Stéphane Bouee wrote: > I receive the mailing list of phylogeny and morphometric geometry and I post > the same message to both lists.
I am replying on the R-sig-phylo mailing list -- will you please forward this to the other relevant lists? > This is about doing phylogeny with morphometric geometry, a subject that > could bring an anathema on me because 1) quantitative data are generally > considered to be unsuitable for doing phylogeny, especially with a cladistic > method, 2) among quantitative data, morphometric ones are considred to be > one of the worst. > Although I understand and agree for the second reason, I never really > understood why cladistic could not be performed with quantitative data. I > understand that there are technical reasons but I think there are other > historical reasons such as the long war between phenetic and cladistic, that > still bring a doubt on studies doing phylogeny with quantitative data. It may depend on who you are talking to. People who consider parsimony methods the only good ones tend to be very unhappy with quantitative measurements and consider discrete characters the only valid ones. But other people, such as statistical phylogeny people who work on comparative methods, use continuous scales enthusiastically. > Let's come to the point, I recently discovered 2 articles that, to my > opinion, threw a stone in the pond as we say in French: > 1. Phylogenetic morphometrics (I): the use of landmark data in a > phylogenetic framework. Santiago A. Catalano, Pablo A. Goloboff, Norberto P. > Giannini, Cladistics, > <http://onlinelibrary.wiley.com/doi/10.1111/cla.2010.26.issue-5/issuetoc> > Volume 26, Issue 5, pages 539?549, October 2010 > 2. Phylogenetic morphometrics (II): algorithms for landmark > optimization. Santiago A. Catalano, Pablo A. Goloboff, Norberto P. Giannini, > Cladistics Article first published online: 28 JUN 2010 DOI: > 10.1111/j.1096-0031.2010.00318.x This is part of a move away from the discrete-characters-only position on the part of the Willi Hennig Society types. Actually in some of the early papers by Farris and Kluge (1969, 1970) continuous scales were used in parsimony methods. Afterwards there was an anathema within the WHS against them. The miasma is only beginning to clear now. > Since several years I try to use my morphometric data in a phylogeny > perspective. I used several methods, including the maximum likelyhood method > with Philips software of Mister Felsentein and the cladistic method of TNT > software of Mister Goloboff, one of the author of the 2 papers. You spelled Goloboff's name correctly ... > I just wanted to express my opinion about this matter and I think those 2 > papers will bring the pylogeny perspective in a new paradigm. The use of morphometric methods together with phylogenies is an important area of work. Fred Bookstein and I are working on some methods that may improve on Goloboff and Catalano's method. For now, their approach, which makes a Procrustes fit of the forms and then infers ancestral states using distances in that space, is not bad ... but we think we can do better. I talked on this at the Evolution2009 meeting (for 13 minutes, anyway) and we hope to have more methods available soon. Obviously I prefer statistically-based methods such as likelihood (or Bayesian) ones instead of parsimony, but I won't bore people with that. Joe ---- Joe Felsenstein j...@gs.washington.edu Department of Genome Sciences and Department of Biology, University of Washington, Box 355065, Seattle, WA 98195-5065 USA _______________________________________________ R-sig-phylo mailing list R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
