[Taxacom] Taxonomy &DNA barcoding

Tue Aug 18 00:06:38 CDT 2009

Dear Doug, 

Just a few thoughts from an "insider" working at the Canadian Centre for DNA Barcoding ;-). I am not sure if most taxacomers have a detailed understanding of what DNA barcoding is and where it is going... so I'm just giving my perspective as a taxonomist, and instead of looking at barcoding as a massively funded initiative competing with other taxonomy projects, I find it more constructive to look into how this large international project can actually be benefitial for taxonomists. It must be of interest, otherwise why would there be so many taxonomists contributing to barcoding projects or creating their owns?

I observe daily that the data generated here (and it's certainly true for any lab in the world producing barcodes or any other DNA marker used for taxonomic purposes) are quite useful to taxonomists who had no or limited access to molecular data so far. 
More precisely, within the community of taxonomists I am directly collaborating with - that is mostly lepidopterists, both "hobbyists and pros" - I can see the combination of the DNA barcodes and morphology slowly making its way into a number of taxonomic projects and then to publications (some not peer-reviewed - as inherent to most taxonomic journals; I can send some references to those interested, it goes from a formal analysis of the sequences in relation to other characters to the mere citation of the sequence for the type and some statements about genetic distances).

This use of DNA barcodes by taxonomists constitutes IMHO a direct support from DNA barcoding to the taxonomic community and its work. I don't think we can expect much more within the current context, especially since major (but not all) barcoding fundings aim at building a reference library, not at solving taxonomic problems... the later is - I think - a cool by-product of the former, and this is appealing enough already for many taxonomists to send us thousands of samples and invest a huge amount of their time doing that. 

Of course, it would be even more positive if DNA barcoding proves, in the future, to be an efficient leverage to be used by taxonomists in their own grant applications. If a "fashion effect" can make applications more succesful, then it's great - I don't think that adding barcodes or other molecular markers to one's toolbox is a too big price to pay. Can someone talk from the experience with phylogenetics? Can we say that the development of molecular phylogenetics somewhat permitted some morphological works to be carried out? I would assume yes, if only for the controversy inherent to the approach while still new (and DNA barcode has its share ;-)). As for myself, my PhD project had a molecular component in it (small one, 10% DNA / 90% morphology), but it was mostly to make it more sexy since what I really wanted to do was morphology ; I am sure it helped getting the scholarship at that time.

About the use of type specimens in barcoding studies and a broader "legacy" concept:

First, a number of campaigns here are developped by taxonomists and/or in close collaboration with expert taxonomists; they are one of the main source of the samples we receive, meaning that the identification coming with them is based on their conception of a 'species name' by reference to what they know of the name-bearing type. Sometimes - hopefully quite often - it ain't to be questioned and all is just fine with barcodes then correctly associated to that name in the reference library.
As such, it looks to me as if the records in this library represent a new "legacy" (based on new material, right) but with an indirect reference to the type via the taxonomist expertise.

Second, when things are not so simple and when f.i. possible cases of cryptic diversity or synonymy are revealed through analysis of DNA barcodes, there are different options:
(a) taxonomists, strong of that new information can get back to the specimens (including the types or the OD at least), look for overlooked differences, look at their distributions, etc... In a number of cases the group can be better understood and new taxa described or raised from/sunk into synonymy (see f.i. our paper in Zootaxa last year: http://www.mapress.com/zootaxa/2008/f/zt01923p036.pdf)

(b) more directly connected to your point: there are protocols (at least for lepidoptera -other groups will certainly follow) permitting to obtain barcodes from very old specimens. The complete barcode region was recovered for specimens as old as 160 years old or more, and partial sequences for specimens older than that, but still providing enough data to link the type to some recent specimens through their barcodes. This approach is used in a paper in press in Zootaxa (Hausmann et al.; should be available within the next few weeks) and the protocol will be published separately. When the type, because of its sex f.i., or because of its poor condition, is of no help to solve a taxonomic/nomenclatural question, then barcodes can be a good alternative over designation of a neotype (which is quite a complex process if the type isn't lost or destroyed). On the other hand, the approach was seen by some colleagues in charge of collections as a way to preserve the integrity of the type by removing a leg instead of dissecting its abdomen.

Hopefully, the "no problem" or "problem solved through re-examination of material" options are usual enough so that there won't be a need for point (b) in « >95% of the existing holotypes ».

The protocol used for old types is still quite costy when compared to fresh specimens, and it isn't integrated in high-throughput workflows (though the 96-wells format and robotic liquid-handling can be used as well, as long as you get very cautious against contamination...) - the good side of it is that a single specimen, the holotype, is certainly worth a hundred fresh ones once sequenced...
BTW, I would add that for insects (tissue samples from dry museum specimens) the routine protocols currently used here can recover with a very reasonable success (70-80%) DNA barcodes from specimens collected between 1980 and 2000 or so, which is not so bad and gives access to a fairly large amount of material in museum collections. Obviously, the fresher the better, but when necessary, for material difficult to find or re-collect, it's good to know that there are ways around it.

Finally, as far as leps are concerned, all records of primary types are made public on BOLD (http://www.barcodinglife.org/views/login.php; click on the published projects link, top left corner), including DNA sequences and all voucher information (institution storing, collecting data, image). There are only 175 records so far, 112 being sphingid moths, but species newly described by collaborators are systematically added and old types as well on a case by case basis, along with the progress of some revisions which have been including barcodes of old types. This may not be a direct support to the institution hosting the types, but in a way it brings an additional value to the collections that their curator may be able to exploit.

Damn, it's long... and late... Let's say this was my "propaganda" for taxacomers to look into the potential of DNA barcoding as a tool for taxonomists, rather than as a competing initiative.

Cheers,

Rodolphe

>Quoting Robert Hanner <rhanner at uoguelph.ca>: 
[snip] 
>> In fact, the barcode community is responsibe for the first concerted 
>> effort to make structured linkages between GenBank Accessions and 
>> Museum Voucher Specimens. 

This is NOT the same as work based on "legacy" museum specimens; the typical morphological taxonomist uses MOSTLY material which is from 30-100 years old, but DNA barcoding does not - unless you have references showing barcoding projects based primarily on OLD specimens, in which case many of us here would be very interested in seeing them. Since most of the material in the world's museums is from 30-100 years old, the barcoding community and the museum community are not entirely "on the same page" here. Most pointedly, if one cannot do a barcode for >95% of the existing holotypes, that leaves very little overlap between the *foundation* of traditional taxonomy and nomenclature - the type specimen - and the barcoding methodology. 

The original quote, alluded to by Stephen, was in this context: the idea that basing names on old museum specimens was quaint and outdated, and barcoding represented a chance to start from scratch; i.e., that legacy material would no longer occupy the place of honor in systematics once everyone saw how barcoding worked, became convinced that it was indeed a panacea, and then (presumably) we all admitted that holotypes that cannot be barcoded serve no purpose, since no one can ever know what taxa they truly represent. 

Maybe you can allay our concerns by answering the following question: how much of the funding for CBoL projects (such as the one you mention by Packer) goes to the maintenance and curation of un-sequenceable legacy material? 

If CBoL support extends, blanket-wise, to the care of legacy material, and if this represents a real increase in support to institutions with large holdings of such legacy material, then maybe there isn't as much of an issue with "competing" for funding dollars as it would appear. But if the money is going to build *new* collections composed of *new* specimens, while existing museums elsewhere are closing their doors, then the "zero-sum" problems are genuine, and worrisome. 

Sincerely, 