see https://www.inaturalist.org/observations/9558997
just some notes to myself or if someone wants to pick this up, since I'm logging off soon.

-check taxonomic authorities
-check spelling, swap C. giesekiana into the correct spelling if needed
-if need to split, create new C. rotundifolia, atlas both species
-check for extent of possible cultivation
-create taxon split

cc: @wdvanhem

Love, A and D. Love. 1965. Series in Biology No. 17: Taxonomic remarks on some American alpine plants. University of Colorado Press, Boulder, CO. 43 pp.

Roquet, C., L. Saez, J.J. Aldasoro, A. Susanna, M.L. Alarcon, N. Garcia-Jacas. 2008. Natural Delineation, Molecular Phylogeny and Floral Evolution in Campanula. Systematic Biology 33(1), pp. 203-217.

The former reference suggests C. giesckiana is a tetraploid taxon limited to arctic-alpine habitats in western North America and Arctic North America. C. rotundifolia is present at lower latitudes and elevations.

The latter reference considers C. rotundifolia and C. gieseckiana conspecific, which is weird if their ploidy differs.

Also notable is that a number of references consider C. gieseckiana to = C. rotundifolia ssp. groenlandica

Love & Love! :)

ITIS spells it C. giesekiana and says "See Shetler (1982:35) regarding the orthography, omitting the c in 'giese(c)kiana' ... Although Shetler (1982, and pers. comm. 2012) argues that the nearctic harebells should all be considered several variable ecotype 'races' of a single species (as Campanula rotundifolia L.), Lammers (2007) restricts C. rotundifolia to the "Europe to S. Siberia, Manchuria [...], & Sakhalin" (and "naturalized in New Zealand, Tierra del Fuego & Falkland Isl."), and recognizes a number of N. American taxa separately. ITIS is following Lammers in this, and changed distributional information accordingly, pending further evaluation"

VASCAN: Splits, spells it C. gieseckeana

The Plant List: Accepts split, but both C. gieseckeana and C. giesekiana lol

Plants of the World Online accepts split, spells it C. giesekiana

Sounds like we should go with C. giesekiana, but I'm not so sure how easy atlasing is going to be...

You can at least review them all here if you wanted: https://www.inaturalist.org/identifications?user_id=wdvanhem&taxon_id=543632&current=true

I've been on here since 2012, the year I started really learning about plants...so a lot of my old IDs are probably total crap...

Here are some comments from an Ontario perspective (from an email to NHIC biologist staff).

As you know, what was traditionally called Campanula rotundifolia in Ontario was recently changed to C. gieseckeana by VASCAN. The taxonomic authority for this change is Lammers, T.G. 2007. World checklist and bibliography of Campanulaceae. Royal Botanical Gardens, Kew. I have not seen this publication but have requested through interlibrary loan via the MNRF library.

Both ITIS (Integrated Taxonomic Information System) and VASCAN follow Lammers (2007) and I have also used this name in our subnational Biotics Elements Table.

Periodically I compare VASCAN’s Ontario species list with our list in Biotics and when I did so recently I was surprised to see another unfamiliar Campanula name, C. intercedens, listed for Ontario and also following the Lammers (2007) checklist. This got me looking at specimens and doing some online research and literature review and though I didn’t find out too much, I thought I’d share what I learned.

The Campanula rotundifolia group is a taxonomically complicated circumpolar complex of species with multiple ploidy levels. It has been studied a fair bit in Europe but very little, at least in recent decades, in North America. In the strict sense C. rotundifolia is restricted to Europe and does not occur anywhere in North America. Campanula gieseckeana is a circumpolar arctic species and likely occurs in Ontario only in the Hudson Bay Lowland (HBL). The Pan-arctic Flora (PAF) lists it as “scattered” in Hudson Bay – Labrador (as C. rotundifolia ssp. gieseckiana), but I don’t know what its distribution and status is in the HBL relative to C. intercedens. Riley (2003) in his HBL flora lists C. rotundifolia as widespread and occasional and C. “rotundifolia” was recorded at about 25 Far North Biodiversity Project (FNBP) locations during recent fieldwork (see attached map; interestingly only one FNBP record west of 85 degrees Longitude).

Unfortunately we didn’t collect many C. “rotundifolia” specimens during FNBP fieldwork and there are only two Far North specimens in the NHIC herbarium (plus two from southern Ontario and one from Rainy River District). One of the Far North specimens in the NHIC herbarium (from a beach ridge at Radar Site 415 in the Cape Henrietta Maria area) is superficially quite different looking from typical southern Ontario C. “rotundifolia” and I suspect is C. gieseckeana based on the characters below. I haven’t found a key to North American members of the C. rotundifolia complex and interestingly the best identification information I’ve found comes from a 1903 review of a 1902 monograph in German of the C. rotundifolia group (see Rydberg 1903, attached). Apparently C. gieseckeana “is distinguished from … the others of the C. rotundifolia group by the short and broad hypanthium, which in flower is much broader than high. It is usually a low plant, densely leafy below and naked above, usually one-flowered. The lower stem-leaves are often spatulate and obtuse. C. Giesekiana is an arctic plant growing in Europe and Asia, as well as in America, where it has been collected in Greenland, Labrador and on the islands of the Baffin's Bay region.” The other Far North specimen, from the gravel floodplain of the Ekwan River, is more like C. intercedens in not being densely leafy below and having multiple flowers per stem. Scans of the suspected C. gieseckeana Far North specimen, plus a typical southern Ontario (Huron Co.) specimen of C. intercedens are attached.

I assume that all C. “rotundifolia” south of the HBL (and likely some C. “rotundifolia” in the HBL) are C. intercedens, which apparently is the common eastern North American member of the complex (south of the Arctic). I have added C. intercedens to Biotics and ranked it S5. I have changed the rank of C. gieseckeana from S5 to SU until we have a better understanding of its distribution and status in northern Ontario.

Add Sutherland and Galloway (2018) to your considerations of the circumboreal range of the polyploid complex of Campanula rotundifolia
https://bsapubs.onlinelibrary.wiley.com/doi/pdfdirect/10.1002/ajb2.1162

Is this split happening? C. rotundifolia is still showing lots of observations where there shouldn't be under the taxonomy the site is currently using.

This species is being split into more than just C. rotundifolia and C. giesekiana. C. intercedens for example, and several others.

C. alaskana, C. kladniana, C. macrorhiza, C. moravica, C. nejceffii, C. petiolata, C. ruscinonensis, and C. willkommii should be the others.

All species split off from C. rotundifolia should be in the Harebell Complex.

An update: all species are now in the « Complex Campanula rotundifolia ».
513 identifications are « Outside of all atlases » and 22170 identifications are in « Overlapping atlases », and all these identifications will be moved to the Complex Campanula rotundifolia if the split is committed now.
That's why it is important to check all the atlases before committing the split.
And if anyone has better information about the distribution, that would be great! (I used atlases from POWO, but they are not very accurate)

@bouteloua @wdvanhem @reuvenm @michael_oldham @raymie @sejohnson ?
Do you still agree with the split? Can you check the atlases? Can you share any information that you have on the distribution?

I have no strong feelings about the split. I still wish there was a reference that coherently described the species with a key. Most identifications seem to be along geographical lines, which makes me skeptical.

A lot of observations of these species are stuck in taxonomy disagreement right now, so I'd say the sooner this split can be committed the better to try and help resolve this issue.

I don't really know the ranges well enough to know super precisely, but if they match the POWO ranges I imagine you're fine.

Yes, just checked and all looks good.

I think this split is a mistake. I do not know of any genetic studies in support of the split. The taxonomy mostly based on leaf characters that are highly variable. Most plants are renamed simply because of their locality, not based on phenotypic or genetic characters. Multiple species grow together and many cannot be identified to species. Have fun reidentifying.

it does still say "Taxon change commit may still be processing so be aware that some records aren't yet updated" on the top of the taxon change page even though it was committed 5 days ago

https://www.inaturalist.org/taxon_changes/134716

@loarie is there something hung up?

I have redrafted the taxon split here : https://www.inaturalist.org/taxon_changes/141283

Oh ! Sorry : I have published my comment too quickly and my second phrase is missing. I have re-drafted the split to show that the first split wasn’t fully committed. But I never intended to commit the second split. I have deleted it.