I am posting this here because I don’t want to log down Darren’s blog with my comment, where he has posted his perspective on the new paper, so it’s going here.
Valentine Fischer, Michael Maisch, and a host of other authors have a new paper out in PLoS ONE dealing with the perennial issue of ichthyosaur taxonomy. There, they erect a new taxon Acamptonectes densus for an ophthalmosaurine ophthalmosaurid located in the Early Cretaceous, defying previous regards that ophthalmosaurids didn’t survive the end of the Jurassic. It is derived on a combination of features which are otherwise not held by an array of other taxa, including numerous autapomorphies; the morphology at hand is so distinctive that additional specimens from across Europe have been referred to the Acamptonectes umbrella, or to the type species, including a paratype from Germany, where the holotype derives from England. Darren Naish, one of the coauthors, has his own write-up at TetZoo. In their phylogenetic analysis, Fischer et al. (2012) find a topology for ophthalmosaurids in which two forms appear paraphyletic, Ophthalmosaurus and Platypterygius. This is supported by the inclusion in the latter of two taxa (Athabascasaurus bituminensis and Caypullisaurus bonapartei) within the array of erstwhile Platypterygius species hercynicus and australis, and in the former by the presence of Acamptonectes densus sister to Ophthalmosaurus natans, which is a sister-taxon to icenicus, the type species.
But if that were all, this wouldn’t be much of a post, and I fear I must break into a repeat of my eternal bogeyman, the argument about the Linnaean System and its effect on systematics and nomenclature.
The authors write:
Our analysis failed to recover monophyletic Ophthalmosaurus and Platypterygius. The polyphyletic, wastebasket nature of Platypterygius has already been noted by other authors , , . Our consensus tree suggests that O. natans should be given a distinct generic moniker since it does not group with the type species of Ophthalmosaurus. As is well known, the name Baptanodon is already available for it. However, O. natans is united with A. densus due to a single homoplastic synapomorphy (reduced crown striations, char. 1) and we do not yet consider this sufficient evidence to resurrect use of the name Baptanodon. Because of the obvious differences between Acamptonectes and the Ophthalmosaurus species, we do not consider it preferable to include Acamptonectes within an expanded version of Ophthalmosaurus. Further study should help clarify the affinities, and hence taxonomy, of these taxa.
[italics in the original; 34 is Fischer et al. (2011) and 61 is Druckenmiller & Maxwell (2010), both referenced below; 57 is an unpublished ms submitted to Acta Palaeontologica Polonica, and is available here -- note that the referenced date in the current paper, "2011," is no longer valid.]
Now, my issue may or may not be clear at this point, but just in case, I again don’t care about the nomenclature, but about the rather screwed up mixture here of phylogenetic nomenclature:
The main topological novelty of our analysis lies at the base of the ophthalmosaurid lineage, which diverges into two clades: Ophthalmosaurinae, which is quite similar to the basal Arthropterygius; and Platypterygiinae, which is markedly derived from the ancestral Arthropterygius stock (see below for a phylogenetic definition of these clade names). Both these subfamilies have been informally used in the literature in the past , but have never been defined or recovered in cladistics analyses until now.
[italics in the original; 79 is Arkhangelsky (2001), referenced below]
It should be noted that paraphyly is only present so long as the use of extra “genera” for various taxa is chosen, and when the taxa are treated as clades with confined definitions that cannot include the other species under their umbrella — or if one is treating the ranks as viable elements of the nomenclature. Paraphyly is not an issue when all clades are given equal weight or utility; while one ranked taxon like a “genus” cannot contain another, an unranked clade named Ophthalmosaurus can contain an unranked clade named Acamptonectes, but I do not believe this is what the authors mean. The authors clearly use rank-based taxonomy to support use of some of these taxa, while at the same time seem to offer a non-generic function for the name Ophthalmosaurus, one which can include supposed “genera,” which is to suggest either they are going to treat it as a non-genus rank (i.e., a clade based on a former genus name), or that they are going to play the infra-rank game and try to shuffle the ranks around to make this make sense. A potential solution could have been solved by employing Baptanodon for natans, and eschewing the argument that Ophthalmosaurus was in any way paraphyletic or, in that shocking twist of events, name densus as a species of Ophthalmosaurus and enforce the lack of paraphyly in the form.
The commentary in the paper on the referral of the taxon and its relationships seem like just as much information could be had were the authors to have named a new species of Ophthalmosaurus, rather than a new binomial genus-species couplet. The value of the binomen is, of course, increased recognition and utility in further phylogenetic analysis without having to restate referral of the taxon to this or that genus, although the problem arises no matter when more and more clades are coined to include ever fewer “genera.” The problem is that, while aware of the issue, the authors appear to exacerbate it by mixing their taxonomic schema to allow ranks a viable option, and phylogenetic nomenclature in qualifying these ranks according to a cladistic analysis.
Other than that, the paper details re-assessing the taxonomic complexes of Ichthyosauria and its range through time, using metrics of species density to time to track whether there were actual extinction events around the Jurassic-Cretaceous (J/K) boundary. The prevalence of ophthalmosaurid taxa across this boundary implies that the “event” in ichthyosaurian context didn’t actually exist, and that this is related to recovery biases (the material is simply not being recognized) and preservation biases (the material is not becomes preserved or found). This may relate to poor preservation in marine strata in the Neocomian (early and middle Early Cretaceous).
A Not So Short Piece on “Acamptonectes densus” Versus “Ophthalmosaurus densus“
From comments in Fischer et al. (2012) and in other papers, it is clear that Ophthalmosauridae as traditionally conceived is homoplastic in its distribution of characters among sub-members, which group three clades, two of which are grouped into the Platypteryginae. These conceptions begin breaking down when the taxa used to represent them become more “generic” in regards to the concepts for which they were originally coined. Because of this, one must think that as traditionally conceived they may not be useful. The opposite tack is often preferred by systematists attempting to preserve the functional relationships of ranked nomenclature (the Linnaean System) while developing the rank-free systems (especially phylogenetic nomenclature, and certainly the PhyloCode). It is one of my biggest disagreements with the group, as the baggage these taxa represent, including the effect that functional use of -aceae and -idae has on systematists in general. These represent “families,” either botanical or zoological (respectively), and the terms are even used by those who prefer phylogenetic nomenclature and the PhyloCode over the Linnaean System. Their prevalence is ingrained in our systematics.
The authors attempt to resolve the shakiness that results in phylogenetic analysis by placing apparent unique taxa off in their own little “genera.” This solves the problem of what you do when shifting arrangements may require new reports of new combinations (“comb. nov.”) for every time one finds a novel array of species, which gets tedious, but it produces a problem in that what results is merely a moving of the goal posts. This bit concerns the Linnaean System and phylogenetic nomenclature. Eventually, this solution results in a plethora of fossil species each with their own unique binomen, and then the neontologists start laughing at you; they point to taxa like Coleoptera and Cichlidae, the latter which contains about 2,000 species in over 200 recognized “genera,” and an average of over 9 species per “genus.” Such a value overwhelms paleontology, which are used to thinking that any species that they recognize is unique enough to warrant useful taxonomy, but I think for the sake of recognition erect a “genus” for it. I’m not saying neontologists and zoologists have it right, recognizing more species in fewer “genera,” as they are also constrained by the trap of treating the “genus” as a viable element of their taxonomy. But the paleontological community has a sense that it is more useful, and preferable, to name what they think are two different taxa, and this leads to problems in itself.
My preference for dealing with “obvious” rank names like this is to simply discontinue their use, define new clades which approximate their content, and move on. This often happens in similar circumstances when a taxon, like a species, is no longer useful for what it historically contains, but the inverse: A species may be no longer useful because its content changes, specimens formerly diagnostic of the species are remove, made types of new species, or reassigned, but the species is still useful and the name may be shifted to something functionally different, as was the case with Iguanodon anglicus (former type species of Iguanodon, which is now bernissartensis [ICZN, 2000]) and Coelophysis bauri (which was based on specimens that may have been quite different from their current more famous Ghost Ranch-based content, and which received the name Rioarribasaurus colberti [Hunt & Lucas, 1991] to emphasize this [ICZN, 1996]). Both of these cases are different, but involve this sense of shifting a useful name or material to preserve the idea of the animal it represents. Instead, it may be necessary to simply abandon the use of names whose baggage leads to the perfunctory inclusion of ranks in remarking about them, such as when saying “new genus and species,” when what one really means is a new taxon, using a representative binomen.
When Fisher et al. (2012) regard the possibility that Ophthalmosaurus may contain the species densus, they are considering the option that Ophthalmosaurus is a genus, and densus a species; they are not necessarily concerned over Ophthalmosaurus as a clade that may contain another clade named Acamptonectes. But the intriguing phrasing leads one to wonder that that could be a solution, and that Ophthalmosaurus, itself almost a functional synonym of Ophthalmosaurinae, may not be quite as useful as it seems in the formulation it is given. The authors already consider it paraphyletic, practically a wastebasket taxon. Currently, Ophthalmosaurus is the container for about four species: icenicus (the type species, Seeley, 1874), natans (the type species of the preoccupied Sauranodon Marsh, 1879, and which is considered to contain the material in what Marsh called Baptanodon, in four species [n1]), saveljeviensis (the type species of Paraophthalmosaurus Arkhangelsky, 1997) and yasikovi (the type species of Yasykovia Efimov, 1999) and periallus (Fernández, 1999). Such an arrangement makes you wonder why we aren’t simply talking about different taxa alongside Ophthalmosaurus, which only contains icenicus.
I don’t think the taxa as recognized are so useless as to require renaming every species to new “genera” (or recognizing those that have been established. I wonder, instead, whether the authors could (should) have named densus to Ophthalmosaurus, and increased the robustness of their analysis further to enable resolution, as well as applying various tests to the data matrix to find problems with it and evaluate this with respect to the resulting cladograms. Doing so might as qualified the decision to make new nomenclature, but rather the authors leave this to further analysis. This is a case of “name first, question later” save that once established, it seems these “genera” become rather hard to get rid of, and condensing species back into complexes without special nomenclature is an action that tends to be eschewed (see my and Mickey Mortimer’s reactions to Greg Paul’s latest book here and here). Lumping, in this case, a project fraught with treading on men’s toes, for you risk hurting their intellectual territory, while the splitting off of new taxa can only at the least bring the ire of one irate blogger. If I had a choice, I’d name the new form in a binomen but only with respect to the field in which its contained: if the new form might be easier used as a “species” in a “genus” which is already available for it, don’t rock the boat and claim paraphyly by coining a new “genus;” just name a new species, and move on. After all, the authors say “[f]urther study should help clarify the affinities, and hence taxonomy, of these taxa.”
Were we to treat “genera” as just any other type of clade, instead of a special clade, the above phylogeny might be interesting, but of course it would be very, very different in execution when applied to all known taxa. The alternative to this, somewhat of an extreme, is that each species would deserve a new “generic” container, so that Ophthalmosaurus natans and Platypterygius australis would require new taxa named for the species. The nomenclature applied above takes the least effort to work, but of course it qualifies as special categories the already named monospecific forms. One may then be tempted to simply remove these “generic” labels in favor of where the species were originally referred when the “genera” aren’t paraphyletic. But, it is harder to get rid of “genera” than to coin them.
Whether Clades Should Have No Name
There is value to not naming clades, include immediate supraspecific taxa, commonly “genera.” This includes peppering the literature with more and more essentially redundant taxa. Above, virtually all ichthyosaur “genera” contain merely one species, or a single valid species with a large number of rather doubtful potentially nonviable species. Naming species, as unique entities, is fine, but these aren’t clades in the essential sense, they are merely containers of a population-like group of fossils, or to which fossils may be referred. It tends to be unconventional, however, for systematists to include new species or “genera” within other erstwhile “genera.” As above, Opthalmosaurus would be a clade that includes a type species and a subclade called Acamptonectes, while Platypterygius is a clade that would include two subclades, named Caypullisaurus and Athabascasaurus, respectively, and two other unassigned subclades. In today’s world, this would almost amount to “subgenera,” and each of those other species would also require a new name. But there is no substantive reason why we should fill up the gaps in the tree above save to “balance” the appearance of the tree, or have something “special” to use to refer to. That point, of course, tends to come at the expense of the use of the species name, so that one only discusses Acamptonectes, not densus. I make a measurable attempt to avert this habit on this blog.
[n1] Marsh took, in his sincerest humility, the effort to distinguish a new “order” of reptiles for this group of ichthyosaurs, Sauranodonta, but for which Ichthyosauria has been used to approximate (under the premise that — using the Linnaean Systems organization of ranks — Blainville’s name, predating Marsh’s, had priority). Marsh has a preference for distinguishing many “orders” of reptile. It was almost his thing.
Arkhangelsky, M. S. 1997. On a new ichthyosaurian genus from the Lower Volgian Substage of the Saratov, Volga Region. Paleontological Journal 31:87–90. [English edition of Палеонтологический Зурнал.]
Arkhangelsky, M. S. 2001. The historical sequence of Jurassic and Cretaceous ichthyosaurs. Paleontological Journal 35:521–524. [English edition of Палеонтологический Зурнал.]
Druckenmiller, P. S. & Maxwell, E. E. 2010. A new Lower Cretaceous (lower Albian) ichthyosaur genus from the Clearwater Formation, Alberta, Canada. Canadian Journal of Earth Sciences – Revue canadienne des sciences de la Terre 47:1037–1053.
Efimov, V. M. 1999. Ichthyosaurs of a new genus Yasykovia from the Upper Jurassic strata of European Russia. Paleontological Journal 33:91–98. [English edition of Палеонтологический Зурнал.]
Fernández, M. S. 1999. A new ichthyosaur from the Los Molles formation (Early Bajocian), Neuquén Basin, Argentina. Journal of Paleontology 73:675–679.
Fischer, V., Maisch, M. W., Naish, D., Kosma, R, Liston, J., Joger, U., Krüger, F. J., Pérez, J. P., Tainsh, J. & Appleby R. M. 2012. New ophthalmosaurid ichthyosaurs from the European Lower Cretaceous demonstrate extensive ichthyosaur survival across the Jurassic–Cretaceous boundary. PLoS ONE 7(1):e29234.
Fischer, V., Masure, E., Arkhangelsky, M. S. & Godefroit, P. 2011. A new Barremian (Early Cretaceous) ichthyosaur from western Russia. Journal of Vertebrate Paleontology 31(4):1010–1025.
Hunt, A. P. & Lucas, S. G. 1991. Rioarribasaurus, a new name for a Late Triassic dinosaur from New Mexico (USA). Paläontologische Zeitschrift 65(1/2):191-198.
International Commission on Zoological Nomenclature. 1996. Opinion 1842. Coelurus bauri Cope, 1887 (currently Coelophysis bauri; Reptilia, Saurischia): lectotype replaced by a neotype. Bulletin of Zoological Nomenclature 53(2):142-144.
International Commission on Zoological Nomenclature. 2000. Opinion 1947. Iguanodon Mantell, 1825 (Reptilia: Ornithischia): Iguanodon bernissartensis Boulenger in Beneden, 1881 designated as the type species, and a lectotype designated. Bulletin of Zoological Nomenclature 57(1):61-62.
Marsh, O. C. 1879. A new order of extinct reptiles (Sauranodonta) from the Jurassic Formation of the Rocky Mountains. American Journal of Science 17:85–92.
Seeley, H. G. 1874. On the pectoral arch and forelimb of Ophthalmosaurus, a new ichthyosaurian genus from the Oxford Clay. The Quarterly Journal of the Geological Society of London 30:696–707.