Roger Hyam

I was just putting my talk together for TDWG2010 and I needed a well known European species with restricted range to make a point about distribution maps. I chose Edelweiss as the name alway makes me want to break into song – plus it as a limited distribution but is in more than one country. I’ll post my full talk next week but in researching the distribution of Edelweiss I came across this quote (Werner Greuter Willdenowia 33 – 2003 page 244)

Leontopodium nivale subsp. alpinum (Cass.) Greuter, comb. & stat. novi ≡ Gnaphalium leonto- podium L., Sp. Pl.: 855. 1753 ≡ Leontopodium alpinum Cass. in Cuvier, Dict. Sci. Nat. 25: 474. 1822. – Note: As the monographer Handel-Mazzetti (in Beih. Bot. Centralbl. 44(2): 140. 1927) noted, his choice to maintain Leontopodium nivale (Ten.) Hand.-Mazz. as a separate species was heavily influenced by his reluctance to change the name of the edelweiss for reasons of priority. Whether Tutin had a similar motive when he proposed the incorrect combination L. alpinum subsp. nivale (Ten.) Tutin (in Bot. J. Linn. Soc. 67: 283. 1973) I do not know, but he was certainly right in treating the widespread genuine edelweiss and the aberrant populations of the Central Apennines and Central Balkans as subspecies of a single species. I have considered proposing conservation of the name Leontopodium alpinum against Gnaphalium nivale Ten. (Fl. Napol. 1: xlviii. 1811), but abandoned the idea for two main reasons: first, granting that the species is important and widely known, its familiar designation in many languages, the vernacular edelweiss, is much more popular than its scientific binomial (you will find a 40 : 1 ratio in favour of the former when searching the Web); and second, I like the epithet nivale, which characterises admirably well both the plant and its habitat, and hesitate to suggest its being rejected in favour of the rather trivial appellation alpinum.

Now I could paraphrase this as saying: “I thought about keeping the scientific name stable but no one uses it so we (taxonomists) can do what we want. Everyone in the real world (Google) uses Edelweiss.”  You may disagree with me.  The fact remains that there are thousands of resources out there tagged with Leontopodium alpinum that are now “wrong” and there will be thousands of resources going forward that will be tagged with Leontopodium nivale subsp. alpinium and we must build big complex systems in order to try and keep track of these nomenclatural changes and not keep track of the biodiversity – which seems to have disappeared between the cracks.

I presume (not being an expert in this group) that the plants in the whole range (Alps, Apennines and Balkans) have been considered one species by many people and called L. alpinium or Edelweiss for a long time. If we could ignore the ICBN and just make up names that match a view that the Apennine and Balkan species should be considered a separate subspecies then one would ideally keep  L. alpinium and create two entirely new names for the two different taxa we recognise within it. This would maintain data integrity in that things tagged as L. alpinium would still be correct but the separate subspecies could be recognised where necessary going forward – as specimens are re-determined. Unfortunately the rules of priority prevent this and just make a muddled mess of recycled old names.

I have been around and around this thing over the past few years and it all comes down to a simple use-case. I thought I would present it graphically here:

The Simplest Use-Case for Taxonomy?

David is a decision maker. He wants to use data from two studies, Helen 2002 and Harry 1998. Is it safe for him to combine studies if:

1. they both use the same taxon name? Maybe – but only if David can be certain the circumscriptions of the taxa used in the two studies were sufficiently similar to satisfy  the needs of his own study. i.e. this can’t be known a priori to David’s study.
2. they use different names but the names were considered synonyms of each other? Again it is David’s judgement whether the two concepts used were sufficiently similar – independent of the names used.

It looks like David has to stop being a decision maker and do some taxonomy in every case where source data studies have not gone out of their way to explicitly state that they have used the same identification methods.

How does the work we are doing with persistent identifiers and building big names/nomenclatural databases help solve David’s problem? We can help him find material to make his decision but if the result is just textual descriptions and images we can’t automate his decision at all. Google and other search based approaches are pretty good at finding material based on scientific names.

Helen and Harry could have helped out if they had cited the identification key they had used – this doesn’t need a database of all identification keys it is a simple literature citation.

David could make the assumption that anything that has the same name or is a synonym of something with that name is the ‘same’ but this involves taking a sensu lato approach to every species.

If there were some absolute way to describe the taxa then perhaps the process could be automated. If, just as an example, Helen et al and Harry had produced DNA barcodes then David could make the assumption that anything that had the ‘same’ (or sufficiently similar) barcode would be treated as being the ‘same’ in his study. Without this kind of mechanism (and we don’t have it in plants) I am at a loss to see how we can help the poor guy out – he must become a taxonomist at least part of the time. Any suggestions?