The character/taxon ratio

The large reptile tree currently has 415+ taxa and it keeps on growing. It also has 228 characters, the same number as when the taxon list was barely over 100. Paleontologist like to see a 3:1 character:taxon ratio, which was the average according to Sanderson and Donoghue (back in 1989). Perhaps unfortunately, the large reptile tree has a character:taxon ratio of about 0.5:1 or about six times smaller than ‘the ideal’.

This brings to mind a famous phrase
‘In theory, theory and practice are the same. In practice, they rarely are.’

This 3:1 ratio is apparently important, because referees often bring it up. At the same time they discount the fact that the tree is fully resolved, which makes me wonder…

Scotland et al. (2003) wrote: “Although the number of characters needed for accurate phylogeny reconstruction is difficult to estimate, the number of characters needed in simulation studies to recover accurate trees is an order of magnitude greater than that available from morphology.” They also claim, “the low character/taxon ratio in many morphological studies itself precludes high support values.”

Of course, in practice (in the large reptile tree), this is proven false. There are high support values throughout (unless two taxa are very incomplete and nest as sisters). So the 3:1 character:taxon ratio has to go the way of the dodo. It’s useless. Based on an average reported back in 1989, it has since then earned some sort of mythical status.

And the same goes for the CI (Consistency Index). 
Paleontologists like to see a high CI. How high is high? Good question. In the large reptile tree it is higher in every subset of the tree (so there’s a clue). I was recently chided for having a very low CI of 0.094 in the large reptile tree (coupled with a completely resolved tree, which, if you think about it, is quite a feat). As you already know, the low CI of the 415 taxon study is due to its very large size and the fact, a well-recognized fact, that homoplasy is rampant within the Amniota with many clades sharing many traits by convergence. In counterpoint, the referee pointed to another large study (Conrad 2008) of squamates that he felt was pert-near ideal. Unfortunately the Conrad study also had a very low CI of 0.1499. That fact must have escaped under the radar.

Well… that’s what happens in large amniote analyses.
It can’t be a bad thing in one case, but acceptable in another.

References
Conrad JL 2008. Phylogeny and systematics of Squamata (Reptilia) based on morphology. Bulletin of the American Museum of Natural History 310:182 pp.
Jenner RA 2004. 
Value of morphological phylogenetics. Accepting Partnership by Submission? Morphological Phylogenetics in a Molecular Millennium. Systematic Biology 53333-359.
Sanderson MJ and Donoghue MJ 1989.
Patterns of variation and levels of homoplasy. Evolution 43:1781-1795.
Scotland RW, Olmstead RG and  Bennett JR 2003. Phylogeny Reconstruction: The Role of Morphology. Systematic Biology 52:539-548.

Advertisements

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 )

Google+ photo

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

Connecting to %s