Barcoding (animal) life

[pierre deleporte]

A 02:09 04/02/2003 -0600, Andrew Mitchell wrote :

>Nobody is proposing a species description that reads: !! !! !   ! !...
>ACGGTAT... without at least a few good pictures of the beast (including
>genitalia or other important features where appropriate, perhaps?) and
>collection data, etc. (although that was indeed a witty comment someone
>made—sorry I can’t find the pipe character on my PC keyboard to do it
>justice). Clearly detailed morphological work by a trained expert is
>preferable, but how long must we wait? There’s no need to rehash the
>figures we’ve seen on this listserv on the numbers of undescribed taxa
>versus the number of taxonomists—we all know it’s a dire situation. And
>just how much time do we have to catalogue the earth’s biodiversity before
>it is gone? But there’s no arguing the point that DNA sequencing is faster
>than good morphological work. The “barcoding” solution can never and will
>never be the final solution but it sure as hell will give us a usable
>roadmap. The DNA sequences and resulting gene trees will form the backbone
>for exploration of uncharted territory. And THAT is the point. As I see it
>this is not a “molecules versus morphology” argument (if it is to you, then
>get over it already... that debate fizzled out in the last decade!).
>Barcoding will advance systematics, period.

My point with the "IIII II III - ATGGTAT" joke was just to emphasize that 
molecular data cannot reasonably stand by themselves without loosing all 
practical utility, which point was repeatedly made in other posts. We must 
answer the question : "a sequence of what?" some intelligible way (the 
preserved specimen, or some specimens from likely the same population for 
tiny animals, and some morphological description). And on that basis, some 
elementary naming will likely come together (an insect, an orthoptera, a 
possible acridid...).

The point at stake, in my view, is: "fast publication on a rough 
description basis". Barcoding can make sense as some convenient retrieval 
system before a classic binominal is given to the critter in question. Or 
even as an extra, permanent retrieval system lasting even when the species 
will be named. But in my view fast publication and barcode have no specific 
relation to DNA sequencing. A rough morphological description can play the 
trick as well. I fear that a limited series of molecular data will 
generally do no more no less than a limited series of morphological ones 
for, e.g., phylogenetic inference. Thus, fast publication on a limited 
morphological basis is just a little worse than the same plus some rough 
molecular data.

As for rapidity: one sequence from one biopsy can be extracted, amplified, 
read and posted  very fast indeed. And then another one. Till the moment 
when you run out of funding.
In the real life, there's likely some precious time spend by the 
morphologist to tediously ascertain homologies while some precious time is 
waisted by the molecularist waiting for the next grant.

Now, I have little objection to considering some molecular data as a useful 
extra layer (not the first layer) of description of deposited specimens, 
and fast/rough publication (or posting) as a possibly useful emergency 
strategy. Would make sense if these rough data could be of some use. I 
guess they could at least help avoiding redundant investigations on already 
"discovered" taxa, if nothing else. Or pointing rapidly to rare and 
endangered species. The very roughness of the descriptions (and inferred 
phylogenies) would thus make them possibly misleading, but better than 
nothing (?).


Pierre


Pierre Deleporte
CNRS UMR 6552 - Station Biologique de Paimpont
F-35380 Paimpont   FRANCE
Téléphone : 02 99 61 81 66
Télécopie : 02 99 61 81 88