Flagger Content Author Content Reason Flag Created Resolved by Resolution
astrobirder Willows (Genus Salix)

Add subgenera and sections

Sep. 19, 2020 22:58:36 +0000 loarie

see comments

Comments

At the very least we need subgenera to be added. Even just having Salix subgenus Salix and S. subg. vetrix could help to clean up a lot of northeastern observations. Subgenera present in North American can be found in FNA: http://beta.floranorthamerica.org/Salix

Posted by jharkness over 3 years ago

Is there a reference that assigns all species (globally) to each subgenus?
https://www.inaturalist.org/pages/curator+guide#nodes

Posted by bouteloua over 3 years ago

This reference I had when this flag was made is also just for North America. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0121965
It cites a source for Russia and Asia (in Russian). Not sure how many species that would leave out.

Posted by upupa-epops over 3 years ago

I'm not aware of any one resource that includes all subgenera globally; I think to include all of these would mean piecing together taxonomy from different floras.

Posted by jharkness over 3 years ago

Is it helpful to put them all in a spreadsheet like this? https://docs.google.com/spreadsheets/d/1ZyIyso9OonTZRhhne1cdZNWjUCRZP3IUX42J5EvuXAo/edit
I think they'll all be covered.

Posted by upupa-epops over 3 years ago

yeah actually, because then there's the potential that staff (Scott) could batch edit them to assign them to their respective parents

Posted by bouteloua over 3 years ago

How do we find out what the parent species of hybrid taxa on iNat are?

Posted by upupa-epops over 3 years ago

That can be a slog, but IPNI is a good start. It often will explicitly list the parents, or it should list the publication in which it was named, which should list the parents.

Posted by bouteloua over 3 years ago

Do we need to get all the information for species not yet in iNat as well?
I might post this in the forum to ask for help.

Posted by upupa-epops over 3 years ago

Do we need to get all the information for species not yet in iNat as well?

Yes because that helps inform whether we can even do subgenera/sections on iNat. As well, Scott can mass import the missing POWO species so that everything is lining up better.

Posted by bouteloua over 3 years ago

Hi, I've tried to assign willows to their respective infrageneric ranks many years ago on Biolib.cz webpage. So maybe it will help: https://www.biolib.cz/en/taxon/id38917/. It may be little bit outdated in places and there are unfortunately no links to literature sources but it may serve as a basis for sorting.

Posted by michal_hrones over 3 years ago

Would it be possible to only organize the willows we have good taxonomic information on into subgenera and sections, while leaving the others classified only under the genus? My guess is that there are some species that can't be cleanly placed into infrageneric ranks because of limited and/or conflicting literature.

Posted by jharkness over 3 years ago

Going down through the list alphabetically, most of the species I've skipped were because they weren't in the main papers I'm referencing or were labeled as synonyms in those papers. Sometimes they're hybrids or spelling differences.
There have been a couple that aren't mentioned in those papers because they were described after those were published. What if they weren't assigned a subgenus in their description?

Posted by upupa-epops over 3 years ago

Ope I just found these sources which already had everything organized:
http://www.fao.org/forestry/16387-064a6c432aa6ad67b71111974322d476f.pdf#page=81
www.fao.org/3/a-i2670e.pdf#page=32
They are still missing some species though and they don't say what subgenera the Chinese sections are in.

Posted by upupa-epops over 3 years ago

@michal_hrones it looks like you were able to find subgenera for all the Chinese sections except Fulvopubescentes? Do you happen to remember how to found them or have sources for that?

Posted by upupa-epops over 3 years ago

115 species and 64 hybrids left.

Posted by upupa-epops over 3 years ago

Okay we've gone through all the species (not all the hybrids yet). There are about 90 left without subgeneric taxonomy because either they need curation (hybrids and/or synonyms), I don't have access to the literature that discusses them, or they were more recently described without taxonomy.

Do we need to get all of them? It won't be possible for the latter group. (Edit I see Quercus and Carex don't have all of them covered)

Posted by upupa-epops over 3 years ago

I'm not sure if there is anything included in this that you haven't already gone over, and I'm not sure how much of the taxonomy is up-to-date, but this appears to be a pretty comprehensive list of willows by subgenera and sections: https://species.wikimedia.org/wiki/Salix

I don't see anything wrong with leaving some species/hybrids excluded from infrageneric ranks; even if only the commonest Salix species could be identified to infrageneric ranks, it would certainly help to clean up the observations on the site.

Posted by jharkness over 3 years ago

Thanks, I tried to use academic papers as much as possible so my sources give similar results.

Also I've basically done as much as I can with regards to the spreadsheet now. I'm not sure exactly what the formatting needs to be?

Posted by upupa-epops over 3 years ago

@bouteloua do the flags need to be dealt with beforehand?

Also just trying to figure this out since I got different opinions (compared to what you said here https://forum.inaturalist.org/t/big-ol-backlog-of-taxon-flags/4464/10)... if a taxon has observations and/or IDs then I should swap it, but otherwise inactivate it?

Posted by upupa-epops over 3 years ago

It's a balance to weigh - are there any IDs? (need to check inaturalist.org/identifications?taxon_id= as they don't always appear on the taxon page) does any of that data (place check lists, common names, range map, atlas, etc) exist on the input taxon? does the output already have the input listed as a synonym? should that data be transferred to the output taxon? I got the idea of inactivating without swapping from Scott, so if you're getting pushback it is probably worth a forum post or email.

I don't think the flags need to be dealt with before adding sections, but it would be nice to resolve them as part of this process.

Posted by bouteloua over 3 years ago

I don't think it's pushback so much as just assuming that that's always the best way to do it because then the old taxon gets listed as a synonym etc.

Okay cool, yeah I agree. In that case what else needs to be done with the spreadsheet before sending it to Scott?

Posted by upupa-epops over 3 years ago

I would try to make it as clean as possible and resolve as many flags as I could before handing anything to staff - looks like there are 54 taxa that are on iNat but not in POWO, for example. I moved a few that have been resolved to a tab called archive.

Instead of having blanks in any of the columns for subg, section, or subsection, I would put some sort of note like "not applicable", "spans two sections", or "unknown". To distinguish whether it was researched and you could not find the answer or it really isn't applicable because it is a hybrid that spans two sections.

Consider whether we actually want subsections - do we know how complete they are compared to the sections? In the sense of https://www.inaturalist.org/pages/curator+guide#nodes for example the section Glaucae only has one subsection listed (Glaucae). There are 3 species in that section assigned and 10 species unassigned to a subsection.

Add a column for the direct parent taxon_id (e.g. to distinguish Salix at rank=section from Salix at rank=subgenus or genus)

Posted by bouteloua over 3 years ago

Got it.

For subsections, we have them for all the Eurasian and North American species but not the Chinese species. Not sure if it's better to not do them entirely or go case by case? Like most of the sect. Helix, Vetrix, and Arbuscella species have subsections but most of the sect. Glaucae and Hastatae don't.

Does the taxonomy get created before import or during? If the latter, the only parent taxon ID I'll have is genus.

Posted by upupa-epops over 3 years ago

I've inactivated/swapped the simpler issues and moved them to the archived tab. The rest I'm not confident in how to deal with them.

Several examples like this where the taxon is a synonym of several other taxa https://www.inaturalist.org/flags/496046
Some examples like this where the output taxon isn't in iNat yet. Why do you have the output taxon inactive here? https://www.inaturalist.org/flags/518476
And where there were swaps/merges that affected species with a decent number of observations I made the taxon change but didn't commit it.

Posted by upupa-epops over 3 years ago

Add a column for the direct parent taxon_id (e.g. to distinguish Salix at rank=section from Salix at rank=subgenus or genus)

@bouteloua does this mean the taxonomy needs to be set up before import?

Posted by upupa-epops over 3 years ago

If the description of a species says what the closest related species is, but not what section it's it, can it be assumed to be in the same section as the related species?

Posted by upupa-epops over 3 years ago

I've definitely done that - assigned it to the section of the related species, even though section wasn't explicitly stated.

Posted by bouteloua over 3 years ago

The best set-up might have some format I'm not aware of, so might be best to talk to @loarie now -

Scott, there's a Google Doc of 500+ willow species with the subgenus and sections assigned. Currently willows are all hanging off the genus on iNat, but the community feels it will be helpful to have infrageneric taxa, and it's also moderately complete (not a ton of unknowns at the subgenus level, and some research still to be done). It's still a work in progress but we'd like to make sure it's formatted correctly to see if they can be assigned to new parent taxa in one go.

https://docs.google.com/spreadsheets/d/1ZyIyso9OonTZRhhne1cdZNWjUCRZP3IUX42J5EvuXAo/edit#gid=2030241887

1 - Some are not assigned a subgenus/section, either because it's unknown or because it spans two subgenera, as a hybrid. They're currently listed in Column B as "unknown" or "2 subgenera" depending on the situation. (Edit: ignore subsections, see below. I think we can ignore subsection right now since there might be too many unknowns there. Up to @upupa-epops.)

2 - do you need the parent taxon_id in a new column or is this current set-up sufficient?

3 - Column G calls out species that are in POWO but not yet in iNat. Is it possible for you to add and assign these in one go as well? There are a ton of additional hybrids listed in POWO that I don't think we need to add to iNat right now.

After we do this we can lock Salix to better manage the taxonomy of incoming taxa.
thanks!

Posted by bouteloua over 3 years ago

I think there are few enough species in the 3 sections that have subsections that splitting them up to that extent probably isn't necessary for utility.

Posted by upupa-epops over 3 years ago

I would agree. I don't really see a need to have subsections, as most of them contain so few species.

Posted by jharkness over 3 years ago

just looking at this now. Seems the tasks are:
1) create the new nodes
2) move the taxa to the new nodes
3) log these internodes in a deviation linking to this flag

Folks should be able to do this through the interface. If you want me to help, I just need a table of
taxon_name, parent_name
e.g. Salix kikodsei, Vetrix

Also not to derail this thread but looks like the whole willow family is a bit of a mess (to the extent that it doesn't line up well with POWO and here aren't deviations to signal curators that we're intending to not follow POWO for whatever reason). Any objections to making swaps to at least bring Salicaceae a little more in line with POWO?
Chosenia -> Salix
Chosenia arbutifolia -> Salix arbutifolia
Chosenia bracteosa -> Salix arbutifolia
Chosenia splendida -> Salix arbutifolia
Guidonia -> Casearia
Guidonia icosandra -> Casearia laetioides
Guidonia latioides -> Casearia laetioides
Hecatostemon -> Casearia
Hecatostemon completus -> Casearia completa
Laetia -> Casearia
Laetia corymbulosa -> Casearia corymbulosa
Laetia crenata -> Casearia laetioides
Laetia micrantha -> Casearia panamensis
Laetia povedae -> Casearia povedae
Laetia procera -> Casearia bicolor
Laetia thamnia -> Casearia thamnia
Phyllobotryum -> Phyllobotryon
Phyllobotryum spathulatum -> Phyllobotryon spathulatum
Samyda -> Casearia
Samyda dodecandra -> Casearia dodecandra
Samyda glabrata -> Casearia kigeri
Samyda icosandra -> Casearia laetioides
Samyda macrocarpa -> Casearia macrocarpa
Samyda mexicana -> Casearia mexicana
Samyda rubra -> Guarea guidonia
Samyda spinulosa -> Casearia spinulosa
Samyda villosa -> Casearia villosa
Samyda yucatanensis -> Casearia yucatanensis
Bennettiodendron brevipes -> Bennettiodendron leprosipes
Bennettiodendron lanceolatum -> Bennettiodendron leprosipes
Populus monticola -> Populus tremula
Populus rasumowskiana -> Populus × rasumowskiana
Populus rotundifolia -> Populus tremula
Populus smithii -> Populus × smithii
Populus talassica -> Populus macrocarpa
Populus tomentosa -> Populus × tomentosa
Populus trichocarpa -> Populus tristis
Populus wettsteinii -> Populus × wettsteinii
Ryania casiquiarensis -> Ryania speciosa tomentosa
Salix boseensis -> Salix tetrasperma
Salix calodendron -> Salix gmelinii
Salix canariensis -> Salix pedicellata canariensis
Salix chrysocoma -> Salix × pendulina f. salamonii
Salix erythroflexuosa -> Salix × pendulina f. erythroflexuosa
Salix fimbriata -> Salix berberifolia subsp. fimbriata
Salix fluviatilis -> Salix melanopsis
Salix glaucophylloides -> Salix myricoides
Salix humboldtiana -> Salix chilensis
Salix kamtschatica -> Salix udensis
Salix kitaibeliana -> Salix retusa
Salix koreensis -> Salix pierotii
Salix libani -> Salix pedicellata
Salix matsudana -> Salix babylonica
Salix occidentalis -> Salix caroliniana
Salix onusta -> Salix myrtilloides
Salix pedunculata -> Salix × pedunculata
Salix pentandroides -> Salix × hermaphroditica
Salix recurvigemmis -> Salix recurvigemmata
Salix richardsonii -> Salix lanata richardsonii
Salix rubra -> Salix interior
Salix schumanniana -> Salix × schumanniana
Salix sericans -> Salix gmelinii
Salix serpyllum -> Salix fruticulosa
Salix wallichiana -> Salix disperma

Posted by loarie about 3 years ago

Folks should be able to do this through the interface. If you want me to help, I just need a table of
taxon_name, parent_name
e.g. Salix kikodsei, Vetrix

@loarie does the spreadsheet linked above qualify for that or does it need some reformatting?

I'm not familiar at the moment with all the other changes, except the Salix species issues at the end look the same as what we determined in the spreadsheet (i.e. when a species wasn't in POWO we noted what species/hybrid POWO has it as).

Posted by upupa-epops about 3 years ago

Chosenia is in the spreadsheet as a section of Salix so that works.

Posted by upupa-epops about 3 years ago

is it possible to make a clean spreadsheet (or paste something in here) where its
taxon_name, taxon_rank, parent_name, parent_rank
e.g.
Salix kikodsei, species, Vetrix, subgenus
Vetrix, subgenus Arbuscella, section
...
for every taxon that you want moved (and the idea being if the parent doesn't exist it will be created)?

Posted by loarie about 3 years ago

Yep I can make a new page on the Google Sheet with that format.

Posted by upupa-epops about 3 years ago

Okay made that table. For inter-section hybrids and incertae sedis species should I just make the cells blank? (you should be able to edit the sheet as well if that makes anything easier)
https://docs.google.com/spreadsheets/d/1ZyIyso9OonTZRhhne1cdZNWjUCRZP3IUX42J5EvuXAo/edit

Posted by upupa-epops about 3 years ago

For taxon_rank, should hybrids be "species" or "hybrid"?

Posted by upupa-epops about 3 years ago

hybrid please

Posted by loarie about 3 years ago

Got it, how does it look now?

Posted by upupa-epops about 3 years ago

@loarie is anything else needed?

Posted by upupa-epops about 3 years ago

there should just be 4 columns: taxon, taxon rank, parent and parent rank
I tried to flatten it and was busy removing duplicates (there shouldn't be any duplicates)
but looks like oyu have both
Vetrix subgenus Villosae section
and
Vetrix subgenus Vetrix section
can you go through and make sure there's no duplicates and each taxon is only represented once with one parent?
Also there are some rows with no parents - every taxon needs a parent
also are there rows for every taxon that needs to be created or moved below Salix - e.g.assuming Section Vetrix doesn't exist, then:
Vetrix section Salix genus
https://docs.google.com/spreadsheets/d/1ZyIyso9OonTZRhhne1cdZNWjUCRZP3IUX42J5EvuXAo/edit#gid=594741879
thanks

Posted by loarie about 3 years ago

@loarie ohh I see, just the direct parent. Makes sense! I'll do that now.
The rows without a parent will have the genus as the parent.

Posted by upupa-epops about 3 years ago

Okay reduced everything to 4 columns and got rid of duplicates.

Posted by upupa-epops about 3 years ago

OK I added/moved everything in the table and created a big deviation for the subgenera/section
there's still 277 "relationship unknowns" in the family
https://www.inaturalist.org/taxa/47567/taxonomy_details
it would be great to either swap these away or document whats going on with deviations
thanks all

Posted by loarie about 3 years ago

Awesome. Thank you so much!

Posted by upupa-epops about 3 years ago

There are a ton of Salix subspecies, varieties, and forms that I don't know what to do with; I've never worked with taxon relationships before. Also a handful of unplaced species names.

Posted by upupa-epops about 3 years ago

Add a Comment

Sign In or Sign Up to add comments