2 Hyla genera from ITIS

[mdoering]

There are 2 accepted Hyla genera from ITIS, both with species:
https://www.checklistbank.org/dataset/9910/taxon/4ZYL
https://www.checklistbank.org/dataset/9910/taxon/4ZYM

[DaveNicolson]

Hyla imitator in ITIS is linked directly to Hylidae, intentionally... Its placement is 'incertae sedis'. It should not be linked to an "implicit name", it should link to the family.

Similarly, Hyla nicefori is linked directly to the hylid subfamily Cophomantinae (again, as 'incertae sedis'). Again, it should not be linked to an "implicit" genus name, it should link to the subfamily.

I'm not sure how ChecklistBank and/or COL normally handle such things, but nothing from ITIS should be creating "implicit" parent names, as all ITIS records link to a valid/accepted "explicit" name.

[mdoering]

I thought so already. The convention is that implicit genera will be created in a "sync" to COL when they are accepted species. But not if they are provisionally accepted as the recommendations are for species of uncertain placement:
https://github.com/CatalogueOfLife/coldp/blob/master/README.md#species-with-an-uncertain-genus

Is there a chance that ITIS add the provisional=true flag for these kind of species in the ColDP export? I guess this is a task for @gdower ?

[mdoering]

Meanwhile @yroskov might want to add a decision to make these species provisional

[mdoering]

It caused issues in the GBIF Backbone: gbif/checklistbank#273

[mdoering]

There are also accepted genus duplicates in ITIS directly:
https://www.checklistbank.org/dataset/2144/duplicates?category=uninomial&codeDifferent=false&limit=100&minSize=2&rank=genus&rankDifferent=false&status=accepted

Are the ones without authorship expected @DaveNicolson ?

[mdoering]

You can find more duplicate genera due to missing provisional flags on species using the duplicate tool and search for accepted genera with different code and having ITIS (one should finally do this across any source, not just ITIS):

https://www.checklistbank.org/catalogue/3/duplicates?catalogueKey=3&category=uninomial&codeDifferent=false&limit=100&rank=genus&rankDifferent=false&sourceDatasetKey=2144&status=accepted

The pairs that come up with one genus having no author and no source are the implicitly generated ones

[DaveNicolson]

@mdoering , ITIS does contain homonyms and unverified records (potentially homonyms or errors), especially in some parts of the database. Genera without authors in ITIS are generally unverified, as we have to include authors when editing genera (genus to subspecies etc. require authorship to edit, above genus they are not required). Not sure if that answers your question...

[DaveNicolson]

I would say that COL should not generate implicit genera for any ITIS records, period. Questionable hierarchy placement should not be conflated with questionable status as a species, so I would not expect such names to be flagged as provisional. ITIS indicates questionable hierarchy placement with an "uncertain_parent_indicator" element, which forces the ITIS website to show the name as "incertae sedis" in the parent taxon. In that case, the placement could be regarded as provisional, but not the entire taxon/name record...

[DaveNicolson]

OK, I have an update... @gdower , you will need to set the provisional flag in COLDP for all TSNs that have the [taxonomic_units].[uncertain_parent_indicator] set (the values and/or name for that element may vary, but that indicator is set for TSN 1095266, if you want to look at an example to be sure).

It bothers me that we are forced to flag a name as "provisional" just to get it handled properly, but David Mitchell tells me that ITIS' implementation of COLDP exports is doing what is here suggested (setting the provisional flag if the uncertain parent indicator is set). So that's apparently what you will need to do too. It should avoid the "implicit genus" issue.

But as I already suggested, ITIS's hierarchy linkages are all explicit and should be used wherever possible (preferably always).

[gdower]

It should be fixed on the next update.

[yroskov]

After Geoff' and Markus' fixes, ITIS incertae sedis species look in CoL now like this (there is no second entry for the genus; temporarily used combination is flagged as Provisionally Accepted):

The same case in AC23 looked like that:

@DaveNicolson & @mdoering, is a new data presentation acceptable for both of you?

[yroskov]

@DaveNicolson, CoL Provosionally Accepted status indicates temporal application of binomial name in this case.

[yroskov]

Meanwhile @yroskov might want to add a decision to make these species provisional

@mdoering, I cannot resolve duplicates in ITIS, because CoL uses 34% of ITIS accepted species. We need ISSUES & TASKs reports for only taxa included in CoL (i.e. in sectors). Otherwise, it's too much work for nothing.

[mdoering]

@DaveNicolson & @mdoering, is a new data presentation acceptable for both of you?

It's perfect!

[mdoering]

Meanwhile @yroskov might want to add a decision to make these species provisional

@mdoering, I cannot resolve duplicates in ITIS, because CoL uses 34% of ITIS accepted species. We need ISSUES & TASKs reports for only taxa included in CoL (i.e. in sectors). Otherwise, it's too much work for nothing.

That happens on the project duplicate page as I used it to detect the other cases:
https://www.checklistbank.org/catalogue/3/duplicates?catalogueKey=3&category=uninomial&codeDifferent=false&limit=100&rank=genus&rankDifferent=false&sourceDatasetKey=2144&status=accepted