[Taxacom] FW: cladistic analysis for morphological characters -- UPGMA is not cladistics

Ashley Nicholas Nicholasa at ukzn.ac.za
Tue Nov 13 08:39:51 CST 2012

Agreed -- if it is acknowledge that the reulting phenogram is assummed to be an approximation of an evolutionary tree then I have no problem. The words assummed and approximation are important. Many paper do not acknowledge the uncertainty that surrounds their phenetically produced hypotheses because of these assumtions and approximations. I am not saying don't use these analyses, but I do feel we should own up to the flaws inherent in all methodologies.

Classifications produced using only morphology are always treated as too subjective (and I agree -- they are) however, many other methods especially those that are use for molecular systematics ignore the assummtions they also make. Such as evolution always being parsimonious, that molecular sequences are never analagous, that evolution is only cladogenic (ignoring anagenic evolution, hybridization [esp. in plants], tokogeny and polygenic evolution [where metapopulations of a widespread species is undergoing multiple divergence and synchronised but quite separate episodes of isolation]). Monospecific genera are automatically nested into the clade below -- and any unique feature on any other level of manifestation is just ignored. I know of labs that don't bother to run control lanes when running gels. So called evolutionary trees based on one gene (usually a plastid gene) are regarded as representing the evolution of those organisms when in fact they are just an evolutionary tree of that gene. Horizontal gene transfer and epigenetics are also ignored. And, then, there is the problem of sample size with large genera that may have as many as 300 species being used to generate cladograms/hypotheses based on  1% to 2% sample size.

I embrace both morphology and genetics (and other levels of manifestation) -- but it is getting just a little annoying that some disciplines are acting more eletist than others; especially when a close examination of their fields shows them to also be flawed in some ways. Not that we want funding bodies to know this of course.

Cladograms, in the end, are also just hypotheses.


-----Original Message-----
From: taxacom-bounces at mailman.nhm.ku.edu [mailto:taxacom-bounces at mailman.nhm.ku.edu] On Behalf Of Richard Jensen
Sent: 13 November 2012 15:46
To: taxacom at mailman.nhm.ku.edu
Subject: Re: [Taxacom] FW: cladistic analysis for morphological characters -- UPGMA is not cladistics

Ashley's comments are clearly reminiscent of what Sneath & Sokal wrote in their seminal texts (Sokal & Sneath 1963; Sneath & Sokal, 1973) on numerical taxonomy.  They referred to the 60 character minimum as well as to the "matches asymptote" - as the number of characters increases, the measure of similarity (or dissimilarity) among OTUs is likely to stabilize.  They were also explicit in stating that, while phenograms are not intended to reflect evolutionary history, they will, in many cases, be a good first approximation of such.

UPGMA phenograms are based on (dis)similarity matrices, and there are a number of choices for determining  pairwise OTU relationships.  If one wishes to use a UPGMA phenogram as an approximation of evolutionary relationships, then one can choose a measure that is consistent with the kinds of data and a clustering algorithm appropriate both for the similarity measure used and the relative numbers of OTUs in different subsets of the "ingroup". Colless, Estabrook and others discussed various approaches that may yield good approximations.


Dick J

On 11/13/2012 7:01 AM, Ashley Nicholas wrote:
> John you are right,
> UPGMA is a phenetics method and is not eplicitly evolutionary. It only measures similarity, and similarity is not always a good indicator of descent from a common ancestor. This is especially true in flowering plants where convergent evolution/homoplasy is rife.
> Analogous (rather than homologous) base pair sequences are probably less common than in morphology -- so maybe molecular systematists can get away with approximating it to an evolutionary tree. However, in the end it is not an explicit evolutionary tree -- and this needs to be acknowledged rather than ignored (which is what usually happens). However, no matter what, the resulting phenogram is a hypothesis. This hypothesis is as valid as any other hypothesis (until falsified) -- and probably carries some interesting insightes and may generate some interesting questions for further explorations.
> The text books say a minum of 60 characters is needed but I would think the number of characters needed would depend on the size of the group being analysed. Some statistician has probably established this??
> Regards
> Ashley
======= Please find our Email Disclaimer here-->: http://www.ukzn.ac.za/disclaimer =======

More information about the Taxacom mailing list