Origin of the Class Crinoidea
Bill Ausich (Department of Geological Sciences, Ohio State University)
Ohio State University, Columbus, Ohio
21 January 1999
Who cares about the origin of crinoids?
One of the most intractable problems in paleontology and evolutionary biology is understanding the origin of groups.
One example: a recent Science (November 27, 1998) has an article on a new angiosperm that pushes back the origin of angiosperms; the cover of the issue illustrated the fossil. This is a hot topic. The origins of things is really tough to figure out. Charles Darwin, in his 1859 Origin of Species, described the origin of angiosperms as an abominable mystery. We still haven’t solved it in 1998. Origins are one of the tough, tough issues in evolutionary biology.
Even the origins that we thought we figured out, like the origin of birds being Archaeopteryx - that is out of the window now. There are all sorts of questions now about whether that is the transitional fossil, or whether it has the transitional morphology or not.
We want to know where groups came from. It is intellectually challenging. The most highly funded areas in paleontology is human paleontology, because we want to know where we came from.
Another of the hot topics in paleontology today is macroevolution, which is trying to find out different things that control evolution above Darwinian natural selection; are there different groups of organisms, clades or evolutionary lineages, that have a propensity for success versus others that don’t? Are there characteristics of clades that can give you clues about success in the evolutionary record? We are worried today about global change and environmental change and climatic change, but if we don’t understand the clades, understanding climatic change isn’t really going to help us predict the future. What are the key things of taxa and groups that allow you understand patterns in the evolution of life, and why things are happening? If we don’t understand the origins, we don’t understand the clades, and we can’t really understand these evolutionary changes, so it’s important for us to sort out these problems.
Why crinoids? Because that is the group that I like. Crinoids are unarguably the most important echinoderm group in the Paleozoic and they are arguably one of the more important groups in the Paleozoic, in terms of the volume of rock that consists of echinoderm and crinoid skeletons.
There are 5 living classes of echinoderms, though some people say it is four. There are starfish, ophiuroids, crinoids, echinoids, and holothurians. This is just a whisper of a once rich class-level diversity of echinoderms There were something like 21 classes of echinoderms if you look at the whole history of life on Earth. Crinoids are one of the more important Paleozoic groups.
Let’s look at what crinoids are. They are basically a starfish on a stick. The business end of the crinoid is the crown, and the guts and gonads are up in the calyx or cup. There are arms that engage in respiration and feeding. The crown is elevated above the bottom by a stalk or a column, and it’s typically attached to the bottom by some sort of holdfast.
I have had the opportunity to go to the ends of the Earth to try and find crinoids. If you want to find the modern Paleozoic analogue for crinoids, you need to hop into a submersible and go down 100-200 meters to the bottom of the ocean, and there they will be. Modern crinoids can be observed off Grand Bahama Island. I’ve found crinoids in Antarctica, such as on the McMurdo Iceshelf. Crinoids are living in a benthic habitat beneath the ice. Things that come up through the ice include crinoids. Living crinoids can be most easily observed by OSU people, perhaps, by going to the Bahamas, to San Salvador Island, such as Snapshot Reef. Living comatulid crinoids can be observed there. The South Pacific is a good place to see living crinoids. Good sites are the Great Barrier Reef and Lizard Island. Crinoids there are really, really common, and abundant. The modern shallow-living crinoids have lost their stalk. Instead of a stalk, they just have a few cirri at the base of their cup or calyx, and they attach to objects. They can crawl around with their arms, and some can also swim. The ones that appear during the day are typically very brightly colored, which means “I taste bad” or “I’m poisonous to eat” to potential predators. In the Caribbean, crinoids are typically nocturnal, so they hide during the day, and they come out to do their feeding at night.
Crinoids begin in the Early Ordovician with a 4-circlet crinoid. The phylogeny of crinoids has the disparids coming from rhombiferan cystoids, and they go to cladids, and the cladids give rise to camerates and flexibles early on. At the end of the Paleozoic, as far as we know, 1 species survived the Permian-Triassic extinction event. They then radiate into the modern crinoids.
Showed some examples of typical Paleozoic crinoids from the Ordovician of Tennessee, Mississippian of Indiana.
Some crinoids were pseudoplanktonic, and were floating or hanging down, attached to floating logs. Example: Jurassic Holzmaden of Germany. Showed a crinoid example from the Pennsylvanian of Illinois.
One of the wonderful things about crinoids that makes them better than almost any other invertebrate group, in terms of doing ecology, is that their feeding apparatus is a hard-part skeleton, unlike a brachiopod or a bryozoan or a coral, where the feeding apparatus is soft-parts.
Showed more crinoid examples from the Mississippian of Indiana, including an unusual form with a helically twisted stem, rather than round, life-savers or cheerios type of stem. Steve Riddle of OSU did a Master’s thesis understanding the functional morphology of these unusual stems.
Another crinoid example, from the Devonian of Ohio - Arthroacantha carpenteri - has nubs that were spine bases where spines attached, like an echinoid or sea urchin; very unusual. Some of the more unusual crinoid forms appear about the same time that fish become important predators. In the Paleozoic, crinoids go nuts with spines, probably an anti-predation feature.
More crinoid examples, including one showing a long anal tube, with an anus at the tip, probably for sanitation purposes, as far as we can tell; it has no hydrodynamic advantage.
The origin of crinoids - the last major summary of crinoids was the Treatise of Invertebrate Paleontology, published in 1978. Their idea of the origins of all the major groups of crinoids has no connections - all dotted lines - who knows? What I’ve been trying to do is to connect those lines, and in doing so, one of the questions you have to ask is what their origin is. Is this a problem unique to crinoids? Modern popular historical geology textbooks show figures about the evolution of plants and mammals with lots of dotted lines. It’s difficult trying to connect these groups and understand their origins. I’ve already mentioned one good example of an origin story that we thought we had - Archaeopteryx from the Jurassic Solnhofen Limestone - it is no longer a good example of a transitional fossil.
Like a good scientist, if I’m concerned about the origin of crinoids, I need to have multiple working hypotheses. Sir Walter Scott in a poem suggested that the origin of crinoids was Saint Cuthbert making beads. The folklore from northeastern England was that crinoid columnals (or Indian beads) were actually St. Cuthbert beads (part of a rosary). And, in a story analogous to Rip Van Winkle, when one hears thunder, it is not Rip Van Winkle bowling, but it is St. Cuthbert forging out beads. Maybe. Or, maybe not. I have another option. Showed a bouquet of sea-lilies.
I’ve made a lot of waves lately, and we’ll see where it all ends up. I want to talk about 3 different ideas that try to understand the origin of crinoids.
Recent proposals concerning the origins of crinoids: 1) Echmatocrinus is not a crinoid (once accepted by many as a crinoid, but now accepted by few); 2) ancestral crinoid has 4 plate circlets; 3) rhombiferan echinoderms are crinoid ancestors.
Echmatocrinus - The first question here - if we’re to understand the early phylogeny of crinoids, we’ve got to understand where they came from, and then you can connect the lines. You have to have an outgroup; you have to have a starting point, or a zero point. Echmatocrinus has been the starting point for crinoids since 1973. Jim Sprinkle published this beast from the Middle Cambrian Burgess Shale. There were two specimens - the holotype and one other specimen. He described it then as a questionable crinoid. In the 1978 Treatise, it got basically blazoned as the oldest crinoid, and so it becomes the outgroup. We decided to see if we agree with that. Loren Babcock and I published a work on Echmatocrinus just this past year. It is from the famous Burgess Shale locality. The holotype is attached to a worm tube. You can’t really tell what the details of the holdfast are, but we have some sort of holdfast. The body is conical; it is very, very flimsy; it is smashed flat as a pancake; its body was very thin. It is probably composed of plates, but boy it’s tough to tell what they are. If anything, they are imbricated, or maybe sutured a little bit, but more or less imbricated. If you look at this in greater detail, you’ll see a sort of a sandy texture of the surface. This was thought to be stereomic microstructure. I disagree with that. Basically, it has this conical structure. It lacks a stem, unlike other crinoids. It does have appendages that have plates. But the problem is that there are only 8 of these. Echinoderms have pentameral symmetry. So, an 8-fold symmetry doesn’t sound too good. Another interesting feature is that there are soft part things preserved in alternating fashion off the sides of these appendages. These were thought, in 1973 and 1978 and by a few people today, to be remnants of the water vascular system - the tube feet of an echinoderm (well seen in starfish or echinoids suckered up against the glass wall of an aquarium).
Let’s do some tests Echmatocrinus to see if it is a crinoid, or even an echinoderm. We’ll look at synapomorphies, which are unique features that a group has to have, that were inherited from its ancestors, and that nobody else have. I would argue that Echmatocrinus lacks all echinoderm synapomorphies (1. water vascular system; 2. pentameral symmetry; 3. calcite endoskeleton; 4. stereomic microstructure). It lacks a water vascular system. I have a better explanation for those appendages. It lacks pentameral symmetry. You can’t tell for sure if it has a calcite or not a calcite endoskeleton, because Burgess Shale material has all been replaced by a phyllosilicate. So, we don’t know what the original mineralogy was. It did have a skeleton. Did it have an endoskeleton or an ectoskeleton? That is a key question, but we can’t tell. It definitely doesn’t have stereomic microstructure. Echmatocrinus doesn’t have any of the synapomorphies of an echinoderm. How can it even be a crinoid?
Well, let’s look at crinoid synapomorphies, and compare them with Echmatocrinus. Crinoid synapomorphies include:
1) pentameral symmetry
2) a holomeric or meric column
3) a clear distinction between the column and the cup
4) separate, sutured plates
5) well organized plates that are in circlets that are offset by 36˚
6) at least 1 plate in a calyx that is in a radial position (an arm comes down, and it is attached to a plate in that position)
7) erect, uniserial arms
Comparison with crinoids: Echmatocrinus lacks pentameral symmetry; it has 8-fold symmetry. It does not have a stem at all. At best, it has imbricate plates; plates are completely irregular; there are no plates in a radial position anywhere. And, it has appendages. So, it must be a crinoid, right? Well, there are lots of things that have appendages. Why do I say it is not an echinoderm? Comparing with Burgess Shale Gogia (an eocrinoid) and Burgess Shale edrioasteroids - these are preserved in three-dimensional detail as a mold and cast, and they look like any other echinoderm preserved in this kind of rock anyplace else in the Paleozoic. Therefore, no one should be confusing Echmatocrinus with echinoderms. Furthermore, if you look at broken plates on Burgess Shale edrioasteroids (Walcottidiscus), you can actually see on the inside stereomic microstructure, which echinoderms are supposed to have. Echmatocrinus doesn’t have this. So, Loren Babcock and I have concluded that on the basis of these features, Echmatocrinus is not an echinoderm and it is not a crinoid. So, what is it then?
Well, we would like to propose that it is an octocoral, which is a long, long way away from being an echinoderm. The synapomorphies for octocorals are:
1) 8-fold symmetry
2) pinnate arms - they have appendages, which are called arms or tentacles which are pinnate (they have branches)
3) unpaired mesenteries
The mesenteries and septae are soft parts, and you can’t see those. But, aren’t all octocorals colonial, and none of them have plates? Well, yes for 99+% of all living octocorals. There is one living species of solitary octocoral. And, there are several species that are pseudosolitary, with one big feeding zooid and some small ancillary zooids that are hard to see. Furthermore, there is a family of living octocorals that live in the deep sea off New Caledonia called the primnoid octocorals that are plated. The primnoids are colonial - they’re not solitary, but they have imbricate plates, and some of them have uniserial plates along the arms. The fabric or microstructure of the surface overlaps with what we see on Echmatocrinus. I’m not suggesting that Echmatocrinus is a primnoid octocoral. Loren and I would conclude that given all of these characteristics, it makes an awful lot more sense to conclude that Echmatocrinus is some sort of an octocoral. It has 8-fold symmetry; it lacks echinoderm skeletons; it lacks a column; it lacks any sort of organization in the plating, etc.
So, if we get rid of the irregular, multiplated, non-stemmed, 8-fold symmetry as the starting point of crinoids, it changes our whole idea of the origins and early evolution of the crinoid group.
I happen to believe in the fidelity of the stratigraphic record. Some people don’t. But, let’s go back to the oldest crinoid. If Echmatocrinus is out of the picture, what’s the next oldest crinoid? It turns out to be Aethocrinus moorei from the Tremadoc of France. It has a column; it has differentiation of the column and the calyx; it has arms, etc. It is a crinoid. But it is a problematic crinoid. It’s problematic because between the clear column and the starting points of the arms, there are 4 circlets. This has caused problems, because all crinoids have a maximum of 3 circlets of plates. This has resulted in debate over the years. In 1969, it was argued that one has to start at the stem and count upward, and the arms shift over, which doesn’t happen in any other crinoid. Others said this is incorrect, and that one has to begin at the starting place for the arms and count downward, and the last circlet is really part of the stem. In reality, it is simply a crinoid with 4 circlets of plates. It is very simple (to me). I’ve named the new circlet of plates lintels. The homologies of plates that I’ve put together based on the earliest crinoids having 4 circlets is consistent when considering many factors, including lumen angles (when column lumen is not round, the shape has angles) and ontogeny (microcrinoids, which are crinoids in size below 2 mm). It turns out that there are other 4-circlet crinoids, but no one ever thought to interpret them that way before.
Back to this phylogeny - start with a 4-circlet crinoid origin. One branch features the loss of one interradial circlet, which produces the disparid crinoids. The loss of another interradial circlet, the lowest circlet - the lintels, which produced all the other groups of crinoids.
Where did Aethocrinus come from? This is basically a dilemma. See Sprinkle & Guensburg (1997).
I would prefer to find the origin of crinoids among other stalked echinoderms. This seems logical to me. Maybe even a group that even has 4 circlets of plates. The choices now seem to be among the following: eocrinoids, paracrinoids, some rhombiferan cystoids, maybe some diploporan cystoids. And then one goes off into more improbable things, like edrioasteroids or holothurians or echinoids. It’s really hard to imagine origins from the latter groups.
The conclusion that I arrive at is that crinoids originated from some rhombiferan cystoid that looks something like Scoliocystis (very unusual - does not have erect arms; has a stem; the anus is on the side, instead of on the top; it has some specialized rhomb structures; some of the ambulacra, instead of being on the arms, are on the surface of the calyx) or from something like an early ontogenetic phase of Caryocystites, which has a 4-circlet calyx. Both of these are possible ancestors. Paedomorphosis or neoteny of a juvenile form of one of these could be the origin of crinoids.
The synapomorphies that are present in both these rhombiferans and the earliest crinoids that support this idea are a 4-circlet calyx, and a C radial that is smaller than the other radials, a clear distinction between cup and calyx; there are also other characters as well.
The other thing I’d like to discuss today focuses on a list of specialized rhombiferan characters that occur in some crinoids. These are characteristics that all good echinoderm experts would say are rhombiferan or cystoid in character. These characters include: pore structures, recumbent ambulacra, anal vent on the side of the calyx, four plates in the lowest circlet for a very specialized type of column. Such features occur in a variety of early crinoids. I’d like to suggest that this is a clue to their origin.
Some examples of this: Tetracionicrinus has four plates in the lowest circlet and pore rhombs (rhombiferan characters, not crinoid characters). Porocrinus has pore rhombs (this is a rhombiferan character - where the body wall is corrugated; it’s thought to be a respiratory structure, just like your lungs, which have a convoluted wall that increases the surface area and makes for efficient gas exchange). Crinoids don’t need pore rhombs for respiration because they have arms. Recumbent ambulacra - Hybocystites has erect arms and recumbent arms, like a cystoid. Self-respecting crinoids shouldn’t have these features, but some do. Some crinoids retain their four plates in the basal circlet. Some have an anus on the side. These aren’t what is expected in crinoids. What we’re seeing here are evolutionary reversals (reappearances). Once these characters reappeared again, they get passed on and inherited. Putative reversals occur 18 times in 16 genera, which is 16% of all Early Ordovician and Middle Ordovician genera. Inherited reversals occur 12 times in 9 genera (9% of EO & MO crinoid genera). Total reversed characters: 31 times on 23 genera, almost a quarter of all the taxa in the Early & Middle Ordovician. This is an incredible contribution to the morphological disparity of early crinoids.
Doesn’t the idea of evolutionary reversals violates Dollo’s Law, which states that evolution is irreversible? Evolution is not supposed to be reversible. We know that that is now false. A lot of work has been done on this point on living groups where such reversals have been documented. You can’t recreate Tyrannosaurus rex, but you can have reversal of characters. Explanation: the genotype/genetic code is switched off, so it is not expressed phenotypically. But the genetic code is still present, but it is not expressed. Suppressed or silenced genes can remain in the genotype for ~10 my. Reversals are thus expected to occur early on in a group’s history.
When we plot up these character reversals on a family level tree of crinoid phylogeny, we see that the 4 plates in the basal circlet (lintels) appears in early crinoids 4 different times in different places; recumbent ambulacra occur 3 different times; an anal vent on the side occurs 2 or 3 different times; pore rhombs occur 4 different times; a xenomorphic column occurs 5 different times. These cystoid characters occur scattered all across the family-level crinoid phylogeny. The most logical, or parsimonious, answer for what is going on here (the principle of least astonishment) does not allow all of these complex structures to have independently evolved several different times in distinct groups. It is easier to turn off or on a silenced genetic code in the genotype. One sees clustering of reversal characters in some groups, such as in the flexibles and the disparids (the hybocrinids) - usually see scattered occurrences of these reversals, but every once in a while, there is a cluster in a family. Statistical tests show that this is not a random distribution. Certain groups have a propensity for reversals, which makes sense. If a genetic code gets turned on, downstream of that code might be other characters that are linked, that will also get turned on. Often times, you see 2 or 3 cystoid characters in 1 crinoid genus, not just 1 reversal character. This is consistent with this hypothesis.
Rhombiferan cystoids are the most likely ancestor of crinoids.
Another way of doing this is cladistic analysis, using the parsimony technique, and we get trees. Used crinoids, eocrinoids, diploporans, and 2 groups of rhombiferans. Used Gogia as an outgroup. Got 9 equally parsimonious trees. It would be nice to get just 1 possible tree. But, this never happens. Well, only rarely. Get rhombiferans as ancestors of crinoids on the consensus trees.
Not everyone agrees with this scenario. One thing is certain, though. Echmatocrinus is not a player in this story, but several people insist that it is.