porifera, except that module size is significantly
larger than that usually associated with these sponges. If they
are corals, then they may be severally related as separate cal-
cification events to a group or groups of anemones that have
left no other descendants and are themselves extinct.
However, I regard the affinities of these forms as far from
settled.
There are also a number of predominantly solitary organ-
isms regarded as coralomorphs, whose affinities are uncertain.
Several are grouped into the class Hydroconozoa of Korde
(1963),
of which Hydroconus is a small, simple cone with a
thickened base in which plates radiating from an axial canal
can be distinguished (Fig. 18G, H). These plates have been
compared with septa. Although Hydroconus has no strong
similarity with any later established corals, it could be antho-
zoan and is more likely to be cnidarian than related to any
other known phylum (Scrutton 1979; Lafuste et al. 1990;
Rozanov & Zhuravlev 1992). However, the simplicity of the
structure leaves assignment open to doubt, as with the other,
less well known, hydroconozoan Dasyconus. Zhuravlev et al.
(1993) consider the other hydroconozoan genera of Korde,
such as Gastroconus and Tuvaeconus, to be synonyms, nomina
nuda and nomina dubia, but suggest that Yakovlevites and
Cambroporella (previously considered to be a green alga)
might be placed here. From the illustrations of the latter two
genera provided by Sayutina (1985), neither is an anthozoan
and the algal assignment would appear to be much more likely.
Korde described the Hydroconozoa as an independent class of
Cnidarians, which is a reasonable provisional assignment
(Scrutton 1979; Oliver & Coates 1987) and is accepted here
(Fig. 1).
Aploconus (Debrenne etal. 1990a) has a very simple, conical
skeleton, not unlike a hydroconozoan but lacking the differ-
entiated base. There are short, irregular longitudinal ridges or
spines within the lumen which suggest septal processes, but no
tabulae. Other solitary forms include Cysticyathus Zhuravl-
eva, an irregular conical structure completely lacking septal
processes, generally 5-10 mm diameter, rarely <30 mm diam-
eter, and with a few tabulae (Fig. 18E, F). It may have scat-
tered pores in the external wall, suggesting that the skeleton
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 197
Fig. 18 Selected Lower Cambrian coralomorphs. A,B. Flindersipora bowmanni, Moorowie Formation (Botomian); Moorowie Mine, Flinders
Ranges, South Australia, cross- (A) and longitudinal (B) sections, X4. C. Labyrinthus soraufi, Forteau Formation (Toyonian); Southern
Labrador, Canada, cross-section, X 4 D. Rosellatana jamesi, Rosella Formation (Botomian); Kechika Mountains, northern British
Columbia, Canada, cross-sections, xl.5. E,F. Cysticyathus tunicatus, Tommotian Stage; Lena River, Siberia, Russia. E, longitudinal
sections, X8. F, cross-section, X4. G,H. Hydroconus mirabilis, Atdabanian Stage; Bol'shoi Shangan River,Tuva, Russia (Korde
1975,
pi.
1,
figs 1-2). G, cross-section, X8. H, longitudinal section, X12. Photographs courtesy of Dave Gravestock (A,B)> Brian Pratt (C, D) and
Andrei Zhuravlev (E,F), and reproduced by kind permission of Kira Korde (G,H) and Plenum Press (E).
was completely invested in soft tissue (Kruse et al. 1995).
Altogether this would be a somewhat unusual coral although
the possibility cannot be discounted. With all of these pre-
dominantly solitary forms, the very simplicity of their structure
offers little to support a positive classification, but their
identification as cnidarians seems, at the moment, as likely as
any other assignment. The further step, to place them with the
zoantharian corals, reflects little more than the dominance of
calcified conical skeletons within that group among the cnidar-
ians later in the Phanerozoic.
4.4.
Cambrian zoantharian corals
The first of the more convincing Cambrian corals to be
described was the genus Tabulaconus from the Lower Cam-
brian of western Canada (Handheld 1969; Scrutton 1979)
(Table 5b). Although originally described as a solitary cylin-
drical form, Debrenne et al. (1987) were later able to demon-
strate that some examples at least were dendroid (Fig. 19F).
The corallites are large (<16 mm diameter), thin walled, asep-
tate or with rudimentary septa, but with well developed
tabulae. More recently, the Early Cambrian archaeocyathid-
calcimicrobe reefs of the Flinders Range, South Australia,
have yielded two modular genera of undisputed corals,
Moorowipora and Arrowipora (Savarese et al. 1993; Fuller &
Jenkins 1994,1995; Sorauf & Savarese 1995) (Figs. 19 G-M).
Both are cerioid, but Moorowipora, with smaller corallites
(<5 mm diameter), often with up to 20 short laminar septa, and
flat to gently arched tabulae, shows a tendency to fasciculate
form at colony margins, possibly as a result of sediment
fouling. This individuality of the corallites, supported by
typical cerioid wall structure (Sorauf & Savarese 1995), is a
strongly coral-like feature (Scrutton 1987). Arrowipora has
very large prismatic corallites (<14 mm diameter) with rather
thick, irregular walls, also cerioid in structure, septa absent to
very short and numerous, and vesicular tabulae reminiscent of
those of the Carboniferous tabulate coral genus Michelinia, or
a rugosan cystimorph. In Arrowipora, extensive partial mor-
tality occurs, probably due to sediment fouling of the colony
surface, with recovery based on a few surviving corallites.
Neither of the two Cambrian genera has mural pores and both
appear to show axial or peripheral parricidal increase, which
198 C. T. SCRUTTON
Fig. 19 Selected Cambrian corals. A,B- Cothonion sympomatum, group of corallites (A), septal arrangement and fossula on under-surface of the
operculum (B), both X7. C-E. Lipopora lissa, fragment showing branching (C), x5, irregular cylindrical coraliite (D), X3,calical view
showing septal ridges (E), X8. Both from early Middle Cambrian, Coonigan Formation; Mootwingee district, N.S.W., Australia. F.
Tabutaconus kordae, Adams Argillite (Botomian); Tatonduk River, Alaska, longitudinal section of a branching corallum, xl.5. G-J.
Moorowipora chamberensis, longitudinal sections (G,H) showing several examples of parricidal increase, x3, cross-section (J) showing
cerioid corallum with corallites separating to become phaceloid at the margins, X3. K-M. Arrowipora fromensis, cross-section (K), x4;
longitudinal section (L) and polished surface (M), both xl.5, showing highly vesicular tabulae and parricidal increase. Both from
Moorowie Formation (Botomian); Moorowie Mine, Flinders Ranges, South Australia. Photographs courtesy of John Jell (A-E), Francoise
Debrenne (F) and Margaret Fuller
(G~M),
and reproduced by kind permission of the Australasian Association of Palaeontologists (A-E),
the Paleontological Society (F) and the Royal Society of South Australia (G-M).
is more characteristic of the Rugosa than the tabulate corals
with which they have been compared. Although lateral
increase in addition has been claimed for both genera, it
cannot be substantiated in Moorowipora from published
illustrations. Sorauf & Savarese (1995) have suggested
Moorowipora to be a member of the Tabulaconidae Debrenne
et al. (1987) and I think Arrowipora could, with less certainty,
be placed here as well.
The structure of the skeletal elements in many of these
corals has been analysed by Zhuravlev et al. (1993) and shows
a wide range of textures, the significance of which is not yet
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 199
clear. Some are unlike those found in later corals and may mit-
igate against direct descent, but most are likely to be at least
partly altered, as suggested for the complex wall structure
described in Tabulaconus by Sorauf & Savarese (1995); none
denies an anthozoan relationship.
The genera Cothonion and Lipopora were described from
the Cambrian of Australia by Jell & Jell (1976) (Fig. 19A-E).
Identification of the Lower-Middle Cambrian boundary is
difficult in Australia and the horizon yielding these corals is
most likely earliest Middle Cambrian (Peter Jell, pers. comm.,
1996).
Cothonion consists of solitary and modular clusters of
operculate cups, with a clear bilateral symmetry and well
developed radial septa with a prominent fossula on the under-
surface of the opercula. Jell & Jell (1976) included a useful dis-
cussion of the problematical disposition of polypal soft tissue
in operculate corals. They were unable to detect a pattern to
the septal insertion, but similarities with later Siluro-Devon-
ian Rugosa possessing operculae, such as Goniophyllum and
Calceola, led them tentatively to assign Cothonion to that
order. Stolarski (1993) even proposed a direct phylogenetic
relationship between Cothonion and the later operculate
corals, separating them from the Rugosa (see above). The
Cambrian genus was placed in its own family by Jell & Jell
(1976),
and this has since been raised to ordinal level as the
Cothoniida (Oliver & Coates 1987), which more correctly indi-
cates its status in my view (see below). Less attention has been
paid to Lipopora, a dendroid branching modular colony of
small tubular corallites up to 3 mm diameter with eight single
or paired septal ridges but no tabulae. Jell & Jell (1976) tenta-
tively assigned it to the Tabulata and it is a very acceptable
coral. They compared Lipopora to Coelenteratella (Korde
1959),
from the base of the Middle Cambrian in Siberia, but
this is considered a nomen dubium by Rozanov & Zhuravlev
(1992).
A more likely comparison is with Rackovskia
Vologdin, described as a khasaktid by Sayutina (1980), which
is a bushy branching modular form with corallites up to 2 mm
diameter, seven or eight prominent 'septal' ridges and no
tabulae.
4.5. Affinities of the Cambrian corals
Many authors in describing Cambrian corals have suggested,
with varying levels of certainty, relationships to one or other
of the main Palaeozoic clades, the Rugosa or the Tabulata. This
is despite the generally recognized gap in the record between
these largely Early Cambrian records and the first accepted
members of the Rugosa and Tabulata in the Ordovician. It is
accepted that the fossil record is incomplete and that future
work may yield new records of Cambrian corals. However, it
seems unlikely that direct descent will ever be established
between any of the currently most convincing Cambrian corals
and the Ordovician clades.
My own view is that assignment to either Rugosa or Tabu-
lata is misleading. These Cambrian records indicate at most
that the same, or closely related, ancestral groups of non-skele-
tonized anemones were around in the early Cambrian as in the
Ordovician, and thus in some cases skeletal morphologies are
to some degree convergent. Direct evidence for anemones is
sparse (Scrutton 1979), but trace fossils interpreted as
anemone resting traces occur from the Vendian onwards
(Crimes 1992). The Cambrian corals represent discrete clades,
of which there may be many. They represent a series of itera-
tively evolved skeletonized descendants of various anemone
lineages, each of which was of relatively short duration and
which became extinct in the Cambrian. Some of these
achieved grades of organisation not seen in the earliest
members of the later Palaeozoic stocks. Furthermore,
although only the Rugosa and Tabulata are usually considered
as comparators, it is now clear that the Actiniaria/Corallimor-
pharia were in existence in the Ordovician and may extend
back to the initial radiation of the Cnidaria in the late Pre-
cambrian. The possibility also exists, particularly for some of
the questionable corals, that other anemone groups produced
skeletonized descendants in the Cambrian but not subse-
quently, perhaps because the anemones themselves became
extinct.
Whilst comparisons with later Palaeozoic corals are useful
as an indirect indication of the likely antiquity of various
ancestral anemone groups, well defined clades of Cambrian
corals should be given ordinal status, as with the Cothoniida
and as I propose here for the Tabulaconida (embracing the
Tabulaconidae of Debrenne et al. (1987) plus Moorowipora
and, probably also, Arrowipora). The Coralomorpha remains
a useful sack grouping for unassigned and doubtful Cambrian
coralline organisms, but ultimately will hopefully, although
perhaps doubtfully, be resolved into discrete clades.
5. PALAEOZOIC HISTORY OF CORALS
5.1.
Extinction and survival
Many of the Cambrian coral clades disappeared in the first
Phanerozoic extinction events, caused by a late Botomian
anoxic excursion followed by a sharp, end Toyonian regres-
sion, towards the end of the Lower Cambrian (Zhuravlev &
Wood 1996). Potential mound-building metazoans were slow
to rediversify in the later Cambrian (Zhuravlev 1996), when
reefs were largely thrombolite-stromatolite constructions.
Apart from the few scattered Middle Cambrian and doubtful
Upper Cambrian records, skeletonized corals did not reappear
in any numbers until the Ordovician, when their rapid diver-
sification coincided with the great increase in diversity of cal-
cified metazoans in general on the extensive shallow
carbonate-producing shelves.
I have previously briefly reviewed the overall post-Cambrian
history of Palaeozoic corals in terms of their structural organis-
ation (Scrutton 1988). The main Palaeozoic coral clades
appeared in the Ordovician, first the Tabulata in the early
Ordovician, followed by the Rugosa in the mid-Ordovician
(Fig. 20). Both underwent significant diversification, but at first
the Tabulata dominated, with all varieties of modular organis-
ation known in that group represented by the mid- Ordovician.
The Early Palaeozoic was characterized by greenhouse con-
ditions, with the major extinction event of the late Ordovician
(Hirnantian) correlated with a sharp ice-house excursion
(Brenchley 1989; Brenchley et al. 1994; Armstrong 1995). Both
groups of corals show around 70% generic extinction at that
time (Scrutton 1988; Kaljo 1996). However, the database was
formerly crude and is only now being improved in detail. For
example, in the east-central USA, a distinctive low diversity
fauna of rugose and tabulate corals, dominated by solitary
streptelasmatids, characterized the latest Ordovician
(Gamachian) and earliest Silurian (earliest Llandovery). This,
the Edgewood Assemblage, coincided with the glacial regres-
sion and initial stages of the succeeding transgression
200 C. T. SCRUTTON
(McAuley & Elias 1990; Elias & Young 1992; Young & Elias
1995).
It had some similarities with contemporary faunas in
southern Sweden and South China, but was strikingly different
from the preceding diverse Richmondian faunas of the North
American continental interior, and may have migrated in from
the continental margin (for example Anticosti Island). The
Edgwood Assemblage also had very few solitary coral species
in common with the succeeding Silurian assemblage, although
some tabulate corals show closer affinities. Copper (1989) notes
that the main extinction event for reef and peri-reefal biota was
late Richmondian or earliest Gamachian on Anticosti Island.
Although the Gamachian Ellis Bay Formation has traditionally
been equated with the Hirnantian, it is possible that only
Member 7 of the formation is of Hirnantian age (Melchin et al.
1991) and that these extinctions are therefore of slightly earlier
Ashgill age. Bioherms of latest Hirnantian age on Anticosti
contain a coral and stromatoporoid fauna of predominantly Sil-
urian aspect (here earlier than where it succeeds the Edgewood
Assemblage), although some typical Ordovician genera, such
as Calapoecia (Bolton 1981), extend into the earliest Silurian.
From Estonian data, Kaljo (1996) records almost identical
extinction rates among tabulate corals at the onset of glaciation
and ice-house conditions (end of the Pirgu Stage; 60%) as
during the rapid rise in sea level as the trend was rapidly
reversed (end Hirnantian Stage; 67%).
Coral generic diversity slowly recovered through the Llan-
dovery, until by the Wenlock it had exceeded late Ordovician
levels (Scrutton 1989; Kaljo 1996). However the rate of generic
appearances was more than double the extinction rate in
rugosans in the Llandovery, whilst tabulate coral generic
turnover remained high; for the first time generic diversity in
the Rugosa exceeded that in the Tabulata and remained higher
throughout the rest of the Palaeozoic (Fig. 20). Nevertheless
L 1 M 1 U L IMI U L
1
M
1
U L 1 U L
1
Ml U L IU
Ordovician Silurian Devonian Miss Penn Permian
Ordovician Silurian Devonian Carboniferous Permian fasciculate auioporoid
incommunicate D
Fig. 20 A,B Generic standing diversity of the Tabulata (A) and Rugosa (B) subdivided by corallum form and modular organisation, and drawn
to the same scale. The effects of the late Ordovician, mid Carboniferous and particularly the end Devonian extinction events are clear, as
is the limited post-Devonian recovery of the Tabulata. C. Turnover in numbers of genera of Rugosa and Tabulata, illustrating the contrast
in evolutionary activity in the two groups. D,E. Structural terms used to describe corallum form and modular organisation in the Tabulata
(D) and Rugosa (E). Based on Scrutton (1988) using data from Hill (1981). Although this database is out of date and many more genera
have been subsequently described, no later analysis of all Palaeozoic corals by the same authority is available. It is likely that the relative
features of this data set are still pertinent.
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 201
both groups increased in diversity throughout the Silurian and
Devonian until both reached their all time acmes in the Middle
Devonian. The late Devonian extinction was a long drawn out
affair for the corals, beginning in the Givetian and proceeding
in a series of steps, with some recovery in the Frasnian, before
culminating in an extended diversity low in the Famennian
(Sorauf & Pedder 1986; Scrutton 1988; Sorauf 1988; Oliver
1990;
Oliver & Pedder 1994). The rugose coral record is much
the better known in detail and is summarized in Figure 21.
Diversity peaked in the Eifelian and started dropping in the
Givetian, with the most notable decrease around the level of
the varcus conodont Biozone. In the Frasnian, a significant
recovery in platform genera occurred, largely within a single
suborder, the Columnariina (Sorauf 1988), before near total
extinction at the end of the gigas Biozone. However, the dis-
tinctive small, solitary, basinal rugose corals were virtually
unaffected by the end Frasnian event and persisted, even mod-
estly diversified, through the Famennian to a peak of
extinction and origination at the Devonian-Carboniferous
boundary (Sorauf & Pedder 1986; Scrutton 1988). A signifi-
cant number of genera in this ecological group were confined
to the Famennian. It was also in the Famennian that the
Heterocorallia first flourished (but see above for a possible
earlier heterocoral). Oliver & Pedder (1994) have separately
analysed patterns of origination and extinction for rugose
corals of the Old World and Eastern Americas realms, largely
at stage level, through the Devonian. Their results show sig-
nificantly similar patterns in the two biogeographic areas, sup-
porting the idea that causal factors operated on a world-wide
scale. These late Devonian extinctions in the corals are con-
sidered to have been essentially environmentally driven
(Sorauf & Pedder 1986; Scrutton 1988). This was a time of
cooling climates (Copper 1977, 1986) as well as repeated
anoxic incursions across the shelf linked to cycles of sea-level
rise and fall (House 1985; Oliver 1990). Both probably played
a part, with the basinal and platform corals responding to
different environmental stresses. The mid-late Givetian
Taghanic Event (of House 1985) saw the widespread drown-
ing of the highly successful stromatoporoid-algal-coral reefs
with their associated platforms, and the temporary disappear-
ance of the stromatoporoids in many parts of the world. The
end Frasnian Kellwasser Event was the peak of extinction for
most invertebrates and has attracted the most debate in terms
of causes. The end Famennian Hangenberg Event was a major
Fig. 21 Ranges of rugose corals from the end Eifelian (Devonian) to the earliest Tournaisian (Carboniferous). Solitary coral data subdivided to
show differing responses in basinal and platform faunas. Data from Sorauf & Pedder (1986), Scrutton (1988) and Oliver & Pedder (1994).
Ranges in grey, derived from raw data in Oliver & Pedder (1994) modified and proportioned using information from various sources, are
illustrative rather than numerically exact. Extinction events are those of House (1985). Terms describing modular organisation in rugose
corals illustrated in Figure 20. Abbreviations: C, Carboniferous; E, Eifelian; Fras, Frasnian; Fam, Famennian; T, Tournaisian; d, disparilis;
a, asymmetricus; At, Ancryognathus triangularis;
gig,
gigas; Pt, Palmatolepis triangularis;
v,
velifer;
c,
costatus;
s,
sulcata; Phar, Pharciceras;
Manticos, Manticoceras; Ch, Cheiloceras;
P,
Platyclymenia;
C,
Clymenia;
W,
Wocklumaria; G, Gattendorfia. For explanation see text.
202 C. T. SCRUTTON
anoxic excursion with devastating consequences for deeper
shelf faunas. However, precise causality is far from clear and
some evidence exists for an impact event coinciding with the
relatively sharp end Frasnian extinction, which might explain
the prolonged diversity low among shelf faunas in the
Famennian (Oliver & Pedder 1994).
The precursors of the Carboniferous coral faunas diversi-
fied in the late Famennian (Fedorowski 1981). However, after
the effects of the end-Famennian pulse of extinction, Tour-
naisian recovery was slow. Not until the Visean did both
Rugosa and Tabulata achieve close to their maximum diver-
sity post-Devonian. However, whereas the Rugosa almost
matched their mid-Palaeozoic diversity, the Tabulata
remained relatively weakly differentiated (Fig. 20). A signifi-
cant extinction occurred in the mid-Carboniferous (Ser-
pukhovian; near the Eumorphoceras-Homoceras zonal
boundary) attributed to widespread regression and climatic
cooling related to the onset of glaciation (Kossovaya 1966).
Both groups were affected, with the disappearance particu-
larly of the classic rugose coral fauna of the Visean. Tabulate
coral faunas showed little evolutionary activity post-Ser-
pukhovian, with the survival principally of a small number of
long-ranging genera into the late Permian. Limited recovery
in the Rugosa started in the mid-Bashkirian, but with con-
tinuing evolutionary turnover diversity steadily declined to a
final more rapid drop in the later Permian and extinction at
the very end of the period (Fliigel 1970; Fedorowski 1989).
Coral faunas in the Permian were divided into an equatorial
Tethyan Realm and a high latitude Boreal
(Cordillera-Arctic-Uralian) Realm separated by the Pan-
gaean landmass and the Palaeo-Pacific Ocean. Almost all of
the 19% by genera common to the two realms were long-
ranging members of the cool, deeper water Cyathaxonia
fauna of small solitary corals (Fedorowski 1989). A general
trend in rugose coral extinction during the later Permian was
widespread, with complex modular corals disappearing first,
followed by fasciculate forms, with simple solitary rugosans
and michelinid tabulate corals surviving to the last. Corals
had all but disappeared from the less diversified Boreal
Realm by the Guadalupian (Maokouan), with the last strag-
glers extinct in the early Dzhulflan (Fedorowski 1989).
Broadly, each step in the sequence of faunal extinctions
occurred about a stage later in the Tethyan Realm than in the
Boreal Realm, although there are still problems of dating and
correlation to be resolved. Ezaki (1994), however, has
described how in South China, both fasciculate and massive
forms extend up into the latest Permian (Changxingian).
Overall, this sequence of disappearance is regarded as reflect-
ing generally, but heterochronally, deteriorating environ-
ed
>
Q
C/)
c
cd
"o
'§
O
CD _
"co CO
.2
E
a) C >»
+5
3 JZ „.
oE
"18 if o
TD
O
o
o
CO
cO
E
o *-
c o
(I)t
C CD
CD Q.
+-»
CO
o
CD
.2
co
" CO
CD
„. CO
0) o
CO c
o
E
"o
p
(/)
frj O
4r o
CD
•»-•
co
i—
o
1=
CO
Q. CD
3 c/j
o o
CD §
o c
percent
massive
0 25 50 75
T
Fig. 22 Long term trends in various types of modular organisation in tabulate corals (illustrated in Fig. 20). After Scrutton (1988), based on data
from Hill (1981). The principal extinction events affecting the coral record are indicated (note that time resolution is to series level). For
explanation see text.
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 203
merits, in most cases involving climatic change, long-term
regression followed by late Permian transgression and anoxia,
together with local tectonic effects. Erwin (1993) has dis-
cussed the likely chain of events evolving through the late
Permian and resulting in extinction of increasing severity,
with the elimination of various refugia, such as South China,
by the end of the period. More recently, however, improved
correlation of key sections has suggested that the extinction
consisted of two distinct pulses, one in the Maokouan, the
other at the end of the Changxingian (Jin et al. 1994).
Although representatives of many groups did survive the
extinction in refugia, for example the calcareous algae and
calcisponges that are prominent when reefs first reappear in
the mid-Triassic (Stanley 1988), the rugose and tabulate corals
were not among them.
5.2.
Patterns in evolution
The study of trends in the evolution of rugose and tabulate
corals is hampered by the difficulty of erecting satisfactory
phylogenetic relationships within these groups (Hill 1981).
General trends have been noted in features such as a tendency
to evolve modular colonies from solitary ancestors (Rugosa,
Heterocorallia), a trend towards increasing integration in
modular structure (Rugosa, Tabulata), and multi-element
trends in an increasing complexity of septal, septal associated
and horizontal structures (Rugosa, but not particularly appar-
ent in Tabulata). Further discussion in detail is outside the
scope of this paper, but attention can be drawn to simple long
term patterns in their modular development (Scrutton 1988;
see also earlier discussion by Coates & Oliver 1973). The
Rugosa are predominantly solitary corals. Overall only 36% of
genera are modular and less than 6% achieve a level of inte-
gration represented by the loss of walls between adjacent coral-
lites.
They show two matching trends in increasing modularity
and increasing integration, reset by the end Devonian (Fig. 21)
and terminated by the end Permian extinction events. In both
cases,
the most highly integrated (amural) modular forms show
superior survivorship up to the terminal stages of the extinc-
tions,
although ultimately it was the solitary forms, mainly
simple in structure, small to very small in size and tolerant of
marginal conditions, that survived to rediversify in the Carbon-
iferous and were the final representatives of the Rugosa at the
end of the Permian. This pattern is also present in the late
Ordovician extinction event, but with the limited diversifica-
tion of the Rugosa at this time, less strikingly expressed.
The exclusively modular Tabulata show the successive
dominance, in percentage terms by genera, of massive
coenenchymal, massive cerioid communicate, and branching
communicate organisation (illustrated in Fig. 20), the
changeover points coinciding with the late Ordovician and the
end Devonian extinction events respectively (Fig. 22). In
addition, there is a steady decrease in the percentage of
massive genera throughout the history of the group except for
a slight reversal in the Permian, and the majority of massive
tabulates have notably large corallites post-Devonian. These
trends probably have complex underlying reasons, but suggest,
particularly post-Devonian, a response to coping with sedi-
ment movement and aggradation on the shelves. In this
respect, the loss of the stromatoporoid reefs, providing more
elevated niches, and the lack of extensive late Palaeozoic reefs
for tabulates to colonize, may have been significant (Scrutton
1988).
6. CONCLUSIONS
1.
The following coral orders of the subclass Zoantharia, class
Anthozoa, are currently recognized in the Palaeozoic:
Rugosa (mid. Ord.-Perm.), Tabulata (early Ord.-Perm.),
Heterocorallia (?mid. Dev., late Dev.-mid. Carb.), Tabula-
conida nov. (early Camb.), Cothoniida (mid. Camb.), Kilbu-
chophyllida (mid. Ord.), Numidiaphyllida nov. (Perm.). In
addition, further orders may be defined in the future among
the Cambrian corals.
2.
The Rugosa are regarded as a well defined, cohesive, prob-
ably monophyletic group, originating in the mid-Ordovician
by the acquisition of a skeleton in an anemone precursor
belonging to or related to the Zoanthiniaria. The Rugosa
became extinct at the end of the Permian and were not
ancestral to the post-Palaeozoic Scleractinia.
3.
The Tabulata have been less clearly defined and have
embraced from time to time a heterogeneous group of
organisms. The chaetetids had been removed and assigned
to the sponges previously. The tetradiids are removed here.
The remaining tabulates, at least in large part, are regarded
as corals (not sponges) and a monophyletic group, arising in
the early Ordovician by the acquisition of a skeleton in a
modular anemone ancestor probably related to the Zoan-
thiniaria. Although with a shared common ancestor with the
Rugosa, the latter are not thought to be directly descended
from the tabulate corals.
4.
The Heterocorallia are a small, monophyletic coral clade
with a distinctive, but disputed, pattern of septal insertion,
and uncertain relationships. However, they are not thought
to be directly descended from the Rugosa. The Kilbu-
chophyllida are a rare and localized group of corals, thought
to have arisen in the mid-Ordovician by the acquisition of a
skeleton in an actiniarian/corallimorpharian ancestor and
thus indirectly demonstrating the antiquity of this broad
anemone lineage. They appear to have become extinct
almost immediately. The Numidiaphyllida are a small
Permian scleractiniamorph clade, probably derived from
among the anemone stock ancestral to the Scleractinia.
5.
An increasingly large number of Cambrian, mainly early
Cambrian coralline organisms have been described, which
have been placed in the informal grouping, the Coralomor-
pha. Some can be dismissed as cnidarians, others may be
non-zoantharian, or even non-anthozoan cnidarians and a
small number of genera are accepted as true zoantharian
corals. Among these, two orders have now been recognized,
the Cothoniida and the Tabulaconida. These and other
Cambrian corals are considered to represent separate clades
originating by the acquisition of skeletons by anemone
ancestors, which in most cases may have belonged to the
same groups that subsequently gave rise to the post-Cam-
brian corals. However, no direct descent is envisaged
between any Cambrian and post-Cambrian corals.
6. Rugose and Tabulate corals were strongly affected by the
mass extinction events of the Palaeozoic, particularly the late
Devonian event which severely reduced the diversity of both
groups, and the end Permian event during which they
became extinct. Tabulate corals were the dominant group in
the Ordovician, but subsequently the Rugosa were the more
diverse, particularly after the late Devonian extinction event
when the Tabulata failed to recover their earlier significance.
Acknowledgements. I am very grateful to Bill Oliver (USGS,
Washington, DC), Mike Romano (University of Sheffield) and
204 C. T. SCRUTTON
Howard Armstrong (University
of
Durham), who kindly commented
on
an
earlier version
of the
manuscript, which
has
been much
improved
as a
result. Many friends
and
colleagues have helped
me
with discussion, particularly Bill Oliver, and/or by providing slides
for
the Address and/or photographs for this paper, including Paul Copper
(Laurentian University, Sudbury), Francoise Debrenne (Museum
National d'Histoire Naturelle, Paris), Yoichi Ezaki (Osaka City Uni-
versity),
Margaret Fuller (University
of
Adelaide), Dave Gravestock
(South Australian Bureau
of
Mines
&
Mineral Resources), John Jell
(University
of
Queensland), Yves Plusquellec (Universite Bretagne
Occidentale), Brian Pratt (University
of
Saskatchewan), Jim Sorauf
(Binghamton University), Paul Wignall (University
of
Leeds), Rachel
Wood (University
of
Cambridge) and Andrei Zhuravlev (Palaeonto-
logical Institute, Moscow). The photographs were printed
by
David
Hutchinson (University
of
Durham).
REFERENCES
ALPERT,
S. P.
&
MOORE,
J. N. 1975. Lower Cambrian trace fossil evi-
dence for predation on trilobites. Lethaia, 8,223-230.
ARMSTRONG,
H. A. 1995. High-resolution biostratigraphy (conodonts
and graptolites)
of
the Upper Ordovician and Lower Silurian
-
a chronometric framework
in
which
to
evaluate the causes
of
the mass extinction. Modern Geology, 21,1-28.
BASSLER,
R.
S. 1950.
Faunal Lists and Descriptions of Paleozoic Corals,
Geological Society
of
America, Memoir, 44.
BENGTSON,
S. &
CONWAY MORRIS,
S. 1992. Early radiation
of
biomin-
eralizing phyla.
In:
LIPPS,
J. H. &
SIGNOR,
P. W.
(eds)
q.v.,
447-481.
BERKOWSKI,
B. 1995. Calyxocorallia, their relationship
to
Heteroco-
rallia
and to
Rugosa.
VII
International Symposium
on
fossil
Cnidaria and Porifera,
Madrid,
September
1995,
Abstracts,
6.
BIRENHEIDE,
R.
1965. Haben
die
rugosen Korallen Mesenterien
gehabt? Senckenbergiana lethaea, 46,27-34.
BIRENHEIDE,
R.
1974.
Zur
Herkunft
der
devonischen cystimorphen
Rugosa. Senckenbergiana lethaea, 54,453-473.
BOLTON, T.
E. 1981. Late Ordovician and Early Silurian Anthozoa of
Anticosti Island, Quebec.
In:
LESPERANCE,
P. J. (ed.)
Field
meeting, Anticosti-Gaspe, Quebec, 1981, volume II: Stratigraphy
and paleontology. Universite
de
Montreal, Montreal, 107-136.
BOLTON.
T.
E. &
COPELAND,
M. J. 1963. Cambrotrypa
and
Bradoria
from the Middle Cambrian of western Canada. Journal of Pale-
ontology, 37,1069-1070.
BRENCHLEY,
P. J. 1989. The late Ordovician extinction. In:
DONOVAN,
S.
K. (ed.) Mass Extinctions: Processes and Evidence. Columbia
University Press, New York, 104-132.
BRENCHLEY,
P. I,
MARSHALL,
J. D.,
CARDEN,
G. A. R,
ROBERTSON,
D.
B.
R.,
LONG,
D. G. F.,
MEIDLA,
T.,
HINTS,
L &
ANDERSON,
T. F.
1994.
Bathymetric and isotopic evidence
for a
short-lived Late
Ordovician glaciation
in a
greenhouse period. Geology,
22,
295-298.
CAIRNS, S.
D. 1982. Antarctic and subantarctic Scleractinia. Biology of
the Antarctic seas
XI.
Antarctic Research Series Washington, 34
(1),
1-74.
CARRUTHERS,
R. G.
1906. The primary septal plan
of the
Rugosa.
Annals and Magazine of Natural History, (7), 18,356-363.
CHEN,
C. A.,
ODORICO,
D. M.,
TEN
LOHUIS,
M.,
VERON,
J. E. N. &
MILLER,
D.
J.
1995. Systematic relationships within the Antho-
zoa (Cnidaria: Anthozoa) using the
5'-end
of
the 28S rDNA.
Molecular Phylogenetics and Evolution, 4,175-183.
COATES,
A. G.
&
OLIVER,
W. A.,
JR.
1973. Coloniality
in
zoantharian
corals.
In:
BOARDMAN,
R. S.,
CHEETHAM,
A. H. &
OLIVER,
W.
A.,
JR.
(eds) Animal Colonies, Development and Function Through
Time. Dowden, Hutchinson and Ross, Stroudsburg, 3-27,
COPPER,
P.
1977. Paleolatitudes
in the
Devonian
of
Brazil
and the
Frasnian-Famennian mass extinction. Palaeogeography,
Palaeoclimatology, Palaeoecology, 21,165-207.
COPPER,
P.
1985. Fossilized polyps
in
430-Myr-old Favosites corals.
Nature, London, 316,142-144.
COPPER, P.
1986. Frasnian/Famennian mass extinction and cold-water
oceans.
Geology, 14,835-839.
COPPER,
P. 1989. Enigmas
in
Phanerozoic reef development. Memoir
of the Association of Australasian Palaeontologists, 8,371-385.
COPPER, P.
&
PLUSQUELLEC,
Y.
1993.
Ultrastructure of the
walls,
tabulae
and "polyps"
in
Early Silurian Favosites from Anticosti Island,
Canada. Courier Forschungsinstitut Senckenberg, 164,301-308.
COSSEY, P. J.
1995. Spiny heterocorals from Lower Carboniferous reef
limestones
of
Central England
- a
thorny problem
for
preda-
tors.
VII
International Symposium
on
fossil Cnidaria
and
Porifera,
Madrid,
September
1995,
Abstracts, 16.
CRIMES, T.
P. 1992. The record
of
trace fossils across the Proterozoic-
Cambrian boundary. In:
LIPPS, J.
H. &
SIGNOR, P. W.
(eds), q.v.,
177-202.
CUIF,
J.-P. 1977. Arguments pour
une
relation phyletique entre
les
Madreporaires paleozoi'ques
et
ceux du
Trias:
implications sys-
tematique
de
l'analyse microstructural des Madreporaires tri-
asiques.
Societe Geologique
de
France n.
s.
56, Memoire 129,
1-54.
CUIF, J.-P.
1981. Microstructure versus morphology in the skeleton of
Triassic scleractinian corals. Acta Palaeontologica Polonica, 25,
361-374.
DEBRENNE,
F. M.,
GANGLOFF,
R. A. &
LAFUSTE,
J. 1987. Tabulaconus
Handfield: microstructure and its implication
in
the taxonomy
of primitive corals. Journal
of
Paleontology, 61,1-9.
DEBRENNE,
E,
GANDIN,
A.
&
GANGLOFF,
R. A.
1990A.
Analyse sedi-
mentologique
et
paleontologie
de
calcaires organogenes
du
Cambrien inferieur
de
Battle Mountain (Nevada, U.S.A.).
Annates
de
Paleontologie, 76(2), 73-119.
DEBRENNE,
R,
LAFUSTE,
J. G. &
ZHURAVLEV,
A. Yu. 1990ft. Coralo-
morphes et spongiomorphes
a
l'aube du Cambrien. Bulletin du
Museum National d'Histoire Naturelle, 4e serie, section C: Sci-
ences de la Terre, 12(1), 17-39.
DEICI,
G.,
RUSSO,
A. Russo, F. &
MARCHI,
M. S. 1977. Occurrence of
spicules
in
Triassic chaetetids
and
ceratoporellids. Bollettino
delta Societd paleontologica italiana, 16,229-238.
DUNBAR,
C.
O.
1927. Pseudopercula
in the
tabulate coral Favosites.
American Journal
of
Science, 13,101-114.
DUERDEN, J.
E. 1902. Relationships of the Rugosa (Tetracoralla) to the
living Zoantheae. Annals and Magazine of Natural History, (7)
9,381-398.
DUERDEN, J.
E. 1904. The antiquity
of
the zoanthid actinians. Annual
Report of the Michigan Academy of Science, 6,195-198.
DUERDEN, J.
E.
1905. The morphology
of
the Madreporaria, VI. The
fossula
in
rugose corals. Biological Bulletin, 9,27-52.
DUERDEN, J.
E.
1906.
The morphology of the Madreporaria
-
VIII.
The
primary septa
of
the Rugosa. Annals and Magazine of Natural
History, (7) 18,226-242.
EDWARDS,
H. M. &
HAIME, J.
1850.
A
monograph
of
the British fossil
corals,
Part 1. Palaeontographical Society Monograph. London.
ELIAS,
R.
J.
1984. Paleobiology of solitary rugose corals, late Ordovician
of North America. Palaeontographica Americana, 54,533-537.
ELIAS,
R.
J.
&
LEE, D.-J. 1993.
Microborings and growth in late Ordovi-
cian halysitids
and
other corals. Journal
of
Paleontology, 67,
922-934.
ELIAS,
R.
J.
&
YOUNG, G.
A. 1992. Biostratigraphy and biogeographic
affinities
of
latest Ordovician
to
earliest Silurian corals
in the
east-central United States. In:
WEBBY,
B. D.
&
LAURIE,
J.
R.
(eds) Global Perspectives
on
Ordovician Geology. Balkema,
Rotterdam, 205-214.
ERINA,
M.
V.
&
KIM,
A. 1.1981. On some Ordovician Scleractinia-like
corals from
the
south Tien-Shan. Acta Palaeontologica
Polonica, 25,375-379.
ERWIN, D.
H. 1993. The Great Palaeozoic
Crisis:
Life and Death in the
Permian. Columbia University Press, New York.
EZAKI, Y.
1994. Patterns and paleoenvironmental implications of end-
Permian extinction
of
Rugosa
in
South China. Palaeogeogra-
phy, Palaeoclimatology, Palaeoecology, 107,165-177.
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 205
EZAKI, Y. 1997.
The Permian corai Numidiaphyllum: new insights into
cnidarian phylogeny and scleractinian origins. Palaeontology,
40,1-14.
FEDONKIN,
M. A. 1983. Organic world
of
the Vendian. Itogi Nauki
i
Tekhniki, seriya Stratigrafiya, Paleontologiya,
12,
1-127.
[In
Russian]
FEDONKIN,
M. A. 1992. Vendian faunas
and the
early evolution
of
Metazoa. In:
LIPPS, J.
H. &
SIGNOR,
P. W.
(eds), q.v., 87-129.
FEDOROWSKI,
J. 1981.
Carboniferous
corals:
distribution and sequence.
Acta Paleontologica Polonica, 26,87-160.
FEDOROWSKI,
J.
1985. Diffingiina,
a
new suborder
of
the rugose corals
from
SW
Texas. Acta Paleontologica Polonica, 30,209-240.
FEDOROWSKI,
J. 1989. Extinction
of
Rugosa and Tabulata near
the
Permian/Triassic boundary. Acta Paleontologica Polonica, 34,
47-70.
FEDOROWSKI,
J. 1991a. Dividocorallia,
a
new subclass
of
Palaeozoic
Anthozoa. Bulletin de I'lnstitut Royal des Sciences Naturelles de
Belgique, Sciences de la Terre, 61,21-105.
FEDOROWSKI,
J. 1991ft. Principles
of
early ontogeny
in the
rugose
corals:
a
critical review. Hydrobiologia, 216/217,413-418.
FINKS,
R. M. 1986. 'Spicules'
in
Thamnopora. Fossil Cnidaria, 15,1.2,
22.
FISCHER,
J.-C. 1977. Biogeographie
des
Chaetetida
et des
Tabu-
lospongida post-paleozoiques. Memoire
de
Bureau
des
Recherches Geologique et Minieres, 89,530-534.
FLOWER,
R.
H.
1961. Montoya and related colonial corals. Memoir of
the New Mexico Institute of Mines and Technology, 7,1-97.
FLUGEL,
H. W.
1970. Die Entwicklung der rugosen Korallen im hohen
Perm. Verhandlungen
der
Geologischen Bundesanstalt,
1,
146-161.
FLUGEL,
H. W.
1975. Skelettentwicklung, Ontogenie, und Functions-
morphologie rugoser korallen. Palaontologische Zeitschrift, 49,
407-431.
FLUGEL,
H. W.
1976. Ein Spongienmodell fur die Favositidae. Lethaia,
9,405-419.
FRITZ,
M. A. 1947. Cambrian bryozoa. Journal
of
Paleontology, 21,
434-435.
FRITZ,
M. A. 1948. Cambrian bryozoa more precisely dated. Journal
ofPaleontology, 22,373.
FRITZ,
M. A. &
HOWELL,
B. F. 1955.
An
Upper Cambrian coral from
Montana. Journal
of
Paleontology, 29,181-183.
FRITZ,
M. A. &
HOWELL,
B. F. 1959. Cambrotrypa montanensis, a
Middle Cambrian fossil of possible coral affinities. Proceedings
of the Geologists' Association
of
Canada, 11,89-93.
FULLER,
M. K. &
JENKINS,
R. J. F. 1994. Moorowipora chamberensis, a
coral from the Early Cambrian Moorowie Formation, Flinders
Ranges, South Australia. Transactions
of
the Royal Society
of
South Australia, 118,227-235.
FULLER,
M. K. &
JENKINS,
R. J.
F.
1995. Arrowipora fromensis, a new
genus and species
of
tabulate-like coral from the Early Cam-
brian Moorowie Formation, Flinders Ranges, South Australia.
Transactions of the Royal Society ofSouth Australia,W9,l'5-82.
GLAESSNER,
M.
F.
1984. The Dawn
of
Animal Life:
a
Biohistorical
Study. Cambridge University Press, Cambridge.
GLAESSNER,
M.
F.
&
WADE,
M.
1966.
The late Precambrian fossils from
Ediacara, South Australia. Palaeontology, 9,599-628.
GRAY,
D.
1.1981. Spicule pseudomorphs in a new Palaeozoic chaetetid,
and its sclerosponge affinities. Palaeontology, 23,803-820.
GUNIA,
T.
1967. Cambrotrypa (Tabulata)
z
metamorfiku Sudetow
Zachodnich. Rocznik Polskiego Towarzystwa Geologicznego,
37,417-428.
HANDHELD,
R. S.
1969. Early Cambrian coral like fossils from
the
Northern Cordillera of Western Canada. Canadian Journal of
Earth Sciences, 6,782-785.
HARTMAN,
W.
D. &
GOREAU,
T.
F. 1972. Ceratoporella (Porifera: Scle-
rospongiae) and the chaetetid 'corals'. Transactions of the Con-
necticut Academy of Arts and Sciences, 44,133-148.
HARTMAN,
W.
D.
&
GOREAU,
T. F. 1975.
A
Pacific tabulate sponge,
living representative
of a
new order
of
sclerosponges. Postilla,
167,1-13.
HILL,D. 1935.
British terminology for rugose corals. Geological Maga-
zine,!!,
481-519.
HILL,
D.
1956.
Rugosa. In:
MOORE,
R. C. (ed.) Treatise on Invertebrate
Paleontology, Part F. Geological Society
of
America and Uni-
versity
of
Kansas Press, Boulder, Colorado
and
Lawrence,
Kansas,
233-324.
HILL,
D.
1981.
Rugosa and Tabulata. In:
TEICHERT,
C. (ed.) Treatise on
Invertebrate Paleontology, Part
F
(Supplement 1). Geological
Society
of
America and University
of
Kansas Press, Boulder,
Colorado and Lawrence, Kansas.
HILL,
D. &
STUMM,
E. C.
1956.
Tabulata. In:
MOORE,
R. C. (ed.) Trea-
tise on Invertebrate Paleontology, Part F. Geological Society of
America
and
University
of
Kansas Press, Boulder, Colorado
and Lawrence, Kansas,
444-477.
HOUSE,
M. R.
1985.
Correlation
of
mid-Palaeozoic ammonoid evolu-
tionary events with global sedimentary perturbations. Nature,
London,
313,17-22.
HUDSON,
R.
G. S.
1936.
The development and septal notation
of
the
Zoantharia Rugosa (Tetracoralla). Proceedings
of
the York-
shire Geological Society,
23,68-78.
HYMAN,
L. H.
1940.
The Invertebrates: Protozoa through Ctenophora.
McGraw Hill, New York.
IL'INA,T.
G.
1965.
Chetirekhluchevie koralli pozdnei Permi
i
rannego
Triasa Zakavkaz'ya. Trudy paleontologicheskogo instituta, 107,
1-105.
[In Russian]
IL'INA,
T.
G.
1984.
Istoricheskoe razvitie korallov. Trudy paleonto-
logicheskogo instituta,
198,1-184.
[In Russian]
IVANOVSKII,
A. B.
1972.
The evolution
of
the Ordovician and Silurian
Rugosa. 23rd International Geological Congress, Proceedings of
the International Paleontological Union,
2,69-78.
JELL,
J. S.
1984.
Cambrian cnidarians with mineralised skeletons.
Palaeontographica Americana,
54,105-109.
JELL,
P.
A. &
JELL,
J. S.
1976.
Early Middle Cambrian corals from
western New South Wales. Alcheringa,
1,181-195.
JENKINS,
R. J. F.
1992.
Functional and ecological aspects
of
Ediacaran
assemblages. In:
LIPPS,
J.
H. &
SIGNOR, P. W.
(eds), q.v.,
131-176.
JIN
YUGAN, ZHANG
JING
&
SHANG QINGHUA 1994.
Two phases of the
end-Permian mass extinction. In:
BEAUCHAMP,
B.,
EMBRY,
A. &
GLASS,
D. (eds) Pangea: Carboniferous
to
Jurassic. Canadian
Society
of
Petroleum Geologists, Memoir,
17,813-822.
JULL,
R. K.
1973.
Ontogeny and hystero-ontogeny in the Middle Devon-
ian rugose coral Hexagonaria anna (Whitfield). In:
BOARDMAN,
R. S.,
CHEETHAM,
A. H. &
OLIVER,
W. A. JR. (eds) Animal
Colonies. Dowden, Hutchinson & Ross, Stroudsburg,
59-68.
JULL,
R. K.
1976.
Septal development during hystero-ontogeny
in
the
Ordovician tabulate coral Foerstephyllum. Journal
of
Paleon-
tology,
50,380-391.
KALJO,
D.
1996.
Diachronous recovery patterns
in
early Silurian
corals, graptolites and acritarchs. In:
HART,
M. B. (ed.) Biotic
Recovery from Mass Extinction Events. Geological Society,
London, Special Publications,
102,127-133.
KAZMIERCZAK,
J.
1979.
Sclerosponge nature
of
chaetetids evidenced
by spiculated Chaetetopsis favrei (Deninger
1906)
from
the
Barremian
of
Crimea. Neues Jahrbuch
fur
Geologie
und
Palaontologie, Monatshefte 1972
(2), 97-108.
KAZMIERCZAK,
J.
1984.
Favositid tabulates: evidence
for
poriferan
affinity. Science,
225,835-837.
KAZMIERCZAK,
J. 1989.
Halysitid tabulates: sponges in coral's clothing.
Lethaia,
22,195-205.
KAZMIERCZAK,
J.
1991.
Further evidence
for
poriferan affinities
of
favositids. In:
REITNER,
J.
&
KEUPP,
H. (eds) Fossil and Recent
Sponges. Springer, Berlin,
212-223.
KAZMIERCZAK,
J. 1993.
Sclerite-bearing alveolitid favositids from the
Devonian
of
central Poland. Palaontologische Zeitschrift, 67,
27-44.
KAZMIERCZAK,
J. 1994.
Confirmation
of
the poriferan status of favosi-
tid tabulates. Acta Palaeontologica Polonica,
39,233-245.
KIM,
A. I.
1974.
O
filogenii
i
polozhenii
v
sisteme nekotorykh tabuly-
atomorfnykh korallov.
In:
SOKOLOV,
B. S. (ed.)
Drevnie
Cnidaria, 1. Nauka, Moskva,
118-122.
[In Russian]
206 C. T. SCRUTTON
KIRKPATRICK,
R.
1912.
On the
nature
of
stromatoporoids. Nature,
London, 89,607.
KOBLUK,
D. R. 1979.
A
new and unusual skeletal organism from
the
Lower Cambrian
of
Labrador. Canadian Journal
of
Earth Sci-
ences, 16,2040-2045.
KOBLUK,
D.
R.
1984.
A
new compound skeletal organism from
the
Rosella Formation (Lower Cambrian), Atan Group, Cassiar
Mountains, British Columbia. Journal
of
Paleontology,
58,
703-708.
KORDE,
K. B. 1959.
Problematicheskie ostatki
iz
kembriiskikh
otlozhenii yugo-vostoka sibirskoi platformy. Doklady
Akademiya Nauk SSSR, 125,625-627.
KORDE,
K. B. 1963. Hydroconozoa
-
novy klass kishechnopolostnykh
zhivotnykh. Paleontologicheskii Zhurnal, 1963,20-25.
KORDE,
K. B. 1975. Kembriyskie Tselenteraty. In:
SOKOLOV, B.
S. (ed.)
Drevnie Cnidaria, 2. Nauka, Moskva, 53-56. [In Russian]
KOSSOVAYA,
O. L.
1996.
The mid-Carboniferous rugose coral recovery.
In: HART,
M. B. (ed.) Biotic Recovery from Mass Extinction
Events. Geological Society, London, Special Publications, 102,
187-199.
KRUSE,
P.
D.,
ZHURAVLEV,
A. Yu. &
JAMES,
N.
P.
1995. Primordial meta-
zoan-calcimicrobial reefs: Tommotian (Early Cambrian)
of
the
Siberian Platform. Palaios, 10,291-321.
KUNTH,
A. 1869. Beitrage zur Kenntnis fossiler Korallen,
2.
Das Wach-
stumsgesetz
der
Zoantharia Rugosa
und
fiber Calceola san-
dalina. Zeitschrift der Deutschen Geologischen Gesellschaft, 21,
647-688.
LAFUSTE,
J. 1979. Asymetrie
de
1'appareil
septal des Het^rocoralli-
aires.
Compte Rendu sommaire Societe geologique
de
France,
1979 (3),111-113.
LAFUSTE,
I,
DEBRENNE, F.
&
ZHURAVLEV,
A. 1990. Les fuscinules, type
nouveau de biocristaux dans le squelette d'Hydroconus Korde,
1963,
corallomorphe
du
Cambrien inferieur. Comptes Rendus
Hebdomadaires des Seances
de
I'Academie des Sciences Paris,
310,1553-1559.
LAFUSTE,
L,
DEBRENNE,
F,
GANDIN,
A. &
GRAVESTOCK,
D.
1991.
The
oldest tabulate coral and the associated Archaeocyatha, Lower
Cambrian, Flinders Ranges, South Australia. Geobios,
24,
697-718.
LAFUSTE,
J. &
PLUSQUELLEC,
Y. 1976. Kerforneidictyum n. gen. (Tab-
ulata, Devonien). Morphologie et microstructure. Bulletin de la
Societe geologique de France, 18,1699-1711.
LAFUSTE, J.
&
PLUSQUELLEC,
Y. 1990. Les genres Utaratuia, Tabellae-
phyllum, Michelinia et la distinction Tabulata-Rugosa, Annates
de Paleontologie, 76,13-39.
LIPPS,
J.
H. &
SIGNOR, P. W.
(eds) 1992. Origin and Early Evolution
of
the Metazoa. Plenum, New York.
MCAULEY,
R.
J.
&
ELIAS,
R. J. 1990. Latest Ordovician to earliest Sil-
urian solitary rugose corals
of
the east-central United States.
Bulletin of American Paleontology, 98,1-82.
MELCHIN,
M. J.,
MCCRACKEN,
A. D. &
OLIFF,
F. J. 1991. The Ordovi-
cian-Silurian boundary on Cornwallis and Truro islands, Arctic
Canada: preliminary data. Canadian Journal of Earth Sciences,
28,1854-1862.
MISSARZHEVSKY,
V. V. 1974.
Novye dannye
o
drevneishikh
okamenelostyakh rannego kembriya Sibirskoi platformy. (New
data
on
Early Cambrian ancient fossils
of
the Siberian Plat-
form.) In:
ZHURAVLEVA,
I. T.
&
ROZANOV,
A. Yu. (eds) Bios-
tratigrafiya
i
paleontotogiya nizhnego kembriya Evropy
i
svernoi Azii. (Lower Cambrian biostratigraphy and paleontol-
ogy
of
Europe and Northern Asia.) Nauka, Moskva, 179-189.
[In Russian]
MISTIAEN,
B. 1989. Importance de
la
symetrie d'ordre douze chez les
Tabulata. Comptes Rendus Hebdomadaires des Seances de I'A-
cademie des Sciences Paris, 308,451^456.
NEUMAN,
B. E. E. 1984. Origin and early evolution
of
rugose corals.
Palaeontographica Americana, 54,119-126.
NEUMAN, B.
E. E. 1988. Some aspects of life strategies of early Palaeo-
zoic rugose corals. Lethaia, 21,97-114.
OEKENTORP,
K. 1985. Spicules
in
favositid Tabulata
-
remarks
to J.
Kazmierezak's interpretation. Fossil Cnidaria, 14 (1), 34-35.
OEKENTORP,
K. &
STEL, J.
H. 1985. Favosites
- a
true coral. Remarks
to
P.
Copper's discoveries
of
fossilized polyps. Fossil Cnidaria,
14 (2), 28-29.
OLIVER, W. A., JR.
1979. Review: Sponges they are not. Paleobiology,
5,188-190.
OLIVER, W. A., JR.
1980. The relationship of the scleractinian corals
to
the rugose corals. Paleobiology, 6,146-160.
OLIVER, W. A., JR.
1986. Favositids are corals
-
further remarks. Fossil
Cnidaria, 15 (1.2),
19-21.
OLIVER,
W. A., JR. 1990. Extinctions
and
migrations
of
Devonian
rugose corals
in the
Eastern Americas Realm. Lethaia, 23,
167-178.
OLIVER, W. A., JR.
1996. Origins and relationships
of
Paleozoic coral
groups and the origin of the Scleractinia. In:
STANLEY, G. D., JR.
(ed.) Paleobiology and Biology
of
Corals. The Paleontological
Society Papers,
1.
Paleontological Society, Lawrence, Kansas,
107-134.
OLIVER, W. A., JR.& COATES,A.G.
1987. Phylum Cnidaria. In:
BOARD-
MAN,
R. S., CHEETHAM,A.
H. &
ROWELL,A. J.
(eds) Fossil Inver-
tebrates. Blackwell, Oxford, 140-193.
OLIVER, W.
A.,
JR.
&
PEDDER,
A. E. H. 1994. Crises
in
the Devonian
history
of
the rugose corals. Paleobiology, 20,178-190.
OLIVER, W.
A.,
JR.
&
SANDO, W.
J. 1977. Tabellaephyllum Stumm is a
Michelinia (Carboniferous,
Tabulata).
Journal of Paleontology,
51,422-423.
PANDOLFI,
J.
M. 1989. Phylogenetic analysis
of the
early tabulate
corals. Palaeontology, 32,745-764.
PLUSQUELLEC,
Y,
LAFUSTE,
X &
WEBB,
G. E. 1990. Organisation de
type tetracoralliaire
de
rides septales
de
Palaeacis (Cnidaria,
Carboniferous). Lethaia, 32,385-397.
POTY,
E.
1978a. Donnees nouvelles
sur les
Heterocoralliaires
du
Dinantien Beige. Annales de la Societe Geologique de Belgique,
100,233-243.
POTY,
E. 1978ft. La croissance de Heterophyllia
ornata:
un modele pour
l'ontogenie des heterocoralliaires. Comptes Rendus Academic
des Sciences Paris, 287 (7), 687-688.
POTY,
E. 1981. Recherches sur les Tetracoralliaires
et
les Heteroco-
ralliaires du Viseen de la Belgique. Mededelingen Rijks Geolo-
gische Dienst, 35-1 (1-3), 1-161.
POTY,
E. 1995. The systematical position
of
the Heterocoral-like col-
onial corals
of
the lower Carboniferous
of
China.
VII
Inter-
national Symposium on fossil Cnidaria and Porifera,
Madrid,
September
1995,
Abstracts, 68-69.
ROMANO, S.
L. 1995. Evolution
of
two architectural strategies among
scleractinian corals inferred from phylogenetic analysis
of
DNA sequences.
VII
International Symposium
on
fossil
Cnidaria and Porifera,
Madrid,
September
1995,
Abstracts, 11.
ROMANO, S. L.
&
PALUMBI, S. R.
1996. Evolution of scleractinian corals
inferred from molecular systematics. Science, 271,640-642.
RONIEWICZ,
E. &
MORYCOWA,
E. 1993. Evolution
of
the Scleractinia
in the light
of
microstructural data. Courier Forschungsinstitut
Senckenberg, 164,233-240.
ROZANOV,
A. Yu.
&
ZHURAVLEV,
A. Yu. 1992. The Lower Cambrian
fossil record
of
the Soviet Union. In:
LIPPS, J.
H. &
SIGNOR,
P.
W.
(eds), q.v., 205-282.
R6ZKOWSKA,
M.
1981. On Upper Devonian habitats of rugose corals.
Acta Palaeontologica Polonica, 23,597-611.
SAVARESE,
M.,
MOUNT,
J. F.,
SORAUF,
J. E. &
BUCKLIN,
L. 1993. Paleo-
biology
and
paleoenvironmental context
of
coral-bearing
Early Cambrian reefs: implications
for
Phanerozoic reef
development. Geology, 21,917-920.
SAYUTINA,
T. A. 1980.
A
new early Cambrian Family Khasaktidae
n.
fam.
-
possible Stromatoporata. Paleontological Journal, 14,
13-30.
[Translation
of
SAYUTINA,
T. A.
1980. Rannekem-
briyskoye semeystvo Khasaktidae n. fam.
-
vozmozhnye stro-
matoporaty. Paleontologicheskii Zhurnal, 1980 (4), 13-28.]
SAYUTINA,
T.
A.
1985.
K
revizii roda Yakovlevites Korde, 1979.
In:
SOKOLOV,
B.
S.
&
ZHURAVLEVA,
I. T. (eds) Problematic pozd-
nego dokembriya
i
paleozoya. Trudy Instituta geologii
i
geofiziki Sibirskoe oldelenie Akademiya Nauk SSSR,
632.
Nauka, Moskva, 70-74.
PALAEOZOIC CORALS: ORIGINS AND RELATIONSHIPS 207
SCHINDEWOLF,
O. H. 1941. Zur Kenntnis
der
Heterophylliden, einer
eigentumlichen palaozoischen Korallengruppe. Palaontologis-
che Zeitschrift, 22,213-306.
SCHINDEWOLF,
O. H. 1942.
Zur
Kenntnis
der
Polyeoeiien und Plero-
phyllen. Erne Studie iiber
den Bau der
"Tetrakorallen"
und
ihre Beziehungen
zu den
Madreporarien. Abhandlungen
des
Reichsamts
fur
Bodenfurschung, Neue Folge, 204,1-324.
SCHINDEWOLF,
O. H.
1967. Rugose Korallen ohne Mesentherien?
Senckenbergiana Lethaea, 48,135-145.
SCHOUPPE,
A.
VON
&
STACUL,
P. 1968.
Zur
Bildung
der
Septen
und
Mesenterien bei den Pterocorallia. Neues Jahrbuch fiir Geolo-
gic und Palaontologie, Abhandlungen, 130,47-77.
SCRUTTON,
C. T. 1968 Colonial Phillipsastraeidae from the Devonian
of south-east Devon, England. Bulletin
of
the British Museum
(Natural History), Geology Series, 15,181-281.
SCRUTTON,
C. T. 1979. Early fossil cnidarians. In:
HOUSE,
M.
R.
(ed.)
The Origin
of
Major Invertebrate Groups. Academic, London,
161-207.
SCRUTTON,
C. T.
1983. Astogeny in the Devonian rugose coral Phillip-
sastrea nevadensis from northern Canada. Memoir
of the
Association of Australasian Palaeontologists, 1,237-259.
SCRUTTON,
C. T. 1984. Origin and early evolution
of
tabulate corals.
Palaeontographica Americana, 54,110-118.
SCRUTTON,
C.T. 1985. Cnidaria. In:
MURRAY,
X W.
(ed.) Atlas of Inver-
tebrate Macrofossils. Longman, Harlow, 11-36.
SCRUTTON,
C. T. 1987.
A
review
of
favositid affinities. Palaeontology,
30,485-492.
SCRUTTON,
C.T. 1988. Patterns of extinction and survival in Palaeozoic
corals. In:
LARWOOD,
G. P. (ed.) Extinction and Survival
in the
Fossil
Record.
Clarendon, Oxford, 65-88.
SCRUTTON,
C. T. 1989. Corals and stromatoporoids. In:
HOLLAND,
C.
H. &
BASSETT,
M. G. (eds)
A
Global Standard
for
the Silurian
System. National Museum of
Wales,
228-230.
SCRUTTON,
C.T. 1990. Ontogeny and astogeny in Aulopora and its sig-
nificance, illustrated by
a
new non-encrusting species from the
Devonian
of
southwest England. Lethaia, 23,61-75.
SCRUTTON,
C. T. 1992. Flindersipora bowmani Lafuste, and
the
early
evolution
of
the tabulate corals. Fossil Cnidaria, 21 (2),
29-33.
SCRUTTON,
C. T. 1993. New Kilbuchophyllid corals from the Ordovi-
cian
of
the Southern Uplands, Scotland. Courier Forschungsin-
stitut Senckenberg, 164,153-158.
SCRUTTON,
C. T. 1996. The scleractiniamorph coral Kilbuchophyllia
clarksoni
at
Kilbucho. Scottish Journal
of
Geology, 32 (1),
91.
SCRUTTON,
C.T. &
CLARKSON,
E. N. K. 1991. A new scleractinian-like
coral from the Ordovician
of
the Southern Uplands, Scotland.
Palaeontology, 34,179-194.
SCRUTTON,
C. T.,
JERAM,
A. J. &
ARMSTRONG,
H. A. in press. Kilbu-
chophyllid corals from the Ordovician (Caradoc)
of
Pomeroy,
Co.
Tyrone: implications for coral phylogeny and for movement
on
the
Southern Uplands Fault. Transactions
of the
Royal
Society
of
Edinburgh: Earth Sciences.
SEDLAK,
W.
1977. Some aspects
on
the stratigraphy and taxonomy of
Cambrian fauna found
on
Lysa Gora
(the
Swietokrzyskie
Mountains, Central Poland). Memoires
du
Bureau
de
Recherches Geologique et Minieres, 89,42-48.
SEILACHER,
A. 1989. Vendozoa: organismic construction
in the
Pro-
terozoic biosphere. Lethaia, 22,229-239.
SOKOLOV,
B.
S.
1950. Sistematika
i
istoriya razvitiya paleozoyskikh
korallov Anthozoa Tabulata. Voprosy Paleontologii, 1,134-210.
[In Russian]
SOKOLOV,
B. S. 1952. Tabulyaty paleozoya evropeiskoi chasti SSSR,
chast 4: Devon Russkoy platformy
i
zapadnogo Urala. Trudy
Vcecoyuznogo Neftyanogo Nauchno-Issledovatel'skogo
Geologo-Razvedochnogo Instituta, Novaya Seriya, 62,
1-292.
[In Russian]
SOKOLOV,
B.
S.
1955. Tabulyaty paleozoya evropeiskoi chasti SSSR,
Vvedenie. Trudy Vcecoyuznogo Neftyanogo Nauchno-Issle-
dovatel'skogo Geologo-Razvedochnogo Instituta, Novaya
Seriya, 85,1-528. [In Russian]
SOKOLOV,
B.
S. 1962. Gruppa Chaetetida; Podklass Tabulata; Podklass
Heliolitoidea. In:
ORLOV,
U. A. (ed.) Osnovy paleontologii,
2.
Gubki, Arkheotsiaty, Kishechnopolostnye, Chervy. Izdatel'stvo
Akademii Nauk, Moskva, 169-176;
192-265;"
266-285.
[In
Russian]
SORAUF,
J. E. 1981. Biomineralization, structure and diagenesis
of
the
coelenterate skeleton. Acta Paleontologica Polonica,
25,
327-343.
SORAUF,
J. E.
1988. Rugosa
and the
Frasnian-Famennian extinction
event:
a
progress report. Memoir
of
the Association
of
Aus-
tralasian Palaeontologists, 8,327-338.
SORAUF, J.
E.
1993.
The coral skeleton: analogy and comparisons, Scle-
ractinia, Rugosa
and
Tabulata. Courier Forschungsinstitut
Senckenberg, 164,63-70.
SORAUF,
J. E. &
PEDDER,
A. E. H. 1986. Late Devonian rugose corals
and the Frasnian Famennian crisis. Canadian Journal
of
Earth
Sciences, 23,1265-1287.
SORAUF,
J.
E. &
SAVARESE,
M. 1995.
A
Lower Cambrian coral from
South Australia. Palaeontology, 38,757-770.
STANLEY,
G.
D.,
JR.
1988. The history
of
early Mesozoic reef com-
munities:
a
three-step process. Palaios, 3,170-183.
STEL, J.
H. &
DE COO,
J. C. M. 1977. The Silurian Upper Burgsvik and
Lower Hamra-Sundre Beds, Gotland. Scripta Geologica, 44,
STEL,
J. H. &
OEKENTORP,
K. 1981. Desmidopora
and
Nodulipora:
misfits
in the
coral world. Acta Paleontologica Polonica, 25,
403^117.
STOLARSKI,
J.
1993. Ontogenetic development
and
functional mor-
phology
in
the early growth-stages
of
Calceola sandalina (Lin-
naeus,
1771). Courier Forschungsinstitut Senckenberg,
164,
169-177.
STRUSZ,
D. L.
1968.
On
Cyathophyllum mansfieldense
Dun 1989:
Lower Devonian, Loyola, Victoria. Proceedings
of
the Royal
Society of Victoria, 81,11-18.
SUGIYAMA,
T. 1984. Heterocorallia from
the
Akiyoshi Limestone,
Southwest Japan. Part 1, Systematic Paleontology. Bulletin
of
the Akiyoshi-dai Museum of Natural History, 19,27-67.
SUGIYAMA,
T. 1995. New observations
on
some Carboniferous Hete-
rocorallia. VII International Symposium on fossil Cnidaria and
Porifera,
Madrid,
September 1995, Abstracts, 91.
SUTHERLAND,
P. K. 1965. Henryhouse rugose corals. Bulletin
of
the
Oklahoma Geological Survey,
109,1—92.
SUTHERLAND,
P.
K.
&
FORBES,
C. L.
1981. Septal development
of
Oligophylloides pachythecus from
the
Famennian
of
Poland.
Acta Palaeontologica Polonica, 25,497-504.
SYTOVA,
V.
A.
1977.
On the
origin
of
rugose corals. Memoires
du
Bureau de Recherches Geologique et Minieres, 89,65-68.
TERMIER,
H.
&
TERMIER,
G. 1948. Etude sur Calceola sandalina Linne.
La Revue Scientifique, 86,208-218.
TOURNEUR,
F.
&
HERRMANN,
R. 1995. Among the earliest heteroco-
rals:
new genus from the Middle Devonian
of
Spain. VII Inter-
national Symposium
on
Fossil Cnidaria and Porifera,
Madrid,
September 1995, Abstracts, 94.
TYNAN,
M. C. 1983. Coral-like microfossils from the Lower Cambrian
of California. Journal of Paleontology, 57,1188-1211.
VALKOV,
A.
K. 1982. Lower Cambrian biostratigraphy
of
the East
Siberian Platform (Uchur-Maya area). Nauka, Moskva.
[In
Russian]
VAUGHAN,
T. W.
&
WELLS,
J.
W. 1943. Revision
of
the Suborders,
Families,
and
Genera
of
the Scleractinia. Geological Society of
America, Special Papers, 44.
VERON,
J. E.
N. 1995. Corals
in
Space
and
Time: Biogeography and
Evolution
of
the Scleractinia. Cornell University Press, Ithaca.
VERON,
J. E. N.,
ODORICO,
D. M.,
CHEN,
C. A. &
MILLER,
D. J. 1996.
Reassessing evolutionary relationships
of
scleractinian corals.
Coral Reefs, 15,1-9.
VOLLBRECHT,
E.
1928.
Die
Entwicklung
des
Septalapparates
bei
Semaiophyllum. Ein Beitrag zur Entwicklung des Septalappa-
rates
der
Rugosen. Neues Jahrbuch fiir Mineralogie, Geologie
und Palaontologie, Beiligeband, 59,1-30.
VOLOGDIN,
A. G. 1932. Arkheotsiaty Sibiri, vypusk 2. Fauna kembri-
iskikh izvestnyakov Altaya. Trudy vsesoyuznogo Geologo-
Razvedochnogo ob'edineniya NKTP, SSSR, 1932,1-106.
208 C.T. SCRUTTON
WEBB,
G. E. 1995. Middle Ordovician Tetradium microatolls and poss-
ible bathymetric gradient
in
tetradiid morphology.
VII
Inter-
national Symposium on fossil Cnidaria and Porifera,
Madrid,
September
1995,
Abstracts, 98-99.
WEBBY,
B. D. 1986. Early stromatoporoids. In:
HOFFMAN,
A. &
NITECKI,
M.
H. (eds) Problematic Fossil Taxa. Oxford Uni-
versity Press, New York, 148-166.
WEBBY,
B.
D. &
SEMENIUK,
V. 1971.
The Ordovician coral genus Tetra-
dium Dana from New South Wales. Proceedings of the Linnean
Society
of
New South Wales, 95,246-259.
WELBY,C.W.
1961.
Occurrence of Foerstephyllum in Chazyan rocks of
Vermont. Journal
of
Paleontology, 35,391-400.
WELLS,
J. W. 1956. Scleractinia. In:
MOORE,
R. C. (ed.) Treatise on
Invertebrate Paleontology, Part
F.
Coelenterata. Geological
Society
of
America and University
of
Kansas Press, Boulder,
Colorado and Laurence. Kansas, 328-444,
WELLS,
J. W.
&
HILL,
D. 1956. Anthozoa - general features. In:
MOORE,
R. C. (ed.) Treatise
on
Invertebrate Paleontology, Part
F.
Coe-
lenterata. Geological Society
of
America
and
University
of
Kansas Press, Boulder, Colorado
and
Laurence, Kansas,
161-165.
WENDT,
J. 1989. Tetradiidae
-
first evidence
of
aragonitic mineralogy
in tabulate corals. Palaontologisches Zeitschrift, 63,177-181.
WENDT,
J. 1990a. The first aragonitic rugose coral. Journal
of
Paleon-
tology, 64,335-340.
WENDT,
J. 19906. Corals and coralline sponges. In:
CARTER,
J.
G. (ed.)
Skeletal biomineralization: Patterns, Processes
and
Evolution-
ary Trends, Volume
1. Van
Nostrand Reinhold, New York,
45-66.
WEST,
R. R. &
CLARK,
G. R., II 1984. Paleobiology and biological
affinities
of
Palaeozoic chaetetids. Palaeontographica Ameri-
cana, 54,337-348.
WEYER,
D. 1972. Zur Morphologic der Rugosa (Ptercorallia). Geolo-
gic 21(6), 710-737.
WEYER,
D. 1974. Das Rugosa-Genus Antiphyllum Schindewolf, 1952
(Unternamur, Oberschlesisches Steinkohlenbecken) Casopis
pro Mineralogii a Geologii, 19,345-365.
WEYER,
D. 1980. Die
alteste Rugose Koralle Europas.
Wis-
senschaftliche Beitrdge de Martin-Luther-Universitat Halle-Wit-
tenberg, 30 (7), 51-77.
WEYER,
D. 1991. Pseudopetraia Soshkina 1951 (Anthozoa, Rugosa)
aus
dem
Unterdevon
des
Thtiringischen Schiefergebirges.
Abhandlungen und Berichte fiir Naturkunde und Vorgeschichte,
15,9-24.
WEYER,
D. 1995«. Heterocorallia aus Famenne-Cephalopodenkalken
im Rheinischen Schiefergebirge und Tafilalt. Abhandlungen
und Berichte fiir Naturkunde, 18,103-135.
WEYER,
D.
1995ft. News about Famennian Heterocorallia from
Germany and Morocco. VII International Symposium on fossil
Cnidaria and Porifera,
Madrid,
September
1995,
Abstracts, 104.
WEYER,
D.
1995c. Hyposepta and diplosepta within the septal appar-
atus of Rugosa. VII International Symposium on fossil Cnidaria
and Porifera,
Madrid,
September
1995,
Abstracts, 104-105.
WEYER, D.
&
POLYAKOVA, V.
E. 1995. Heterocorallia aus dem Oberen
Serpukhovian
des
Donez-Beckens (Unterkarbon, Arnsber-
gian; Ukraine). Abhandlungen
und
Berichte fiir Naturkunde,
18,143-159.
WOOD,
R. 1990.
Reef-building sponges. American Scientist,
78,
224-235.
WOOD,
R. 1995. The changing biology
of
reef-building. Palaios,
10,
517-529.
WOOD,
R.,
COPPER,
P. &
REITNER,
J. 1990. "Spicules" in halysitids.
Lethaia, 23,113-114.
WRIGHT,
A. J. 1969. Notes on tetracoral morphology. Journal
of
Pale-
ontology, 43,1232-1236.
WRZOLEK,
T.
1981. Coral growth
in
Oligophylloides pachythecus
Rozkowska, 1969. Acta Palaeontologica Polonica, 25,513-517.
WRZOLEK,
T. 1993. Affinities
of
the Heterocorallia. Acta Palaeonto-
logica Polonica, 38,119-120.
YABE,
H.
&
SUGIYAMA,
T. 1940. Notes
on
Heterophyllia and Hexa-
phyllia. Journal
of
the Geological Society
of
Japan, 47 (557),
81-86.
YOUNG,
G. A.
&
ELIAS,
R.
J.
1995. Latest Ordovician
to
earliest Sil-
urian colonial corals
of
the east-central United States. Bulletin
of American Paleontology, 108,1-153.
ZHURAVLEV,
A. Yu. 1996. Reef ecosystem recovery after
the
Early
Cambrian extinction. In:
HART,
M. B. (ed.) Biotic Recovery
from Mass Extinction Events. Geological Society, London,
Special Publications, 102,79-96.
ZHURAVLEV,
A. Yu.,
DEBRENNE,
F. &
LAFUSTE,
J. 1993. Early Cam-
brian microstructural diversification
of
Cnidaria. Courier
Forschungsinstitut Senckenberg, 164,365-372,
ZHURAVLEV,
A. Yu. &
WOOD,
R. 1996. Anoxia as the cause of the mid-
Early Cambrian (Botomian) extinction event. Geology,
24,
311-314.
Manuscript received: 29th January, 1997
