plazi / community

This repo is intended to serve as a help desk for TreatmentBank-users.
6 stars 1 forks source link

10 insect genera spotted in plants #65

Open mdoering opened 3 years ago

mdoering commented 3 years ago

These 10 genera coming from Plazi come in as plants, most as Fabaceae, when in fact they are insects. The first dataset I verified seems to have mixed genera from Animalia and Plantae so I suspect this is a species interaction problem, feeds on, host plant, etc.

mdoering commented 3 years ago

These misplacements across kingdoms have a considerable impact on the quality of the GBIF taxonomy. Can we think of a way to avoid them? Maybe flag datasets for manual review in case they span several kingdoms? I assume entire journals should often even be restricted to a certain kingdom or at least nomenclatural code. Is that a way to flag problematic datasets for manual review?

myrmoteras commented 3 years ago

this is interesting: just the last case, there is one that is wrong, need to explore more to understand what makes this happening.

http://tb.plazi.org/GgServer/srsStats/stats?outputFields=doc.uuid+doc.articleUuid+tax.name+tax.phylumEpithet+tax.classEpithet+tax.orderEpithet+tax.familyEpithet+tax.genusEpithet&groupingFields=doc.uuid+doc.articleUuid+tax.name+tax.phylumEpithet+tax.classEpithet+tax.orderEpithet+tax.familyEpithet+tax.genusEpithet&FP-doc.articleUuid=BB06FF93EB76FFFBFF8E2E36FF8CFFDB&format=HTML

http://tb.plazi.org/GgServer/srsStats/stats?outputFields=doc.uuid+doc.articleUuid+bib.title+tax.name+tax.phylumEpithet+tax.classEpithet+tax.orderEpithet+tax.familyEpithet+tax.genusEpithet&groupingFields=doc.uuid+doc.articleUuid+bib.title+tax.name+tax.phylumEpithet+tax.classEpithet+tax.orderEpithet+tax.familyEpithet+tax.genusEpithet&FP-doc.articleUuid=3F24823C545CFFF8C15A2E33FF9A1177&format=HTML

http://tb.plazi.org/GgServer/srsStats/stats?outputFields=doc.uuid+doc.articleUuid+bib.title+tax.name+tax.phylumEpithet+tax.classEpithet+tax.orderEpithet+tax.familyEpithet+tax.genusEpithet&groupingFields=doc.uuid+doc.articleUuid+bib.title+tax.name+tax.phylumEpithet+tax.classEpithet+tax.orderEpithet+tax.familyEpithet+tax.genusEpithet&FP-doc.articleUuid=FA7E2A39FFC4FFF3FFD3067BFFF2FFC5&format=HTML

myrmoteras commented 3 years ago

is fixed Trichromoithrips from Trichromothrips genus-group (Thysanoptera, Thripidae) from China, with descriptions of three new species and ten new records http://tb.plazi.org/GgServer/summary/BB06FF93EB76FFFBFF8E2E36FF8CFFDB

Stosicthrips from New taxa and new records of Australian Panchaetothripinae (Thysanoptera, Thripidae) http://tb.plazi.org/GgServer/summary/872EFFDBA55F897DFF9FFFCD992BD60B

mdoering commented 3 years ago

is phylum Tracheophyta for Stosicthrips szitas correct?

myrmoteras commented 3 years ago

of course not - thank for pointing it out -is fixed now

Just wondering: You write above about considerable impact. I wonder, whether there is a difference between when we enter new species, new genera like in the above case of _Stosicthrips szitas_where we are the sole provider of the time being of these taxa, and just a taxon for which you already have a taxon in your backbone.

mdoering commented 3 years ago

It depends on how we identify a name as being the same name. The classification mostly plays not much role, but we need to decide if a name could be a homonym and cannot in all cases count on an authorship. So the kingdom does play a kew role for us in that decision. In the case of Stosicthrips for example we created a redundant plant genus: https://www.gbif.org/species/search?q=Stosicthrips&rank=GENUS

The classification below is (currently) not that much of an issue - unless as you say the name is coming in via Plazi only. In that case the classification is taken from you guys further down the hierarchy.