Markus Döring

Yes, this is by design as the Catalogue of Life (CoL), which we use as the main source for the upper hierarchy and other major sources, only provide that level...

There are some often reported missing taxa that exist in occurrences but fall through the major Linnean ranks that should be included in the Backbone: - Radiolaria - Oligochaeta -...

species should be ranked higher than infraspecifics

No idea how this happens. There is even a test for it: https://github.com/gbif/checklistbank/blob/master/checklistbank-solr/src/test/java/org/gbif/checklistbank/index/service/NameUsageSearchServiceSuggestIT.java#L72 boosting by rank: https://github.com/gbif/checklistbank/blob/master/checklistbank-solr/src/main/java/org/gbif/checklistbank/index/service/SolrQueryBuilder.java#L74

I have no explanation, this is a bug. There are also synonyms in the list, e.g. the variety, but all should have a species/speciesKey

There appears to be wrong authorships involved, right? Its the same combination but with 3 different authorships and one that claims its an original combination: UNITE: - Alnicola sphagneti (P.D.Orton)...

not really. NULL values lead to the property being dropped. And pointing to oneself as the accepted I never felt is right. But yes, I guess we need to take...

We mark all new species from a nomenclator as doubtful. ZooBank comes first in our priority list, hence the usage gets created as a doubtful one. Unfortunately we forgot to...

as you can see in GBIF it comes from ITIS: https://www.gbif.org/species/102134437/verbatim https://www.itis.gov/servlet/SingleRpt/SingleRpt;jsessionid=C32BB9AC327CFCD0BB9A359DB031A9FD?search_topic=TSN&search_value=204200#null T.o. Barton as the [other homonym](https://www.gbif.org/species/search?q=Turbinaria%20ornata&rank=SPECIES) comes from the GBIF Type specimen dataset which is not very trustworthy....

The problem boils down to our Language enumeration which only tracks 2 letter codes: https://github.com/gbif/gbif-api/blob/master/src/main/java/org/gbif/api/vocabulary/Language.java#L36 The API does not use strings but this enumeration