Barcoding (animal) life

Andrew Mitchell MitchellA at NU.AC.ZA
Tue Feb 4 17:18:22 CST 2003


Richard Zander wrote:
>Great statement by A. Mitchell . . . we needed a clear political
exposition
>on the subject.
Well you know what they say about sarcasm as a form of wit!

>Note the figure mentioned of 1 billion dollars.
Do you think the cost is excessive? The human genome project is
estimated to have cost $3 billion over 15 years. Why not $1 billion over
20 years as suggested. This is money that might otherwise be spent on
tax cuts for the rich or something like that (I assume that few
taxacomers fall into that category).

>This is not for sequencing
>all the genes of all taxa, but for sampling.
Sequencing all the genes of all the taxa? THAT would indeed be a waste
of time and money. Why determine a few billion nucleotides from each
species when a few thousand will do the job? 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.

>Samples don't include all the
>genetic info we need and don't include important associated
phenological,
>ecological, biogeographical info.
As I stated, how long should we wait for people to gather all that
other info while biodiversity is disappering at an unprecedented rate?
And why shouldn't we proceed with DNA sampling in the mean time?
Molecular and morphological/phenological/ecological etc. work can
proceed simultaneously.

>What (selected) small samples do (and this is indeed marvelous!) is
>approximately match our general intuitive ideas of evolutionary
>relationships. But that's all.
On the contrary, I find that "(selected) small samples" more often than
not yield a few surprises, leading to the collection of further data
which either corroborate the initial novel finding or simply underscore
the importance of taxon sampling density (which is why at least $1
billion would be needed).

>Demonstrable reliability is lacking for DNA
>results in most cases for resolving actual problems, such as (1)
certain
>questions of relationship not soluble with morphology and (2)
selection of
>one solution when relationships implied by morphology and DNA, or by
>different genes, conflict.
>
>This is in addition to the usual horrors of sample error (wrong
sequences),
>dubious alignment, model choice when there are reasonable
alternatives, and
>differential gene histories, all of which can reduce the chance of
being
>correct.

Your points about taxon sampling density, horizontal transfer, etc. are
well taken and have to be considered in the planning stages of a
"barcoding" project, but they do not in themselves constitute reasons
for not starting such a venture. Nobody pretends that DNA data is
perfect. I hope you wouldn't pretend that morphological data is perfect.
However, your dismissal of DNA data is cavalier. Sure there are some
poor molecular studies in the published literature (as indeed there are
bad morphological studies) but many are preliminary studies that have
gone on to produce very robust phylogenies once more data has been
gathered. Those "progress report" papers are necessary to assess
progress, generate continued funding if appropriate, and they provide
useful heuristics for further research of both molecular and
morphological persuasions.

>Our main problem is that molecular analysis remains a crap shoot. Take
any
> <deleted>

Yaaaaawn... as I said previously, I won't be drawn into an outdated
line of argument that polarizes molecules and morphology (4 legs good, 2
legs bad!).

Andrew

------------------------------------------------------------------------
Andrew Mitchell
Senior Lecturer, Molecular Phylogenetics
School of Molecular and Cellular Biosciences
University of Natal
Private Bag X01
Scottsville, 3209
SOUTH AFRICA

Tel:  +27 (0)33 260 5815
Fax: +27 (0)33 260 5462

http://www.nu.ac.za/biogen/mitchell.htm




More information about the Taxacom mailing list