[Taxacom] BioNames

Tony.Rees at csiro.au Tony.Rees at csiro.au
Mon Jun 3 20:30:44 CDT 2013

Thanks, Rod, for the reply.

You wrote:

>>- is this likely to be a rare or common event (is it a more general class of error)

> I notice it a few times bit, although the GBIF classification improves with time. Would be useful to quantify it.

Maybe one way to approach quantifying this type of error would be to start with presently known (to me) genus level homonyms and see how well they are handled. My lists are available at (e.g.) http://www.marine.csiro.au/mirrorsearch/ir_search.list_homonyms?hlevel=genus&start_letter=A for which the first 10 entries currently read:

    Aa H.G. Reichenbach, 1854 (Plantae-Magnoliophyta-Liliopsida-Asparagales-Orchidaceae)
    Aa Baker, 1940 (Animalia-Mollusca-Gastropoda-Stylommatophora-Helicarionidae) .. (Current name: Philonesia) Remarks: Currently valid as subgenus of Philonesia, refer Cowie et al., 1995.

    Aaaba de Laubenfels, 1936 (Animalia-Porifera-Demospongea-Poecilosclerida-Crellidae) .. (Current name: Crellastrina) Remarks: Spelling as in Hooper & van Soest, 2002, cf. Aaba in Nomenclator Zoologicus (original not seen).
    Aaaba Bellamy, 2002 (Animalia-Arthropoda-Insecta-Coleoptera-Buprestidae) Remarks: Replacement name for Alcinous Deyrolle, 1865.

    Aaata Semenov, 1906 (Animalia-Arthropoda-Insecta-Coleoptera-Buprestidae) Remarks: Authority cited elsewhere as Semenov-Tian-Shanskij, 1906.
    Aaata de Laubenfels, 1930 (Animalia-Porifera-Demospongea-Poecilosclerida-Microcionidae) .. (Current name: Clathria) Remarks: See Anaata de Laubenfels 1932. (Nomen. Zool.).

    Aaleniella Plumhoff, 1963 (Animalia-Arthropoda-Ostracoda-Podocopida-Cytherideidae)
    Aaleniella Conti & Fischer, 1981 (Animalia-Mollusca-Gastropoda-Archaeogastropoda-Trochidae)

    Abacella Stechow, 1920 (Animalia-Cnidaria-Cnidaria (awaiting allocation)-Cnidaria (awaiting allocation)-Cnidaria (awaiting allocation))
    Abacella V.P. Maslov, 1956 (Plantae-Chlorophyta-Chlorophyceae-Bryopsidales-Codiaceae)

    Abacodes Thomson, 1858 (Animalia-Arthropoda-Insecta-Coleoptera-Carabidae) .. (Current name: Buderes)
    Abacodes Jeannel, 1948 (Animalia-Arthropoda-Insecta-Coleoptera-Carabidae) .. (Current name: Chaetodactyla) Remarks: As Chaetodactyla (Abacodes) in Hallan, 2000-.

    Abakaniella V.A. Chachlov, 1939 (Plantae-Algae (awaiting allocation)-Algae (awaiting allocation)-Algae (awaiting allocation)-Algae (awaiting allocation))
    Abakaniella Betekhtina, 1966 (Animalia-Mollusca-Bivalvia-Bivalvia (awaiting allocation)-Bivalvia (awaiting allocation))

    Abalius Cabanis, 1861 (Animalia-Chordata-Aves-Aves (awaiting allocation)-Aves (awaiting allocation))
    Abalius Kraepelin, 1897 (Animalia-Arthropoda-Arachnida-Uropygi-Thelyphonidae) .. (Current name: Abaliella) Remarks: See Abaliella Strand 1928. (Nomen. Zool.).

    Abama Adanson, 1763 (Plantae-Magnoliophyta-Liliopsida-Dioscoreales-Nartheciaceae) .. (Current name: Narthecium) Remarks: An illegitimate, superfluous name for Narthecium Huds. (GRIN). A homotypic synonym of Narthecium Hudson 1762 (nom. cons.) according to Index Nominum Genericorum.
    Abama Rafinesque, 1837 (Plantae-Magnoliophyta-Liliopsida-Dioscoreales-Nartheciaceae) .. (Current name: Narthecium) Remarks: (Pre APG placement: Liliaceae).

    Abantiades Herrich-Schäffer, 1856 (Animalia-Arthropoda-Insecta-Lepidoptera-Hepialidae) Remarks: n.n. pro Abantiades Broun, 1914 (Nomenclator Zoologicus). Authority cited elsewhere as Herrich-Schaeffer, 1856.
    Abantiades Fairmaire, 1894 (Animalia-Arthropoda-Insecta-Coleoptera-Tenebrionidae) .. (Current name: Diphyrrhynchus) Remarks: n.n. pro Abantis Fairmaire 1892) (See Neoabantis Gebien 1910. (Nomen. Zool.).
    Abantiades Broun, 1914 (Animalia-Arthropoda-Insecta-Coleoptera-Curculionidae) .. (Current name: Abantiadinus) Remarks: See Abantiadinus Schenkling 1929. (Nomen. Zool.).


(Note I interpret "homonym" loosely, including misspellings and cross-code homonyms but with known nomina nuda and later usages excluded from this list). The indicated taxonomic placements, along with the names themselves, come from my IRMNG genera database in which as you may see, not all names are yet placed to family, though the proportion will improve over time. Of course the plant-only names would be out of scope for your BioNames compilation at present, though somebody might like to check...

Anyway - if all 10 of these (or even most) are correctly handled in the GBIF taxonomy (and/or your mapping to it), that would be a very good result and possibly be more informative than a single bad instance.

Any takers?

Regards - Tony

From: Roderic Page [mailto:r.page at bio.gla.ac.uk]
Sent: Tuesday, 4 June 2013 8:05 AM
Cc: Neal Evenhuis; Stephen Thorpe; Rees, Tony (CMAR, Hobart); Markus Döring
Subject: Re: [Taxacom] BioNames

Hi Tony,

On 3 Jun 2013, at 21:45, <Tony.Rees at csiro.au<mailto:Tony.Rees at csiro.au>> wrote:

OK... so returning to the "Cepa" thread as identified by Chris Thompson ...

Looking at http://bionames.org/search/Cepa, the problem arises on account of Cepa being a genus-level homonym (twice in both animals and plants), which can be discovered via various sources as previously noted, however this is not reflected in the GBIF backbone taxonomy (http://ecat-dev.gbif.org/search?q=Cepa&rkey=1) to which Rod has pinned a lot of his classification information. So some questions might arise as follows:

- why does the GBIF taxonomy mangle this group of names (what is specific to this case)

Because the input data is messy, and the aggregation algorithms are not perfect. Also some problems (like homonyms) only become apparent at a global scale.

- is this likely to be a rare or common event (is it a more general class of error)

I notice it a few times bit, although the GBIF classification improves with time. Would be useful to quantify it.

- what could assist GBIF to (a) detect this and equivalent problems, and (b) fix them

Gee, how about a database of names linked to the primary literature, ICZN decisions, etc., so GBIF (or anyone else) could investigate further and fix the problem?

- what could assist Rod or other external user of GBIF data to do the same (i.e. warning signs)

See above ;) Also we could have some simple rules to flag obvious issues (e.g., GBIF can have species with different generic names placed in the same genus which is clearly inconsistent. There are lots of cases of multiple spellings of the same name (you see this in classifications where you have a run of taxa with similar names). Another warning flag is a node, such as a genus, with no descendants.

The problem with fixing is that GBIF assembles the classification anew from multiple "checklists" and without manual intervention. I've discussed this a little with Markus Döring, who suggested having a checklist that was highly weighted and included corrections that would then be applied to the new aggregation. Elsewhere I (and others) have argued that if we placed the GBIF classification under version control we could edit it, merge and fork it like a software project.

- does BioNames choose wisely in using the GBIF taxonomy, bearing in mind the latter's disclaimers i.e.:

"Nub Disclaimer: The GBIF Backbone Taxonomy (Nub) is an automatically synthesised management classification with limited manual curating. Information presented here does not represent a consistent taxon but may conflict with other nub "usages" in many cases to a trained taxonomists eye. The information presented on this page was aggregated from the data found in the sources below."

Why GBIF? It's the single biggest classification available that I'm aware of, and it's connected to data that is central to a lot of biodiversity research (see http://www.mendeley.com/groups/1068301/gbif-public-library/ ). People use GBIF to do science,  it would be nice if the classification it had was clean and internally consistent. Exposing it to scrutiny and linking it to the primary literature is one way to do this.



More information about the Taxacom mailing list