GBIF backbone taxonomy wish list

GBIF Backbone Taxonomy issues of concern to OTOL project, approximate priority order

September 2013 (migrated from opentree repo wiki)

No license click-through for taxon pages or backbone taxonomy
(This section was present in Sept. 2013, but since then the GBIF DUA click-through has been removed)

"Chromista" and "Protozoa" are archaic taxon concepts
Recommendation: adopt a modern 'upper' ('trunkward'?) classification of life (three domains). Workaround: we will attempt manual surgery in an attempt to clean up the top of the tree. We are also starting to look at Silva as a possible provider of upper taxonomy. If we come up with a decent solution we'll formulate a recommendation for GBIF or possibly even figure out how to set up a good upper taxonomy as a source taxonomy for the backbone.

Better harmony with NCBI taxonomy
Some of our curators say that NCBI taxonomy provides a better classification than the GBIF Backbone in many parts of the tree. We run into annoyances when we try to merge the two taxonomies. It would be nice if the two were more similar. (This problem is so broad as to probably intractable but I thought I should list it.)

IRMNG not up to date (GBIF backbone has a 2011 version) issue 32
New taxa aren't there (e.g. Osteolepiformes) + many taxa are spuriously marked 'accepted' in the old version, fixed in new version. The problem is general of course; e.g. Index Fungorum is also stale.

Numerous problems in birds. issue 34
Could be fixed by replacing what's there with IOC World Bird List (recommended by one of our curators).

Mycobank would be nice

Redundant copies of taxa gbif-ecat issue 100
I can probably write a processor that identifies these and folds them together, but would prefer that this be fixed in the GBIF backbone.

Spelling variants - with/without diacritic; o vs. oe , etc. issue 14
What should be the same taxon has two entries in the GBIF backbone, e.g. Dieffenbachia bürgeri (there are lots of these)

Disappeared taxa gbif-ecat issue 101

Invalid taxa from IPNI (Stephen) There seems to be another problem coming through in taxonomy that Doug has pointed out. The taxon in question is Flacourtiaceae which is defunct (split correctly in NCBI into Salicaceae and Achariaceae). On GBIF, this is just coming from IPNI as far as I can tell. There are other references in there but that one sticks out. IRMNG correctly labels this family in GBIF as not accepted. Anyway, we definitely want to completely ignore IPNI because it is a superset including good and bad names without distinction.

Curators are requesting indication of which taxa are extinct. issue 38
This information is in GBIF's sources, I think (IRMNG and PDB), just not carried over into the DwC files.

Source taxonomies should have ids gbif-ecat issue 100
We'd like to use GBIF's metadata about the source taxonomies (checklists), and maybe associate new metadata with them, but they only show up as names in the DwCA. It would be nice to have unique ids for them, and a file full of the metadata (we know the table exists, it's just not being put in the DwCA). Workaround: Use the textual names as ids.

Richer metadata for the backbone In particular: a list of all input checklists, with write date and/or other version information for each. E.g. it ought to be possible to find out that IRMNG was an input, and which version (date) of IRMNG it was.