Heads up: Some or all of the identifications affected by this split may have been replaced with identifications of Allium tricoccum. This happens when we can't automatically assign an identification to one of the output taxa. Review identifications of Allium tricoccum 55634

Comments

@loarie and @rynxs , and possibly @kitty12 , what more needs to be done for the taxonomic changes to be committed to?

Posted by pdabell 11 months ago

The Allium burdickii atlas is a bit too complex. Also the Allium tricoccum atlas is at state level so having the Allium burdickii atlas at county level won't gain us anything. Can you unexplode the Allium burdickii atlas to state rather than county?

Posted by loarie 11 months ago

@loarie , if I understand you correctly the result for states and provinces such as North and South Dakota and Ontario will result in huge numbers of Allium tricoccum sensu stricto that are currently not IDed to their variety of var. tricoccum losing their less refined IDs and being IDed no lower than at the level of the complex.

Posted by pdabell 11 months ago

correct - in places where the atlas overlaps, IDs of Species Allium tricoccum will be replaced by IDs of Complex Allium tricoccum. You can read more here https://inaturalist.freshdesk.com/support/solutions/articles/151000015337-section-d-how-to-respond-to-a-flag-requesting-to-split-a-taxon
there's really no way to split a taxon when the outputs overlap without rolling back IDs - my preference would be to just focus on doing the hard work of adding new IDs post split rather than worrying too much about rolling back IDs

Posted by loarie 11 months ago

RG Allium tricoccum in Ontario to be downgraded to Complex Allium tricoccum in the taxon split: https://inaturalist.ca/observations?locale=en&lrank=species&place_id=6883&preferred_place_id=57637&quality_grade=research&subview=map&taxon_id=55634 (1263 RG observations)

RG Allium tricoccum sensu stricto (A. tricoccum var. tricoccum) in Ontario to be swapped into A. tricoccum in the taxon swap: https://inaturalist.ca/observations?hrank=variety&locale=en&lrank=variety&place_id=6883&preferred_place_id=57637&quality_grade=research&subview=map&taxon_id=55634 (683 RG observations)

Needs ID Complex Allium tricoccum (lowest rank = complex), each with suggested IDs from A. burdickii and A. tricoccum (var. tricoccum)
currently in Ontario: https://inaturalist.ca/observations?hrank=complex&locale=en&lrank=complex&place_id=6883&preferred_place_id=57637&quality_grade=needs_id&subview=map&taxon_id=1457593 (7 verifiable observations in need of IDs)

The above current scenarios make sense but once the taxon split happens, the number of verifiable observations of Complex Allium tricoccum in Ontario will increase from 7 to 1270 and the 1263 observations of A. tricoccum sensu lato will no longer be RG when elevated to the complex. Considering the significantly few observations in Ontario that are potentially of A. burdickii are all contested (I admit a bias here since I am the primary user contesting the originally-suggested IDs and instead suggesting A. tricoccum var. tricoccum in most if not all cases), the taxon split will be a regressive change for observations of A. tricoccum there. Putting in the comparatively small effort to atlas the counties and similar administrative regions in Ontario that are within the range for A. burdickii recognized in A. G. Jones (1979) study* will be much simpler and potentially regressive only in those municipal regions.

Something else to consider: although I can't verify this, I suspect that many users haven't bothered to suggest an infraspecific ID for observations of A. tricoccum sensu stricto since there has been a lot of confusion promoted by print and online botanical authorities as to whether there are two species or one species with two varieties. And if there are two species with no infraspecifics, then the need to suggest var. tricoccum would have been seen as unnecessary, and possibly be a confusing addition to some users.

*Jones, A. G. (1979). A Study of Wild Leek, and the Recognition of Allium burdickii (Liliaceae). Systematic Botany, 4(1), 29–43. https://doi.org/10.2307/2418663

Posted by pdabell 11 months ago

@loarie , further to my above comments, it states at the link you shared above that "if the output taxa [in a taxin split] overlap spatially, atlases might not be helpful." Consider what would further happen if Canada's maritime provinces were part of Quebec or all the New England states were just one state, iNat would unnecessarily and unhelpfully roll back an even larger number of RG IDs of Allium tricoccum sensu stricto to the complex.

Although I think the taxon split will be regressive in Vermont, where the two atlases essentially overlap (save for in one county, I think), the atlas boundaries are at least largely in agreement with the boundaries in Jones's (1979) study (referenced above) and not decided upon because using county-level detail would be too complex.

Although I agree there should be a taxon split, proceeding with it in the way it is currently being proposed seems premature (even forced due to the accepted taxon split in mid February to Allium tricoccum var. tricoccum into A. burdickii--isn't helpful.

Posted by pdabell 11 months ago

@loarie , would it rectify the above issues simply, or at least somewhat, if you had access to a shape file of the range by county (and similar adminiatrative regions in Ontario and Quebec) for Allium burdickii?

Posted by pdabell 11 months ago

No, next steps are to coarsen the Allium burdickii atlas, commit the split, and then hope over time IDs that got rolled back will get rolled forward by the community making new IDs

Posted by loarie 11 months ago

If the taxon split was to be accurate by the atlased distribution of Allium burdickii* closely following the natural range of the species recognized by a botanical authority such as A. G. Jones aforementioned study or Flora of North America and not using the geopolitical boundaries of states and provinces (a significant shortcoming in POWO's distributions mapped by state and province), I'd support the split. Not, though, as it will currently proceed.

*As mentioned in an earlier comment of mine with my flag for the taxon swap of A. burdickii, it is not as necessary for A. tricoccum sensu stricto since, with exceedingly few exceptions, it's natural range encompasses the smaller one of A. burdickii. That being said, if needed, the range by county for A. tricoccum can be put into a spreadsheet and mapped/atlases, too.

Posted by pdabell 11 months ago

ok - I don't to object to abandoning the split, since I'm not in a good position to lead in the direction of my preference

Posted by loarie 11 months ago

The taxonomic split has to eventually be made but at least for now, @loarie , I think your abandoning the process is good.

@rynxs and @kitty12

Posted by pdabell 11 months ago

OK, at least we have the complex to catch affected IDs. Are we leaving var. tricoccum until the eventual split?

Posted by rynxs 11 months ago

If Alium tricoccum var. tricoccum swaps (merges?) into A. tricoccum sensu lato (as it currently is) before that taxon is split, how will new observations added that are A. tricoccum sensu stricto be IDed as such? If they can't be, then when the split occurs and they are within common range of A. burdickii and A. tricoccum sensu stricto, they will roll back into the complex. If var. tricoccum remains until the split, at least those RG observations will be swapped into A. tricoccum sensu stricto and they will remain RG. Only having one variety of a species is of course confusing.

Posted by pdabell 11 months ago

@pdabell I already know that, just confirming so I could add a temporary deviation to prevent other curators from swapping it before we split.

Posted by rynxs 11 months ago

@pdabell I can't help with the Nova Scotia question you messaged me about - maybe @rynxs can help?

Posted by loarie 11 months ago

@rynxs , I'm not sure if you are wanting a response from me on your comment about adding a temporary deviation to Allium tricoccum and perhaps A. tricoccum var. tricoccum. If so, it certainly makes sense to me to do so. @loarie , do you have any thoughts on this?

Posted by pdabell 11 months ago

@loarie , OK, thanks. @rynxs , I'll send you a message about my question that @loarie mentioned. I didn't add it here because I don't consider it specifically relevant, and it is one I have brought up in the discussion with the flag I made for the Allium burdickii taxon swap but I don't want to add any more to that already too long discussion if I don't have to.

Posted by pdabell 11 months ago

So...I'm confused, is a split happening or not?

I am getting a bunch of notifications for IDs that disagree with my sensu lato A. triccocum IDs, which is annoying.

Posted by bouteloua 11 months ago

@bouteloua , please see the above comments, which will hopefully help you understand why the split is on hold.

Regarding the suggested IDs you are being notified about, I know I am responsible for some if not most of those but many of mine were also in agreement with your suggested IDs of Allium burdickii. Hopefully before the split more observations that are on the fence in Complex Allium tricoccum will be moved to either the A. burdickii side or the A. tricoccum sensu stricto side before so that they don't end up remaining in the complex after the split with all the A. tricoccum sensu lato in areas with an overlapping range with A. burdickii.

Posted by pdabell 11 months ago

my preference would be to:
1) coarsen the atlases to the state level
2) commit the split
3) identifiers focus on IDing obs sitting at Complex Allium tricoccum to roll them forward to Species Allium tricoccum or Species Allium burdickii

I agree with bouteloua that this is what our policy on narrowing a taxon says and we should stick to it
https://help.inaturalist.org/support/solutions/articles/151000015337-section-d-how-to-respond-to-a-flag-requesting-to-split-a-taxon

But not knowing anything about these two taxa (I'm not an observer or IDer of either) I'm not interested in leading this split so I'm not in a good position to make this split happen. I also don't want to spend too much time on discussing this split and it's been taking up a lot of time already. But in short, I agree with bouteloua - it would be fantastic if we could focus on steps 1,2,3

Posted by loarie 11 months ago

If you use the recognized ecological range and not an arbitrary state- and province-wide one for the atlas for A. burdickii, then I will support the split. But, as mentioned above, basing the atlas on such geopolitical distributions will cause a sizable mess to IDs of Allium tricoccum in some jurisdictions. This will be particularly the case in Ontario, where the majority of RG A. tricoccum is outside of the very small overlapping ranges of A. tricoccum sensu stricto and A. burdickii. Of course, users will be notified of the taxon split and be made aware that they can update their IDs, but how much does that happen? Here's the link again to the RG. A. tricoccum in Ontario not IDed to its variety: https://inaturalist.ca/observations?lrank=species&place_id=6883&quality_grade=research&taxon_id=55634 . That being said, please don't put the necessary split on hold on my account. I support it. Just not as it is currently proposed with coarsened, state- and province-wide ranges in the atlast for A. burdickii.

Posted by pdabell 11 months ago

I unexploded the states in the A. burdickii which unbroke it (atlases can't handle hundreds of atlas places). Now if you run the Analyze ID tool
it looks like most IDs (94%) get rolled back to Complex

Total IDs of input taxon: 23327
Number of IDs Destination Atlas
1346 Species Allium tricoccum Atlased
3 Species Allium burdickii Atlased
913 Complex Allium tricoccum Outside of all atlases
21065 Complex Allium tricoccum Overlapping atlases

If this is now what you want, please make minor adjustments to the atlases. Its fine to explore a state or 2 but please don't make atlases with hundreds of counties

Posted by loarie 11 months ago

@loarie , participating in a taxon change is new to me. Unless that is normal and accepted practice on iNat, a 94% roll back to Complex of mostly RG observations seems to be something that should be prevented. Now, of course, everyone who added an ID will be notified and given the option to update their IDs if they want to, which, if they choose to, will resolve the issue a little. It's understandable when the roll back occurs to Allium tricoccum sensu lato in truly overlapping ranges but not in areas where the ranges do not overlap.

Perhaps the user who committed one species in the complex to a taxon swap wants to finish what they started and not semi-force taxon changes to the rest of the complex by other users. That swap to A. burdickii was done hastily and this one, although not being done without prior discussion, seems to me will be done poorly if so many RG A. tricoccum get rolled back to complex. That being said, at least many--though, far from the majority--A. tricoccum sensu stricto have been IDed to var. tricoccum, so not all RG observations of A. tricoccum sensu stricto will be lost.

Posted by pdabell 11 months ago

there's really no way to avoid rolling back IDs when a taxa is split and the outputs overlap geographically. Its not a huge deal, obs can sit at complex and if the photos show characters that can separate the outputs IDers will hopefully eventually roll them forward to species.

If you can alter the atlases to reduce the amount overlap without exploding everything down to county that would be great. If not, lets commit as is and keep moving forward

Posted by loarie 11 months ago

@loarie , OK, then I support moving forward with the taxon split as you and @rynxs have described it will be done. Can you re-open this abandoned one? And who commits it?

@rynxs , further to @loarie 's immediately-preceding comment, "if you can alter the atlases to reduce the amount overlap without exploding everything down to county that would be great," could you please do that?

Posted by pdabell 11 months ago

If the range for Allium burdickii was up to me and not as described by 1) Jones (1979) in the author's aforementioned and referenced study of it and A. tricoccum and even 2) Flora of North America, I'd omit all of Canada and New England from it. It would solve the problem of 100s of RG observations of A. tricoccum in those regions being rolled back to Complex Allium tricoccum. But it's not up to me. So I'll point to those sources, and even POWO's province- and state-based range if any users are annoyed by the result of the split in those locations.

Posted by pdabell 11 months ago

In opening the atlas, it looks like someone already unexploded it? I'm not sure how to roll back the atlas @loarie, but Illinois, Iowa, North Dakota, South Dakota, Wisconsin, Minnesota, Indiana, Michigan, New York, Ohio, Vermont, New Hampshire, and Quebec are all OK to unexplode.

Posted by rynxs 11 months ago

I've rolled back the atlas (ie unexploded states) - if you want to explode a state or 2 can you do that here https://www.inaturalist.org/atlases/84632
you'll have to also explode the corresponding states for the other atlas 2 because it the atlas of one output is pres/abs in Illinois and the other atlas is pres/abs in counties of illinois it will get treated the same as if the other atlas is pres/abs in illinois. Does that make sense? Currently the other atlas has no counties (ie its all at the state level)

Posted by loarie 11 months ago

I don't think that's possible for A. tricoccum because we don't have a county-level map for it, so let's just go ahead with the state level one

Posted by rynxs 11 months ago

Although it sounds like it wouldn't be useable, if you want a spreadsheet like I created for A. burdickii showing the counties and similar administrative districts in Ontario and Quebec, and based on the distribution shown by A. G. Jones (1979) study of both species of wild leek, I can do one of Jones's map showing the distribution of A. tricoccum. I won't be able to get it to you, though, until sometime next week. I realize @loarie that iNat doesn't use such detailed data for its atlases (yet?) but I'm offering anyway.

Posted by pdabell 11 months ago

@pdabell it would be useful for the states I didn't list above

Posted by rynxs 11 months ago

@loarie , I'm confused as to whether you can use a county-level atlas showing pres/abs for A. tricoccum sensu stricto. The reason I offered to create a spreadsheet with the data showing pres/abs was because it is implied--in my opinion, at least--in your last comment that it would be useful, but in some of your earlier comments and at the link you shared about the methodology behind taxon splits, I understood that you could only use coarse state- (and province-)wide data for the atlases. Please confirm the need for the data I mentioned in my immediately preceding comment before I prepare a spreadsheet with it.

Posted by pdabell 11 months ago

@pdabell an atlas can include both coarse and fine places, for many states we should be able to leave them coarse, but for the few which are overwhelmingly A. tricoccum sensu stricto but include a little bit of A. burdickii's range we can use counties. However, county maps for both species would be necessary.

Posted by rynxs 11 months ago

@rynxs , replying to your last two comments, your not needing county level detail for the states and one province you mentioned earlier makes the work simpler. Since the range of A. tricoccum sensu stricto in the majority of states and Ontario extends beyond--sometimes well beyond--the range for A. burdickii (e.g. Kentucky and Ontario) and includes portions of states and provinces that are not even in Jones's range for A. tricoccum (e.g. Georgia and New Brunswick), for the purpose of this completing this taxon split, your needing county-level information for the range of A. tricoccum in the states and provinces you didn't mention, doesn't seem necessary. In my opinion, it would be useful, though, to add the county-level detail to the range for A. tricoccum sometime after the split is committed to. I offered on Friday to list this detail for you and @loarie "sometime next week," but, unfortunately, it's going to have to wait until later.

@loarie , regarding the number of currently RG observations of A. tricoccum outside of the range for A. burdickii that will no longer be A. tricoccum following the split and will have their RG rolled back: 1) I now think that, despite its being messy with 100s of new complex-level observations to re-suggest species-level IDs for, it will at least present an opportunity to revisit the observations and reconsider what lower taxon they should be IDed in. If one or two users each volunteer to do it for a state or province, since there are only two accepted species in the complex, I doubt that the process will take too long.

Posted by pdabell 11 months ago

looks good - ready to commit?

Posted by loarie 11 months ago

@loarie , yes, I am ready for this taxon split to be committed to.

Posted by pdabell 11 months ago

Yes, ready to commit.

Posted by rynxs 11 months ago

@loarie , is this taxon split going to be happening soon? There seemed to be momentum 2 weeks ago when @rynxs and I were ready to commit to the split, but the process seems to have stopped, or something not communicated is stalling it.

Posted by pdabell 11 months ago

I committed it now

Posted by loarie 11 months ago

@loarie , that is good to hear.

Posted by pdabell 11 months ago

@loarie , why is the output taxon's ID of 55634 for Allium tricoccum sensu stricto the same as the input taxon's ID of 55634 for A. tricoccum sensu lato? Although I noticed this before the split, since it is how it is written above, not until after it did it seem confusing. I think that some iNat users will be confused by how IDs originally suggested to be A. tricoccum 55634 that were rolled back to the complex can have that suggested ID added back again. If the ID was different for A. tricoccum sensu stricto, it would be less confusing. Adding a note about the input/old taxon being sensu lato and the output/new taxon being sensu stricto might also be clarifying. For an example of what I am referring to, see the following observation at https://www.inaturalist.org/observations/6979026#activity_identification_d6dfaa59-f73b-48ca-b244-205869d34d29 .

Posted by pdabell 11 months ago

One aspect that is further confusing to people like me (far from experts) is that the notice box posted at the top of the dashboard for this taxon split states "Wide Leek" as opposed to "Wild Leek".

Posted by mimi_c 11 months ago

By clicking on the "Update your content" tab/icon at the top right of this page, it goes to another page, "Commit Taxon Change 124543 for [user ID]" and has a tab/icon there that reads, "Change your matching identifications for others to this taxon." I have two questions regarding this: 1) Although the taxon ID is the same for Allium tricoccum before and after the split, does that simply re-identify those identifications as A. tricoccum sensu stricto instead of A. tricoccum sensu lato, which it formerly was but wasn't indicated as such (and still isn't indicated as being sensu stricto)?; and 2) is there a similar tool for changing my identifications for others (and not other users' IDs) that were A. tricoccum that were rolled back to the complex during the split back--of sorts--to A. tricoccum. And if that is possible, can it be done for specific locations, such as just for my identifications on others' observations in Ontario or Vermont. I would be very hesitant about doing this for my identifications in states such as Illinois and Kentucky and would instead manually change any of those IDs, if necessary.

Regarding @mimi_c 's above comment, the common name for A. tricoccum is confusing since it's difficult to find "wide leek" used in a Google search. Wild leek is common and so is ramps. Use Ramps for the complex, wild leek for A. tricoccum and the currently-used common name of narrowleaf wild leek for A. burdickii. I doubt that such names would be confusing for anyone--experts and non-experts included.

Posted by pdabell 11 months ago

why is the output taxon's ID of 55634 for Allium tricoccum sensu stricto the same as the input taxon's ID of 55634 for A. tricoccum sensu lato

You can read more here https://www.inaturalist.org/blog/40417-using-a-taxon-split-input-as-an-output and also here https://help.inaturalist.org/support/solutions/articles/151000015337-section-d-how-to-respond-to-a-flag-requesting-to-split-a-taxon

1) Although the taxon ID is the same for Allium tricoccum before and after the split, does that simply re-identify those identifications as A. tricoccum sensu stricto instead of A. tricoccum sensu lato, which it formerly was but wasn't indicated as such (and still isn't indicated as being sensu stricto)?

pdabell, because you've not opted out of "Automatically update my content for taxon changes", most of your IDs were already rolled back to complex A. tricoccum by this split. So the only IDs showing up in the 'update your content' section are ones (e.g. from Massachussets) where the split left them alone because they didn't need to be replaced. So, assuming you trust how the split went, you don't need to touch any of those. That tool is mainly for people who opt out. As you say, if you used that tool to replace and ID of Allium tricoccum 55634 with one of Allium tricoccum 55634 you could do that but it wouldn't change anything.

2) is there a similar tool for changing my identifications for others (and not other users' IDs) that were A. tricoccum that were rolled back to the complex during the split back--of sorts--to A. tricoccum.

I'd recommend using the ID tool with these settings (reviewed=any, lrank=complex, taxon = complex A. tricoccum) to find obs that would benefit from being rolled forward (there's around 1.7k of them) https://www.inaturalist.org/observations/identify?reviewed=any&verifiable=true&taxon_id=1457593&lrank=complex
not sure if thats exactly what you're looking for though.

Posted by loarie 11 months ago

@loarie , thank you. That helps.

The input and output taxon issue makes sense--I think!--but I do wonder if this issue is a little different since A. burdickii's range is largely overlapped by A. tricoccum's (using Jones (1979) study and FNA, a little; and using GBIF occurrences, significantly) as opposed to having distinct ranges as in the examples. And the examples use ecological ranges as opposed to the ones that iNat has defined geopolitically, which results in A. burdicki's range being much larger in some states and in Quebec and especially Ontario than recognized by Jones and FNA.

About the automatic updates to my content in taxon changes, I had never considered that issue before, as until now, I hadn't participated in any discussion regarding one. It won't help in this case but I have now unchecked the box and have opted out of them.

Thanks for the ID tool settings. I will try them. Probably by province and state, though. [Edit: trying this a minute ago, since I have suggested IDs for a good (absurd?) number of those observations (almost 3k of the almost 9k total), changing "reviewed=any" to "reviewed=true" gives me more than enough to deal with! :) ]

Posted by pdabell 11 months ago

@loarie , in light of the abovementioned issue, is it considered acceptable on iNat to restore the ID that was changed due to a taxon change, such as I did on observation 141840568 since, in this case, the taxon I originally suggested didn't change in the split? Or is it considered more acceptable to instead add a new ID of the originally suggested taxon, such as I've done on observation 140002697? If the former option is acceptable, it makes the process of re-suggesting an ID a little simpler. In my opinion, it also doesn't look as confusing to the average iNat user that isn't a botanist or other expert.

Posted by pdabell 11 months ago

Add a Comment

Sign In or Sign Up to add comments