[Taxacom] barcode of life wins Ebbe Nielsen Prize

[Doug Yanega]

David Schindel wrote:

>Taxacomers may (and alas, probably will) argue over the merits of 
>minimalism versus the robustness of whole genome data.

That's not the argument here. I think for many of us it's more a 
matter of people using a mitochondrial marker - with all of the 
associated non-Mendelian pitfalls - as an arbitrary delimiter of the 
biological entities we know as species. Other people seem to be using 
COI to work out higher-level phylogeny, as well, where a very 
fast-evolving gene like COI is not really appropriate. If there are 
merits to be argued, it's COI's merits versus non-mitochondrial genes 
in the former case, and versus slow-evolving genes in the latter 
case. Both cases offer good arguments against COI, too.

>   Unfortunately, there is still a real-world trade-off between depth 
>and breadth of taxonomic coverage of sequencing and it will likely 
>persist for a while.  Should we really conclude that despite 
>elevated extinction rates, we should wait for whole genomes to 
>register species in GenBank?

I don't think that is what Donat was saying, nor have I heard any 
other folks raise such an objection; the objections are usually along 
the lines of overemphasis on a SINGLE organismal character - the COI 
sequence - as if it, and it alone, could and *should* tell us what 
was or was not a species.

It seems not very long ago when people who discovered differences 
between geographically isolated populations of one species were often 
quite happy to use "subspecies" to describe this variation. Now it's 
as if an entire generation of taxonomists have completely forgotten 
that this taxonomic category ever existed, and almost act as if it is 
*impossible* for geographically isolated populations to have genetic 
differences between them and still belong to the same species. Did 
the rest of us all miss the memo telling us that subspecies don't 
exist any more? When one person can conclude that we are dealing 
with, say, a single species that has 30 identifiable genetic 
variants, and another can conclude (from the exact same evidence) 
that this organism is actually 30 cryptic species, then we have a 
fundamental difference in opinion on what a "species" is. If those 
two people then have to make their case to the public and political 
community regarding, say, the conservation of species, the first - 
and possibly the ONLY - thing that any non-scientist is going to 
notice is that these two scientists do not agree, and if the 
scientists don't agree, then it's unlikely *either* side will be 
taken seriously. Or, given human nature, if the two opinions a 
politician hears is "This is a common, widespread species, and it is 
not threatened by this planned solar power array" versus "This is a 
complex of 30 species, each of which occurs in such a small area that 
any one of them could go extinct at any time, and 3 of them will 
probably go extinct if this solar power array is built", then the 
latter scenario is not very likely to be what anyone wants to hear. 
Just look at the "debate" over climate change. It's no longer just a 
purely academic exercise at that point.

Ultimately, the claim that:

"CBOL regards barcode sequences as amazingly reliable data (not 
perfect, not fool-proof, just amazing) for species identification - 
more reliable than any other character of similar dimension."

is *entirely* dependent on the associated belief that species limits 
can be defined arbitrarily (e.g., as 4% sequence divergence, rather 
than 5%, or 3%, or some other number). Where is the published 
evidence that organisms with a 3% sequence divergence can interbreed, 
but those with a 4% divergence *can't* interbreed? How can one prove 
or disprove, say, that a COI haplotype moved from one species to 
another via hybrid introgression? It's as if people believe that COI 
just *happens* to evolve at EXACTLY the same rate that speciation 
occurs - *never* faster (because then one species could have many 
different COI haplotypes, including variants exceeding 4% 
divergence), and *never* slower or laterally (because then more than 
one species could share the same COI haplotype, or haplotypes with 
divergence below 4%). How realistic is that??

If one does not share that belief, then the claim that COI is a 
reliable species-level marker looks like a massive tautology. What, 
after all, is the *independent test* that can be used to support or 
refute the hypothesis that any given "species" has a distinct *and* 
unique COI profile? Building a tree of haplotypes is not a test of 
species boundaries. Isn't science supposed to be based on testable 
hypotheses??

Sincerely,
-- 

Doug Yanega        Dept. of Entomology         Entomology Research Museum
Univ. of California, Riverside, CA 92521-0314        skype: dyanega
phone: (951) 827-4315 (standard disclaimer: opinions are mine, not UCR's)
              http://cache.ucr.edu/~heraty/yanega.html
   "There are some enterprises in which a careful disorderliness
         is the true method" - Herman Melville, Moby Dick, Chap. 82