Evolving Thoughts

Evolution, culture, philosophy and chocolate! John Wilkins' continuing struggle to come to terms with impermanence... "Humanus sum, nihil humanum a me alienum puto" - Terence

Tuesday, May 03, 2005

Barcoding species

Cripes, I'm on a roll today. It's because I am finally set up on the Mac, and am catching up on all kinds of reading. That, and I am off to see Hitchhiker's Guide to the Galaxy tonight (I do not think Adams was God, so I can accommodate the compromises. Whether the film is any good is another matter). Anyway...

My good mate Malte Ebach, a fellow student of Gary Nelson's, is coauthor of a critical letter of a new movement in systematics call "DNA Barcoding" that was published in Nature recently. A response appeared in the next edition, from T Ryan Gregory.

DNA Barcoding is supposed to allow researcher to identify the number of living species through the use of "phylotypes" based on the sequence of a mitochondrial gene cytochrome oxidase I (COI). The idea is that this short sequence can be quickly sequenced using a microchip, so identifying the existence of species in, say, an ecosystem.

Now Malte is not shy of unpopular positions. He is, for instance, an unapologetic pattern cladist. He opposes (as do I) the Phylocode. And here he and his colleague Craig Holdridge attack DNA Barcoding as not being a substitute for taxonomy. Gregory replies that it isn't - the sources of funding are different. Are they? Isn't this just a substitute for taxonomy?

Malte thinks so. Most taxonomy is done in woefully underfunded and understaffed museums. If there is a Big Science project funded by, say, the NSF,that ignores the most important facts about organisms - their morphology, developmental cycle, ecological habitat and behaviors, and so on, in favor of what is effectively just another identification key, we are going to make some very big mistakes about the future and state of organisms.

And Malte thinks (in private email) that it is a mistake to think that that sort of funding won't take away from "alpha taxonomy" - the enterprise of describing species. Taxonomy is already massively in trouble, to the point where people are talking about it needing to be rescued. Some even think it is in danger of extinction itself.

This goes in part to the relative lack of interest in the role of classification in science by methodological discussions. Popper barely mentions it. It is considered, somehow, de trop. The real buzz is in theory construction, or sociology of science, or experimentalism. But, in the nineteenth century, they knew that classification was part of the deal. And in bioinformatics, the trend is to reinvent, and make the same old mistakes, of the past.

Cladistics, or phylogenetic systematics, has been a major player in evolutionary biology and systematics in general, to be sure, but most scientists treat this as a kind of Methodological Black Box. First you build up a data set of characters (never mind how), and then you run that through a program like PAUP* or Phylip. Then you have your classification/history. All's well with the world...

But even in cladistics there has been a conflation of the identification of taxa, and the description of them, at times. The best cladists don't make this mistake, but the field biologists and geneticists tend to. Barcoding is another symptom of that malaise.

In computing terms, a DNA barcode is at best a record ID. The details of the record - the rest of the fields describing the name, address, and personal details of the species - still need to be recorded. Otherwise, all you have are empty records...