[Taxacom] FW: DNA only descriptions

[Thomas Simonsen]

________________________________

From: taxacom-bounces at mailman.nhm.ku.edu on behalf of Kim van der Linde
Sent: Mon 16/08/2010 17:38
To: Taxacom
Subject: Re: [Taxacom] FW: DNA only descriptions





On 8/16/2010 12:25 PM, Richard Zander wrote:

> Of course, morphology alone is usually useful for identification, and DNA isn't.

Well, DNA barcoding is exactly developed for doing just that, with the
added advantage that you do not need to know all your species.

For insects, we only need one leg (even from a living individual) to
make an ID, without even knowing what family those individuals are in.


Well, Sergio has already made several good points so I will only comment on the claim above.

1) There are several good examples in insect systematics where DNA barcoding simply doesn't work: Schmidt & Sperling (2008) on Grammia, and Lumley & Sperling (2010) on Choristoneura spring to mind, but there are several others. These papers also show that you have to be very, very careful using barcodes to delimit species (i.e. identifying NEW species).

2) For DNA barcoding to work the first prerequisite is that you have an elaborate DNA library of all the species in question. By elaborate I mean 100s or even 1000s of sequences from all over the species rage, and not just a couple of dozen sequences from one part of the range (unless it's a seriously restricted-range species) - how can you otherwise be sure to cover the entire intraspecific variation? And all the specimens from which these sequences came HAVE to be confidently identified by other means than COI sequences (i.e. morphology, ecology, etc.) - otherwise the argument becomes perfectly circular. Despite the claim above, we still need to be able to identify specimens and species "old fashioned way".

DNA taxonomy relying on multiple loci, SNIPs, and so on is a different ball game altogether, not least because it generally relies on analysing the phyloGENETIC relationships of the groups in question instead of the phyloPHENETIC N-J analyses often used in barcoding. In reality, basing a species delimitation on a single locus (or rather half a locus), a (fixed) percentage difference, and an N-J analysis is comparable to numerical phenetics. 

Don't get me wrong, I do find DNA barcoding very useful within its limits (as is of course the case with all our taxonomic and systematic tools). It is particularly useful to quickly pinpoint specimens or clusters of specimens that can then be singled out for closer examination (using a wide selection of tools).

Thomas

----------------------------------------- 

Thomas J. Simonsen, PhD. 
Researcher in Lepidoptera 
Department of Entomology 
The Natural History Museum 
Cromwell Road, London 
SW7 5BD, United Kingdom 
+44 020-7942-6548
-----------------------------------------