Monday, June 24, 2013

A note on the taxonomy used in this blog

The current consensus of the phylogenetic relationships of Tyrannosauridae. Pairs of sister species are subordinate members of a common ancestor that establishes the genus name. This approach maximizes the information content of the Linnean binomial names, where phylogenetic (sister group relationship) and morphological similarity (synapomorphy) is emphasized.

I am alone among my fellow tyrannosauroid colleagues, where I do not use the generic names Gorgosaurus or Tarbosaurus. My reason is that taxonomic names (i.e., genera) should be used to reflect phylogenetic relationships between pairs of sister species. This approach maximizes the information content of the name.
The cladogram above shows the present consensus on tyrannosaurid phylogenetic relationships, where Albertosaurus libratus and A. sarcophagus are sister species, and Tyrannosaurus rex and T. bataar are also sister species.  If A. libratus is bumped up to the genus Gorgosaurus, and if T. bataar is bumped up to Tarbosaurus, then the information content regarding its sister species relationship is lost. Also, the generic status implies a greater morphological and phylogenetic distance between the sister species than actually exists.
Notice the taxa that are not paired with sister species (Bistahieversor, Alioramus, Teratophoneus, Daspletosaurus), which mark successive lineages that contain multiple species (more than two). In these cases, generic status for each species is reasonable to emphasize their phylogenetic and morphological distinctiveness.
It has been argued that bestowing a genus name for every species makes communication more efficient. However, this is not truly expedient if the sister group relationship still needs to be spelled out. It also runs against the grain of using binomials in the first place.
In short, I'll adjust the taxonomy the day it is shown that A. libratus and A. sarcophagus, or T. bataar and T. rex, are no longer sister species. Until then, I stand apart with - I think - a defensible conviction.


  1. It is good to get your philosophy out there on the first foot, Tom.

    Problem is, names like Bistahieversor run counter to this already. The philosophy you espouse should work at all levels of a cladogram: A "genus" is merely another clade name when two or more taxa are combined into a node or stem. Under that philosophy, Bistahieversor would only work if there was another species aside from its type included in the umbrella. Otherwise, each species can have a binomial name when it comes to fossil taxa. Each species-species pair gets a new name, and so forth ad infinitum. This accomplishes the same goal of maximal information in a name, minimal problems when dealing with "What's a genus?" questions.

    Still, it is good to set your position up front, so you don't have to debate it constantly!

  2. Thomas, well said! Early on I did stick with Tyrannosaurus bataar and Albertosaurus libratus. The reason I reverted to the other generic names is that at least some trees of my own and others popped other taxa intermediate between these species and the type species of the genera.

    Indeed, Greg Paul's old model (with Albertosaurus essentially as here, and Tyrannosaurus as the same as most workers' "Tyrannosaurinae", is a logical use.

  3. I have to agree with Holtz here. If we discover a species closer to rex than to bataar, it would belong to Tyrannosaurus using your philosophy, but suddenly our reason for including bataar in Tyrannosaurus is gone.

    Further, I'd question any claim of sister species being more morphologically similar to each other than the next most derived/basal species. It's very possible that (using your cladogram) sarcophagus would have a ton of autapomorphies, making libratus more similar to Alioramus, even though libratus and sarcophagus share a few characters that make them a clade. I'm not saying this is true of any tyrannosauroid example, just that it's not ruled out by any cladogram.

    My philosophy has been that since genus status is subjective anyway, it's most useful just to follow the consensus. Makes communication simpler. Ideally, we'd just switch to uninomials, but there's no chance of that happening soon.

    1. Of course, consensus shifts and is difficult to define.

      At various times I have toyed with the idea of sidestepping the whole pointless debate by always using the original combination. But people are so used to the thought of the praenomen having taxonomic meaning that I'm sure it would go over about as well as a lead balloon. (Hello again, Megalosaurus wetherilli....)

  4. Then, sir.. Do you consider the genuses like Suchomimus or Mapusaurus as invalid, too?

  5. Taxonomy might not be too difficult for you to transcribe to your own words but perhaps you need to put your best effort onto understanding and solving those issues. type an essay online