Fwd: Re: [TAXACOM] genetic vs morphological trace of phylogeny

John Grehan jgrehan at TPBMAIL.NET
Wed Apr 7 21:25:49 CDT 2004

>Date: Wed, 07 Apr 2004 21:25:25 +1200
>To: Derek Sikes <dsikes at UCALGARY.CA>
>From: John Grehan <jgrehan at tpbmail.net>
>Subject: Re: [TAXACOM] genetic vs morphological trace of phylogeny
>>DNA data are not phenetic. Phenetics is a method of analysis, not a type of
>According to my understanding - yes and no. I agree its a method of
>analysis, that method being the analysis of characters without the
>characters being restricted to shared derived states.
>>Distance data are normally analyzed with phenetic methods (eg NJ or
>>UPGMA) - and the early Chimp/Human/Gorilla work was based on questionable
>>DNA hybridization data which are indeed distance only, no characters. Later
>>studies have used sequence (character) data, however. DNA sequence data can
>>be converted to distances but they normally are not - each site is equal to
>>a character of 4 states. Many morphological characters have 2 states only so
>>DNA is theoretically richer in states than many morphological characters.
>My understanding (which may be wrong) is that the difference between
>phenetics and cladistics is not in the number of character states. Anyway,
>its seems to me that morphological characters can have more than two
>states and some of the orangutan characters are of this kind.
>>Also, please compare these datasets:
>>taxon1  AAATGCC
>>taxon2  AATTGCG
>>taxon1  1112344
>>taxon2  1122343
>>Both have characters that can be analyzed with either parsimony or
>>statistical methods like maximum likelihood, combined or independently.
>Yes they can, but this does not make them cladistic (the example is a
>two-taxon exercise and for a cladistic analysis the example will need to
>be a minimum of four taxa).
>>also that the first contains DNA sequence data and the second contains
>>morphological data. Both datasets, however, have *identical* information.
>I guess I'm a bit dense as I cannot make sense of your morphological
>characters. By the first character '1" I take it this is present in both
>taxa. The second character '1'  - is this a different character? If so why
>label it as a '1'? The third character is also labelled as a '1' but then
>as a '2' in the other taxon. Is this to say it is absent. If these are
>binary characters could you represent them as such? Here's my
>interpretation below (which is probably incorrect as I obviously do not
>understand your notation) placed into a four taxon statement. Could you
>also repeat the genetic data in a four taxon statement as that would be
>most helpful.
>Character 1234567
>taxon1     +++++++
>taxon2     ++0+++0
>taxon3     000+++0
>taxon4     0000000 (outgroup)
>>Although there is no reason to believe morphology is less informative than
>>DNA on a per character basis, it is clear that DNA provides more characters
>>= larger sample sizes and thus less worry about random error producing
>More characters would make a difference if they were of the same status -
>I will wait to see your four taxon statement for the DNA sequences
>(although Ken Kinman has dismissed the numbers game altogether).
>John Grehan
>>Morphology is excellent to combine with DNA because morphological
>>data are a wide, and potentially random, sampling of the nuclear genome
>>(assuming the characters are heritable) - whereas sequence data are usually
>>taken from a single or a few genes of the genome, i.e. they are not widely
>>scattered across the entire genome.
>>Derek S. Sikes, Assistant Professor
>>Division of Zoology
>>Department of Biological Sciences
>>University of Calgary
>>2500 University Drive NW
>>Calgary, Alberta, Canada, T2N 1N4
>>dsikes at ucalgary.ca
>>phone: 403-210-9819
>>FAX:  403-289-9311
>>"Remember that Truth alone is the matter you are in Search after; and if you
>>have been mistaken, let no Vanity reduce you to persist in your mistake."
>>Henry Baker, London, 1785

More information about the Taxacom mailing list