The end of the line for taxonomists - or is it?

[Felix Sperling]

Barry,

Although I don't know of any articles that are exactly what you're
asking for, here is one that is close:

Funk, D.J. and K.E. Omland. 2003. Ann. Rev. Ecol. Evol. Syst. 23: 397-423.

This is a large scale review of species-level monophyly in animal
mitochondrial DNA. Based on many hundreds of studies, 23% of 2319
assayed species show species-level paraphyly or polyphyly. This was
quite widespread, with 44% of all genera having at least one species
that were not monophyletic with respect to mtDNA.

This does not mean that "barcoding" is useless, but rather that it is
currently being seriously oversold. The most important message of the
Funk and Omland review is that it is crucial to sample variation
across the range of species, and to have credible sample sizes. This
is exactly what I don't see the barcode enthusiasts doing. Setting up
a complex barcoding scheme based on sampling just one or a handful of
specimens from one locality (as is the case in the much-cited papers
that supposedly establish the effectiveness of barcoding) or small
subset of the species range is just asking for trouble later.

I'm reminded of media stories reporting the latest cure for cancer.
There is often something of value behind it, but it is a small,
incremental advance, not the end of all our problems. Meanwhile, real
harm can be done in diverting funding away from the careful taxonomic
work that builds the foundation for us all.

Felix Sperling


>Yes.  Nevertheless, because some end-users of taxonomy are likely to
>buy into the scheme (perhaps even eventually writing it in as a
>requirement for grants and contracts), we need a strong statement in
>print concerning its unreliability:  a cogently stated,
>not-too-long, article, preferably multi-authored, in a journal with
>good "street cred," that we taxonomists could show our clients when
>the barcode bugbear arises.  Does anything like this exist at
>present?  Even a short bibliography of existing papers that
>demonstrate Doug's point (with salient quotes provided) would be
>helpful.
>
>Barry Roth