For me we could accept POWO treatment

I use T. officinale in sensu lato; makes sense to just merge with the section.

@srall @susanhewitt @bouteloua @loarie @crbjork @snedergaard, @blue_celery, @sambiology, @davidenrique

It must be remembered that any taxonomic presentation is a hypothesis, subject to testing, validation and of course later revision as more evidence is introduced and analyzed. One thing botanists have been very good at is preserving evidence (type specimens and literature) which facilitates re-evaluation of hypothesis. Botanists have also generally been open-minded and flexible, allowing for differing conclusions based on the evidence. As a result, options are available to users who interpret the same data differently. Whether Taraxacum officinale qualifies as a species or not will always be a matter of opinion and will probably always be debated among specialists and non-specialists alike. The system as it exists now does not preclude anyone from applying a sectional name to any observation if they so chose. But merging the name into section Taraxacum or another section (and eliminating it) will preclude everyone from being able to apply a species concept to the weedy, apomictic, triploids that occurs across the world. This is contrary to the long tradition of flexibility in taxonomy. And once that name is eliminated, a tremendous of amount of information will be lost, hindering later re-evaluation. As it is now, it is relatively simple to construct any scheme one chooses based on the available data. But once the specific name is eliminated, only one scheme will be possible to apply from the available data.

Affecting either change proposed here will have tremendous disruption. There are currently over 83,000 observations carrying the name Taraxacum officinale Because there is so much debate and ambiguity around what is a species and how to apply the concept to Taraxacum officinale, I think it is unwise to force the entire world into one conclusion and permanently eliminate the possibility for alternative interpretations. A consumer of science (all iNat users included) will rightly question why scientific principles are being applied arbitrarily and unevenly.

An example of this is Persicaria coccinea which has variously been considered a species, subspecies, variety or plain synonym of Persicaria amphibia. I have been studying these plants for over a decade and have been accumulating data to inform a rational and scientifically sound hypothesis about these plants. Several years ago, someone committed a taxon merge to lump the name Persicaria coccinea into Persicaria amphibia because that is what Flora of North America did. Luckily, there exists a varietal name (preserved in iNat and elsewhere) that has enabled the identification of both entities distinctly. I had examined most of the thousand or so herbarium specimens held in major North American herbaria, but had not come to a point of formulating a firm hypothesis. With the availability of over 2,000 iNaturalist observations, most with good (often high-resolution) diagnostic photos and precise georeferencing and notes, it was possible to discern patterns that were obscure before and add a whole new line of evidence (geography and ecology) to bear on the problem. See the journal post here. If the names Persicaria coccinea and Persicaria amphibia var. emersa had been deleted and merged into Persicaria amphibia, none of this new analysis would have been possible. After the person saw the new evidence, they wrote to me and apologized for committing those name changes and offered to reverse them. I responded that to do so would cause yet more disruption prematurely. It was still possible to distinguish both entities with a name to facilitate further analysis and so a change at that time was not absolutely necessary. I said the change should be committed when a solid body of evidence had accumulated and been peer-reviewed. I am honest enough to admit that I have biases and opinions, but I am a scientist and I will not impose those on others. At some point I will present a hypothesis which the world will be free to review, test and interpret as they see fit (The journal post above is a draft). But at no time will I preclude others from accessing and using the data as I did or repeating my methods. I also know that additional evidence will come to light (such as molecular genetics) and it must be possible to integrate that data into the body of existing data in the most flexible and rigorous way.

In the interest of scientific integrity, I oppose both changes. The name Taraxacum officinale must be preserved and available for use.

Is any information lost by committing this merge? As far as I know T. officinale is considered roughly equivalent to section Taraxacum by both "sides". The disagreement is mainly just what label to use. If you prefer the macrospecies label, you don't recognize any other species in the section so having the section is redundant. Using POWO's taxonomy, identifying to section is exactly as informative. The name "Taraxacum officinale" will remain in the iNat system as a synonym and thus still usable in a way.

If I understand the situations correctly, the Persicaria situation is looking at a clade with two lower clades and asking whether the species designation should apply at the first or second level. If it was used at the first level and no sub-specific labels were applied to the two lower clades, valuable information would be harder to collect. I'm not sure that this situation is analogous.

Information is lost if an observation can only be identified as either T. officinale or a microspecies of section Taraxacum.

Let's say we can all agree that an observation can be identified as T. officinale sensu lato (there is no such thing as T. officinale sensu stricto by the way), or T. officinale agg. Let's say we could also agree that the observation keys to the microspecies T. ekmanii.

Currently there is no mechanism in iNat's taxonomy for both identifications to apply at the same time; since both "species" are currently set as if they are different clades within section Taraxacum. I don't think anyone thinks that is how it should be?

we should be extremely wary of doing taxon merges unless there are very clear and compelling reasons to do so. Conversely, we should be very willing to err on the side of making taxon splits. There's just an inherent asymmetry there that makes merges much more problematic.

@davidenrique because of the strange situation going on here, that's not really what's happening. The actual curatorial action being taken is a single merge, but it wouldn't be entirely inaccurate to describe the general taxonomic change happening here to be splitting the species into several hundred species. "Common Dandelion" changing to "Common Dandelions". The weird situation is that all those other species already exist in the system and are competing with the one species.

So my call would be to keep officinale and erythospermum as "informal" complex-level taxa. That can be done by simply editing the taxa at https://www.inaturalist.org/taxa/47602/edit and changing "species" to "complex", without taxon swaps. Then, allow entries to be made for microspecies headings under the correct sections where needed. It's a trend already used by Rubus fruticosus among other aggregate microspecies complexes.

The problem here is that it seems that T. officinale is a binomial that should not be used even for plants found in the area where it was described for some issues related to the type.
Apart this, maybe we could try to see the matter from the newbies' point of view. Seeing that the name T. officinale is allowed here (and almost always proposed by the AI) would just make them think that it is a good name and that they have photographed the real T. officinale. This means allowing a false truth or a cliché to be spread among people. We have the opportunity to try to make people reflect on what they find rather than taking for granted that every common dandelion is T. officinale.
In the end I think that there are already too many generalist website that take for granted these clichés (at least here in the old world almost every wild rose is Rosa canina according to many websites, etc...).

We have to live with genus IDs for plenty of common insects, plants, lichens, etc. Why is the name "Common Dandelions" so much worse than "Common Dandelion"?

It's true that some organisms are difficult to identify from photographs (or even DNA), but that does not mean they are not species as most people think of the term (with valid names). It just means we don't have enough evidence to consistently and accurately apply a particular name or concept. The situation with the common Dandelion is different. It is entirely possible and indeed happens all the time that one can apply the name "Taraxacum officinale" to a plant based on decent visual characters and to do this consistently across space and time. As I said, some may think this is is too broad a category, but it defines a well-known and very consistent assemblage of plants (albeit with internal variation) that is generally understood to be a "species" and has a whole suite of well-defined character states (morphological, chemical, genetic and behavioral) that are consistent from generation to generation (albeit with internal variation). This is the classic definition of a species and it can be applied with remarkable accuracy and consistency to these plants. The concept of "Common Dandelions" encompasses a constantly changing array of entities that are only vaguely defined on the basis of continuous and often very ambiguous characters. One person's concept of "Common Dandelions" is therefore subjective and can differ substantially from the next person's. This reduces botany to a matter of personal taste and opinion which is not scientific.

Using the name Taraxacum officinale in no way hinders the research on or discussion of the complexity of these plants and the concept of microspecies. Those choosing to are fully empowered to apply the name "Taraxacum" sect. Taraxacum or any other taxon name (like a microspecies name) to the plants without hindrance. But elimination of the name will severely handicap research and discussion of these plants for reasons I have explained. What is the purpose of this change and why the urgency?

I teach my students that plants are very pragmatic and cooperative and it is possible to know them intimately and to trust in their goodness and wisdom. It is we who act insane!

@upupa-epops said:

"We have to live with genus IDs for plenty of common insects, plants, lichens, etc. Why is the name "Common Dandelions" so much worse than "Common Dandelion"?"

The reason this is a different problem is because these two viewpoints are not conflicting, but different views on taxonomy to begin with. This is not a disagreement on whether observations can and should be identified at species level or otherwise, though I still argue that classifying these names is a benefit for reasons given above. But rather the founding of microspecies follows a different biological and taxonomical path, because it recognizes distinctions that are otherwise synonymized or considered inconsequential in "traditional" taxonomy.

iNaturalist's goal is to have a consistent (as possible) global taxonomy. And either direction here causes iNaturalist to then become inconsistent with one or the other approach.

So the only solution I see is allowing the framework for both viewpoints simultaneously, which is:

1- keep T. officinale and erythrospermum as complex headings. Whether T. officinale is kept at genus or under section Taraxacum is for debate.

2- allow entry for microspecies headings under the correct sections. This allows the specific entry for those who are able to use it.

T. officinale and erythrospermum are already informally used as a cluster name and have been (mis)used this way for decades. But it's come to a point that it's actually become beneficial enough to warrant keeping it around to distinguish these observations from "other" sections and species. That's why dismissing these on iNaturalist's end offers no solution, because other spaces are going to continue using it regardless of what the microspecies taxonomy follows.

In no way did I mean to imply that any of us here were acting insane. I'm sorry if anyone took it that way. I just meant that the plants are doing what plants do (very successfully, I might add) and it is we humans that are confused- myself included!

I really don’t want to interfere in the debate, but here is a viewpoint from across the ocean.
The taxon T. officinale has been out of use among botanists (professionals and amateurs) for more or less a century in my part of the old world, and I´m pretty sure that this “loss” has done nothing to diminish enthusiasm among naturalists. With iNat this “species”, which has not existed among us for a long time, has been reintroduced in very high number. It covers almost half of all new finds, and it seems to be identified more or less randomly. As far as I can tell, the knowledge and effort needed to correctly identify genus Taraxacum is approximately the same as needed for T. officinale. The most experienced expert in my country (not me) has spent a life-time studying dandelions. He is able to identify most (perhaps 90%) of all collected and dried specimens from the country and surroundings into species (or microspecies as termed by some). I once asked him whether some of the material he had seen could be T. officinale. His answer was: “I don´t know, I don’t know that species – and no one does”.

Some various points...
-- There is no urgency, I just flagged the taxon to note that the POWO change had happened and to spark the discussion.

-- Some of the objections here are to the recognition of microspecies, but hundreds of microspecies are already in iNat dandelion taxonomy and I assume they're going to stay? They are getting good use by a few observers in Europe (none of whom are career dandelion experts as far as I know). I assume iNat also has microspecies for Heiracium, I don't know what the situation is for other apomictic groups. Assuming the microspecies and section Taraxacum are staying, keeping T. officinale would be having two competing taxonomies at the same time.

-- T. erythrospermum is a valid species; in the strict sense I believe it's a non-weedy sexual Mediterranean dandelion, so it's not going away. (according to a French source I found, it's similar to T. rubicundum but with lots of pollen, araneous peduncles, and a bit smaller) However, I wouldn't be surprised if there are no observations of it on iNaturalist in the strict sense. I still don't understand why observations can't just be identified to section, since that wouldn't lose any information and it would allow the species to be used in the strict sense so those observations aren't lost among the thousands of others.

-- Many Erythrosperma microspecies do not have reddish achenes (their colour is important for keying them to species), and some members of other sections including Naevosa, Spectabilia, Crocea, and Celtica do have reddish achenes. In Europe it doesn't seem to be considered a reliable ID method for section.

-- As far as I know the concept of section Taraxacum has been fairly stable, although it has gone through some name changes. There are specific consistent characters used to identify plants to section. I'm less clear about the concept of T. officinale. Formally it's equivalent to section Taraxacum, but in practice, at least in North America, it's less clear because the Flora of North America key doesn't necessarily match how sections Taraxacum and Erythrosperma are defined in Europe, and was made before sections Borea, Boreigena, Celtica, Hamata, and Naevosa were discovered in North America (in the BC paper). Are all of those by default implicitly included in T. officinale?

Honestly the more I think about this situation the more it occurs to me the issue is less the taxonomy but more that the coverage of the known resources are practically missing or unexplored. I think if there was more widespread understanding on the microspecies things would be far easier. For instance, I'm 100% sure that the "erythrospermum" I see in the US, Australia and elsewhere on lawns is the same species which is likely Taraxacum subbracteatum or a relative.

Until the necessary research on specimens and populations has been done, it's going to be floundering all over. Personally I'm willing to put in the effort to ID these plants but the resources are just not available, and a lot of the US content is probably undescribed.

I'm also reviewing my talks about data and I'm no longer feeling much of a weight to defend T. officinale. It's too arbitrary, it's obsolete. I'm not going to lose sleep over it.

@danielatha I appreciate your comments. Variations in morphological expression are certainly correlated with environmental variables, and this holds for every single species of the genus (including the non-weedy ones). However, thanks to the existence of i) type material from the original collection provided by the author of the species, and ii) reported results from cultivation experiments to test and verify the robustness of the distinguishing characters (as also pointed out by @blue_celery), it is possible for any researcher to examine, evaluate, and challenge the dataset behind a given (micro)species name. This is not the situation for T. officinale.
@silversea_starsong: Thanks for your latest comment. You have nailed it!
A further remark: With the current taxonomy, I´ll be reluctant to help identifying species found in North America. I have tried it a few times, where it was possible to identify the section, but the situation is awkward. Let´s say you have a find of T. officinale with research quality. There are several sharp photos of different parts of the plant allowing for an ID, which turns out to be sect. Hamata. To me, sect. Hamata is the correct and indeed more precise ID for the plant than T. officinale. However, in the mind of the finder, sect. Hamata would be less precise as he/she already has made a verified ID at species level. I feel that my interference would only serve to cause embarrassments and confusion by (apparently) bringing it to a higher taxonomic level.

@silversea_starsong here’s a start: https://www.inaturalist.org/journal/upupa-epops/48329-resources-for-identifying-dandelions-to-section

Are sections not globally defined? I assumed that the characters that distinguish the sections apply anywhere in the world, by nature of their characteristics.

I think the first step to having useful local resources is to have a global one prepared. From there, it can be repurposed into country or local flora contexts.

The UK keys don’t include some sections that are found elsewhere in Europe, so it would be nice to at least have one that includes all of those.

I just skimmed through this discussion, so forgive me if I am repeating anything already said.

I have to wonder if the POWO taxonomy is still in flux here, and maybe we should wait to see where it settles out.

The reason I wonder that is that the current state of their taxonomy does not follow the International Code of Nomenclature. Any genus or subdivision of a genus has to have a species as its type, and must include at least one species.

As currently constructed in POWO, the taxonomic concept "Taraxacum sect. Taraxacum" may have a type, but it contains no accepted species, only non-accepted species listed as synonyms. It is an "empty" taxon.

Yes, this is the first instance I have encountered of POWO using any classification between genus and species. And they are not even doing it right (yet, anyway).

I suppose they may have taken the position (understandable) that species taxonomy in Taraxacum is impossible, and just abandoned any attempt. The iNat taxonomic system certainly is set up in a way that we could still reflect what POWO has. But I have an issue with reflecting incorrect taxonomy (a section containing no species), even if it is POWO that is promulgating it.

From a purely nomenclatural/taxonomic viewpoint, one alternative for POWO is to recognize Taraxacum section Taraxacum (and other sections in similar situations) at species rank using the oldest available species name in the section. I understand that such a species may be meaningless biologically, but it would at least conform with ICN.

Or another alternative might be to "de-synonymize" the names currently included in the section, and accept them as species members of the section, even if the taxa they represent can't be circumscribed beyond the holotype of each name. Then we would just have to accept that most or all other specimens (and observations) can't be identified as any one of those species, and can only be identified at the rank of section. But then at least we would be using a valid section that contains at least one species.

@jdmore Wouldn't it also work for them to just not mention the sections, and accept all the species within the sections, like they've done for every other genus that uses sections? I think they're already doing that basically: http://powo.science.kew.org/taxon/urn:lsid:ipni.org:names:328762-2

From a quick look, all the synonyms they have under section Taraxacum are old synonyms of T. officinale, except one which I think is an old synonym of T. palustre. If it weren't for the fact that the section existed before they synonymized T. officinale with it, I would think that they only have it for that synonymization to work.

Your first suggestion is mentioned in this paper, which says:

Richards and Sell (1976) overcome the problem illegitimately by employing a macrospecies name equivalent to the section (and by implication the type of it) which is the earliest agamospecies name of certain status. … This practice seems contrary to the Code and should be ignored unless formal lectotypification has occurred.

I don't know the Code though so I don't know why the author says that method is illegitimate, since he doesn't go into more detail that what I quoted as far as I can tell.

In 2011 the type of T. officinale was intentionally designated to a specimen that is not identifiable past section level: https://www.jstor.org/stable/41059837

This was done to allow people to use either the T. officinale sensu lato macrospecies or the section Taraxacum section/microspecies taxonomy and have both be valid. It seems like a pretty convoluted decision but I guess the authors thought it was the least worst solution to the mess.

However, you can't functionally do both at the same time like iNat currently, and POWO formerly, was trying to do. It doesn't make sense to accept both officinale and a microspecies within officinale (e.g. T. ekmanii) as species.

I'm sharing a link to the forum discussion too just so there's connection between the various Taraxacum-related discussions.
https://forum.inaturalist.org/t/overlooked-dandelion-diversity-in-bc-and-everywhere-in-north-america/3808

@blue_celery @danielatha Not sure I'm ready to kneel before it or anything like that, but I am in awe by how much a genus like this can get people talking about taxonomy! :-)

For what its worth, this is the most observose plant on iNat that's not accounted for in POWO directly or via a deviation.

It would be great to either get rid of this taxon or set up a deviation.

Short term, can we at least swap the two ssp in to the species? It doesn't look like they're being used much and that would at least simplify this a bit

thanks - I committed https://www.inaturalist.org/taxon_changes/95650

The species names, as they currently are, are basically meaningless anyway. Most users are identifying everything dandelion-like as T officinale or T erythrospermum, at random, depending on which one the CV suggests first. And then they immediately get confirmed to research grade by some other user who thinks it's a simple species to ID. I don't know how many observations have any kind of correct identification, but I suspect it is a very small percentage.

I'd really like to see them all combined into a complex.

I think it's because POWO has to list a source to explain why they aren't using a name, and in their template that's listed under the title "Not accepted by". For this case it doesn't work to have both T. officinale and sect. Taraxacum at the same time (because they're synonyms; you can only use one or the other). So they have sect. Taraxacum and then reference that source to explain why they don't also have T. officinale. Does that make sense?

Any outcome here is probably not going to help as things stand...

The complication is many if not most of the existing officinale observations are not referrable to section Taraxacum in the (proposed) expanded taxonomy.

iNat isn't a taxonomic database, therefore it should not make its own taxonomic decisions. We should be following our selected taxonomy exactly and not trying to deviate and create our own. We should follow POWO and not use this species.

@blue_celery You are free to follow whatever taxonomy you want - but you are not iNaturalist.

@raymie I am part of iNaturalist and I can act as far as I am allowed by its rules

New forum thread about dandelions, to keep conversations connected: https://forum.inaturalist.org/t/about-those-dandelions/39650

As I said there, I think iNat doesn't really have any option but to follow the European section/microspecies system because it's the only consistent and global taxonomy system that exists for Taraxacum. This is regardless of my personal opinions about microspecies; I think North Americans are lumping too much here and Europeans are probably splitting too much.
T. officinale doesn't exist in that system and is a synonym of sect. Taraxacum, just as POWO has it. The FNA taxonomy only includes a small number of the introduced Eurasian aggregate species and on iNat we can't just decide independently how to classify other species/sections when they show up, or how to make that system cover all of Europe and Asia.

Does anyone have other alternatives that fit with iNat's taxonomy policies?

we try to match parts of our taxonomy to global taxonomic authorities. When that's not possible, we try stitching together regional taxonomic authorities (not recommended because of inevitable regional taxonomic conflicts ...)

"It also may not be self-consistent, if, as said above, not every taraxacum officinale observation is even in 'section taraxacum'"

That is correct, but it mostly means that those ID's are mostly incorrect and should not be in fact Tarraxacum officinale sensu Kirschner et Štěpánek 2011, but e.g., T erythrospermum sensu lato other respective broad sense species. I myself thought that I can reasonably recognize that the Taraxacum I am looking at is Taraxacum officinale=section Taraxacum and then the first time I tried to post something it was actually Taraxacum lacistophilloides that is considered to be sect. Erythrosperma. It is often just better to ID to genus Taraxacum because even "Taraxacum officinale" can easily be incorrect.

Or the name Taraxacum officinale as used here should be taken in the sensu latissimo as really the whole genus Taraxacum. It just should be noted somewhere. For example, this page https://observation.org/taxa/10719/?genus=Taraxacum explicitly uses Taraxacum officinale s.l. (incl. all sec.) and that makes it clear that all sections are included.

Ok, so if those are just normal mis-IDs to section, that is not a problem with inat's taxonomic concept, right? That is just a problem that can be resolved just through the normal community ID process. There are already plenty of situations where we deviate from POWO for good reasons, and as mentioned above it sounds like the POWO taxonomy apparently isn't even internally consistent on this point. I think in hopefully most of the situations where we have a true deviation it is sourced to another secondary source, and we do have other secondary sources we could cite for the deviation, such as the FNA or any other North American secondary source. This is part of the reason we won't be able to ID a split taxon in North America at all, even if we wanted to, because even if primary literature proposes species names in North America they likely won't be accepted by the North American secondary sources in the near future, so a split in North America would most likely just leave all of our taraxacum observations in permanent limbo where there simply is no valid inat species name to apply to them.

If I'm being honest, even knowledge of the impending possibility of the split has already discouraged me from observing taraxacum, even when I observe practically every other flowering plant species at a site. I don't really want to spend my time taking observations that may shortly become useless permanent needs ID filler.

"making it possible to ID as T. officinale while making it at the same time equivalent to either section Taraxacum or genus Taraxacum somehow would be a good solution."

There is a really easy way to do this, which is to call the microspecies subspecies (or varieties, or forms, or whatever you want to call them) of taraxacum officinale. From a technical perspective, that is a compromise that completely fulfills all relevant parties needs; it is possible to refine without disagreeing, and it is possible to achieve RG at either or both ranks without contradiction. It naturally handles the fact that not all the micro species are known, and doesn't interfere with IDing them in the future if new ones are described. It doesn't mess up the computer vision's ability to help new users. It is consistent with keys available in secondary sources in North America. Implementing it could be done by one curator in less than an hour to with almost no re-indexing load on inat's servers. It is fully invertible with no loss of data if North American secondary sources do someday change their mind about recognizing microspecies.

I think this compromise is the best of all worlds, and I struggle to understand why anyone would dislike it from a purely practical inat-focused perspective. The best I can tell is that some people seem to think of a microspecies as much less ecologically important than a true subspecies, but inat doesn't have any less important terminal node to call it than a full 'species'. I think that issue doesn't really matter, because inat is not a resource for taxonomy. Whether inat applies the names 'subspecies' vs 'variety' vs 'form' vs 'microspecies' is just a minor nomenclatural/philosophical issue about the name as it appears on autocomplete suggestions. I think what we should do is just call them a 'subspecies', and all collectively agree that 'subspecies' is not actually what we mean by the concept, it is just the closest available label among the inat preset options.

I think external data users can handle synonymy, or deal with it; we would still list the proposed microspecies names as alternative names.

Are there many other examples of taxa where there is a fundamental intercontinental philosophical disagreement about the species concept? I'm not sure a slippery slope argument makes sense here because you can just as easily make an argument about there being a slippery slope of eliminating more practically identifiable species concepts in favor of super frustrating non-universally-accepted species concepts which will turn off users to the site and extend the average time to reach RG asymptotically to infinity. It is already hard enough to get anyone to ID aseteraceae to genus.

And, as has been said above, it seems undisputed that the microspecies are totally ecologically irrelevant; an entire microspecies going extinct would evidently have the same or less impact on biodiversity than a single individual dying would in any other genus, and introducing a new invasive dandelion microspecies to an area that already has another invasive dandelion microspecies would have no additional impact. So I'm not sure what scientific benefit classifying exported inat observations to microspecies has, other than conforming to classifying inat observations to microspecies.

Inventing names completely is a separate issue (which does also happen, but it would be better discussed there).

It is worth noting that at least as far as North America is concerned the 'microspecies on GBIF' formatting issue is almost purely theoretical, because there is exactly one total inaturalist observation on GBIF for all microspecies in Section taraxacum combined (and it is ID'd purely based on a UK-only key).

What we are talking about here is declaring that there are no extant described section taraxacum species in North America at all, so just either ID them to whichever putative European microspecies happens to look most similar, or actively discourage IDers and observers from trying to be helpful on a very common and easy taxa as they are learning to use the site. I think that is more of an anarchy situation. New users may think being told it is impossible to ID is a defect in the quality of their observation, which it is not; their observation could not be conclusively ID'd to microspecies even if they attached a full genome DNA sequence. I think it is easy to underestimate how discouraging that can be to a new user, and keep in mind that none of them are directly represented in the discussion on this flag.

I think there a couple major difference from Arthropoda where non-specific IDs are par for the course:
1.) The most up-to-date keys from the best available secondary sources do not claim Arthropods are easy to ID
2.) With Arthropods a dissection or DNA sequence would at least be helpful. With Taraxacum a DNA sequence would not even theoretically be helpful, even for regions where putative microspecies are described, because (https://www.researchgate.net/publication/265346629_Towards_a_better_understanding_of_the_Taraxacum_evolution_Compositae-Cichorieae_on_the_basis_of_nrDNA_of_sexually_reproducing_species):
"‘raw’’ nrDNA [...] is not very suitable for phylogenetic analyses in Taraxacum."
"the repeated multiple hybridization events in the history of most taxa make the exploitation of cpDNA for phylogenetic inference in Taraxacum problematic."
"the extent of cpDNA intraspecific variation encompasses the taxonomic limits."
"None of the basic nrDNA clades identified in the present study, however, matches any of the groups delimited by means of the cpDNA study."
"not even the most distinct and stable clades resulting from the nrDNA analyses can be recognised in the combined tree."
"not even the restriction of the analyses to sexually reproducing diploids did increase the congruence of the two data sets"
Whats worse is that that paper was only considering the relationships of 65 total representatives of 39 different sections, not the 3,000 supposedly distinct microspecies. To me, a group with 'repeated multiple hybridization events' across all pairs of taxa, where 'the extent of intraspecific variation encompasses the taxonomic limits' just kind of sounds like what we would normally call a 'species'. I'm not even trying to make an argument about the actual taxonomy though, just the practicality of IDing that version of the concept in practice with any tools that could possibly be available to site users.

I guess one option that is more or less implicitly what Wikipedia and several of the papers I looked at while writing this comment do, is just to assert by definition that the microspecies only occur in Eurasia and that taraxacum officinale sensu lato does occur in North America. I had kind of expected that version of the concept might be less popular, but at least it wouldn't require any hint of making up names, and I guess is also effectively just the status quo.

infrahybrid removes italicization

Not sure this would be such a bad thing, if it's being used for names not sanctioned by the Code of Nomenclature. And infrahybrid may actually reflect the origin of many microspecies.

It is also not true that the microspecies treatment is more global. The 'taraxacum officinale' treatment makes a single specific prediction for what species every relevant observation in the world should be called. Therefore it is both consistent and global. The microspecies treatment is certainly not global, because it assumes most dandelions are some undescribed species that will never be described because the regional taxonomic authorities will never recognize them. It also may not be self-consistent, if, as said above, not every taraxacum officinale observation is even in 'section taraxacum'. Bumping the taraxacum officinale observations even further, back all the way to genus certainly would cause actual loss of the underlying data about other species in the genus produced by identifiers.

I want to clarify what you mean here since I feel like I disagree with all of it but maybe I'm just misunderstanding. It's been a while since I've thought about this topic so that's quite possible...
The microspecies treatment doesn't assume most dandelions are undescribed species; e.g. all the non-native dandelions in North America are presumed to be introduced European species that just haven't been identified in North America yet. Same thing in South America, South Africa, Australia etc.

As @vladimir_fuka mentioned, T. officinale by definition (in the literature) is equivalent to Sect. Taraxacum so any T. officinale observations that don't match Sect. Taraxacum have been misidentified. It can be difficult to identify dandelions even to section without good views of the leaves or at the wrong time of year, so many lower quality observations of dandelions should stay at genus (as is the case with many plant genera) regardless of microspecies discussions. I think the vast majority of identifiers don't know this, so not much (correct) data would be lost.
T. officinale is often used to refer to any of the complicated mess of apomictic weedy dandelions, which following that description does include multiple sections of the genus. However it also definitely excludes many other sections. Like would you call these dandelions T. officinale? Even European/North American members of Section Palustra look pretty clearly different.
https://www.inaturalist.org/taxa/1051010-Taraxacum-leucanthum
https://www.inaturalist.org/taxa/854364-Taraxacum-confusum
https://www.inaturalist.org/taxa/795603-Taraxacum-porphyranthum
https://www.inaturalist.org/taxa/869027-Taraxacum-leucoglossum
https://www.inaturalist.org/taxa/795597-Taraxacum-lilacinum

This segues to another major point which is that T. officinale/Sect. Taraxacum aren't the only taxa we have to deal with here. There are several other dandelion sections introduced to North America which have the same type of conflicts, and as far as I know Sect. Erythrosperma is the only one which has a clearly defined "macrospecies" equivalent name. What do we do with the others (e.g. Sect. Hamata)?

Inat does not even have a rank called 'microspecies' so no matter what the solution will be in some kind of deviation from what is meant by the terms.

A microspecies is still a species, it's not a different rank. The prefix just helps indicate that it's part of the weird apomixis situation where applying standard species concepts produces weird results.

Infrahybrid actually wouldn't all that far off the truth, would it? Per the quotes that they are all hybrids of unknown parentage within Taraxacum?

They're self-reproducing populations descended from hybridization events potentially thousands of years ago (I'd have to check for the specifics), as is the case for the origins of many other species.

However it also definitely excludes many other sections.
T. officinale by definition (in the literature) is equivalent to Sect. Taraxacum so any T. officinale observations that don't match Sect. Taraxacum have been misidentified.

I agree that we probably can't switch our species concept to taraxacum officinale sensu latissimo (i.e., the entire genus, as someone suggested above), which isn't used by either POWO or our commonly available secondary sources. I don't deny that there are probably plenty of bona fide mis-IDs among the >150,000 current taraxacum officinale observations if we take taraxacum officinale to mean just section taraxacum. We already have the ordinary community ID process to handle such mis-IDs, so the existence of a 'mess' of actual mis-IDs are mostly tangent to the discussion of the taxonomy because changing the taxonomy will not fix them.

While I don't think we really need to put all of the things we theoretically could under a macro species name, macrospecies names are available for a lot of the sections (see https://www.jstor.org/stable/1222201), for example:

Section Alpina = T. alpinum
Section Taraxacum = T. officinale
Section Erythrosperma = T. erythrospermum
Section Ceratophora = T. ceratophorum
Section Parvula = T. parvulum (accepted on POWO but missing from inat?)
Section Palustria = T. palustre
Section Mongolica = T. mongolica
Section Obliqua = T. obliquum
Section Spectabilia = T. spectacle
Section Calanthodia = T. calanthodium

Interestingly, all of those macro species names except taraxacum officinale are still listed as accepted by POWO, so I'm not sure we can regard the overall POWO treatment as self-consistent anyway. Hamata does look like a bit of an interesting case because the section was proposed relatively recently and so doesn't seem to really have had the type species name used as a macro species. It looks like every macro-species name that had previously been used to describe the plants in that section has been consumed under Section Taraxacum. Therefore if we don't want to put it under t. officinale, I guess we'd have to either just use the nominate species name despite there not really being a history of that, or leave it.

They're self-reproducing populations descended from hybridization events potentially thousands of years ago (I'd have to check for the specifics), as is the case for the origins of many other species.
all the non-native dandelions in North America are presumed to be introduced European species that just haven't been identified yet.

Right, but dandelions were introduced to the US 300-400 years ago. If every single hybridization/sexual reproduction event between agamospecies produces a new microspecies lineage, and even just 1 in a trillion dandelion reproductions are hybridization events, then there should probably already be a 300-400 completely undescribed dandelion microspecies in the US, assuming about a trillion dandelions grow in the US per year (say, 10 million suitable acres at an average density of 100,000 dandelions per acre? which is probably kind of low). This paper estimated that the hybrid seed production rate was actually more like 1/2000 when a sexual and apomictic species were in contact so 1 in a trillion feels pretty conservative. https://link.springer.com/article/10.1007/s10592-017-1014-y Even if the rate of producing an identifiably different lineage is somehow still much lower than that, it still feels like regarding them as basically infrahybrids isn't really all that far off.

But anyway the point of this discussion in the flags on inat isn't to resolve the overall taxonomic issues of genus taraxacum, its to come up with a workable ID scheme for site users. It doesn't really matter how many generations of apomictic reproduction you think it should take before you can call a 'hybrid/infrahybrid' a bona fide 'microspecies', just whether using the rank in that way interferes with any ID or export use cases, which I think it does not.

I can't find any papers with even rough estimates but I'm not sure it can really be much slower than that. The UK apparently has over 200 microspecies and you've said most microspecies are believed to mostly be extremely regional in Europe (and can't be ID'd at all without location because the morphological differences are too minute), and the climate in the UK was totally different until maybe 10,000 years ago due to the last ice age. So we could guess the UK has formed maybe 1 microspecies per 50 years. The US (not even including the rest of North America) has 40 times the area to work with so probably far more total dandelions to have a chance to hybridize, so forming 1 new microspecies per year sounds like it would be about the right ballpark. Because there wouldn't have been competition for niches to adapt to the new ones likely wouldn't have gone extinct as fast as they would in a place with an established population.

But yeah the goal here is specifically not to do taxonomy, its to create a useful ID framework for the website. So we would only need to do the macro-species that are known to occur in the part of the world where they would be regarded as useful, and maybe not even all of them, maybe just the Taraxacum and Erythrosperma, or perhaps those two plus the ones were the macro species name is still listed as accepted by POWO; we don't need to designate types (there isn't a mechanism to anyway).

@raymie I'm in agreement but to answer your question, the reason is that North American botanists don't have a good understanding of what introduced dandelion diversity exists here or how to identify it (and often don't recognize its existence). Since most iNaturalist users are North American, there's a lot of resistance to adopting a global system since that system isn't very workable here so far. Curtis Bjork is apparently working on adapting the Eurasian system to what we have in North America, but it's going to be a years-long project and I don't think it's his highest priority project.

I think there should really be no urgency to remove one of the top 20 most observed species on the site without replacing it with anything; I really doubt its existence bothers anyone who has not already commented on this thread. For practical purposes removing it would mostly just be an experiment in how chaotic computer vision suggestions can get if the correct answer for 100s of thousands of observations no longer exists as an option it can propose. The computer vision chaos will create an insane amount of work for actual IDers to fix continuously forever, and for that reason there is a good chance the change would eventually have to be reversed just to manage finite IDer resources; i.e. the non-existence of this taxon will bother people who haven't commented in this thread.

As I said before the infra-hybrid solution would address all practical concerns on both sides and is probably a better reflection of the actual truth of the matter. I really can't see a downside to it.

Interestingly, all of those macro species names except taraxacum officinale are still listed as accepted by POWO, so I'm not sure we can regard the overall POWO treatment as self-consistent anyway.

@wildskyflower this is because T. officinale doesn't refer to a particular microspecies whereas all those other names you listed do. POWO accepts the microspecies and not macrospecies (since those are mutually exclusive). I can't access the Richards 1985 paper right now but as far as I can tell, he was listing the type species (microspecies) for each section, not naming equivalent macrospecies. Using those names for macrospecies requires redefining those names to no longer refer to specific microspecies, and instead refer the sections as a whole.

For example this is the established taxonomy:

Genus Taraxacum
~ Section Palustria
~ ~ Species T. palustre (type species of section)
~ ~ Species T. crocodes
~ ~ etc.

Converting that into this:

Genus Taraxacum
~ Species T. palustre
~ ~ Subspecies T. palustre ssp. palustre
~ ~ Subspecies T. palustre ssp. crocodes
~ ~ etc.

...would be creating new taxonomy. Or am I misunderstanding your suggestion?

i think the infraspecific idea is more along the lines of

genus Taraxacum
- section Taraxacum [?]
- - species Taraxacum officinale
- - - infrahybrid Taraxacum aberrans
- - - infrahybrid Taraxacum accedens
- - - infrahybrid Taraxacum acre
- - - infrahybrid Taraxacum acroglossum
- - - etc.

not changing the names of the microspecies, just changing their rank and nesting them within T. officinale

@upupa-epops
"and we also can't create new macrospecies concepts that don't exist in the literature." I think that sections work well for this scope.

"I'm not sure about keeping the microspecies as subspecies or forms?" please NO, I think that the combinations do not even exist.

It is unlikely but there will always be the possibility that someone would identify to species a dandelion. So, I think we should guarantee this possibility keeping all the names of the microspecies.

@wildskyflower
"I think there should really be no urgency to remove one of the top 20 most observed species on the site without replacing it with anything; I really doubt its existence bothers anyone who has not already commented on this thread. For practical purposes removing it would mostly just be an experiment in how chaotic computer vision suggestions can get if the correct answer for 100s of thousands of observations no longer exists as an option it can propose"
Sad but true. There is still the possibility to ID observations at the section rank and, in the case, to explain users that T. officinale is not a valid name. Researchers could just consider all the T. officinale observations as section Taraxacum. Anyway, there is also the possibility that many T. officinale observations are a section different from Taraxacum (e.g. Erythrosperma).

Ahh okay thanks, I wasn't familiar with the infrahybrid rank.

The AI will be able to ID Taraxacum, just not to species level, right? Or is that incorrect?

I picked the species I mentioned based on the first CV suggestions that weren't taraxacum officinale or taraxacum erythrospermum from a random sample of a few observations. I think they are representative the kind of mistakes the CV could make. When I turn off 'expected nearby', I get taraxacum platycarpum most often. Maybe you are right and eliminating taraxacum officinale will coincide with a rapid North American invasion of taraxacum platycarupum instead, I don't know.

I don't think the CV can add the naked genus as a node for suggestions if taraxacum officinale and erythrospermum are eliminated but taraxacum platycarpum et al are not, because I think it can only have a genus as a CV node if it has no descendants that are CV nodes.

I also don't think downstream impacts on the CV should influence taxonomy decisions... If taxonomy causes problems with the CV then that's an issue with the CV.

I'd replace the word 'CV' with 'site user's needs' in that statement and check how it sounds. If the staff had unlimited manpower and resources they could probably code a taraxacum-specific workaround, but they don't and also don't like coding narrow workarounds that they have to maintain. For example, when a few avian hybrids caused a problem with the CV model that started causing mis-IDs another one of the top-20 most observed species (mallards), they just removed all hybrids from the CV model. That was 3 years ago, and they still have not had the bandwidth to work on adding any hybrids back to the model, even though the lack of hybrids in the model causes other problems elsewhere. It is just that big of an issue for the site to be consistently messing up even one super-observose species. And that was for a problem with mallards, which do not suffer from any real shortage of IDers, much less the kind of shortage of IDers that asteraceae does.

In the interest of discussion progress, I just picked a random microspecies (Taraxacum subtile), and moved it into how it would look in my proposed compromise: https://www.inaturalist.org/taxa/747413-Taraxacum-subtile. Please take a look and see how you think it looks.

Taraxacum subtile has no observations and the taxon page has no info, so I hope that using it for a demo for a few days is ok.

@wildskyflower I know that many users who post the "average Cichorieae observations" (only the capitulum from above) are young. And it is just for this reason that I am concerned. Are we feeding the "lie" that all such plants are Taraxacum (officinale)?
After all, I think that one of the most important problems is the settings of the CV. I do not know if it is techically feasible but in many cases it should suggest only the family or, at most, the tribe, without any other possibility.
The same often happens for Rosa. Here if single-flowered they are often suggested as R. canina, if double-flowered as R. chinensis or, alternatively, as R. gallica. And, unfortunately, people are more and more confident in artificial "intelligence", while less and less in humans.

I have posted a "test observation" in which everything for the correct section identification (achenes colour, inner leaves shape, outer phyllaries shape) is provided:
https://www.inaturalist.org/observations/204900224
Yet the CV suggests T. officinale as first choice.

With observations lacking all the necessary characters even more so it should limit its suggestion to a higher rank.

I am actually also concerned about T. erythrospermum. It seems to me that many RG observations of this species are in a clear opposition to the determination key in our national flora made by well-known specialists (Kirschner, Štěpánek, Trávníček, Vašut) although I also just may be not qualified to judge such a thing.

@danielatha no offense but all those observations may just demonstrate that in that area there could be many (micro)species.
Sometimes it is hard to discriminate between red or brownish-red achenes with photos. It may largerly depend on light and on camera settings.
In all that observations of fruiting capitula how can you say that the outer phyllaries have no horns? On the contrary, here they are clearly horned: https://www.inaturalist.org/observations/30449058
As regards, just the outer phyllaries count for this latter character.
Just to make things even more complicated, there are species attributed to section Erythrosperma (which is just a name) that have light brown or pinkish achenes.
Plants with light brown achenes and growing in anthropic habitats are more likely belonging to section Taraxacum which has not necessarily less divided leaves. It is quite common to find them mixed with other of the section Erythrosperma and this should not suggest that they are just the same.
I agree that there is usually a large variability but this often occurs intraspecifically. Measuring the variability inside a "population" may be risky since such "populations" may actually be a mixture of many microspecies.
In the end, it could be useful to take a look also at the literature from the old world.

@snedergaard @davidenrique @raymie @jdmore @silversea_starsong @nathantaylor @loarie @maxkirsch @graysquirrel @sambiology @danielatha @srall @sedgequeen @dysm @charlie @jasonhernandez74 @comradejon @holyegg @blue_celery @vladimir_fuka
I would appreciate any more feedback on the infrahybrid compromise that I have demoed with T. subtile (https://www.inaturalist.org/taxa/747413-Taraxacum-subtile). You can review how the taxon page looks, and please feel free to test how it works for ID purposes with this test observation I made (https://www.inaturalist.org/observations/204926150).

For anyone just tuning in who doesn't want to read this whole thread, the compromise proposal is as follows: keep T. officinale as it is, and agree to interpret the existing taxon rank of 'infrahybrid' to mean 'microspecies' for this purpose on inat. Then we move all of the existing microspecies to infrahybrids of the (macro)species T. officinale.

The advantages:

1.) It is possible to refine IDs to research grade at microspecies without disagreeing with T. officinale IDs
2.) It is possible to ID to T. officinale in the conventional sense and reach research grade at T. officinale, without DQA votes
3.) Interpreting the microspecies as infrahybrids is close-ish to the truth in either taxonomic hypothesis, so it does not require any significant distortion of the conventional meaning of the word 'infrahybrid', or doing any new taxonomy on inat
4.) There is no standardized naming convention for infrahybrids, so we can just use the bare microspecies names as the display names, which will make the unusual taxonomic structure essentially invisible to most inat users
5.) T. officinale will remain in the computer vision, and thus can be suggested to Seek users and on upload, to prevent the chaos of 10s of thousands of CV mis-ID'd uploads
6.) The microspecies will not be eligible to enter the CV, avoiding a future scenario where just one of the hundreds of microspecies gets enough IDs to be CV-eligible and then the CV starts IDing all T. officinale as that microspecies

The only cons that I can think of:
1.) It is a deviation from POWO
2.) The 'species' tap in the 'explore' webpage will lump all observations as T. officinale for the purpose of counting occurrences. It is still possible to get this information other ways, see https://www.inaturalist.org/flags/655973.

If we do decide to go with the infrahybrid compromise, I think we should source the deviation to:

Kirschner, Jan, and Jan Štěpánek. “Typification of Leontodon Taraxacum L. (≡ Taraxacum Officinale F.H. Wigg.) and the Generic Name Taraxacum: A Review and a New Typification Proposal.” Taxon, vol. 60, no. 1, 2011, pp. 216–20. JSTOR, http://www.jstor.org/stable/41059837. Accessed 2 Apr. 2024.

Which is the source that proposes the current lectotype for T. officinale and says:

Non-specialists can continue in using the name T. officinale in the generally and traditionally adopted way: it is equal to what is called common dandelions

Note that despite it's endorsement of the continued usage of T. officinale for non-specialist purposes, this source is actually the primary source POWO cites for not accepting T. officinale, so it is undisputedly a respected and relevant source.

I think we should continue allowing use of Taraxacum officinale. Technically it's not correct (99% of the time, at least) but (1) anybody researching Taraxacum will understand this problem and be able to cope with it and (2) people want a species name and being unable to get one for as simple, common, obvious as plant as the dandelion would be confusing and discouraging. I would argue for the pragmatic approach -- even though I regularly ID all dandelions as Taraxacum (genus) and sometimes explain why.

@sedgequeen We leave lots of taxa at genus level - what makes dandelions different? Not every observation can be IDed to species level, and that's okay.

I feel no loyalty to POWO specifically, but in my opinion we should pick a taxonomy and stick with what it does. POWO is the plant taxonomy iNat has chosen, and IMO this means we should be following it to the letter. We're following our chosen taxonomies exactly for many other groups, and I see no reason plants should be any different.

I'm not exactly outraged to be clear, I'm just really confused why people seem so intent on keeping this taxon. Especially since authorities in Europe pretty much universally consider it to be invalid and iNat users generally seem to be using it as nothing more than a wastebasket.

iNaturalist is a fantastic tool that has vastly expanded the availability of biological data for research, conservation and education. It is not where scientific hypotheses are presented and adjudicated. That is the purpose of scientific journals with peer review. Without a doubt, the elimination of Taraxacum officinale as an available name for observations on iNaturalist will cause great confusion and immediately and drastically reduce the number of observations of this group, thereby compromising iNaturalist's utility as a research and learning tool.

@raymie.... Identification of an observation at genus rank because the evidence is insufficient to identify it at a lower rank does not imply a judgement on the validity of the constituent taxa.

@raymie @danielatha Again, please just put a pause on any back and forth for a bit to keep this section of the thread condensed for easy future reference.

The advantage of following POWO is because it's basically the only option where we aren't just inventing our own taxonomy.

@comradejon this change would not affect any of the other sections; it only affects section Taraxacum. The others would remain their own sections and their species and microspecies would remain as formal species on iNaturalist, as far as I understand it (some sections have more clearly defined species that make sense, others have similar confusion to sect. Taraxacum but they'd be even more complicated to resolve).

There are species and sections in Taraxacum that are clearly distinct from T. officinale and section Taraxacum, e.g.:
https://www.inaturalist.org/taxa/795603-Taraxacum-porphyranthum
https://www.inaturalist.org/taxa/854364-Taraxacum-confusum
https://www.inaturalist.org/taxa/1076477-Taraxacum-dissectum
https://www.inaturalist.org/taxa/793808-Taraxacum-pacheri
https://www.inaturalist.org/taxa/1047467-Taraxacum-crocodes

Flora of North America lumped sect. Taraxacum together as T. officinale and sect. Erythrosperma together as T. erythrospermum. Since then my impression is that North American botanists having been trying to force any non-native dandelions into one of those two buckets, but there are other non-native sections present as well (based on the BC paper). So this change helps with some of the confusion but sect. Erythrosperma and other sections that don't have a designated macrospecies would be staying.

Thanks for the explanations! So where do you all think @wildskyflower's proposal would leave us in terms of appropriately ID'ing dandelion observations (especially in the United States)? Would we basically be in the same situation we are now, where we have insufficient data to go any further than genus-level on nearly all observations?

So many good points and counterpoints here, without understanding them all fully I hesitate to pick a side. So the following is offered only as a thought experiment:

Suppose for the name "Taraxacum officinale" that we follow POWO to the letter (almost). How would that look, and what would be the practical consequences? That would mean that we:

(1.) create an empty species-rank taxon called "Section taraxacum". (Trying to call the species "Taraxacum section Taraxacum" as POWO does, at a rank that is expecting a binomial, could be problematic in iNat, or if successful could create confusion with iNat's actual section-rank taxon of the same name.)

(2.) make it's parent taxon the section-rank taxon Taraxacum section Taraxacum already present in the iNat taxonomy.

(3.) swap the existing species-rank taxon Taraxacum officinale into the empty species-rank taxon Section taraxacum, bringing with it its common names and scientific synonyms (to now include Taraxacum officinale as in POWO).

(4.) leave all the other microspecies of Taraxacum section Taraxacum right where they are in iNat, at species rank within the section-rank taxon.

The results I can foresee:

(5.) People can still identify their dandelion observations using "common dandelion," Taraxacum officinale, or whatever name they are already accustomed to using, and iNat will automatically assign them to the species Section taraxacum.

(6.) Anyone confident enough to identify one of the microspecies can still do so unhindered. Any disagreements with Section taraxacum would cause the community taxon to be Taraxacum section Taraxacum.

(7.) Since we'd be using a regular iNat taxon change process, CV should follow along and start recommending Section taraxacum wherever it previously would have recommended Taraxacum officinale.

Personally the biggest drawback I'm seeing is the potentially confusing species name Section taraxacum. But I think that is mitigated by having the familiar common names attached to it. If needed to further reduce confusion, common names could be modified to add a parenthetic (Taraxacum "officinale") for clarity.

OK, tell me what I'm missing here...

EDIT to add a "visual" representation of how the taxonomy would look in iNat:

genus Taraxacum
section Taraxacum
species Section taraxacum <== Taraxacum officinale, common dandelion, etc.
species Taraxacum microspecies belonging to the section...

I don´t have any major objection to the change proposed by @wildskyflower, or, at least I don´t think so. I gather that the result will be that any Taraxacum species should fall into either of two categories: Those belonging to sections other than sect. Taraxacum will (continue to) be “species”, and those of sect. Taraxacum will be at another (kind of lower?) taxonomic level, “infraspecies”. This distinction seems a bit odd to me as European, but it appears to have little or no practical consequences other than allowing T. officinale to merge with sect. Taraxacum while still preserving its rank as a “species” – if I understand it correctly.

@jdmore Species names in inat have to be proper binomials. I was just describing the weird quirk of how POWO has done the database entry there; part of the reason it has to do that is because the POWO database does not even have an existing rank of Section, whereas inat does, and partly because I assume they want to find a way to both respect the nomenclatural priority of the name T. officinale and keep the microspecies that works in their framework.

For what it's worth I reached out to Curtis Bjork to see if he had any suggestions for us and he said he had no idea. I'm pretty sure he would prefer to have the microspecies accepted as full species but he seems a bit cynical about iNaturalist in general including taxonomy discussions here. You can see his opinion on the microspecies here: https://www.inaturalist.org/observations/108199765

That is a very insightful comment. I think it is the most succinct summary I have seen of why the microspecies can be conceptually useful as tools.

Note that Curtis really has more like 50-60 IDs, he just posts them almost exclusively as comments instead of official IDs; he has ID'd the vast majority of sedgequeen's observations that are ID'd to microspecies, for example. I do wonder what his reservations about inat because I think it could be is a good example of where supporting both kinds of use cases can be useful to accomplishing scientific progress via. I suspect it is in part because the microspecies just can't usually be ID'd from photos (most of his comment IDs are on observations where someone has sent him a physical specimen).

@wildskyflower

Species names in inat have to be proper binomials.

If by proper binomial you mean a binomial that is effectively, validly, and legitimately published according to the rules of nomenclature, I agree 100%. I was just trying to think through what it would mean to not deviate from POWO, which in this case would mean using an invalid species name to "contain" Taraxacum officinale Auct., as POWO currently does.

Unfortunately, it would not be the first time iNaturalist used an invalid binomial because it was nevertheless accepted by POWO.

@wildskyflower well, as it turns out, my scenario would not work anyway, because the word "section" is automatically stripped out of taxon names by iNaturalist - it being the name of a rank.

What would work though (I tested it without committing anything) is doing a taxon change to swap Taraxacum officinale into Taraxacum sect. Taraxacum. That would have about the same effect as my original scenario. People could still use Taraxacum officinale to enter IDs, and iNat would interpret it as sect. Taraxacum. Any microspecies IDs would not create unwanted disagreements. People just wouldn't see their favorite macrospecies name displayed anymore.

As I think about it more, to me that scenario still seems preferable to co-opting an iNat rank intended for nothosubsp. and nothovar. taxa for the species-rank taxa of one section of Taraxacum, while leaving all the other Taraxacum species at species rank.

I also would not mind just leaving the status quo, and letting people use their favorite macrospecies or microspecies as circumstances warrant. If there is a disagreement between Taraxacum officinale and someone's microspecies ID (thus making the Community ID sect. Taraxacum), that's probably a good thing. It would take two additional agreements with the microspecies ID to override the Taraxacum officinale ID and the sect. Taraxacum community ID, which seems like a reasonable threshold in the circumstances.

While the debate continues, I have added a tiny bit of progress in the form of the deviation that @loarie requested a while back in the discussion, so that the current situation is at least mapped to POWO:
https://www.inaturalist.org/taxon_framework_relationships/753820

If the ultimate consensus goes in a different direction, then the deviation can easily be revised or removed.

It would be very easy for iNat to influence taxonomy by leveraging its observers. I believe iNat is being extremely sensitive to the need for taxonomists and others to continue exploring, discovering, and describing. I see no reason there can't be two views of the same data. One view which is primarily morphological and groups what a taxonomist sees as multiple species into a single species that a less savvy citizen like me can understand. And another view which is more deeply scientific and presents a full panoply of described discovery using all science known to all scientists (and some citizens).

Species proliferation due to new technology like simple and cheap DNA sequencing, macro photography, etc should not mean that ordinary citizens lose their understanding of "species". Science on iNat should continue to be accessible to the most humble of citizens. This will mean either careful presentment of increasingly esoteric species distinctions -or- influencing taxonomy to further celebrate the discovery and description of new sub-species such that the species rank only proliferates due to evolution rather than new technology.

"The taxonomy isn't just there to generate papers, it's there for everyone to use" -- isn't that an easy distinction for scientists in any branch to forget?

The Forum discussion on this topic mentioned "evolutionarily significant units." Well, from an ecological point of view, we can just as well think of "ecologically significant units," which may or may not correspond exactly to the species level. I think of the Pacific Chorus Frog, recently split into three species, and then even more recently lumped back into the original one species. In ecological terms, though, hasn't it always been a single ecologically significant unit, that is, filling the same ecological niche across its range? Would an assessment of a wetland's health change according to whether it was home to Pacific Chorus Frogs or Sierra Chorus Frogs?

"i feel like the vastly higher proportion of people who use taxonomy should have influence over the people who create it."

Agree with you Charlie. It's not like we have ill intent. We just want to be able to observably identify the things we are asked to preserve... species.

i think we have to ask ourselves whether microspecies that can't even be reliably differentiated with genetics much less with photos have any place on iNat, and if they do, what that place is. What is missing from the discussion is, what would the microspecies elevated to species level even be used for? Do we think it's possible for people to actually collect iNat observations that can be identified to species under such a regime? If not, why do it? I'm sorry but 'POWO did it' just isn't a compelling answer, even if it is useful for other applications. If POWO jumped off a cliff would you do it, etc...

and what exactly would be simplified by breaking all the existing Taraxacum officionale observations? We tried POWO, and not all of us think it's still going well. So it's certainly valid to discuss whether we should still do so

I'm not sure dandelions make any sense anywhere.

:-)

After I poked Curtis about his dislike of iNaturalist, he gave me a rant about all his problems with POWO as well haha. But I'm not sure how it would improve things to abandon an imperfect system in favour of no system at all...

There are a decent number of sources and people out there who are identifying dandelions to microspecies (I have some sources collected here), they're just not very accessible outside of Europe because they only cover the species found in their respective countries.

"How the taxonomy is being handled right now in iNat, T. officinale is a species that only exists where it's introduced."

That's not what I see on the observation map just now.

@wildskyflower Since you bring it up, I can add that I had a discussion on the T. officinale subject some time ago with Hans Øllgaard. It appears that his view on the matter is in line with the conclusions made by Kirschner and Štěpánek on T. officinale in their 2011-paper. He believes that - for practical reasons - the name T. officinale is unsuitable among botanists, while admitting that its use as synonym for Sect. Taraxacum is valid - but it would then need to be clarified each time. In my opinion, the last part illustrates the problem very nicely, and it will likely persist no matter what is done on iNat. I still support your proposal, though, as I see it as a positive step towards "talking the same language".