Wednesday, October 10, 2007

Fossil Octopodes

Dear Constant Readers,

As mentioned previously in the last post, this is an expansion of a subject not easily crammed in as a tangent. Sure there are only three examples of fossil octopodes (and some indirect evidence), but the story is shockingly complicated. As Persuasive and Analytical Writing has taught me, I really do have a tendency towards unnecessary broadness. Combating that will also probably give me a more respectable post output. So let us explore this little knot of a topic.

On an ironically placed tangent I should note the suspiciously changed title of the previous post. As far as the plural of "octopus" is concerned; "octopi" is hyper-corrected (note: pus/pod not present); "octopuses" is strange sounding and anglicized; and "octopodes" appears to be the proper plural. It is also the most international way of putting it, or so Wikipedia says to me. I try not being a typical 'murrican, I really do. All three (and "octopods") confusingly appear in recent scientific journals but from now on: octopodes.

With that nitpicking aside, yes, octopodes have a terrible fossil record. Way worse that choristoderans even. Despite having the advantage of extant species (and thus testable genetic material), the phylogenetics of "Octopodiformes" (octopodes and vampire squid) are currently quite confusing. Recent genetic analysis (Strugnell et al 2006) dealing with divergence times concluded that Octopoda and Vampyromorpha diverged some time in the Paleozoic; a diagram in the paper showed it as being in the Permian over 250 mya. As noted in the prior post, cirrate and incirrate octopodes diverged in the Jurassic, much earlier than anticipated. The paper also illustrated the major Decapodiform lineages diverging all the way back in the Devonian, perhaps around 400 mya. It should be noted that this test used examples from the fossil record in order to establish minimum ages for certain lineages and divergences. This paper did not estimate estimated the divergence of the major cephalopod divisions of nautiloids, Decapodiformes and Octopodiformes, nor did any other I could find. Things seemed peachy until...

An even more recent study by Yokobori et al 2007 discussed the relations of those major groups. Traditionally, the concept of Octopodiformes (= octopodes + vampire squids) has been morphologically supported but some genetic tests have concluded that Vampyromorpha is more closely allied with Decapodiformes. This reveals the problem with Strugnell et al 2006 of presuming the monophyly of the two traditional coleoid groups. The Yokobori et al study concluded that octopodes and vampire squids did have an mt organization closer to the proposed ancestral cephalopod condition the Decapodiformes. Interestingly, this does not support the notion that octopodes and vampire squids are necessarily sister groups. Yokobori et al conclude that there need to be three major divisions of coleoid cephalopods (Octopoda, Vampyromorpha, Decapodiformes) and that they all diverged rapidly at an ancient event before the Decapodiform radiation. So judging by the apparent Devonian (at least Carboniferous) radiation of Decapodiformes, this has to be an ancient radiation occurring relatively early in the history of cephalopods.

Now, how does this compare to what is known about fossil octopodes?


Pohlsepia mazonensis
Kluessendorf & Doyle, 2000

Most safely described as a coleoid cephalopod from the Pennsylvanian (~300 mya) Mazon Creek fauna of Illinois. Preservation of soft body parts from other cephalopods (and maybe one other specimen) are found in this formation, as is the famous and bizarre Tully Monster (Tullimonstrum gregarium). Despite these spectacular fossils, soft body preservation is still rare even in a formation such as this. The specimen is tiny (max body with 3.5 cm/1.4") with a flattened sub-circular body, two small fins, and small black eyes. The body is sac-like with an ill-defined head, characteristics typical of octopodes. Curiously the specimen does not possess 8 limbs, but has 2 additional arms longer than the rest (modified arms or tentacles). Decapodiformes modify/lose arm pair IV while Vampyromorphans and Octopodes modify/lose arm pair II, but it is never stated which arm pair is modified. This is where we begin to run into trouble. After some speculation that this may be a cuttlefish-like Decapodiform, the authors suggest that it is some sort of octopus, and compared it to cirrates. They classify it as ?Cirroctopoda, a suggestion I don't really buy. This would make cirrates paraphyletic (with incirrate and vampire squid offshoots), twice as old as indicated by genetic evidence, and would mean they re-evolved a shell somewhere along the line. Since the fins of cirrates use the shell as a muscular anchor, I was wondering if the "gut trace/ink sac" may have actually been a shell/gladius of some sort. The authors don't comment on and probably did not take into account the internal structures of "Octopodiformes". So what is Pohlsepia? I think that given the extreme separation of this fossil from any relatives and the apparent closeness to the Octopoda/Decapodiform/Vampyromorpha divergence this cephalopod may not be classifiable in any crown groups. I wouldn't give my opinion too much weight - but I think that more genetic and fossil evidence is certainly needed for this incredibly murky time period.


Proteroctopus ribeti

I tried avoiding the other fossil octopus article (here) but since I can't find the article for this species I guess I'm going to have to consult it. There certainly isn't anything wrong with the article, I just wanted to read things and put them in my own words. This specimen is from the mid-Jurassic (132 mya) of France, a time period near the divergence of cirrates and incirrates according to Strugnell et al 2006. Eyden's excellent article mentions that this specimen has suckers, a sac-like body, 8 equal appendages and fins. Strugnell et al 2006 notes that there was no shell present and they mentioned the possibility that this was a "teuthid" or stem-octopus. Eyden mentions that some authorities consider it a vampyromorph. While this has some superficial cirrate traits, the lack of a shell and cirri once again seems to be a considerable problem. Like Pohlsepia, with all traits considered it really isn't very clear what this specimen is.


Palaeoctopus newboldi

This is a Late Cretaceous (89-71 mya) octopus which bears strong resemblance to modern cirrate species. It has an indistinct head, sac-like body and 8 arms with suckers, fins, internal U-shaped fin supports and an ink sac (the last two from Voight, 1997). Judging from these characters, I would agree with the classification of this as a Cirrate/Cirroctopoda in the family Paleoctopodidae. The time period also strongly hints at this being a cirrate (as opposed to a basal Coleoid), although the ink sac hints at it not being as adapted towards a deep-water environment. While probably fairly basal in Cirrata, this is the least ambiguous fossil by far.


Table of Prominent Characteristics


Shell/Gladius

MA

DH

Ink sac

Fins

Pohlsepia


x

-


x

Proteroctopus


-

x


x

Palaeoctopus

x

-

-

x

x

Cirrata

x

-

-

-

x

Incirrata

-

-

-

x

-

Vampyromorpha

x

x

-

-

x

MA = Modifed arms
DH = Distinct head


Argonauts

Before I leave, I should mention argonauts, usually listed separately from octopodes but certainly derived from them. While most species in the superfamily don't, in a few species the females secrete an egg case from their arms which bear an uncanny resemblance to ammonite shells. Obinautilus is the earliest known fossil argonaut from the Oligocene of at least 29 mya. Strugnell et al 2006 estimate that argonauts originated some time in the Jurassic, not much after the cirrates and incirrates diverged. This indicates that they actually did live at the same time as ammonites and may give hints as to the nature of their bizarre "shell". While not direct preservation of octopodes; Izumonauta, Kapal, Mizuhobaris, and Obinautilus certainly are the majority of fossil octopus species. Darnit, one of these days I'm going to have to do another tangent post (and one for Vampyromorphans too).



So there have it, a post on fossil forms of some of the least likely animals to be fossilized. This certainly is a rather confusing subject and unfortunately there is no definite word to be had on it. This goes to show that while books and webpages may write up some definite phylogenetic relationships, behind the scene there seems to be an immense amount of argument. With the perfection of genetic testing and the finding of additional specimens, the murky origins of octopodes may yet be discerned. Or not.


-Cameron


References:

Kluessendorf, Joanne & Doyle, Peter. 2000. Pohlsepia mazonensis, an early "octopus" from the Carboniferous of Illinois, USA. Palaeontology 43 (5) 919-926. Available: Here

Strugnell, Jan et al. 2006. Divergence time estimates for major cephalopod groups: evidence from multiple genes. Cladistics 22, 89-96. Available: Here

Voight, Janet R. 1997. Cladistic Analysis of the Octopods based on anatomical characters. J. Moll. Stud. 63, 311-325. Available: Here

Yokobori, Shin-ichi et al. 2007. Mitochondrial genome structure and evolution in the living fossil vampire squid, Vampyroteuthis infernalis, and extant cephalopods. Molecular Phylogenetics and Evolution 44 (2) 898-910. Available: Here

5 comments:

Anonymous said...

Instead of tangents you should try secants. They still get you from place to place but makes much more sense.

Octopodes? I hate it. Is the plural of cactus now catopodes?

A rather amusing blog post, lots of clever names...I'm still amazed that an octopus can be fossilized.

Anonymous said...

Wasn't there a theory that argonaut ancestors actually laid their eggs in the shells of dead ammonites? Also, I've seen an old paper claiming that the three living argonaut groups have shells similar to those of different Late Cretaceous ammonite genera. The author interpreted this as evidence of polyphyletic descent of octopods from ammonites (obviously wrong), but it could also fit with the above theory.

Christopher Taylor said...

I've put a link to this post up at Linnaeus' Legacy.

Anonymous said...

The argonaut/ammonite link was proposed by Lewy in the 1990s. Half the paper is potentially interesting, talking about heteromorph ammonites using their shells to incubate their eggs and then dying like argonauts do. The rest didn't fit because (at the time) the earliest argonaut fossils were much later than the latest ammonites. Nor is there any evidence for modification of ammonoid shells by another organism

Lewy, Z. 1996. Octopods: nude ammonoids that survived the Cretaceous-Tertiary boundary mass extinction. Geology, 7:620-630

Hewitt, Roger A, Westermann, Gerd E G. 2003, Recurrences of hypotheses about ammonites and Argonauta
Journal of Paleontology

Anonymous said...

What a great resource!