Liveblogging the conference: Mishler

Brent Mishler is a very nice guy who is wrong on a few things - Phylocode, species, and so on - but he's absolutely right about barcoding. He's talking today about so-called DNA barcoding and species concepts.

He says that species are just the least inclusive taxa for whatever view of taxa one has - there is no species problem per se.

He thinks that species are just the smallest monophyletic group worth diagnosing. And more: since biology should be free of taxonomic ranks, species (as a rank) should be dropped. I can agree or not depending on what interpretation is given here. Brent thinks that the Phylocode is the formal implication of this - I don't, because I think he and others have misunderstood the nature of the Linnean system. Another time...

No... it seems Brent is going to expound the Phylocode logic anyway. He thinks that the ranks imply particular types have a rank that is vulnerable to revision. Which I agree on. But then he makes a nomenclatural point - there aren't enough ranks. I agree also, but mandating some ranks doesn't mean that one cannot intersperse names between ranks.

He dislikes changing the ranks (and having disputes about what ranks are appropriate). But I think again all classifications are subject to revision in the light of new knowledge. One point, that I have made before: ranks are incommensurable - you can't compare ranks across all clades. Brent points out that even systematists make this mistake. Rankfree now!

Brent argues that this applies to the species rank as well. He wants to have monophyletic species, with the Phylocodic uninomial replacing the Linnean binomial. A short discussion on the ways the Phylocode would be employed follows. A terminal clade is merely a state-of-the-art report on diagnosis of the smallest as-yet identified monophyletic group.

On to DNA Barcoding: using a short stretch of the COXI mtDNA gene to diagnose species. Brent rightly says that this is both new and interesting, as Johnson noted, but the new bits aren't interesting and the interesting bits aren't new. Using a stretch of DNA to find a possible new taxon is not a bad idea, but using it to name a new taxon is a bad idea. One gene is insufficient.

Using the Barcode is going back to the Aristotelian view that a species is a type" (no, it isn't, Brent, but he knows my views on this). Barcoding syphons off money and resources from real systematics. It returns to single key classifications (it does). We need an integrated taxonomy (as he, Will and Wheeler said in 2005) to identify species.