DNA Barcoding is intellectually bankrupt. But it works.

One of the most vociferous debates in taxonomy is over a catchy-sounding concept called DNA barcoding.  Since nearly all organisms carry a version of the COI gene in the mitochondrion, the idea is that the DNA sequence of the gene can serve as a standard identification marker.  A barcode, of sorts.  Of course, the practice only works if species have unique COI sequences.  Which they do, much of the time, and the barcoders consequently have been successful in garnering research money and churning out publications.

So what’s the problem? There are two major objections.  The first is largely philosophical.  The intellectual backbone of DNA barcoding is conceptually the same as creationism- that species are unique entities that neither blend with nor give rise to other species-  and about as vacuous.*  A typological concept of species flies in the face of decades of empirical and theoretical research in population genetics documenting all sorts of situations- from hybridization to lineage sorting- where single genes do not reflect species boundaries.  So the very notion of a unique genetic identifier is offensive to a great many biologists.

The second objection is more political and academic. DNA barcoding allows anyone with access to sequencers to identify just about anything.  The horror!  Without any knowledge of mycology, for instance, I can sequence a mushroom and put a name on it.  Specialists in the taxonomy of particular groups are perceived as no longer needed, money and positions cease flowing to traditional taxonomists, and we lose a great deal of knowledge.  I don’t think this is actually happening (positions ceased flowing to taxonomists long before DNA barcoding came along), but it’s understandable that systematists feel threatened.

Having said all that, though, most criticisms of DNA barcoding in practice fall flat.  However vapid the barcode concept is, the fact remains that morphology-based identifications also have a signficant error rate, and it is normally substantially higher than that of mitochondrial DNA-based identification.   Until critics come to grips with the foibles of traditional identifications, they are just shouting into the wind.  What good is pointing out that the barcoding approach fails 5% of the time when the alternative fails 10% of the time?

Still, barcoding is monumentally stupid terminology when applied to biology.  Species aren’t manufactered by make and model, even if implying that they are brings in research dollars.  Can’t we figure out a way to incorporate DNA into taxonomy without resorting to marketeering?

*One of my least favorite science sites is the Barcode Blog, which has all the tone and introspection of a late-night infomercial.

14 thoughts on “DNA Barcoding is intellectually bankrupt. But it works.”

  1. what spurred that rant on? i agree tho, i hate the term barcoding and i hate that people who do it don’t think like population geneticists…especially since most high profile pubs deal with crypsis (i.e. recent speciation events with reproductive isolation but little morphological separation), a situation where population genetic processes are MOST likely to have an effect on inference.

  2. Oh yes. The rant was a response to a Molecular Ecology Resources special issue on DNA barcoding.

    Plus, our street has turned into a lake from all the rain this morning and I didn’t have anything better to do.

  3. oh yeah i saw that issue. blah blah blah. i’m debating making a rush for campus right now….hrm.

  4. You should take a look at Applied DNA Sciences. They are incorporating botanical DNA into anti-counterfeiting markers. No insects, but worth a look… I do biz for them.

  5. DNA Barcoding is no substitute for traditional systematics and taxonomy. However, it can nicely supplement the work that is already being done. In my research, I’m using COI gene sequences from adult insects to associate them with the immature stages, not so that we never have to identify the immatures with more traditional methods. Rather, genetic barcodes added to the other morphological and geographic differences allow me to confirm with much more consistency (and much more quickly) that the larva I have in hand is indeed the immature state of So and-so. I think this is a very useful tool in associations ALONG with other tools.

    I mean, the biomonitoring professionals don’t just use one metric, why should we?


  6. Alex – As you wrote, the strongest objections are political but I don’t think is an issue of fear about loosing a job as a taxonomist once identification becomes molecular and automated. It is more an issue of how the whole DNA-barcode discourse is base on throwing dirt at a discipline that is already looked down upon, and presenting itself as a replacement of taxonomy rather than as a powerful tool in the taxonomist’s arsenal.

    Comparing identification error rates between different markers is all good and useful. But then in DNA-barcode papers, like the one by Packer et al. on that special issue, they go ahead and reduce a whole discipline to just identification keys.

    As you said, it works. It is here to stay. Lets move on. There is no need to trash morphology with titles like “the mediocrity of morphology” to sell DNA-barcodes. They already sell very well, as a cursory inspection of the jobs being advertised in the field will attest.

  7. Why does it have to be one or the other? Simply identifying the species while ignoring the biology(incl morphology) to me seems fairly pointless. Perhaps it can be useful to ecologists with whatever they do, but it’s quite irritating when all those resources are spent on identifying and barcoding things while the biology of some of those wonderful organisms remains limited to a small handful of arcane, ancient publications.

    That said, classification based only on morphology is quite dangerous; convergent evolution seems more common than previously thought, especially among protists (most non-photosynthetic heterokonts were thought to belong elsewhere in the tree prior to molecular data; they share convergence with things as distant and varied as fungi, ‘heliozoa’, small excavates and ciliates!). So as Roberto says above, why can’t it just be used as a extra tool as opposed to replacing others?

    Biology is plagued by endless hypes and fads…and rather disgustingly cliquish at times; hmmm, speaking of which, I’m employed by botanists these days, and aspire to become a developmental cell biologist: therefore I really shouldn’t be reading this blog… shhh! ^_~

  8. I don’t know what the fuss is all abut. Barcoding will have hard times supplant good old morphology, unless some sort of device is invented that can be used in the field (think Star Trek Tricorder™) in which case, morphology will become a supporting discipline. But I think we’re far from that point. Until then, we’re “stuck” with old fashioned specimen collection and often long and strenuous identification.

  9. I don’t think barcoding is all that useless. It may not be a perfect solution, but, like you said, taxonomy is a dieing, if not dead, science, there are still many species yet to be described, and I think there are countless potential uses for it in disciplines such as ecology.

  10. Justin-

    The common perception is that taxonomy is dying, but I don’t think the evidence bears that out. In terms of the number of new publications and species descriptions, taxonomy is undergoing something of a renaissance. We’re more active now than ever.

    I think where the idea comes from is that the current flowering of taxonomic activity is buried by the explosive growth of Big Biology, so that it’s hard for people to notice amongst all the genomes that taxonomy is being done.

    The other problem is that since taxonomists no longer occupy the majority of large university research positions (as they did 100 years ago), they lack the academic visibility they used to enjoy. But that doesn’t mean that scores of students, hobbyists, museum curators, and faculty at smaller institutions aren’t active in the field.

  11. Hi Alex – happened by myrmecos this morning, saw your essay and the comments that have piled up over the past several days and couldn’t help add my two cents.

    1) A barcode is a epistemological tool, an ‘evaluation criteria’ for identification and can act as a catalyst for discovery. It is not an ontological truth that defines a species. Anyone who has proposed barcoding as a concept is clearly incorrect – but has anyone? “We test the epistemological hypothesis that the barcode MOTU can be used as a surrogate for the identification of diversity within and between collection localities.” (Smith et al 2005). To paraphrase Fitzhugh (2008) barcodes can be used as an “epistemological bases for indicating phylogenetic hypotheses”.

    2) Identifying the species and ignoring the biology? Much of what the people that I work with have tried to do have been far more integrative than such a statement suggests (Fisher and Smith 2008, Smith et al 2008, 2007, 2006). Again – I know of no one that would support the idea to ‘ignore biology’.

    3) Many of the pros for integrating DNA into existing taxonomy, ecology, and evolutionary biology programs are accentuated if one of the genes is a standardised marker that permits comparison between localities, researchers and taxa. This is a clear plus to inclusion of a standardized gene – and but clearly stops short of being an argument for stopping all questions after this first pass evalulation.

    4) ‘the barcoders’ – This is a straw man. There is no mob of phylogenetically illiterate cabal of taxonomy-hating gel jockeys. I hope that the moment in time when ‘us, them’ language can be used to characterise, or be made caricatures of, is passing. This is a tool – I know of no contractors who call themselves ‘#2 Robertson’s” because that is their standard first pass screw when constructing a home. There is no them.

    Will a standardized single gene identifier solve all problems in biology? Un-equivically no. Could it make future hypotheses regarding species membership and terminology more transparent and reproducible – yes. Many people’s distaste for the metaphor is abundantly clear, but I’ve got to tell you that what I’ve experienced for the past five years is that when talking to many people on the planet (the non-biologists, the teachers, the government bureaucrat,, the administrators, the… generally majority of humanity) – the edges of the metaphor where it is biologically meaningless become less relevant. What is more immediately relevant is that ‘they ‘understand an IUPAC identifier; this code gives me information about something (most often the price). Is co-opting this societal understanding to add to some degree of biological literacy “marketeering” or is it communicating science?

    Just some thoughts on what could obviously be a much longer discussion! Hope the rain’s stopped there, hasn’t here.

  12. Pingback: DNA Barcoding, a reply from Alex Smith « Myrmecos Blog

  13. Pingback: DNA Barcoding, an unwilling demonstration « Myrmecos Blog

Leave a Reply