Download Evolutionary and ecological significance of

Blackwell Publishing LtdOxford, UKZOJZoological Journal of the Linnean Society0024-4082©
The Linnean Society of London? 2007
150?
743754
Original Article
2007
EARLIEST MULTIRADIATE STARFISHL. G. HERRINGSHAW
ET AL.
Zoological Journal of the Linnean Society, 2007, 150, 743–754. With 4 figures
Evolutionary and ecological significance of Lepidaster
grayi, the earliest multiradiate starfish
LIAM G. HERRINGSHAW1,2*, M. PAUL SMITH1 FLS and ALAN T. THOMAS1
1
Earth Sciences, School of Geography, Earth and Environmental Sciences, University of Birmingham,
Edgbaston, Birmingham B15 2TT, UK
2
Department of Geology and Petroleum Geology, College of Physical Sciences, Meston Building, King’s
College, Aberdeen AB24 3UE, UK
Received April 2005; accepted for publication November 2006
Lepidaster grayi Forbes, 1850, from the Much Wenlock Limestone Formation (Silurian: Wenlock) of England, is the
earliest species of starfish (Echinodermata: Asteroidea) to deviate from pentaradial symmetry, having 13 rays rather
than five. Based on the patterns of supernumerary ray development seen in extant multiradiate asteroids, two possible models are evaluated for the origin of the eight additional rays seen in L. grayi. In the ‘all-in-one’ model, all rays
were added in the same interradius, whereas in the ‘quadrants’ model generations of rays would have been added
in each of four interradii. The smallest specimen of L. grayi, apparently having only nine rays, suggests that the
‘quadrants’ model is most probable for the species. The presence of supernumerary rays in Silurian starfish, coupled
with the existence of numerous other Palaeozoic multiradiate taxa, shows that asteroids have been able to deviate
from pentamerism for most of their evolutionary history, and the variety of methods of supernumerary ray addition
indicates that the multiradiate condition is homoplastic. The ecological significance of multiradiate Palaeozoic starfish is reviewed: the mouth frame of L. grayi had considerably greater flexibility than that of contemporaneous fiverayed species and, in combination with its supernumerary rays, enabled L. grayi to manipulate and consume larger
food items. It is probable that Silurian starfish utilized a similar range of trophic guilds as those exploited by extant
taxa. © 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754.
ADDITIONAL KEYWORDS: Asteroidea – palaeoecology – pentamerism – pentaradial – Silurian – supernumerary rays – Wenlock.
INTRODUCTION
Despite a fossil record stretching back nearly 500 million years, the evolutionary history of starfishes
(Echinodermata: Asteroidea) is not well understood.
This is due primarily to their scarce and sporadic
occurrence as fossils. The asteroid skeleton is formed
of numerous individual ossicles connected by soft
tissues that decay swiftly after death, so that preservation requires exceptionally rapid burial. There is
consequently much controversy over the biology and
ecology of Palaeozoic taxa, with some authors (e.g.
Blake & Guensburg, 1988, 1989, 1990, 1994) interpreting them as comparable with extant forms, but
others (e.g. Gale, 1987; Donovan & Gale, 1990; Gale &
*Corresponding author. E-mail: [email protected]
Donovan, 1992) regarding them as having been ecologically restricted. The earliest multiradiate starfish
(asteroids with more than five rays) are of particular
interest in this regard, as their modern equivalents
are among the most ecologically specialized asteroids,
many being voracious predators.
In a paper describing two Mississippian species,
Blake & Guensburg (1989) examined the palaeobiological implications of supernumerary rays, while Hotchkiss (2000) provided a comprehensive review and
analysis of the possible origins of multiradiate asteroids. These studies apart, there has been very little
work on the origin and ecological significance of early
multiradiate taxa. This paper examines Lepidaster
grayi Forbes (Fig. 1), the oldest known multiradiate
asteroid, from the Wenlock (Silurian) of England. The
evolution of its 13-rayed body morphology is discussed, along with the possible impact of asteroids
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
743
744
L. G. HERRINGSHAW ET AL.
Table 1. Specimen sizes and number of rays seen in Lepidaster grayi Forbes, 1850. Where necessary, diameter calculated from radius measured along least distorted ray.
Specimen NOTNH FS03800 has only 11 rays preserved,
but ossicles of mouth region indicate at least 12, probably
13, originally present; DUDMG 606 not located, measurements based on illustrations of Spencer (1918); OUM
C00515 too poorly preserved to determine size or number of
rays. Institutional abbreviations as follows (all UK): BGS –
British Geological Survey, Keyworth; BU – Lapworth
Museum of Geology, University of Birmingham; DUDMG –
Dudley Museum & Art Gallery, Dudley; NHM – Natural
History Museum, London; NOTNH – Nottingham Museum
of Natural History, Wollaton Hall, Nottingham; OUM –
Oxford University Museum of Natural History; SM – Sedgwick Museum, University of Cambridge
m
Figure 1. Lepidaster grayi, Much Wenlock Limestone
Formation (Silurian), Dudley, England. Specimen BGS
GSM27515 (British Geological Survey, Keyworth,
England), showing oral surface; m = madreporite. Scale
bar = 10 mm.
with such morphologies
ecosystems.
on
Palaeozoic
marine
ECHINODERM SYMMETRY AND
PALAEOZOIC MULTIRADIATE ASTEROIDS
The five extant classes of the Echinodermata – Asteroidea, Crinoidea, Echinoidea, Holothuroidea and
Ophiuroidea – display an array of extraordinary
morphological characteristics. Perhaps the most striking of these is the five-fold body symmetry of adult
echinoderms. The origin and nature of echinoderm
pentamerism has been well debated (see, for example,
Nichols, 1967a, b; Stephenson, 1967, 1974, 1979;
Lawrence, 1988; Lawrence & Komatsu, 1990; Hotchkiss, 1998a, b, 2000) but remains contentious, particularly because a number of early taxa (e.g. carpoids:
Smith, 2005; helicoplacoids: Sprinkle & Wilbur, 2005)
were not pentaradial. All extant classes of echinoderm, however, show five-fold symmetry. Echinoids
and holothurians are the most consistent, always having five ambulacral grooves, although holothurians
and some echinoids may also have a superposed bilateral symmetry. Many crinoids have supernumerary
arms, but these normally occur in multiples of five due
to bifurcation of the five primary brachia. Asteroids
and, to a lesser extent, ophiuroids are the exception:
although most species are five-rayed, many deviations
are encountered, across both time and taxa. Of 34
extant asteroid families, 20 include only five-rayed
forms, nine have both five-rayed and multiradiate species, and five families are exclusively multiradiate
Specimen
Diameter (mm)
No. of rays
NHM 40215
NOTNH FS03800
DUDMG 606
NOTNH FS03795
BGS GSM27515
SM A5496
BU 673
OUM C00515
102
83
∼82
73
68
48
24
?
13
> 11
13
>8
13
6 (incomplete)
>7
>4
(Hotchkiss, 2000). As most of the multiradiate forms
have ray numbers indivisible by five, questions are
raised about the nature of pentamerism, both in starfish and across the echinoderms as a whole.
From their first appearance in the Tremadoc and
throughout the Ordovician, all species of asteroid are
essentially pentaradial (see Shackleton, 2005). Very
occasional six-rayed individuals are known, but these
are believed to be teratological (J. D. Shackleton, pers.
comm.) and it is not until the Silurian that multiradiate species are recorded. Lepidaster grayi from the
Much Wenlock Limestone Formation (Wenlock: Homerian) of England is the oldest, with eight specimens
known (for full systematics see Herringshaw, Thomas
& Smith, in press). The holotype and largest specimen
(NHM 40215, Natural History Museum, London) has
a diameter of 102 mm; it and the three other most
complete specimens all appear to have had 13 rays
(Table 1). Lepidaster has long been the only known
pre-Devonian multiradiate starfish, but a second
undescribed taxon has been discovered in the Kilmore
Siltstone (Ludlow) of Victoria, Australia (D. J. Holloway, pers. comm.). Like Lepidaster, this species has 13
rays.
Absent from the Ordovician and scarce in the
Silurian, multiradiate asterozoans (asteroids and
ophiuroids) are relatively abundant in the Devonian
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
EARLIEST MULTIRADIATE STARFISH
745
Table 2. Variation in ray number seen in Palaeozoic multiradiate asteroids
Age
Species
Rays
Permian
Carboniferous (Namurian)
Carboniferous (Visean)
Carboniferous (Tournaisian)
Devonian (Upper)
Devonian (Upper)
Devonian (Middle)
Devonian (Middle)
Devonian (Lower)
Devonian (Lower)
Silurian (Ludlow)
Silurian (Wenlock)
Ordovician
None known
Lepidasterella montanensis
Schondorfia fungosa
Lacertasterias elegans
Lepidasterella gyalum
Devonistella filiciformis
Arkonaster topororum
Michiganaster inexpectatus
Palaeosolaster gregoryi
Helianthaster rhenanus
Unnamed Australian form
Lepidaster grayi
None known
–
27–35 (Welch, 1984)
10 (Blake & Guensburg, 1989)
19 or 20 (Blake & Guensburg, 1989)
23–25 (Clarke, 1908)
11 (Woodward, 1874)
28 (Kesling, 1982)
17–25 (Kesling, 1971)
25–29 (Bartels et al., 1998)
16 (Bartels et al., 1998)
13 (D. J. Holloway, pers. comm.)
13
–
(Table 2), the most prolific source being the Hunsrückschiefer (Lower Devonian) of Germany. Bartels, Briggs
& Brassel (1998) recognized five species, three of
which – Helianthaster rhenanus Römer, H. rhenanus
var. microdiscus Lehmann and Palaeosolaster gregoryi
Stürtz – they placed in the Asteroidea, the other two –
Medusaster rhenanus Stürtz and Kentrospondylus
decadactylus Lehmann – being interpreted as ophiuroids. Current research suggests all five taxa may be
ophiuroids (A. Glass, pers. comm.), but this remains to
be confirmed. Elsewhere, Michiganaster inexpectatus
Kesling, 1971, and Arkonaster topororum Kesling,
1982, were found in the Middle Devonian of Michigan
and Ontario, respectively, while the Upper Devonian
has yielded Devonistella filiciformis (Woodward, 1874)
from Devon, and Lepidasterella gyalum (Clarke, 1908)
from New York State. Multiradiate asteroids were not
known from the post-Devonian Palaeozoic until Welch
(1984) described Lepidasterella montanensis from the
Bear Gulch Limestone (Namurian) of Montana. Subsequently, Blake & Guensburg (1989) recognized Lacertasterias elegans from the Gilmore City Formation
(Tournaisian) of Iowa and Schondorfia fungosa from
the Haney Formation (Visean) of Illinois. At present,
no multiradiate starfish are known from rocks of Permian age.
ORIGIN OF THIRTEEN RAYS IN LEPIDASTER
The five-part body symmetry of all five extant echinoderm classes indicates that it is ‘rigidly programmed
into the developmental process’ (Lawrence, 1987: 7).
Most species of starfish are five-rayed and this is
tightly regulated developmentally, such that deviation
from pentamerism is rare in five-rayed species
(Lawrence & Komatsu, 1990). However, the occasional
occurrence of non-pentameral individuals in such spe-
cies, combined with the number of multiradiate taxa
known, indicates that starfish have a capability for
symmetry variations not present in other echinoderms. Some authors (e.g. Dawkins, 1996) have used
this variability to argue that pentamerism is not a
fundamental character of crown-group echinoderms,
but a more widely accepted hypothesis is that of
Hotchkiss (1998a, b, 2000). His research into echinoderm pentamerism (Hotchkiss, 1995, 1998b) led to the
proposal of a ‘rays-as-appendages’ model (Hotchkiss,
1998b) for its origin, one of the predictions of which
was that asteroids that are multiradiate as adults
actually begin with five rays, adding supernumerary
rays later in development. The number of rays varies
both within and between multiradiate species:
Lawrence & Komatsu (1990) showed that ray number
is generally constant in taxa with 11 or fewer rays, but
variable above that. Hotchkiss’s prediction was supported by studies of the metamorphosis of extant multiradiate asteroids (see summary in Hotchkiss, 2000)
and led to the proposal of the ‘Five-Plus’ hypothesis
(Hotchkiss, 1998a, 2000), which states that five
primary rays are generated synchronously as a developmentally constrained unit and that separate,
independent pathways produce supernumerary rays.
The nature of these pathways is uncertain, but Hotchkiss (2000) suggested post-generation of rays in the
incompletely developed starfish or intercalated regeneration of rays in the ‘imago’ (sensu Hotchkiss, 2000)
as two possibilities.
The ability to regenerate lost body parts is a feature
seen in all extant classes of echinoderms. Thus, it is
possible that supernumerary rays derive from the
ability of starfish to grow new rays: essentially, multiradiate asteroids could be five-rayed forms that have
generated ‘replacement’ rays without the primary rays
having been lost. With such a small number of speci-
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
746
L. G. HERRINGSHAW ET AL.
mens, and none showing rays of differing lengths that
could be interpreted as undergoing regeneration, it
cannot be established whether Silurian starfish
such as Lepidaster could regenerate rays. However,
Silurian crinoids were capable of growing new arms
(C. Franzén, pers. comm.) and Bartels et al. (1998:
fig. 183) showed that Devonian asteroids could regenerate part, if not all, of a lost ray. As regeneration of
body-parts is a plesiomorphic feature of echinoderms,
it is logical to assume that Wenlock asteroids also had
regenerative capabilities.
Following the ‘Five-Plus’ hypothesis, L. grayi has
five primary rays and eight supernumerary rays. The
locus (or loci) of extra ray addition is not known, but
modern multiradiate asteroids exhibit two broad patterns. The first is the addition of all supernumerary
rays in the same interradius, intercalated between
primary rays C and D (Carpenter letters as assigned
by Moore & Fell, 1966). This is the more common pattern, seen in genera such as Acanthaster, Pycnopodia,
Solaster and Crossaster, although the precise sequence of ray addition is different in each genus
(Hotchkiss, 2000; Fig. 1). The second is the addition of
supernumerary rays in generations in four of the five
primary interradii, a phenomenon observed by
Sanchez (2000) in Heliaster and Labidiaster, two of
the most prolifically rayed extant genera (up to 50
rays present in the Antarctic sun-star Labidiaster
annulatus Sladen: Dearborn, Edwards & Fratt, 1991).
Based on these patterns, two possible models – the
‘all-in-one’ model and the ‘quadrants’ model – are proposed for the acquisition of 13 rays by Lepidaster
grayi.
THE ‘ALL-IN-ONE’
MODEL
The ‘all-in-one’ model (Fig. 2A), based primarily upon
the common sun-star Crossaster papposus (Linnaeus),
hypothesizes that L. grayi added all eight supernumerary rays in one interradius. C. papposus was chosen because it is morphologically similar to L. grayi,
having an R : r ratio of around two (R being the distance from the centre of the disc to the tip of the ray, r
the distance from the centre of the disc to the interradius) and a modal ray number of 13 (57% of specimens:
M’Intosh in Lawrence & Komatsu, 1990). C. papposus
is one of a group of asteroids that add all supernumerary rays in interradius C–D. In extant asteroids the
anus lies in this interradius and can be used as a
marker to observe the exact pattern of supernumerary
ray addition: in Crossaster, supernumerary rays are
all added to one side of the anus, whereas in Acanthaster they are added alternately either side of the
anus. As the anus is not visible in any available specimen of L. grayi (and may have been absent from
Palaeozoic asteroids: Jangoux, 1982), it is not possible
E A
D
B
m
II
VII
VI
V
IV
III
A
I
II
E
I
VIII
A
C
VIII
VII
B
m
D
III
VI
V
C
IV
B
Figure 2. Models for the acquisition of supernumerary
rays in Lepidaster grayi (after Hotchkiss, 2000). A, the
‘all-in-one’ model. B, the ‘quadrants’ model. Five primary
rays denoted by letters A–E; eight supernumerary rays
denoted by Roman numerals I–VIII; m = madreporite.
to assess which of these patterns, if either, is more
plausible.
All supernumerary rays in C. papposus are added
early in ontogeny, forming before the ring canal has
completely developed (Gemmill in Hotchkiss, 2000). A
similar timing occurs in species of Solaster, where all
rays have been added before the starfish is 1.5 mm in
diameter (Carson, 1988). Thus, if L. grayi followed the
‘all-in-one’ model, it is likely that the eight supernumerary rays would have been present when the starfish was no more than a few millimetres across.
THE ‘QUADRANTS’
MODEL
The ‘quadrants’ model (Fig. 2B) predicts that L. grayi
added two rays in each of four interradii. This is based
primarily on the Pacific sun-star Heliaster helianthus
(Lamarck), which reaches a modal number of 33 rays
(Tokeshi, 1991) by adding seven supernumerary rays
in each of four interradii (Sanchez, 2000). Sanchez
showed that H. helianthus adds a first generation of
four supernumerary rays, but then adds rays in generations of eight, such that it begins with five primary
rays, then successively nine, 17, 25 and 33: it does not
have a 13-rayed stage. However, this does not invalidate the model, because L. grayi could have added
either one generation of eight rays (two rays in each of
four primary interradii) or two successive generations
of four rays in these interradii.
The only interradius in which supernumerary rays
are not added is D–E, which contains the madreporite.
The madreporite is an integral part of the asteroid
water vascular system, such that the addition of
supernumerary rays in the madreporitic interradius
would necessitate a fundamental modification of the
body. This is particularly pertinent for L. grayi with a
large madreporite occupying the whole of the D–E
interradius (Fig. 1).
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
EARLIEST MULTIRADIATE STARFISH
747
In contrast to the early addition of supernumerary
rays predicted by the ‘all-in-one’ model, asteroids that
follow the ‘quadrants’ model may continue producing
rays well into adult life. The addition of rays in Heliaster, for example, is known to occur when the asteroid
has a body diameter of up to 100 mm (Clark, 1907).
The absence of shorter rays from any specimen of
L. grayi suggests that all rays were added early in
development, although this would be expected in a
taxon with only eight supernumerary rays.
APPLICATION
OF MODELS TO
LEPIDASTER
GRAYI
In addition to differences in the pattern and timing of
ray addition, there is a clear distinction between the
‘all-in-one’ model and the ‘quadrants’ model in terms
of the end-number of rays: asteroids following the
latter growth pattern commonly have far more rays
than those following the former. Of the eight known
specimens of Lepidaster grayi, four are sufficiently
complete to show the number of rays. Ranging in
diameter from 68 to 102 mm (Table 1), all four had 13
rays, indicating that this was the typical adult number. As such, it might seem more likely that L. grayi
followed the ‘all-in-one’ model, but specimen BU 673
(Fig. 3) suggests that this was not the case. BU 673 is
the smallest example of L. grayi and is imperfectly
preserved, but its diameter of approximately 25 mm
is well within the size range at which Heliaster is still
adding rays, and beyond that at which Crossaster and
Solaster have added all supernumerary rays. This
may be significant because BU 673 does not appear to
have 13 rays: Spencer (1918) thought it had between
eight and ten, and that it was an immature form that
would have added the remaining rays later in development. Although the specimen is distorted, BU 673
certainly had more than seven rays. The interradius
between rays 2 and 3 (Fig. 3) is the least disturbed
and forms an angle of not less than 40°. If this is
the true arc, it indicates that BU 673 had nine rays
(360°/40° = 9). This fits neatly into the ‘quadrants’
model, with L. grayi adding one set of four rays to
reach nine, then four more at a later stage to reach
13. Additional small specimens are required to test
this possibility.
SIGNIFICANCE OF SUPERNUMERARY RAYS
In terms of biomass, the multiradiate condition represents a drastic change from being five-rayed: if similar
ray and central disc proportions are maintained, the
volume of the body is approximately doubled. Assuming a similar metabolic rate, a 13-rayed starfish of the
same diameter as one with five rays will thus need
more food to support its greater body mass: for multiradiate taxa to have become so successful, there must
Figure 3. Lepidaster grayi, photograph and camera lucida
drawing of specimen BU 673 (Lapworth Museum of
Geology, University of Birmingham, U.). Seven visible rays
numbered arbitrarily; scale bar = 5 mm.
be some advantage associated with having supernumerary rays.
The most obvious consequence of having more than
five rays is an increase in the number of tube feet. As
tube feet are used for locomotion, attachment and
feeding, the implication is that multiradiate asteroids
could be more mobile, more able to resist being
detached from the seabed, and/or more successful
feeders. In terms of both mobility and attachment
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
748
L. G. HERRINGSHAW ET AL.
there is no convincing evidence that multiradiate taxa
are better equipped than pentaradial forms (see
Lawrence, 1988), but there is at least some disparity
in feeding behaviour.
Extant asteroids exploit an extremely wide range
of food sources (see Sloan, 1980; Jangoux, 1982). Of
the six feeding categories defined for benthic marine
invertebrates by Walker & Bambach (1974: table 6) –
suspension feeding, deposit feeding, browsing, carnivory, scavenging, and parasitism – all except parasitism are exploited by living starfish (Jangoux, 1982:
table 2). The additional food required by multiradiate
taxa could be obtained either by out-competing
pentaradial species for the same food source, or by
specializing to feed on a separate one. Within the
asteroid trophic groups, multiradiate taxa exhibit
some of the most unusual modes of feeding, suggesting that in many cases it is the latter. Members of the
order Brisingida, for example, capture plankton by
forming a crinoid-like feeding fan with their long,
thin rays (e.g. Novodinia antillensis: Emson & Young,
1994), whilst many other multiradiate specialists are
active carnivores. The crown-of-thorns starfish Acanthaster planci (Linnaeus) is notorious for its ability to
destroy coral communities (Blake, 1979); Labidiaster
annulatus catches krill, amphipods and small fish
(Dearborn et al., 1991); Crossaster papposus preys
commonly on other asteroids, inducing them to autotomize rays which it then consumes (Jangoux, 1982);
Solaster stimpsoni Verrill eats holothurians, but is
itself the chief prey of Solaster dawsoni Verrill
(Mauzey, Birkeland & Dayton, 1968); and Pycnopodia
helianthoides (Brandt) is capable of capturing large,
mobile prey such as crabs and octopuses, as well as
burrowing for infaunal bivalves (Shivji et al., 1983).
It is inferred that supernumerary rays enable multiradiate asteroids to capture and manipulate food that
five-rayed asteroids cannot, although it should be
noted also that multiradiate starfish often have
larger mouth frames (Jangoux, 1982: 143). This is
particularly true of taxa with many supernumerary
rays, where the central disc (and hence its oral surface) is larger as a consequence of accommodating the
additional rays.
At first sight, Heliaster helianthus appears to be an
exception to the rule given that it feeds on the same
prey types – mussels, gastropods and barnacles
(Tokeshi, 1991) – as many five-rayed species. However,
Vermeij (1990) noted that Heliaster is the only asteroid genus known to feed consistently on such organisms in a tropical environment. Sympatric genera feed
either on small prey that can be swallowed whole or on
immobile animals with exposed soft parts, such as corals and sponges. Again, it can be inferred that the multiradiate body shape enables Heliaster to feed in a way
different from pentaradial asteroids.
One further aspect of the multiradiate state is its
closer approximation to radial symmetry than that of
a five-rayed body plan. In simple geometrical terms,
a pentaradial starfish has an arc of 72° between each
ray, whereas a 13-rayed starfish reduces that arc to
27.7°. Like all echinoderms, asteroids have no brain
or true centre of organization and no front or back,
such that the rays act rather like individual elements of a moving colony: each ray has sensory
devices and can lead the animal in a chosen direction. In terms of both prey detection and predator
avoidance, supernumerary rays may therefore be
advantageous in giving more complete coverage of
the area around the starfish, especially as asteroid
rays, unlike those of ophiuroids, are relatively inflexible in a lateral plane.
IMPLICATIONS OF SUPERNUMERARY RAYS
IN LEPIDASTER
As shown by Dean (1999) and Shackleton (2005), there
can be difficulties in assigning asterozoans from the
Early Ordovician to the classes Asteroidea and Ophiuroidea but, by the Ashgill, the diagnostic body plan of
each class had evolved (see also Blake & Guensburg,
2005). Lepidaster is unequivocally an asteroid and its
functional morphology can be interpreted on that
basis.
Gale (1987), Donovan & Gale (1990) and Gale &
Donovan (1992) interpreted Palaeozoic asteroids as
lacking various morphological characters present in
extant starfish, including suckered tube feet and complex ray musculature, such that they were restricted
to deposit feeding, scavenging or predation of small,
immobile benthos. This conclusion was supported, at
least with regard to the earliest asteroids, by Shackleton (2005) who documented a morphological conservatism among Ordovician taxa and suggested that
they ‘probably remained dominantly deposit-feeders
throughout the period’ (Shackleton, 2005: 60). However, Blake & Guensburg (1988, 1989, 1990, 1994) disagreed, arguing that Ordovician asteroids utilized a
range of feeding methods similar to those of modern
forms, including extra-oral predation (see Blake &
Guensburg, 1994). This is the technique observed in
living asteroids of the orders Valvatida, Spinulosida
and Forcipulatida (Jangoux, 1982), whereby the
starfish extrudes its stomach over or into prey too
large to fit into the mouth, and digests the organism
externally.
The basis for much of the controversy over the functional morphology of Palaeozoic starfish is that the
phylogenetic relationships of the Asteroidea are poorly
understood. It is generally agreed that all Palaeozoic
taxa belong to the stem (Blake, 1987; Gale, 1987;
Blake & Guensburg, 1990; Donovan & Gale, 1990;
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
EARLIEST MULTIRADIATE STARFISH
Blake, Janies & Mooi, 2000; Blake & Hotchkiss, 2004;
Shackleton, 2005), with crown-group starfish known
only from the post-Permian, but no consensus has
been reached as to which are the most basal extant
asteroids. Some authors (e.g. Jangoux, 1982; Gale,
1987) have interpreted paxillosids as the most primitive living order, whilst other analyses (e.g. Blake,
1987; Knott & Wray, 2000) have contradicted this.
Hence it is unclear which morphological features are
primitive, and which derived. For example, depending
on which phylogeny is accepted, tube feet can be interpreted as either primitively suckered or primitively
non-suckered, with the tube feet of Palaeozoic asteroids interpreted accordingly. Since Jangoux (1982:
140–141) stated that all starfish with suckered tube
feet feed extra-orally and all with non-suckered tube
feet feed intra-orally, their likely morphology in Palaeozoic taxa has been used as a palaeoecological proxy.
However, a number of additional important factors
need to be considered, some of which have not been
adequately addressed by previous studies. Firstly,
despite the work of Walker & Bambach (1974) and
Jangoux (1982), no explicit definition of the trophic
guilds potentially available to Palaeozoic starfish has
ever been produced. Secondly, most extant starfish are
‘fundamentally polytrophic’ (Jangoux, 1982: 117),
regardless of whether they feed extra-orally or have
suckered tube feet. Finally, as shown in Table 3, there
are few instances in which asteroid morphology can be
associated unequivocally with specific types of feeding
behaviour, making ecological interpretations of extinct
taxa more problematical.
In the absence of close phylogenetic relationships,
morphological similarities between Palaeozoic and
extant forms are attributed to convergence: Gale
(1987) explained the resemblance of the Ordovician
genus Platanaster to living paxillosids in this way, as
did Blake & Guensburg (1994), who noted the similarity of form between another Ordovician taxon,
Protopalaeaster, and modern asteriids. Hence, as
Lepidaster grayi is not closely related to any extant
multiradiate starfish, its morphological similarity to
taxa such as Crossaster papposus must be the result
of convergence. However, Crossaster is an active carnivore capable of extra-oral feeding, behaviour that
Gale (1987) argued was not available to Palaeozoic
starfish because they did not have suckered tube
feet, sufficient mouth or ray flexibility, and were not
able to evert their stomachs. Thus, Lepidaster either
was morphologically but not functionally convergent
with living multiradiate species, or Gale (1987) was
incorrect and at least some Palaeozoic starfish were
active predators. The null hypothesis is the latter:
it must be demonstrated that Lepidaster could not
have used similar feeding techniques to extant
forms.
749
As noted above, a wide spectrum of feeding modes
exists in living starfish: Jangoux (1982) showed that
asteroids utilize five of the six trophic groups defined
for benthic invertebrates by Walker & Bambach
(1974), and refined this into eight asteroid feeding
categories (Jangoux, 1982: table 2). These were as
follows: carnivores on epifaunal macroprey (Group I),
carnivores on small sessile or colonial epifauna (Group
II), carnivores on infaunal organisms (Group III),
deposit feeders on epibenthic or substrate films
(Group IV), deposit feeders on sediment (Group V),
algivores (Group VI), suspension feeders (Group VII)
and scavengers (Group VIII). This is further modified
here (Table 3), with 11 groups recognized, including
six categories of carnivory. The food types available
and morphological features required for each feeding
category are assessed and examples given, where they
exist, of extant multiradiate taxa that utilize each
method.
Most of the trophic groups utilized by living starfish
are available to both intra-orally and extra-orally
feeding species. There are many specializations within
feeding categories (see Jangoux, 1982) but only two
categories – soft-part carnivory of sessile, protected
epifauna, and carnivory of fast-moving epifauna – are
apparently restricted entirely to asteroids with specific morphological adaptations. The critical morphological features identified by Gale (1987) are clearly of
significance to the feeding behaviour of extant taxa,
but few can be correlated directly with particular
trophic groups: most categories include asteroids
using a diversity of feeding methods. In this light, the
potential utilization of each feeding category by Lepidaster is critically examined.
DEPOSIT
FEEDING
Deposit feeding was not documented in multiradiate
asteroids by Jangoux (1982), being a trophic group
characterized by five-rayed starfish primarily living
on soft substrates in deep water (e.g. porcellanasterids). However, it is perhaps the simplest method of
feeding available to asteroids, with the tube feet used
to push nutrient-rich sediment into the mouth cavity
and, as such, would certainly have been available to a
multiradiate taxon such as Lepidaster.
SCAVENGING
No living starfishes are specialist scavengers, but
most scavenge opportunistically or in times of necessity (Jangoux, 1982). Scavenged food can be consumed
intra- or extra-orally, the size of the item being governed by flexibility of the mouth frame. Gale (1987)
argued that Palaeozoic asteroids had inflexible mouth
frames and could only have eaten food items smaller
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
750
L. G. HERRINGSHAW ET AL.
Table 3. Starfish feeding categories (after Walker & Bambach, 1974; Jangoux, 1982)
Specialized features
required
Multiradiate examples
(Jangoux, 1982)
None identified
Common
None identified
None identified
Rare – e.g. brisingids
Sponges, coral,
sea pens
None shared by all taxa
None shared by all taxa
Stomach eversion
Body ciliature
Flexibility to extend and
align rays with currents
Stomach eversion in some
taxa
Bivalves, gastropods
None shared by all taxa
Common – Luidia, Heliaster
Bivalves
Flexible rays
Common – Luidia, Pycnopodia
Echinoderms
None shared by all taxa
Common – Solaster, Crossaster
Bivalves, gastropods
Suckered tube feet, eversible
stomach
Rare – asteriids only
Crustaceans, fish
Pedicellariae, flexible rays
Very rare – e.g. Labidiaster
Feeding category
Example food types
Deposit feeding
Scavenging
Grazing/browsing
Ciliary feeding
Suspension feeding
Detritus
Carrion
Algae
Substrate films
Plankton
Carnivory – sessile
epifauna with exposed
soft tissues
Carnivory – whole-prey
consumption of sessile,
protected epifauna
Carnivory – digging for
infauna
Carnivory – slowmoving epifauna
Carnivory – soft-part
consumption of sessile,
protected epifauna
Carnivory – fast-moving
epifauna and nekton
than the buccal opening. This argument (see also Gale
& Donovan, 1992) was based on the arrangement of
the interradial marginal ossicles relative to those of
the mouth frame. In Palaeozoic taxa, where a single
axillary ossicle is present it forms part of the marginal
series and its proximal surface abuts directly against
the mouth ossicles, indicating a much-reduced flexibility of the mouth frame (Gale, 1987; Gale & Donovan,
1992). In crown group asteroids, this axillary has
moved to a position between the marginal ossicles and
the mouth frame and is interpreted as having provided greater oral flexibility.
Although the ossicular arrangement described by
Gale (1987) and Gale & Donovan (1992) is undoubtedly present in most Palaeozoic asteroids, Lepidaster
is different, having a mouth frame that is distinctly
separated from the axillary marginals. As illustrated
in Figure 4, the best-preserved specimen of Lepidaster
shows that numerous small actinal ossicles of the central disc lie between the fan-shaped array of axillary
ossicles and the mouth frame. Preservational imperfection does not enable the exact arrangement of ossicles to be determined, but it is clear that the mouth
plates do not abut directly against the axillary marginals. As such, an oral flexibility more like that of
extant asteroids is indicated; Lepidaster would have
been able to consume larger food items than sympatric
pentaradial asteroids.
Common – Luidia Acanthaster
t
mp
mp
t
t
Figure 4. Camera lucida illustration of arrangement of
oral ossicles in Lepidaster grayi (specimen BGS
GSM27515, British Geological Survey, Keyworth). Dark
grey ornament indicates mouth ossicles; medium grey
ornament indicates inferomarginal and axillary ossicles;
light grey indicates matrix; mp = mouth-angle plates,
t = tori.
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
EARLIEST MULTIRADIATE STARFISH
GRAZING/BROWSING
Few extant asteroids feed on living plants, but many
consume microscopic algae or decaying plant matter
(Jangoux, 1982). As with deposit feeding, such food
can be passed directly into the mouth using the tube
feet or consumed by eversion of the stomach, the latter
method enabling larger quantities to be eaten. Benthic
marine algae were present in the Silurian and potentially available to Lepidaster, although no multiradiate herbivores were recorded by Jangoux (1982).
CILIARY
FEEDING
Jangoux (1982: 148) defined ciliary feeding as the collection of ‘small particulate food directly from the substratum using ciliature of the ventral body surface
and/or of the everted stomach’. It is not known
whether Lepidaster possessed cilia, but no living multiradiate starfish are known to feed in this way and
the five-rayed exponents of ciliary feeding live mainly
in deep water. The lithofacies of the Much Wenlock
Limestone Formation in which Lepidaster is preserved
are indicative of moderate to shallow water depth (see
Ratcliffe & Thomas, 1999), with the holotype of Lepidaster grayi occurring in the crinoidal grainstone
lithofacies of Ratcliffe & Thomas (1999), a bioclastic
limestone interpreted as having been deposited in
high-energy conditions above fair-weather wave base.
SUSPENSION
FEEDING
Suspension feeding in asteroids was considered by
Jangoux (1982) to be a specialization of ciliary feeding,
where particulate material is collected from the water
column rather than the substrate. Multiradiate asteroids of the order Brisingida are specialized suspension
feeders (Emson & Young, 1994) but their crinoid-like
body morphology of tiny central disc and long, slender
rays is quite unlike that of Lepidaster. However, asteroids with a variety of body shapes are known to use
suspension feeding (see Jangoux, 1982: 149–151) with
the main requirement being flexible rays that can be
aligned with nutrient-bearing currents.
CARNIVORY
In previous discussions of Palaeozoic asteroid palaeoecology, the most contentious subject has been that of
carnivory. Prey available to marine carnivores can be
categorized into five types: sessile prey with exposed
soft parts, sessile prey with protected soft parts, slowmoving prey, fast-moving prey and buried prey. With
regard to carnivorous asteroids, each of these is
defined as a feeding category (Table 3), but with two
methods of predation defined for sessile, protected
prey: whole prey consumption and consumption of soft
751
parts only. The latter technique involves the opening
of the prey’s protective shell and is seen only in asteriids (Jangoux, 1982), which have suckered tube feet
and an eversible stomach.
The flexible mouth frame and supernumerary rays
of Lepidaster would have enabled it to manipulate and
consume relatively large food items. An abundance of
sessile benthos was present in the Much Wenlock
Limestone Formation, many of which would have been
potential prey items (e.g. brachiopods, gastropods)
either by intra- or extra-oral feeding. Sessile epifauna
with exposed soft tissues, such as corals, sponges and
bryozoans, might also have been preyed on: extant
multiradiate starfish are capable of feeding both intraand extra-orally on such organisms.
All extant echinoderms have tube feet: extant phylogenetic bracketing (Witmer, 1992) allows us to infer
their presence in Palaeozoic asteroids. Based on his
phylogeny of crown group asteroids, Gale (1987) interpreted the tube feet of Palaeozoic starfish as not having been suckered, and that they could neither have
gripped prey in the way that extant asteriids do when
attacking bivalves or operculate gastropods, nor have
lived on hard substrates. Tube feet apparently lacking
suckers have recently been described in an asterozoan
from the Herefordshire Lagerstätte (Wenlock) of
England (Sutton et al., 2005) but the tube foot morphology of Lepidaster is unknown. However, even its
tube feet were not suckered, this would not preclude
Lepidaster from being a carnivore. Asteroids of the
families Luidiidae and Astropectinidae lack suckered
tube feet (Vickery & McClintock, 2000), yet the genus
Luidia is ‘rather strictly carnivorous’ (Jangoux, 1982:
118) and includes multiradiate species that are active
predators of large, mobile prey. Furthermore, work by
Thomas & Hermans (1984, 1985), Flammang, Demeulenaere & Jangoux (1994) and Flammang (1995) has
revealed that, in at least some starfish, the tube feet
use chemical adhesion rather than suction to gain
hold of prey items or substrates, suggesting that suckers may not be essential to extra-oral feeding.
Of the infaunal organisms preyed on by extant starfish, primarily bivalves, few existed in the Wenlock,
making the exploitation of that feeding category by
Lepidaster relatively improbable. Carnivory of fastmoving epifauna, such as crustaceans and fish, is
extremely uncommon in asteroids (Jangoux, 1982)
and, given the abundance of sessile or slow-moving
benthos in the Much Wenlock Limestone Formation,
also unlikely to have been utilized by Lepidaster.
CONCLUSIONS
The palaeoecology of Palaeozoic asteroids has proved a
contentious topic in the last two decades, with the
Blake and Gale schools of thought providing different
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
752
L. G. HERRINGSHAW ET AL.
interpretations. However, it is apparent that, regardless of which hypothesis is preferred, most of the feeding categories utilized by living starfish were available
to Palaeozoic forms: other than asteriid-type predation, none is dependent directly on both suckered tube
feet and an eversible stomach. Additionally, even if
it were functionally possible, asteriid-type predation
would probably have been extremely rare in Palaeozoic taxa, given the abundance of small, sessile or
slow-moving benthic organisms available as prey.
However, the greater mouth-part flexibility apparent
in Lepidaster suggests that extra-oral predation might
have been possible in the earliest multiradiate
starfish.
The morphological conservatism of Ordovician
asteroids was interpreted by Shackleton (2005) as
indicating their ecological restriction, but the presence
of Lepidaster in a morphologically varied fauna from
the Wenlock of England (see Herringshaw et al., in
press) suggests that Silurian starfish were more ecologically diverse. Too few data are available presently
to determine whether the asteroids of the Much Wenlock Limestone Formation were an endemic fauna or
representative of a global diversification and, given
the poor preservation potential of asteroids, it is also
possible that more diverse morphologies, including
multiradiate taxa, existed in the Ordovician.
Without an accepted phylogeny of the Asteroidea, it
is difficult to evaluate completely the degree of convergence between Palaeozoic multiradiate asteroids such
as Lepidaster and morphologically similar extant taxa
such as Crossaster. Further investigation is required,
both of post-Ordovician members of the stem Asteroidea and of the clade as a whole, to determine which
characters are homologous and which convergent.
However, Lepidaster grayi is the earliest known asteroid with supernumerary rays, a large central disc and,
most crucially in terms of functional morphology, a
mouth frame dissociated from the marginal ossicles.
This greater oral flexibility is interpreted as having
enabled Lepidaster to consume larger quantities or
items of food, with the extra rays assisting in its
manipulation. The feeding categories it exploited will
probably never be known, but there are no convincing
reasons for regarding Lepidaster as ecologically
restricted: as with modern multiradiate starfish, a
diverse range of feeding behaviour, including active
predation, was potentially available.
ACKNOWLEDGEMENTS
F. H. C. Hotchkiss, J. M. Lawrence, D. B. Blake, S. K.
Donovan and an anonymous reviewer are thanked for
useful comments, discussions and suggestions. L.G.H.
acknowledges a School of Earth Sciences, University
of Birmingham, Research Studentship.
REFERENCES
Bartels C, Briggs DEG, Brassel G. 1998. The fossils of the
Hunsrück Slate – marine life in the Devonian. Cambridge:
Cambridge University Press.
Blake DB. 1979. The affinities and origins of the crown-ofthorns sea star Acanthaster Gervais. Journal of Natural History 13: 303–314.
Blake DB. 1987. A classification and phylogeny of postPaleozoic sea stars (Asteroidea: Echinodermata). Journal of
Natural History 21: 481–528.
Blake DB, Guensburg TE. 1988. The water vascular system
and functional morphology of Paleozoic asteroids. Lethaia
21: 189–206.
Blake DB, Guensburg TE. 1989. Two new multiarmed Paleozoic (Mississippian) asteroids (Echinodermata) and some
paleobiologic implications. Journal of Paleontology 63: 331–
340.
Blake DB, Guensburg TE. 1990. Predatory asteroids and the
fate of brachiopods – a comment. Lethaia 23: 429–430.
Blake DB, Guensburg TE. 1994. Predation by the Ordovician asteroid Promopalaeaster on a pelecypod. Lethaia 27:
235–239.
Blake DB, Guensburg TE. 2005. Implications of a new early
Ordovician asteroid (Echinodermata) for the phylogeny of
asterozoans. Journal of Paleontology 79: 395–399.
Blake DB, Hotchkiss FHC. 2004. Recognition of the asteroid
(Echinodermata) crown group: implications of the ventral
skeleton. Journal of Paleontology 78: 359–370.
Blake DB, Janies DA, Mooi R. 2000. Evolution of starfishes:
morphology, molecules, development, and paleobiology.
Introduction to the symposium. American Zoologist 40: 311–
315.
Carson SF. 1988. Development and metamorphosis of three
species of the asteroid Solaster from the northeast Pacific.
In: Burke RD, Mladenov PV, Lambert P, Parsley RL, eds.
Echinoderm biology. Rotterdam: Balkema, 791.
Clark HL. 1907. The starfishes of the genus Heliaster. Bulletin of the Museum of Comparative Zoology 51: 25–76.
Clarke JM. 1908. A Devonic brittle-star. Bulletin of the New
York State Museum 121: 61–64.
Dawkins R. 1996. Climbing Mount Improbable. London:
Penguin.
Dean J. 1999. What makes an ophiuroid? A morphological
study of the problematic Ordovician stelleroid Stenaster and
the palaeobiology of the earliest asteroids and ophiuroids.
Zoological Journal of the Linnean Society 126: 225–250.
Dearborn JHK, Edwards C, Fratt DB. 1991. Diet, feeding
behaviour and surface morphology of the multi-armed
Antarctic sea star Labidiaster annulatus (Echinodermata,
Asteroidea). Marine Ecology – Progress Series 77: 65–84.
Donovan SK, Gale AS. 1990. Predatory asteroids and the
decline of the articulate brachiopods. Lethaia 23: 77–86.
Emson RH, Young CM. 1994. Feeding mechanism of the brisingid starfish Novodinia antillensis. Marine Biology 118:
433–442.
Flammang P. 1995. Fine structure of the podia in three species of paxillosid asteroids of the genus Luidia (Echinodermata). Belgian Journal of Zoology 125: 125–134.
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
EARLIEST MULTIRADIATE STARFISH
Flammang P, Demeulenaere S, Jangoux M. 1994. The role
of podial secretions in adhesion in two species of sea-stars
(Echinodermata). Biology Bulletin 187: 35–47.
Gale AS. 1987. Phylogeny and classification of the Asteroidea. Zoological Journal of the Linnean Society 89: 107–
132.
Gale AS, Donovan SK. 1992. Predatory asteroids and articulate brachiopods – a reply. Lethaia 25: 346–348.
Herringshaw LG, Thomas AT, Smith MP. in press. Starfish diversity in the Wenlock of England. Palaeontology in
press.
Hotchkiss FHC. 1995. Lovén’s Law and adult ray homologies
in echinoids, ophiuroids, edrioasteroids, and an ophiocistioid
(Echinodermata: Eleutherozoa). Proceedings of the Biological Society of Washington 108: 401–435.
Hotchkiss FHC. 1998a. Discussion on pentamerism: the fivepart pattern of Stromatocystites, Asterozoa and Echinozoa.
In: Mooi R, Telford M, eds. Echinoderms. Rotterdam:
Balkema, 37–42.
Hotchkiss FHC. 1998b. A ‘rays-as-appendages’ model for the
origin of pentamerism in echinoderms. Paleobiology 24: 200–
214.
Hotchkiss FHC. 2000. On the number of rays in starfish.
American Zoologist 40: 340–354.
Jangoux M. 1982. Food and feeding mechanisms: Asteroidea.
In: Jangoux M, Lawrence JM, eds. Echinoderm nutrition.
Rotterdam: Balkema, 117–159.
Kesling RV. 1971. Michiganaster inexpectatus, a new manyarmed starfish from the Middle Devonian Rogers City Limestone of Michigan. Contributions from the Museum of Paleontology, University of Michigan 23: 247–262.
Kesling RV. 1982. Arkonaster, a new multi-armed starfish
from the Middle Devonian Arkona Shale of Ontario. Contributions from the Museum of Paleontology, University of
Michigan 26: 83–115.
Knott KE, Wray GA. 2000. Controversy and consensus in
asteroid systematics: new insights to ordinal and familial
relationships. American Zoologist 40: 382–392.
Lawrence JM. 1987. A functional biology of echinoderms.
Baltimore: Johns Hopkins University Press.
Lawrence JM. 1988. Functional consequences of the multiarmed condition in asteroids. In: Burke RD, Mladenov PV,
Lambert Parsley RL, eds. Echinoderm biology. Rotterdam:
Balkema, 597–602.
Lawrence JM, Komatsu M. 1990. Mode of arm development
in multiarmed species of asteroids. In: De Ridder C, Dubois
P, Lahaye M-C, Jangoux M, eds. Echinoderm research.
Rotterdam: Balkema, 269–275.
Mauzey KP, Birkeland C, Dayton PK. 1968. Feeding behavior of asteroids and escape responses of their prey in the
Puget Sound region. Ecology 49: 603–619.
Moore RC, Fell HB. 1966. Homology of Echinozoan rays. In:
Durham JW et al. Echinodermata 3, Asterozoa–Echinozoa.
Part U of Moore RC, ed. Treatise on invertebrate paleontology. New York: Geological Society of America, and Lawrence:
University of Kansas, U119–U131.
Nichols D. 1967a. Pentamerism and the calcite skeleton in
echinoderms. Nature 215: 665–666.
753
Nichols D. 1967b. The rule of five in animals. New Scientist
35: 546–549.
Ratcliffe KT, Thomas AT. 1999. Carbonate depositional
environments in the Wenlock of England and Wales. Geological Magazine 136: 189–204.
Sanchez P. 2000. The sequence of origin of the postmetamorphic rays in Heliaster and Labidiaster (Echinodermata:
Asteroidea). Revista Chilena de Historia Natural 73: 573–
578.
Shackleton JD. 2005. Skeletal homologies, phylogeny and
classification of the earliest asterozoan echinoderms.
Journal of Systematic Palaeontology 3: 29–114.
Shivji M, Parker D, Hartwick B, Smith MJ, Sloan NA.
1983. Feeding and distribution study of the Sunflower Sea
Star Pycnopodia helianthoides (Brandt, 1835). Pacific Science 37: 133–140.
Sloan NA. 1980. Aspects of the feeding biology of asteroids.
Oceanography and Marine Biology Annual Review 18: 57–
124.
Smith AB. 2005. The pre-radial history of echinoderms.
Geological Journal 40: 255–280.
Spencer WK. 1918. A monograph of the British Palaeozoic
Asterozoa. Monograph of the Palaeontographical Society,
London (Part III for 1916: 109–168).
Sprinkle J, Wilbur BC. 2005. Deconstructing helicoplacoids:
reinterpreting the most enigmatic Cambrian echinoderms.
Geological Journal 40: 281–293.
Stephenson DG. 1967. Pentameral symmetry in echinoderms. Nature 216: 994.
Stephenson DG. 1974. Pentamerism and the ancestral echinoderm. Nature 250: 82–83.
Stephenson DG. 1979. The trimerous stage in echinoderm
evolution: an unnecessary hypothesis. Journal of Paleontology 53: 44–48.
Sutton MD, Briggs DEG, Siveter David J, Siveter Derek
J, Gladwell DJ. 2005. A starfish with three-dimensionally
preserved soft parts from the Silurian of England. Proceedings of the Royal Society B – Biological Sciences 272: 1001–
1006.
Thomas LA, Hermans CO. 1984. Suction is not the answer –
evidence for the duo-gland adhesive system in starfish tube
feet. American Zoologist 24: 15.
Thomas LA, Hermans CO. 1985. Adhesive interactions
between the tube feet of a starfish, Leptasterias hexactis, and
substrata. Biological Bulletin 169: 675–688.
Tokeshi M. 1991. Extraoral and intraoral feeding: flexible foraging tactics in the South American sun-star,
Heliaster helianthus. Journal of Zoology 225: 439–
447.
Vermeij GJ. 1990. Asteroids and articulates: is there a causal
link? Lethaia 23: 431–432.
Vickery MS, McClintock JB. 2000. Comparative
morphology of tube feet among the Asteroidea: phylogenetic implications. American Zoologist 40: 355–
364.
Walker KR, Bambach RK. 1974. Feeding by benthic invertebrates: classification and terminology for paleoecological
analysis. Lethaia 7: 67–78.
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754
754
L. G. HERRINGSHAW ET AL.
Welch JR. 1984. The asteroid Lepidasterella montanensis n.
sp., from the upper Mississippian Bear Gulch Limestone of
Montana. Journal of Paleontology 58: 843–851.
Witmer L. 1992. The extant phylogenetic bracket and the
importance of reconstructing soft tissues in fossils. In: Thom-
son JJ, ed. Functional morphology in vertebrate palaeontology. Cambridge: Cambridge University Press, 19–32.
Woodward H. 1874. Description of a new species of starfish
from the Devonian of Great Inglebourne, Harberton, South
Devon. Geological Magazine decade II 1: 6–10.
© 2007 The Linnean Society of London, Zoological Journal of the Linnean Society, 2007, 150, 743–754