"types" - was Barcoding (animal) life

Michael Bayly MichaelB at TEPAPA.GOVT.NZ
Wed Feb 5 10:24:11 CST 2003


Stephen,
The "Type" concept is a nomenclatural one - and statistics don't really come
into the matter, as I see it.  The concept ties a name to a specimen, and
therefore makes the name definable in an exact sense (in that it refers at
least to that specimen).  How widely you apply that name (e.g. to a range of
other specimens) is a matter of interpretation, and there will often be
seriously conflicting views on such matters.

Modifying the type concept such that you would need multiple (geographically
spread) specimens, would just be asking for trouble.  Two biologist could
easily disagree as to whether the "types" represent one or more biological
entities (species, subsp. etc) - no matter how carefully the work is done.
If conflicting later classifications are proposed, what then happens about
the application of names? You would probably end up (through
lectotypification) with a re-narrowing of type concept for particular taxa -
then aren't you back where you started?

Obviously for any study that seeks to assess the circumscriptions and/or
relationships of taxa (using molecules, morphology, whatever) you need to
include reasonable sample sizes to ensure the soundness of your conclusions.
You also need to know where the "type" sits in relation to your assessment
of variation - so you know how to apply names to the taxa you perceive.  But
there is no need to think of "types" as representative of all (or most)
variation in the taxa you recognise.


Mike


----------------
Dr Michael Bayly
Research Scientist in Plant Biosystematics
Museum of New Zealand Te Papa Tongarewa
PO Box 467
Wellington, New Zealand
Ph: +644 381 7262
Fax: +644 381 7070



-----Original Message-----
From: STEPHEN MANNING [mailto:SDMANNING at ASUB.EDU]
Sent: Wednesday, 5 February 2003 9:31 a.m.
To: TAXACOM at USOBI.ORG
Subject: Re: Barcoding (animal) life


At 11:12 AM 2/4/03 -0600, Richard Zander wrote:
>A.M.:
>By reliably identifying organisms, you mean like forensic DNA typing? Is
>that what taxonomists do? Identify organisms?
>
>I think taxonomists do a lot more. As to money, we taxonomists have 250
>years of the Linnaean endeavor behind us and have generated or mediated
>great economic gains for the human race. A billion dollars is to laugh at.
>(Oh, that's slightly rhetorical, since my 2004 research budget is presently
>zero.)
>
>Sure, I agree with you that molecular and morphological work can proceed
>simultaneously. The fact that it is mainly molecular work that is now
funded
>(grants, positions) indicates that such synergy must be rare and getting
>rarer. The Planetary Biodiversity Inventory now supported by the U.S.'s
>National Science Foundation is a worthy exception.
>
>_Everybody_ ducks my concerns about the statistical basis behind
reliability
>of phylogenetic estimation. Your response, Andrew, ("Yawwn") is at least
>novel.

Richard et al.,

We have some things in common!  First, my research budget for 2004 is also
zero.  Second, I have been troubled for quite awhile by the thought that
phylogenetic analyses even at the family or higher level, let alone at the
species level, based on either DNA or morphology, may be based on single
specimen samples of the taxa represented.  It would seem that elementary
statistics would dictate a sample size of AT LEAST ten randomly selected
specimens be used for either DNA analysis or morphological analysis of the
variation or lack thereof; and a sample of at least 100 specimens per taxon
would be preferable and by no means excessive.

I attribute most of the problem to the fact that taxonomists have not,
traditionally, been statistically oriented.  Even the "type specimen"
concept suffers from this deficiency.  There is no real reason to believe
that a type specimen represents an average genotype or phenotype for the
taxon of which it is the type, with obvious potential for error in
phylogenetic and other analyses even if, as I think someone pointed out
recently in this forum, the type specimen is what should be used to extract
DNA for analysis.  Maybe the codes need to be modified to require a sample
of ten or more specimens, as randomly selected and hopefully as
ecologically and geographically diverse as possible, to qualify as the
"type" of a taxon for future species descriptions, unless fewer than ten
specimens are available.  I acknowledge that it probably won't be easy to
require broadening the type specimen concept to the point of assuring its
statistical validity, but nothing ventured nothing gained.

If we ever do get to the point of using Bar Codes or other means of
categorizing DNA diversity, I hope it can be made essentially mandatory
that a sample of a statistically valid number of specimens (10?) (100?) and
gene sequences per specimen (10?) (100?) be included in the DNA analysis in
order for the analysis to be recognized as valid.  If this is done up front
and becomes the tradition, it will be easier than it probably would to be
to wait until later to do what is necessary for statistical validity, when
less valid approaches would have become the norm.

If  the above were implemented, it ought to be able to generate more
funding for both morphological and molecular taxonomists!

Steve Manning

>R.Z.
>______________________
>Richard H. Zander
>Bryology Group
>Missouri Botanical Garden
>PO Box 299
>St. Louis, MO 63166-0299
>richard.zander at mobot.org <mailto:richard.zander at mobot.org>
>Voice: 314-577-5180
>
>
>
>-----Original Message-----
>From: Andrew Mitchell [mailto:MitchellA at NU.AC.ZA]
>Sent: Tuesday, February 04, 2003 9:18 AM
>To: TAXACOM at USOBI.ORG
>Subject: Re: [TAXACOM] Barcoding (animal) life
>
>We *sample* taxa not only
>because of limits of time and money but because of the law of
>diminishing returns. We want to be able to reliably identify organisms
>and as an added bonus work out their evolutionary history, not count all
>stomata on all specimens in the herbarium just for the fun of it.
>
>
>Molecular and morphological/phenological/ecological etc. work can
>proceed simultaneously.
>
> >

Stephen D. Manning, Ph.D.
Professor of Biology
Mathematics and Science Division
Arkansas State University - Beebe
P. O. Box 1000
Beebe, Arkansas 72012-1000
USA
Tel: 501-882-7162

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Visit the Te Papa website http://www.tepapa.govt.nz
The email message together with the accompanying attachments may be
CONFIDENTIAL. If you have received this message in error, please notify
mail at tepapa.govt.nz immediately and delete the original message. The
views expressed in this message are those of the individual sender,
except where the sender specifically states them to be views of Te Papa.
Te Papa employs strict virus checking measures and accepts no liability
for any loss caused either directly or indirectly by a virus arising
from the use of this message or any attached file.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++




More information about the Taxacom mailing list