Operculum: origin, disappearance, reappearance and disappearance

The fish operculum first appeared
in the sturgeon, Acipenser, among tested taxa in the large reptile tree (LRT, 1870+ taxa; subset Fig. 2). More primitive taxa, from the hagfish (Myxine) to Thelodus, had several gill openings lined up without an operculum covering the serial openings.

Figure 1. Chondrosteus animation (2 frames) in situ and reconstructed in lateral view. This is the transitional taxon linking sturgeons to bony fish + sharks.
Figure 1. Chondrosteus animation (2 frames) in situ and reconstructed in lateral view. This is the transitional taxon linking sturgeons to sharks. The operculum is the lavendar plate between the jaws and pectoral fins.

The operculum first disappeared
and multiple gill openings reappeared in sharks and rays, like Manta and Loganellia.

The operculum first reappeared
in ratfish (Chimaera), then reappeared again in paddlefish (Polyodon) and again in bony fish (Gregorius).

The operculum finally disappeared
in placoderms, and again in moray eels (Gymnothorax) + gulper eels (Eurypharynx) and again in tetrapods (Greererpeton, Fig. 1).

Figure 2. Subset of the LRT focusing on taxa with an operculum.

Frogs and salamanders also have an operculum,
but it is not the same (= homologous) structure. It just has the same name and acts like an eardrum (Fig. 3).

Figure 2. Skull of the frog, Rana
Figure 3. Skull of the frog, Rana. The operculum is the gray series of concentric circles below the squamosal and above the ptergyoid.

At sharks.org, they asked,
“A classical evolutionary question has been how and when these differences arose during Chondrichthyian evolution, and what the common ancestor of Elasmobranchs and Holocephalans looked like. Did sharks gain additional branchial rays after they diverged from a common ancestor who had a chimaera-like pattern, or did chimaeras lose most of their branchial rays after diverging from a common ancestor who had a shark-like pattern?”

Then concluded,
“So elephant sharks have lost the ability to make posterior branchial rays, but still carry the embryonic pattern for where they should form. It seems the common ancestor of Elasmobranchs and Holocephalans had the gills of a shark, rather than the gills of a chimaera,” according to Gillis et al. 2011 in Neil Shubin’s lab. This is confirmed by the LRT (Fig. 2).

The takeaway message is this:
Don’t get caught Pulling a Larry Martin! (= defining a clade based on a single trait or a dozen traits). You’ve just seen how one trait can appear and disappear and reappear and disappear. This is but one example of several hundred I could have shown. Only the last common ancestor method can define a clade. A last common ancestor can be hypothetically recovered in phylogenetic analysis using 150 to 250 multi-state character traits and as many taxa as you have time to do. Convergence and reversal are out there, so beware!

Gillis JA, Rawlinson KA, Bell J, Lyon WS, Baker CVH and Shubin NH 2011. Holocephalan embryos provide evidence for gill arch appendage reduction and opercular evolution in cartilaginous fishes. Proceedings of the National Academy of Sciences USA 108:1507–1512.


11 thoughts on “Operculum: origin, disappearance, reappearance and disappearance

  1. Alternatively you could put in a decent number of meaningful characters (you know- minor things like ability to possess bone, enamel, teeth on the jaw margin and so on; and yes juvenile sturgeon have teeth) and it appears once at the base of the osteichthyes, Chimaeras don’t have an operculum- they just have reduced openings on the brachial pouch, but there is no skeletal structure there (I have dissected enough of the slimy buggers).

    • You know there are 235 characters employed here. Not just one. The appearance and disappearance and reappearance of many traits is well documented. This is but one. With regard to sturgeon teeth, they are keratinized, like those of hagfish, not made of dentine and enamel. The single gill opening of the chimaera (your “reduced openings” is in the literature.

      • Sturgeon (and more heavily studied) paddlefish embryos have normal osteichthyan teeth complete with enamel; they have the same gene control, structure, development and jaw attachment as teeth in other osteichthhans. They are in no way related to the tooth like structures of lamprey etc, and are very different from teeth in chondrichthyans, possibly suggesting independent origin of teeth at least twice. Sturgeon also posses true bone, both dermal and vertebral. Not much of it internally, but it is there in older specimens are clearly osteichthyan type true bone (will all the physiology that goes along with it) and not chondrichthyan tesserae.
        Chimaeras have no operculum; they have a single (if internally divided) gill opening but within this fleshy enclosure the gills are normal chondrichthyan,
        While you refuse to use the plethora of characters that are relevant to ‘fishes’, including the ability to form certain biominerals, dentition, vertebral column, etc and randomly draw on non existent skull bones onto chondrichthyes (and jawless forms) that have none, you are going to end up with a meaningless scatter of taxa. (oh, those ‘sutures’ on the Squalus skull- they are shrinkage marks that occur as the chondrocranium dries- they are not there in fresh dissections (I have dissected skulls from maybe 10 species of squaliformes and it is the same in all).

      • You have given some great traditional arguments here. Unfortunately you have not provided a more parsimonious cladogram of interrelationships. This is what I need with the same taxon list. We need to see where ratfish appeared with a gradual accumulation (or loss) of traits demonstrating microevolution at every node better than the one presented, if not in toto, then, at least, in parts. I have struggled to understand how evolution has shaped every taxon, and continue to do so, making corrections incessantly. So take heart. I am open to suggestion. However, at present your suggestions for the ability or inability to create or lose certain biominerals, the appearance of a trait or two or a dozen does not (at present) unbalance the current cladogram. That can change with your competing cladogram or some other data input. So bring it on.

  2. If a feature survives in embryo, then it’s re-emergence in adults is surely just a case of neoteny? It survives in the wings as it were, just waiting for the right selective pressures to cause it’s re-appearance in adults …

    • Not sure what the embryological data says. The Gillis et al. 2011 paper is the only one I am aware of. Notably Gregorius and Prohalecites, close to the origin of bony fish, are tiny compared to ancestral shark taxa, like Hybodus and Orodus, supporting your point with regard to neotony.

  3. You missed one of charlieju’s most important points – that you’re coding character states for taxa that don’t have the parts in the first place. The tree doesn’t matter when it comes to basic homology assessment. Sharks don’t have dermal cranial bones. Assigning them states for characters expressing dermal cranial bone variation is inappropriate, and the trees you’re getting for that part of the tree can’t be used.

    You need to re-do your analysis with sharks coded as unknown (which is computationally the same as non-codable) for these characters.

    If it causes resolution to collapse, so be it. At least the resulting trees won’t be misled by erroneous homology assessments.

    Suppose you came up with a tree for placental mammals with lots of characters based on the interclavicle, predentary bone, and caudal fin. The tree might have excellent support values, but since no placental has an interclavicle, predentary, or caudal fin, those numbers will mean nothing, and no rational person would consider the tree itself as having any scientific value. This is exactly the situation we face with the relationships of fishes in your trees.

    (You could also try adding more characters, by the way).

    • No. The elements of the shark skull are present, whether ossified or not. Your attempt at analogy is hyperbolic in that it refers to body parts that are not present (except a caudal fin (flukes) which you might remember appears several times by convergence). The nasal cartilage in sharks continues to support the anterior rostrum and naris. The parietal cartilage continues to roof the brain. There are small cartilage shapes both pre and post the orbit. There is a large cartilage, here called the lacrimal, that continues to support the tooth-bearing premaxilla and maxilla zones, wherever teeth are present. Following sharks, when bony fish, like Amia, develop bone during ontogeny these appear in discrete crustal islands on cartilage substrates. Some merge. Others split. The present LRT provides a continuum of development which includes a departure from ossification during the shark series and a return thereafter. At present the LRT provides the most parsimonious solution to the issue of what taxon evolved from what taxon in chordates. But this is science. So I eagerly look forward to a _more_ parsimonious solution that includes all currently included taxa. No exceptions. So, please, bring it on!

      • “The elements of the shark skull are present, whether ossified or not. ”

        No, they are not.

        There are no cartilaginous precursors to any of the bones you indicate. This is a simple biological fact. Please, please, please – spend some time learning more about skull development and homology assessment in fishes.

        I’m not going to argue with you. Your claims about “lacrimal cartilage” and “parietal cartilage” and such are wrong. I am informing you of that. You can continue to run analyses based on incorrect assessments, or you can update your matrix to reflect biological reality.

        I’m honestly not trying to be a jerk. Whether I agree with your tree is a separate issue. This boils down to basic homology assessment. When you code sharks, chimaeras, and cyclostomes for dermal cranial bones, you are wrong to do so.

        Please re-do your analyses with these fishes coded correctly as unknown/uncodable for dermal cranial characters.

        Either that, or present an analysis supporting your homology assessments. The trees are irrelevant here – we’re getting at the data used to construct those trees. Analyses based on pictures will be dismissed as meaningless – you really need to get your hands on an actual shark chondrocranium, preferably fresh and not dried, and demonstrate to the scientific community that homologues of such bones as the lacrimal and parietal, neither of which preforms in cartilage, can be recognized.

        So there – I’ve brought it on. It’s not a matter of debate. Your homology assessments regarding dermal bones on shark chondrocrania are wrong. And now I’m not the only one telling you this.

      • I understand your thoughts represent the current attitudes and traditions regarding shark skulls. I was heretical trying to include cartilaginous fish in some fashion. Perhaps it would have been acceptable if only the biochemistry was different. I also reidentified many bony fish skull elements, giving them tetrapod homologs. That’s also unacceptable in the scientific community. I also work with photos and that’s unacceptable in the community. With no other presented alternatives, other than to ditch the scores, as you suggest, I’m going hold on to the presentation as is and wait for a little time to pass. I’ve been waiting for workers to understand the LRT origins of turtles, archosaurs, mammals, birds, reptiles, pterosaurs and whales for several years, or for workers to present a more parsimonious solution with the same taxon list. I’m ready to change, but only for a better cladogram or a better way to identify cartilaginous elements with tetrapod homologs.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.