Thursday, April 23, 2015

Giving taxonomists recognition through citations

Today we discussed the recent Zootaxa paper Fried spicy Linnaeus (Kottelat, 2015) in our journal club. It is open access and a fun read, well worth checking out.

Kottelat discusses the pros and cons and potential variants of mentioning the taxonomic authority after Latin names of plants and animals. Well, mostly animals, but the same applies in all of biology. In botany, the usual format is as follows. Anemocarpa podolepidium (F.Muell.) Paul G.Wilson means that the species was originally described by the 19th century botanist Ferdinand Mueller in a different genus, as Helichrysum podolepidium F.Muell., and then later transferred into the genus Anemocarpa by Paul Wilson.

Botanical journals usually demand the author of a paper to give this full taxonomic authority the first time any species is mentioned but to leave it out afterwards, which can lead to rather odd looking sentences in which some species have the authority and others don't. In general, Kottelat argues convincingly, these authorities make the text clunky and do not usually add any information relevant to the reader.

On the other hand, many taxonomists are concerned that taxonomists should be given more, not less, recognition for the work they do. In botany at least the authorities as part of plant names do not count as bibliographic references and thus do not increase a taxonomist's number of citations. Because we are unfortunately evaluated based on how often people cite our work (as opposed to, say, how often they use it without citing it), many colleagues would like to see them turned into proper bibliographic references and would certainly not like them to disappear altogether.

Kottelat ultimately comes down in favour of turning them into full bibliographic references in taxonomic research papers and doing away with them in all other publications, a compromise that does not entirely convince me.

However, I really want to make a different point. I believe that the taxonomists who push for a rule requiring a bibliographic reference every time a species name is used vastly overestimate the utility of such a change.

I am aware that there are probably hundreds of thousands of undescribed insect and nematode species. Still, especially in plants and many other animal groups but to a large degree also in insects it seems obvious that nearly all species that actually matter much to humans and are thus likely to find their names in many scientific publications have already been described, and often a long time ago. Medicinal plants, culinary plants, popular ornamentals, timber species, widespread noxious weeds, important agricultural pests, pollinators of major agricultural crops, the major fishery species, etc., will rarely have been missed by previous generations of taxonomists precisely because they are abundant and important to humans.

So what would happen if the papers and books in which they were first described had to be cited whenever their names are used? Well, for the most part it would significantly raise the lifetime citations and h indices of the likes of Carolus Linnaeus (1707-1778), Augustin Pyramus de Candolle (1778-1841) or George Bentham (1800-1884), people who for obvious reasons do not really care about such metrics any more. But any contemporary taxonomist who hopes that it would noticeably improve their chances of getting a grant or tenure should prepare for disappointment.

Instead, I would then turn our attention to the kinds of taxonomic publications that are (a) more recent and up-to-date, (b) more crucial to current scientific practice anyway, and thus (c) probably easier to have cited by the end-users: species checklists, floras, monographs, identification keys, field guides and so on. A good two-step plan for giving taxonomy and taxonomists appropriate recognition for the importance of their work would be this:

(1) Change editorial policies so that every scientific study dealing with species needs to provide information on what resources they used to identify those species or to verify their prior identifications. I am not talking here about the 3,000th paper using well known model organisms like Drosophila melanogaster or Arabidopsis thaliana, but for example about ethnobotanic surveys, bioprospecting, community ecology, pollinator observations, phylogenetic, evolutionary and biogeographic studies and so on.

It should be clear that in those cases the correctness of species identifications is at least as important for the validity of the results as are good laboratory procedures and defensible analyses. There is, unfortunately, a good likelihood that the answer many authors would give to such a question is something on the lines of "we just trusted that the museum specimens we extracted the DNA out of were correctly labelled" or "the third person on our author list put names on the specimens, and we think he knows what he is doing".

When given such an answer, peer reviewers and editors could at least raise a red flag. In the best case, however, the authors would cite the floras and field guides they used to key specimens out, and so those would get the recognition they deserve, just like we have to cite the software we use for our analyses. And those publications would be from the last few years and not from 1880, so that the recognition would accrue to people could actually benefit from it. But to work, the plan also needs its second step...

(2) Include species checklists, floras, monographs, identification keys, field guides and so on in the citation databases. At the moment they are of course not in Web of Science, so even if they are cited it doesn't appear in the stats. Here, however, Google Scholar is helpful because it aggregates everything. For example, even under a system that does not require people to cite the identification keys they used the taxonomic revision that was the core of my dissertation already has nine citations in GS; but it has what can only be called 'not applicable' in Web of Science because it appeared in a monograph series.

Admittedly, an alternative solution would be to do away altogether with the impact factor and citation number mania, which the author of the paper we discussed paints colourfully as "an intellectual fraud and a scientific suicide for institutions, journals and scientists". Well, one can dream.

No comments:

Post a Comment