Another problem is that the entire genus has been synonymized with Orobanche. http://plantsoftheworldonline.org/results?q=Aphyllon Will make separate flag.

Hard to tell given the closed-access literature, but I suspect this would be better addressed with a Taxon Split on iNat than by re-identifying existing records (which is what seems to be happening). If they are sympatric, at least all the records of A. fasciculatum in range will get bumped up to genus and it will allow manual identifiers to refine them to A. franciscanum or A. fasciculatum without having to persuade past identifiers to update their IDs. Since the literature is closed-access, a split would also provide an opportunity for iNat curators (who, in this case, might be the authors of the paper) to explain how to tell these two species apart in a single place instead of in comments on every single obs of A. fasciculatum sensu lato.

Well, it's good news that they're not entirely sympatric. FWIW, you could help out curators a lot by posting a URL to the full text of your paper, e.g. on ResearchGate or on a personal website.

Here's the article if any curator wants make the split: https://drive.google.com/file/d/1Je4JksCLFbVLlJH1eB88KPKINq6Xl_Sn/view?fbclid=IwAR3h6cOfUxwdQow7dWexSiOHsBC_zsv0OUmV60lYsxvpUXg1cgzE8qhF6_k

On this specific issue, POWO does not yet support A. franciscanum, so strictly speaking, that taxon should not have been added to iNat until POWO picks it up. If we really want to support it, we could mark it as a deviation from POWO. Is that what you all want to do?

More broadly, I would not spend any time re-identifying observations that could be dealt with via taxon changes. As you say, it's not very efficient, plus it creates the inevitable mess in cases of sympatry where former IDs get replaced with coarser IDs which end up disagreeing with those manually-added IDs. I would instead consider whether a new taxon should have been activated via a taxon change, and if it should have been but wasn't, flag it to have a curator de-activate it until it can be handled with a taxon change.

Ok, looks like someone else set up a deviation, so I guess we're deviating from POWO. I made https://www.inaturalist.org/taxon_changes/111802 and I tried to set up atlases for these two taxa:

https://www.inaturalist.org/atlases/17463
https://www.inaturalist.org/atlases/53671

@benbenton01, @graysquirrel, @jdmore, and @acschneider, I'd really appreciate it if you could take a look at those atlases and make changes if they're inaccurate, because those are what will determine whether observations get automatically re-identified to species or bumped back to genus.

Thanks, Jim.

I did check some observations just outside the west edge of the A. fasciculatum Atlas in the Tahoe area, and it looks like the two near Truckee and South Lake Tahoe are good fasciculatum, so those two counties would unfortunately need to be added to the area of overlap.

I added all the counties touching Lake Tahoe to the A. fasciculatum atlas.

On the east edge of the franciscanum atlas, the published sampling map doesn't seem to show any locations in Colorado or Wyoming

I was trying to use the localities in Appendix I of the paper, which is why Wyoming is there. Colorado comes from https://www.inaturalist.org/observations/122852899, so maybe that needs to be accessed by the experts.

For that matter, it bothers me that such a huge number of fasciculatum observations across the Intermountain West (mainly NV and UT) are going to revert to genus

It kind of sucks, but we really don't know what those A. fasciculatum IDs really mean right now, right? It's unfortunate that we activated A. franciscanum before executing a split that would automate a lot of this, but I think we were always going to have the headache of needing to reassess records in the range of sympatry.

Agreed, I'm just suggesting that this can be done much more efficiently for Wyoming, Utah, Colorado, New Mexico, and maybe Nevada, by assessing them via the Identify modal outside the split while they can still be filtered as A. fasciculatum, instead of including those states in the franciscanum atlas and having to go through all Aphyllon observations in those areas after the split. The vast majority in those states are, and can remain, fasciculatum.

If you want to remove those states from the franciscanum atlas before committing this and then add them back after, go right ahead. IMO that's pretty fussy, but I'm probably not going to be doing much of the cleanup work outside of CA, so if that helps you and others out, go for it.

1.- Appendix 1 in S&B is only meant to capture the range at the state level. We did miss Colorado, but there is now ample evidence that A. franciscanum is present there, too:

https://www.inaturalist.org/observations/120544801
https://www.inaturalist.org/observations/122718840

2.- Atlas changes to A. fasciculatum
a) exclude from Arizona: Pima, Santa Cruz, Cochise counties
b) exclude from Texas entirely, or limit to northeastern Texas. See S&B:

"The presence of A. fasciculatum in Texas is uncertain. Reports from Montague County (BONAP, Kartesz 2015) and Ellis County (USDA, NRCS 2020) in northeastern Texas do not appear in Turner et al. (2003) and could not otherwise be verified. If present, such populations likely would be Aphyllon fasciculatum due to the similar habitat and geographic proximity to confirmed Oklahoman populations parasitizing Artemisia in dry prairie."

c) expand to include Klamath county (Crater Lake, OR observations are A. fascic.)
d) San Bernardino Co: only the two Mojave desert observations are fascic-- is there a way to more finely divide this area to split these off from the pops around the LA basin, which are all franciscanum? The two species split along biogeographic lines here that are not reflected in political boundaries.
e) add to California: Tulare county (see https://www.inaturalist.org/observations/48506917)

And a couple suggestions for outside the area of range overlap, where the plans are known from a very specific or ecologically unique portion of a much larger state:

f) exclude from Alaska all but "Southeast Fairbanks County, US, AK"
g) exclude from Indiana all but Porter and Lake Counties
h) exclude from Illinois all but Lake and Cook Counties

3.- Atlas for A. franciscansum looks fine to me. The species is known from the Franklin Mountains in El Paso County, TX, but there are no iNat obs from the state yet. Same for New Mexico- known but not on iNat

4.- "selectively" atlasing the taxa makes sense to me if someone wants to do it-- its really California, and to a lesser extent, OR, WA, and BC where the sympatry problem is the worst. In the intermountain region, the vast majority of observations seem to be A. fasciculatum. For example, looking through herbarium specimens at UN-Reno, I'd say 3% or less of A. fasciculatum records from the state were actually A. franciscanum. Alternatively, there are fewer than 10 iNat RG observations from the intermountain states that are A. franscicanum, so you could just make a list of those, and re-ID them manually after the split

5.- Adding section-level nodes would be great. There are names available:
Aphyllon sect. Aphyllon = purpureum, uniflorum, franciscanum, fasciculatum, and epigalium
Aphyllon sect. Nothaphyllon = everything else

Hope this helps- let me know if I missed anything

If anyone wants to help move this forward, A. franciscanum still needs to be atlased. I haven't checked on the sections Jim referred to.

Thanks Jim. I updated the split description so it links to the paper (which is now open, yay!) and quotes the key couplet that should give people some insight into how to separate the two species.

I also think this can be committed now, and will do so on Monday unless someone else gets to it before then or has any reason not to do so. FWIW, I think this is a really cool example of iNat providing useful data for research that then directly impacts iNat users.

Committed! Give it at least a day to grind through all those IDs.