Badly merged higher taxa with double letters - Vallonia gracilicosta has incorrect taxonomy

18718
Reporter: feedback bot
Assignee: mdoering
Type: Bug
Summary: Badly merged higher taxa with double letters - Vallonia gracilicosta has incorrect taxonomy
Priority: Blocker
Resolution: Fixed
Status: Closed
Created: 2016-09-01 19:35:43.42
Updated: 2017-03-03 11:58:05.452
Resolved: 2016-09-07 12:57:24.096
        
        
Description: GBIF ID 2296014 (Vallonia gracilicosta) is a land snail: Kingdom Animalia	Phylum Molluscsa	Class Gastropoda	Order Stylommatophora Family	Valloniidae; not a Dipteran insect as the GBIF profile describes.


*Reporter*: Matthew
*E-mail*: [mailto:matthew.s.bolton@gmail.com]]]>
    

Author: rdmpage
Comment: Strange bug as AFAIK none of the checklists that -prove- +provide+ this name make this link to the dipteran fly family.
Created: 2016-09-01 21:01:27.144
Updated: 2016-09-02 11:33:35.689

Author: trobertson@gbif.org
Created: 2016-09-02 11:51:00.843
Updated: 2016-09-02 11:51:00.843
        
I looked also and came to the same conclusion.

My suspicion is this is a bad algorithmic assembly of the nub itself, due to Valonia vs Val[l]onia.

  _Valonia complicata_ Walker, 1856
  _Valonia planiscutellata_ Frey, 1964

My hunch being that those genera were created in the backbone, and then when the otherwise unseen _Vallonia_ genus  was added it was too aggressively fuzzy-matched and all those snails were brought in under the existing flys of _Valiona_.

Or something along those lines.
    

Author: mblissett
Comment: That seems likely; the more recent changes to the NUB matching allow double letters and single letters to be considered similar, but reduced the similarity of other spelling changes. 
Created: 2016-09-02 12:14:08.204
Updated: 2016-09-02 12:14:08.204

Author: rdmpage
Comment: Good spot, currently both _Valonia_ (one l, fly http://www.gbif.org/species/8707617 ) and _Vallonia_ (two l's, snail http://www.gbif.org/species/2296006 ) are treated as children of the fly family Platystomatidae http://www.gbif.org/species/9504. Perhaps one for the https://github.com/gbif/backbone-patch to resolve...
Created: 2016-09-05 12:44:13.639
Updated: 2016-09-05 12:44:13.639

Author: mdoering@gbif.org
Created: 2016-09-05 14:07:03.829
Updated: 2016-09-05 14:07:03.829
        
Thats likely the cause, yes. And not easy to fix I bet. The reason the 2 genera exist is they differ in authorship.
Maybe we need to restrict the name normalization that reduces double chars to a single one to epithets only and leave the genus as is. I'll add a test and see how that can be adressed. The backbone patch unfortunately won't help here as its names also undergoes the name normalization
    

Author: mdoering@gbif.org
Comment: Backbone building fixed in https://github.com/gbif/checklistbank/commit/176a739b1e7823082d1d302833c7a7cdd4933fbc to just apply character dedoublication for bi/trinomials. Gonna be fixed live with the next nub build
Created: 2016-09-07 12:56:40.637
Updated: 2016-09-07 12:56:40.637

Author: mdoering@gbif.org
Comment: See also https://github.com/gbif/checklistbank/issues/19
Created: 2017-02-21 16:58:07.527
Updated: 2017-02-21 16:58:07.527