Tooth-Based Taxonomy

Taking abreak here to discuss something more germane to this blog’s theme. We’ll get back to oviraptorosaurs and their nearly utter lack of teeth shortly.

Tooth-based taxonomy (or TBT) is something you will see frequently from here on in this blog. It is fundamental to my primary direction of research, in that species named based solely on teeth are limited in their comparability to species not based on teeth or species for which no teeth are known. To resolve this issue, the topic of the importance of isolated teeth (the basis of TBT) will fill this (dense) post. More below.

“What’s in a tooth?”— not Shakespeare

In many cases, teeth enjoy a special place in the taxonomic scheme. They are used predominately to identify mammal and fish fossils, and are very useful in biostratigraphic work when comparing marine and land faunal ages. Fossil sharks and some fairly extensive mammal clades are known solely on the basis of teeth. For mammals, a domineering amount of research has gone into defining and associating tooth types and associating each dietetically and phylogenetically [2,3].

This utility is often anecdotal, however, and has led to some misidentifications when more complete material has come up. In fact, much of the early erroneous identification of mammalian teeth is cleared up almost solely through the recovery of a set of associated teeth and/or a jaw with teeth in situ.

[Note: A section appeared here describing dental morphologies. I removed this because I realized it deserves a more specific treatment; dropping a small section about morphology of teeth into a topic about the messiness of TBT seems clunky and unfocused, and this post is already too much of that. Rest assured, it will come back in a more discussable form.]

Unlike mammals, in reptiles more complex morphologies are less common, and when they crop up, they are distinctive. Especially peculiar and distinctive teeth occur in some crocodilians (such as the multi-cusped, multi-rowed teeth of Chimaerasuchus [1]), hadrosaurian dinosaurs (enamel on one side, tiered and ranked like a Roman phalanx, and bearing articulate ridges, lophs, and marginal and lophal denticulation). Blade-shaped teeth (called ziphodonty) is the most common morphology among crocodilians, and indeed among archosaurs in general, as well as various carnivorous lizards, including monitors; snake teeth are generally cones (termed conodont, although as some note, this term is clunky and confusing), and lizards in general vary a good deal between ziphodonty (not common) and phyllodonty (common), although the difference between these may be so blurry that the term “phyllodont” (leaf-tooth) may be useless. This makes associating tooth morphology to a clade difficult, to say the least. Historically (and [2] & [3] are good examples of this, dentition alone has provided the systematist with the idea to refer a taxon to a given clade: many ziphodont crocs were considered theropod dinosaurs, or the reverse. So why is a seemingly unstable system still being used?

The clue in which TBT has been most useful has been biostratigraphy, associating faunal compositions to a fine degree across formations, countries, and continents (as well as through time), as well as assessing ecomoprhologies (knowing your prey and your predator species, and their relative numbers). Collapsing taxonomy has a related affect in reducing (although sometimes increasing) the comparability of two different faunas or formations, allowing the stratigrapher or the ecologist the ability to simplify his or her argument when comparing two groups. It is arbitrary, but at least the data is all there to assess.

For the most part, TBT is predominately a match game within a broader systematic puzzle: Teeth are referred to one another on a “best-fit” paradigm, in which gross morphology is used to assign specimens to varying types of taxa. Traditionally, these taxa are defined by their rank: It is easier to place varying teeth in families, but not so much genera, and certainly more difficult into species. The scale of the rank of the taxa the teeth are placed into are categorizations of how clear-cut the “best-fit” model works: Conodont teeth with fluting in theropods are generally assigned to Spinosauridae (and even closer, can be placed into Spinosaurinae), and tiny, low-aspect, “coarse” denticulate teeth are referred to Troodontidae. It takes more specific criteria to begin assigning these to genera and species, but for the most part, this is not at all morphological but stratigraphic (see above, although there are exceptions). This renders much of the broader concept of TBT to a parataxonomic process, by which form taxa are used to represent proxies to real taxa.

In the end, theoretical arguments need to be made which regard the morphology of teeth, and questions about their applicability to taxonomy. This requires a framework under which the parataxonomic process can be assessed to “actual” taxonomic processes, and if the two are compatible in the case of teeth.

So where am I going with this? I want to ask some important questions, and certainly the biggest among them are: Does the variation in teeth represent actual taxonomic variation, and if it does, how significant is this? If this variation is not significant what do we do about TBT?

In the general scope of science, comparability among morphological schemes is important. In systematics, for example, the referral of specimens to taxa are taken under certain rules, although most of these are soft and not enforceable. One of the more important of these, and it’s the most by a wide margin, is that when two specimens are not comparable, one should not consider them identically. We have a tendency to overlook this when there are other considerations, and in the hopes of ansering the above questions, I am tempted to simple state that when all is said and done, when two specimens with different names are not comparable to one another, do not refer them to the same taxon. The systematist who does this is taking a leap of faith with little grounding on the principle of assumptiosn that have nothing to do with comparison of material. This will be my primary rule when comparing tooth-based taxa, and my yardstick for segregating material for data points in order to answer the above questions.

[1] Wu X.-c, Sues, H.-D. & Sun A.-l. 1995. A plant-eating crocodyliform reptile from the Cretaceous of China. Nature 376:678-680.
[2] Carroll, R. L. 1988. Vertebrate Paleontology and Evolution. W. H. Freeman and Co., New York. xiv+698 pp.
[3] Romer, A. S. 1962. The Vertebrate Body. W. B. Saunders Co., Philadelphia. vii+627 pp.

Update: I’ve edited this post slightly to reflect removal of a section concerning general morphology of teeth.

This entry was posted in Biology, Philosophy, Taxonomy and tagged , , , . Bookmark the permalink.

8 Responses to Tooth-Based Taxonomy

  1. Jason Testin says:

    I is my current research goal to eliminate as many Tooth Based Taxa as I can, by putting together morphologic standards for many of my major Theropod taxa. A standard that takes into account both ontogeny and regional variation.

    Jason Testin
    Geology and Paleontology Undergrad
    South Dakota School of Mines and Technology

  2. Nick Gardner says:

    A word of caution, maybe those names in concept are commonly used, but those exact names are not used in the reptile literature… it’s a little misleading to present as such.

    To use the term conodont is not a good idea, it begs confusion with conodonts…

    Why not just use what’s more prevalent?

    • qilong says:

      For the most part, adjectival forms like “conical” and “globular” are used, but “ziphodont” and “molariform” are used when it comes to things like lacertilians and crocodilians. “Globodont” is used for some tooth forms, such as the oblate or spherical teeth of some crocodilians and mosasaurs, but this doesn’t not endanger Globidens, the mosasaur, or Globidonta, the crocodilian clade.

      I take your point in regards to “conodont”, but the terminology should be consistent: [form/shape] + -donty is a classic useage outside of mammalian tooth-based description, and the adjectival forms of the terminology (i.e., “conical,” “blade-like”) are used in mammalian terminology for parts of teeth. So while “conodont” is useful to describe a morphology of a tooth or a part thereof, and it also describes a group, I do not think these conflict nearly enough to be problematic; this is especially important when considering that conodonts and conodonty cannot ever be confused, and one was named one the basis of the morphology (despite P and S elements having absolutely nothing to do with typical cones or comprising them).

  3. Brian Beatty says:

    I agree with Nick in concern for the usage of the term conodont, though I would say that in general so many terms are thrown around for teeth that the main point is that terms should always, ALWAYS be presented with proper definitions and in proper context the first time they are used in a paper, or at least in reference to a usage elsewhere that can be cited. Clarity of the meaning for many terms in paleontology is very poorly kept track of, and ideas morph into strange and misleading ways all the time.
    As for TBT, it is a serious problem and needs to be tackled in a variety of ways, some of which are easier than others. The principle matter that needs attention, yet lacks is usually because of the work involved is VARIATION. If more people spent their time looking at larger samples of conspecifics from the same populations (either modern or catastrophic assemblages of fossil taxa), then we would have a much better notion about how to handle fossil taxa like proper biological entities. Some of the topics in variation that REALLY need attention for teeth are:

    1. microstructure

    2. developmental pathways / morphogenesis

    3. functional associations with morphology, particularly how function deviates from form (which is influenced by phylogenetic history) – this is where independent contrasts and phylogenetic regression studies could be useful.

    I am very glad to see this discussion happening, and would like to encourage it to occur in more published, formal ways. If anyone is interested in publishing on these views in a peer-reviewed format, I’d welcome it to It would be nice to see the journal used as a forum for ideas.

    I have only done a little work on the topic of TBT in archosauriforms, but once involved I found it immediately compelling and widely reaching into many facets of vertebrate paleontology. I hope this discussion continues.

  4. qilong says:

    As Nick and Brian note, the issue of conodonty and conodonts is a bit tricky. I’ve edited the post to reflect this, and will return to the topic in greater detail as a result of this discussion. The idea, poorly implemented, was to approach the morphology of teeth in a general manner, and to reflect on its poor fit to describe the variable morphologies in reptilian dentition (among other groups). Because of this, and because I really want to use that as a springboard to describe all the other various dentition morphologies out there, I’ve removed that bit, referred to it in the post, and am commenting here now. Thanks for the discussions.

  5. Bill Parker says:

    TBT can be valid if the teeth are autapomorphic as is the case for taxa such as Trilophosaurus, Revueltosaurus and Troodon. What is often mixed up is assigning these taxa to larger subgroups based on overall similarity in tooth morphology. The goal is not to get rid of all TBT, but rather to find the rest of the animal to figure out where they go in the classification. A prime example of this is Kraterokheirodon colberti, which was thought for awhile to represent a traversodont cynodont (see the Chinle Formation section of Walking with Dinosaurs), when in reality the character states present cannot present a taxonomic assignment higher than tetrapoda. Despite this it is still a diagnosable taxon based on tooth characters.

    • qilong says:

      While I agree with you on the specific, it’s to a point. I rendered two rules for the recognition of taxa, and the absolute uniqueness and possession of apomorphies are one of them that permit recognition; such is the case with Kraterokheirodon and Troodon. I am not so sure about Revueltosaurus. Distinguishing teeth on the basis of independant, non-phyletic means is difficult, and so far only Smith [1] seems to have attempted to attempt a morphometric test of the “eyeball” method. While I am not completely opposed to the idea of naming teeth, I am wary for reasons of the potential viability for a single morphotype, even something like Kraterokheirodon, being explicit to a larger group of taxa than the original; this seems to be true of Revueltosaurus, but it may not be true of Troodon — although this is an historical consideration.

      [1] Smith, J. B. 2002. An examination of dental morphology and variation in theropod dinosaurs: implications for the identification of isolated teeth. Ph.D. dissertation. University of Pennsylvania, Philadelphia, Pennsylvania, 617 pp.

Leave a Reply

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

You are commenting using your 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