Thursday, 12 July 2012

The early evolution of Echinoderms

Echinoderms are one of the majour phyla that exists today and are easily distinguishable from other organisms based on their highly derived morphology. Key features including pentaradial symmetry, a calcitic endoskeleton of stereom (each plate in the theca consists of a single calcite crystal and displays the characteristic cleavage of calcite), and a hydrovascular system controlling the tube feet of these organisms.
Above is a lovely illustration of various examples of the Cystoidea. The bottom left and right creatures are Edrioasteroids. Drawing by Ernst Haeckel. Image taken from Wikipedia.

There are two main groups of living Echinoderms; the Pelmatozoa; which include only Crinoids today, have a dorsal stem (opposite the mouth) extending from the theca (main body) of the organism. This usually acts as a holdfast. Extensions of the water vascular system (arms in crinoids) arise from around the mouth and gather food. The terms dorsal and ventral are fairly arbitrary decisions when talking about echinoderms and there is little clarity in how their axis correspond to those of other phyla; dorsal seems to mean opposite the mouth, ventral corresponding to the mouth surface; but the terms oral and aboral are also used. Extinct classes of Pelmatozoans are generally lumped in the paraphyletic clade; Cystoidea (Cambrian to Permian). Cystoids are basically any Pelmatozoan that is not a member of the surviving superclass; Crinozoa, which almost certainly arose from them. They possess specialized respiratory structures in their theca, these may be pores in between plates in the theca (for example the epispires of Eocrinoids), or external structures on the surface of the thecal plates (rhomboid canal systems of the Rhombifera), pores whithin the plates themselves (diplopores and humatipores of the Diploporita) or foldings of the theca (Blastoidea). Many Cystoids possessed extensions of the ambulacral series (plates associated with the water vascular system and tube feet, usually involved in gathering food and passing it to the central mouth) called brachioles which are derived from the cover plates of the ambulacrum and contain extensions of the water vascular system. The arms of Crinoids may appear similar, but these are direct extensions of the theca and contain extensions of the main body cavity (additionally the gonads, which are within the theca of Cystoids are present in the arms of Crinoids) as well as just the water vascular system. They are also uniserial as opposed to biserial, with the ambulacral system held externally on the surface of the brachials (arm plates).
The above image shows representatives of the 5 Eleutherozoan classes alive today.

The second group; the Eleutherozoa is much more diverse today and includes animals in which the oral surface is either very large and encompasses most of the surface of the organism (Echinoids and Holothuroidaea) or about half of the surface of the organism (e.g. Asteroids and Ophiuroids). Eleutherozoans are generally free living creatures without a stem. They include the extinct Edrioasteroidea which had the mouth and ambulacra facing upwards as in Pelmatozoans. Stellate forms with the ambulacra and mouth facing downwards include carnivorous Asteroids (starfish) and Ophiuroids (brittle stars). The Echinozoa includes groups in which the ambulacra extend over most of the body surface and include today the Echinoids (sea urchins) and the Holothuroidea (sea cucumbers).

However the above groups are by no means the entire diversity of Echinoderms that has ever existed. All modern Echinoderms are crown-group echinoderms with a pentaradial symmetry. In this post I will cover the controversial and enigmatic stem group Echinoderms. I have decided to ignore the Helicoplacoids for now. These early Cambrian Echinoderms seem to show indications of a radial symmetry and have unambiguous ambulacra, indicating that they are the most crownward of the Homalozoa.

Echinoderms belong to a large group or Bilaterian animals called the Deuterostomes. In these animals the blastopore (the first opening to appear during development) of the embryo/ larva becomes the anus and a mouth is secondarily developed on the other end of the animal. This contrasts to Protostomes (another majour bilaterian group) in which the blastopore becomes the mouth. In reality it is less clear cut than this, but that is ignored here.
Above is a simplified cladogram of the Deuterostomes as they are currently understood. From (

Deuterostomes also include us; the Chordates, as well as the Hemichordates (acorn worms, Pterobranchs and the Graptolithina). Echinoderms have been placed in a grouping with the Hemichordates called the Ambulacraria based on molecular evidence. This is important because it has been suggested that the asymmetry seen in Echinoderms is homologous to the asymmetry seen in Cephalochordate and tunicate development (e.g. in the lanclet larva the mouth initially appears on the left hand side of the animal) and that this implies a sister taxon relationship between the Echinoderms and Chordates forming the Dexiothetica (Jefferies 1979); to the exclusion of Hemichordates which show no asymmetry in development at all. However the molecular evidence in support of the Ambulacraria and therefore a sister taxon relatiosip between Echinoderms and symmetrical Hemichordates implies that asymmetry evolved independantly in Cephalochordates, tunicates and Echinoderms.

Above are various representatives of the Vetulicolia from the Cengjiang biotia in China. Image taken from Wikipedia. Artist: Santon.F. Fink.

Perhaps the most basal Deuterostomes we know of were the Vetulicolians from the Cengjiang fauna of early Cambrian age. These animals had a round body which appears to have 5 pairs of gill openings on either side, a circular anterior mouth, a segmented tail perhaps involved in swimming or crawling on the substrate and a terminal anus at the tip of the tail. The pharyngeal cavity and gill slits are shared with Chordates and Hemichordates, suggesting that this feature was present in the ancestral Deuterostome and that Echinoderms origionally had them. Unlike the condition in chordates; no notochord is present. The notochord is also lacking in Hemichordates (older texts may say otherwise, but this is now considered a non homologous structure called the stomochord) and Echinoderms, suggesting it is not basal in Deuterostomes and was not present in basal Echinoderms.

Above is a diagram showing an Enteropneust Hemichordate to the left, and a Pterobranch Hemichordate to the right. Pterobranchs do have a single pair of gill slits (not labelled).

Features shared within the Ambulacraria include a coelom (body cavity) split into 3 pairs down the length of the animal. In Echinoderms this condition undergoes significant torsion in adulthood, but can be recognised in larval form. In the Hemichordates the tricoelomate body plan is visible in adulthood. Even more amazing is the apparent homology between the coelomic cavity in the feeding arms of Pterobranchs and the water vascular system of Echinoderms; which are both derived from the second coelomic cavity. In symmetrical Pterobranchs both left and right cavities form arms, but in echinoderms only the left hydrocoel goes on to form the water vascular system, an example of asymmetry in Echinoderm development.
This is the Dipleurula larva; a hypothetical basal Eleutherozoan larva. Note the tricoelomic body plan. Image taken from Wikipedia.

Next we come to stem echinoderms; and this is where it gets a little trickier. In 2004 Shu, Conway Morris, Han, Zang and Liu described another unusual group of early Cambrian Chengjiang critters. These were the Vetulocystids. They appear to be stem Echinoderms, but if so are very basal; lacking a stereom. They have an egg like theca and a thick, possibly segmented tail; a bipartite body plan shared with Vetulicolians and later stem group Echinoderms; the Homalozoa (see below). The Vetulocystids had 2 plated orofices, one at the end (anterior?) opposite the tail and another nearer the tail, but still on the theca as in Echinoderms. The anterior opening appears to be a mouth and the more posterior was probably an anus. The radially arranged plates around the anus and mouth to me superficially resemble the periproct and peristome of certain Homalozoans. There is also what appears to be a respiratory organ to the right of the anus, consisting of a series of ridges and grooves. It seems that the tail like organ of the Vetulocystids may be homologous to the stele (tail like organ in certain Homalozoa). However this supposition may be incorrect.

Above are shown 2 specimens and interpretative drawings of Vetulocystis from the Chengjiang biotia in China. Image taken from

The Homalozoa form an evolutionary grade of Echinoderms which have an endoskeleton of stereom and lack radial symmetry; most show asymmetry. The forms described below all belong to this evolutionary grade. These animals are also known in some texts as Carpoids. There have been suggestions (especially by Jeffries) that these animals are ancestral to chordates. This is known as the calcichordate hypothesis. It is suggested that the stereom was basal to all Dexiotheticans and was independantly lost in the Cephalochordates, Urochordates and Craniates. I agree with the idea held by some researchers that it is more parsinomous to think that the stereom evolved once in Echinoderms and was not subsequently lost. It is still possible that Chordates could have derived from Homalozoans by loosing the stereom only once, but because Hemichordates are a more closely related to the Echinoderms, the stereom would have to be lost in them as well. I consider the Homalozoa stem Echinoderms, as this requires no subsequent loss of the stereom and is therefore more parsinomous.

Recently a new paper was published in the wonderful open access journal PLoS one ( by Zamora, Rahman and Smith 2012) describing a new taxon; Ctenoimbricata spinosa from the Mid Cambrian of southern Europe. This unassuming little critter meant I am having a real headache coming to tearms with basal Echinoderm evolution. The Vetulocystids appear to be asymmetrical, possessing only one lenticular respiratory organ to the side of the anus. They also lack a stereom; instead, having a flexible unmineralized theca. This implies that the stereom evolved after asymmetry in Echinoderms. However Ctenoimbricata; a possible basal Echinoderm was bilaterally symmetrical and posessed a theca of mineralised stereom plates! This implies that the stereom developed before asymmetry. In the paper this issue wasn't tackled, which left me a little perplexed. There is always the possibility that the Vetulocystids were not stem Echinoderms at all. This paradox may indicate that this is indeed the case.
Above is a diagram showing plate homologies in basal Homalozoans from the recent paper by Zamora, Rahman and Smith 2012. Image taken from

Even more interesting is the fact that Ctenoimbricata lacked any sign of a stele or other tail like organ found in more advanced Homalozoa such as the Stylophora, Cincta and Soluta, as well as more basal Deuterostomes like the Vetulicolians and the Vetulocystids. If Ctenoimbricata is a basal Echinoderm then the bipartite body plan seen in Vetulicolians and Vetulocystids was lost in the base of the Echinoderm stem, leaving only the round theca without a tail. The stereom then evolved, and tail-less forms such as Ctenoimbricata and the Ctenocystoidea branched off the Echinoderm stem. Later, as suggested by Zambora and co workers, a superficially similar tail called the stele evolved from the posterior marginal plates on the theca of Ctenoimbricata or some related form, independantly from that seen in Vetulicolians and Vetulocystids.

Perhaps the asymmetry was similarly lost prior to the divergence of Ctenoimbricata, and then regained in later Homalozoans? The theca of Ctenoimbricata does not seem perfectly symmetrical in the diagrams provided by Zambora and co workers, so perhaps this aparrent bilateral symmetry is secondarily derived. Of course if Vetulocystids were not stem Echinoderms this problem wouldn't exist and the slight asymmetry in Ctenoimbricata was probably due to a trend towards asymmetry rather than away from it.

Ctenoimbricata had an outer ring of marginal plates that were arranged symmetrically across the central body axis. Dorsally and ventrally the surface was covered in smaller central plates. The anus of Ctenoimbricata is hypothesized (the likely area of the anus is not preserved) to have been placed on the posterior part of the dorsal area of central plates on the central body axis as in the similar basal Ctenocystoid; Courtessolea. There is a single anterior opening surrounded ventrally by blade like ctenoid plates and 4 modified anterior marginals and dorsally by imbricate, flat ctenoid plates. Dorsally there is a large suroral plate.

In the Ctenocystoidea (Mid Cambrian to Early Ordovician?) the dorsal imbricate ctenoid plates are reduced and tessolate and the ventral ctenoid plates surround the anterior opening. Advanced forms such as Ctenocystis incorporate the anus into the marginal plate series and a second layer of marginal plates is present, dorsally bounding the anus. The arrangement of marginal plates in this form has become asymmetrical.
Above is a diagram of the Cinctan Trochocystites in dorsal (left), ventral (right) and anterior (bottom). Image taken from

In the Cincta (Mid Cambrian) it appears that the posterior marginals have extended to form the beginnings of a stele. The Cincta are probably the most basal Homalozoans to have a stele since in these animals it is most continuous with the marginal plate series. The Cincta appear to have evolved the stele as either a transport or anchorage mechanism. It seems to have been much less flexible than that of other Homalozoans. 
As in Ctenoimbricata, the marginal plate series is uniserial, and both dorsally and ventrally is a pavement of smaller central plates; loosely arranged on the dorsal surface, but tesselated ventrally. Threre is an anterior opening which is dorsally covered by an opercular plate which could open outwards. This opening probably functioned as a water outlet. Another opening is found in the marginal series to the right of the operculum. This is interpreted here as a mouth because running from it are 2 grooves along the front margin of the theca which may represent ambulacra (Ubaghs 1975). To the left of the operculum is an opening in amongst the antero-dorsal central plates. This forms a pyramidal shape indicative of a preiproct (plates surrounding the anus). It has been suggested that Cinctans acted as "basket filters"; drawing water into their mouths and passing it through a pharyngeal cavity lined with gill openings as in tunicates. Filtered water would then pass out of the body through the atrial opening at the front when the operculum was opened. The anterior position of the anus indicates that these animals had a u-shaped gut as in tunicates today; which also practice this feeding mechanism.

The fact that Ctenocystoids and Ctenoimbricata lacked 2 anterior openings suggests that the single anterior opening may have contained both the atrial opening for expelling water and the mouth opening. The homology of the middle anterior opening of the Ctenocystoids and that of Cinctans is supported by the presence of a median, anterodorsal suroral plate which may be homologous to the Cinctan opercular plate.

Interestingly the Cincta seem to show what some believe to be the possible beginnings of a water vascular system. Extending on the anterior surface of the marginal plate series are a pair of grooves. One groove extends to the left of the mouth below the operculum, the other extends to the right of the mouth. These grooves may have housed the water vascular system (derived from the left mesocoel only), therefore representing ambulacra; and seem to have had a series of small plates associated with them that may represent cover plates. Smith (2005) disagrees and suggests that these grooves housed both the left and right "lophophore tentacles". In other words; Cinctans still retained the left and right 2nd coelomic cavities (mesocoels) developed as feeding arms (lophophores) in Pterobranchs and had yet to loose the 2nd right hand cavity and develop the left one into a water vascular system as in modern Echinoderms. This is supported by the fact that the right hand groove is usually smaller than the left hand groove and is even lost in some members of the group, prehaps foreshadowing the later loss of the right mesocoel in more advanced Echinoderms.

Now we come to the Stylophora (Mid Cambrian to Mid Devonian); consisting of 2 groups; the Cornuta and the Mitrata. Cornutes are highly asymmetrical, with a generally boot shaped theca, bounded by large marginal plates and smaller central plates as in the Cincta. Mitrates show much more symmetrical theca and have increased the size of their marginal plates to encompass the whole/ most of the theca. Stylophora have a much more complex stele than that of the Cincta. The short, proximal section or proxistele was quadriserial and consisted of overlapping plates that would have made it flexible. The median section contained the Stylocone, an enlarged and specialised structure unique to Stylophorans. There is a uniserial series of basal plates and a biserial series of cover plates in the dististele (the most distal part of the stele). The dististele was long and inflexible in Cornutes according to Smith (2005), based on observations of the articulations in Cornute basal elements in the dististele, but a specimen illustrated in Ubaghs (1988) indicates that the stele was able to coil up in Cothurnocystis bifida? A ball and socket articulations in the Mitrate dististele indicates that it was also flexible and able to coil upwards as some specimens show.

There are 2 main interpretations of the stele in Stylophorans. One hypothesis (e.g.Ubaghs 1961, 1968, 1975, 1981) is that this organ was not a locomotory muscular stele at the posterior end, but rather a feeding organ (aulacophore) containing the ambulacrum at the anterior end with the mouth somewhere at its base, with the cover plates capable of opening. Their ability to open is suggested certainly in some Cornutes by their open position in situ. But problems arise when trying to interpret the stele in Mitrates as an aulacophore, since the cover plates of the dististele of some Mitrates are fused together along the midline, making it impossible for them to open. The cover plates also overlap in Mitrates in which they are not fused; making opening them difficult. It seems to me more probable that the Stylophoran appendage is not an aulacophore, but a locomotory stele, homologous to the similar organs found all through out the Homalozoa. This interpretation is supported by the prescence of trail marks associated with the Mitrate, Rhenocystis, from the Hunstruck slate, but these are controversial and may represent marks due to preparation. The open position of the cover plates in some Cornutes may represent a product of post mortem decay and displacement due to loss of support from soft tissues. You know when you have an enigma when palaeontologists argue about which end is the front!
Above is a diagram of the dorsal aspect of Cothurnocystis; a Cornute. On the top left of the theca can be seen a row of possible gill slits (cothurnopores). The mouth is believed to be on the bottom right hand side of the theca in this image. To the left can be seen the tripartite stele. Image taken from Wikipedia.

In Cornutes the mouth is represented by a large opening at the anterior edge of the theca, opposite the stele. The theca consists of a series of marginal plates, confining a region of smaller central plates within (except in Ceratocystis, a primitive Cornute with a theca consisting of roughly equal, medium sized plates). Dorsally in the central region are a series of openings arranged in a linear fasion, with a superficial resembelance to gill slits; which they may well be. These may be simple sutural openings as in Ceratocystis, or more complex structures such as lamellipores or cothurnopores. These pores basically functioned as outlet valves, probably representing atrial openings or gill slits through which water, that had entered through the mouth, would exit the body after being filtered through internal gills (if the opening is atrial). Various posterior openings could well represent the anal opening.
Above; the mitrate Renocystis in what is generally believed to be dorsal (left) and ventral (right). Image taken from

Evidence of paired internal gills can be found in the Mitrates (Jaekelocarpus). These are much more bilaterally symmetrical than the Cornutes (but there is always some asymmetry in plate arrangement). There is no convincing anal or atrial opening in most Mitrates (accorring to Smith 2005), however convincing paired openings in the theca of Jaekelocarpus indicate that Mitrates may well have had paired atrial openings on the theca. It has been suggested that plates of the theca could have opened apart to reveal the atrial opening, which, given their clarity in Jaekelocarpus seems to make this a plausible idea. Assuming most Mitrates lacked paired atrial openings and only had a single large anterior opening; it is suggested that like the Ctenocystoidea; their single anterior opening may well have been a combined mouth and atrial opening, but an anal outlet as well. Interestingly I observed from a diagram of Mitrocystites mitra (a Mitrate of course!) from Smith (2005) that the anterior opening possessed ventrally a series of small elongate plates pointing forewards on its margin like the ventral ctenoid plates in Ctenocystoids. Could this be homologous to the ctenoid plates and indicate that Mitrates are descended from a Ctenocystoid- like organism? I think it is quite possible; especially given the fact that Mitrates are the most symmetrical Stylophorans (which is another basal Ctenocystoid feature). If the recent discovery of Ctenoimbricata does indeed indicate that the basal Echinoderm body plan was Ctenocystoid- like and symmetrical, then that may make Mitrates the most basal Stylophorans (contra Jefferies and Pokop 1972 (cited from Ubaghs 1975) and contra Ruta 2003 who finds Cornutes as paraphyletic and ancestral to Mitrates in his supertree analysis). This idea of mine contradicts the fossil record; with the earliest Stylophorans being Cornutes from the mid Cambrian and the earliest Mitrates appearing later in the early Ordovician. Others believe that the Cornutes and Mitrates were monophyletic sister taxa.

Above is the solute Syringocrinus. Image taken from This is a website well worth visiting for more information.

The final group I will mention today is the Soluta (Mid Cambrian to Early Devonian). This group of Homalozoans is highly asymmetrical. The theca generally consists of small irregular plates. At one end is a large bilateral tripartite appendage very similar to the Stylophoran appendage, to which it is probably homologous. At the opposite; presumably anterior end is a smaller appendage consisting of a biserial series of basal plates and a biserial series of smaller cover plates on top. If this represents a feeding organ (which it must be since it is too small for locomotion) then this strongly indicates that the large tripartite appendage of Solutes and Stylophorans is not a feeding arm, since it is on the opposite side of the body from the feeding organ in Solutes, but is instead a muscular tail as suggested earlier. The small anterior appendage of Solutes is interpreted as an ambulacrum, housing a water vascular system derived from the left mesocoel only (Smith 2005), and tube feet as in modern echinoderms. This is supported by the presence of a plate perforated by tiny holes next to the feeding arm, which is very similar to the hydropore (external opening of water vascular system) of modern Echinoderms. Jefferies suggested that a gill slit was present in Solutes, but others (e.g. Smith 2005) are not convinced and think that the opening may have been created by post mortem disarticulation. An anal pyramid may be found on the right posterior edge of the theca near the stele. The presence of a relatively unambiguous ambulacrum, possible water vascular system and possible loss of gills makes the Soluta the most derived and crownward of the Homalozoa. These animals probably represent one of the last branches in the Echinoderm family tree before pentaradial symmetry evolved.

The lifestyles of these strange and enigmatic animals is unknown, but a few inferences can be made: Firstly the Ctenocystoids;. As suggested by Zamora (2012), these organisms probably behaved as "basket feeders", drawing water and sediment in through their anterior opening; filtering it through gill slits in a pharyngeal cavity and expelling it into an atrial cavity and then back out of the anterior opening. Faeces was expelled from an anal opening at the opposite end of the body. The Cincta probably fed in a similar way, but they separated the atrial opening from the mouth and moved the anal opening to a more anterior position. Stabilizing and anchoring ventral swellings on the marginal plates indicate that these animals lay lengthwise on the sea floor, perhaps using their stele to move to new feeding grounds if food became scarce or to help anchor it to the substrate by pointing it obliquely into the sediment. The 2 ambulacra/ lophophores either side of the mouth would have served to draw water and food into the mouth.
Lingula in feeding position.

The Stylophorans seem to have been much more mobile than the Cinctans because of their more flexible stele.  Unlike Cinctans they show no indications of an ambulacrum/ water vascular system (if you do not believe the Stylophoran appendage to be an aulacophore). In the Cornutes there was a single large anterior opening which may have functioned as a mouth. Water was probably drawn into the boot shaped theca via this mouth and food filtered out by the gills. The presence of stabilizing rims on the presumably ventral side of the theca in Ceratocystis implies that it probably lay flat on the sea bed much like Cinctans. Mitrates are interesting because they only seem to possess one large anterior opening; with no convincing seperate anal, oral and atrial openings (this may be incorrect; see above); suggesting that the anterior opening in the theca held all 3 of these openings (as suggested by Smith 2005). This suggests to me that these Mitrates lay buried in the sediment with their tails pointing downwards into the sediment with the combined anterior anal, oral and atrial orofice opening at the surface of the sediment; drawing in water, filtering it through the paired internal gills and expelling it back out through the front of the theca. This arrangement is very similar to the habits of the burrowing Inarticulate Brachiopod; Lingula. Lingula; like Mitrates has a bipartite body plan with an anterior bivalved shell and a posterior tail for locomotion; and the inhalent and exhalent openings are both at the front of the bivalved shell, opening into the water, free of the sediment to filter for food.

Some Solutes may have lived like Crinoids today; attached by a holdfast at the tip of their stele to the substrate, using thier arm to gather food from the passing current. Other Solutes lacked a holdfast and probably lay flat on the seabed, crawling along using their stele, and probing around with the anterior ambulacral arm to find food.

I hope this post has been interesting and thought provoking. There are many conflicting views on these basal Echinoderms and I have tried to present them, as well as my own ideas as best I can. These organisms really caught my eye due to their enigmatic nature. Because opinions are so varied there is much we do not yet know about them; and as a result much more to discover.

References: (All of these papers are avaliable free online. Typing their titles into Google or Google Scholar should bring them up).

Borradaile, Eastham, Potts, Saunders (1958). The Invertebrata. Third edition. A revision by G.A.Kerkut. Cambridge University Press.

Ubaghs (1975). Early Palaeozoic Echinoderms.

A.B.Smith (2005). The pre - radial history of early echinoderms. Geol. J. 40: 255-280.

Ubaghs and Robinson (1988). HOMALOZOAN ECHINODERMS FROM THE WHEELER FORMATION (MIDDLE CAMBRIAN) OF WESTERN UTAH. University of Kansas palaeontological contributions. Paper 120. (yet again; well worth a look!)

2005 Vol. 50 No. 20 23422354 Shu Degan. On the Phylum Velulicolia. Chinese Science Bulletin

Zamora S, Rahman IA, Smith AB (2012) Plated Cambrian Bilaterians Reveal the Earliest Stages of Echinoderm Evolution. PLoS ONE 7(6): e38296.

A.B.Smith (2005). Fossil Invertebrates/ Echinoderms (other than Echinoids). 334-341.

D.-G. Shu et al (2004). Ancestral echinoderms from the Chengjiang deposits of China. Nature.


Benton (2005). Vertebrate Palaeontology. Third Edition. Blackwell Publishing.

Ruta, M. 2003.Aspecies−level supertree for stylophoran echinoderms.
Acta Palaeontologica Polonica 48 (4): 559–568.