Flagger | Content Author | Content | Reason | Flag Created | Resolved by | Resolution |
---|---|---|---|---|---|---|
upupa-epops | dandelions (Genus Taraxacum) |
what to do about T. officinale and T. erythrospermum |
Apr. 3, 2021 13:43:32 +0000 | upupa-epops |
Move discussion to flags on species pages |
If we wanted to deviate from POWO and go the macrospecies path, there is a 1976 Flora Europaea taxonomy that sort of follows it (https://zenodo.org/record/293764, pg 333). Richards 1985 (https://www.jstor.org/stable/1222201) gives updated type species for each section but says that the 1976 effort was illegitimate according to the Code (I don't understand why).
-- Rename all of the microspecies as subspecies and merge the sections into their type species.
-- Most T. officinale observations cannot be confirmed even to section Taraxacum level because they don't show bracts or leaves well enough. Other sections/macrospecies are common; sections Celtica, Erythrosperma, Hamata, and the fulvicarpum groupare apparently common in parts of BC and are presumably in other coastal and/or alpine areas. There are probably additional Mediterranean (and neotropical?) taxa in the southern US. Most observations are are officinale agg./sect. Taraxacum, but they can't be confirmed...I'd suggest merging the species with the genus and recreating it. However, that's going to be unpopular. The other option is manual mass identification. Anyone manually pushing observations back to genus as a result of this is going to get a lot of pushback.
However, considering that the macrospecies model conflicts with POWO, the literature, and what Europeans are doing, I don't think it makes sense to do that.
If we want to consistently apply the microspecies...
-- The type specimen of T. officinale was redesignated in 2011: https://www.jstor.org/stable/41059837 It was intentionally not identified to microspecies level. As a result, by my understanding, practically speaking T. officinale is synonymous with section Taraxacum and should not exist as an identifiable species within the section.
Based on that I suggest merging T. officinale with section Taraxacum (or preferably genus level, as above).
-- T. erythrospermum is a sexual species from mainland Europe. For some reason sexual dandelions aren't weedy so it doesn't exist in the UK or North America.Because of classic computer vision problems, as far as I can tell erythrospermum has been misused in Europe just as much as in North America, so any changes in the taxon would affect few if any valid observations.
Based on that I suggest merging T. erythrospermum into section Erythrosperma and recreating it. I'm not sure if this is technically a method curators are allowed to use, I've used it once before for a similar issue. It would deal with the issue quickly, but it could be dealt with through manual identification if necessary...
as far as I am concerned I would suggest to
1) put T. officinale under synonymy with section Taraxacum. I would not take for granted that this choice could be unpopular but, conversely, it could raise someone's curiosity to understand why what was taken for granted as T. officinale actually should be named in a different way.
2) T. erythrosperum (sect Erythrosperma) and, just to cite another misused name, T. gasparrinii (sect Scariosa) actually are good species and should be left as they are without any merging into their sections.
Thanks @blue_celery. For clarification, my suggestion wasn't to merge the latter and keep it that way, but rather to merge it and recreate it so it can be used properly in the future. For example we did that with the Narceus complex since both species in the complex are valid species but there's no way to distinguish them. It evidently wasn't particularly effective in that case though...
I guess that is the benefit of manual identification campaigns; re-identifying the thousands of T. erythrospermum obs to section would make everyone aware of the issue more clearly than an automatic taxon change.
I support the idea of merging Taraxacum officinale with the genus, so that "T. officinale" = "Taraxacum sp." Both names should, however, be available for reporting (T. officinale is widely used, but it´s status is dubious, and many people would rather not refer to it, including myself). If both names are allowed, I don´t see why that should be unpopular. Regarding the other option of merging T. officinale with section Taraxacum: please don´t. Most photos of T. officinale cannot be allocated to any section, and quite a lump of them are definetely not section Taraxacum. In general, identification of a section, based on photos, is very difficult (unless you can recognize the species - then it is easy).
How to deal with T. erythrospermum? Photos that clearly show the necessary details (red-colored fruits and/or bract corniculations etc.) = section Erythrosperma. All the others = Taraxacum sp./T. officinale? I guess such revision could be done, and in most cases without doing any harm. But some observers may actually know the species and have reported the name based on knowledge. How to deal with those? Another problem is that the "machine" has learnt this species overly well, and I suppose it will continue to generate it no matter what. Can it be stopped?
"Most photos of T. officinale cannot be allocated to any section, and quite a lump of them are definetely not section Taraxacum"
@snedergaard is right
"Can it be stopped?"
I think it cannot be stopped but I hope to be corrected. Users should rather be stopped from using, at least, the name T. officinale. I asked if there was the possibility to create a system to alert users to be careful to use specific names in Taraxacum (and this could apply also to other very critical genera rich in agomospecies) since, of course, their identification is extremely critical but it seems that it cannot be implemented.
Maybe a possible "emergency" solution is to spread the word about the use of these names with, for example, a journal post.
Both names should, however, be available for reporting (T. officinale is widely used, but it´s status is dubious, and many people would rather not refer to it, including myself).
As far as I can tell if it continues to be an option then people will continue to use it in the place of all Taraxacum and the whole problem will not be solved. I agree that it's still widely used even in literature though...
Can it be stopped?
Theoretically yes, but it would require insane effort and good timing. According to the help page, to be in the CV suggestions "taxa included in the computer vision training set must have at least 100 observations, at least 50 of which must have a community ID."
This would apply while the CV is training, which I think is done a couple times a year and lasts a couple months. So from what I understand we would have to make sure there are fewer than 100 observations of the species for several months in a row. That's probably not realistic.
Not sure how helpful this is, but the International Cichorieae Network portal and Euro+Med PlantBase both have T. officinale as a synonym of section Taraxacum and common misapplied name for genus Taraxacum, as well as T. erythrospermum as a common misapplied name for section Erythrosperma.
I'm curious if there are similar issues with T. palustre and section Palustria, although that would affect a lot fewer observations.
I emailed POWO about why they accept T. officinale but I don't understand the response...
You cannot have a species name for a section, That’s why we use:
http://plantsoftheworldonline.org/taxon/urn:lsid:ipni.org:names:254151-1
The situation is still unclear and perhaps the name should be conserved with a conserved type, but it cannot represent a group of species even though they intended to that, for that you need to use “Taraxacum section Taraxacum”
OK, it is evident that the Cichoriea Network + others have adopted the redefinition made by Kirschner and Stepanek (2011), which states that T. officinale is a synonym for sect. Taraxacum. However, if you cannot have a species name for a section (POWO answer), there seems to be no way of preserving the name T. officinale unless you reject section Taraxacum altogether. Or am I wrong?
Regarding sect. Palustria: I´m not so familiar with it, but it is not an easy one. So, I guess there could be a similar issue with T. palustre.
After clarifying again with POWO, they no longer accept T. officinale: http://powo.science.kew.org/taxon/urn:lsid:ipni.org:names:1003018-2
(I noticed a couple weeks ago that they renamed "T. officinale aggr." to "T. sect. Taraxacum", which is what prompted my recent email for clarification)
See new flag here: https://www.inaturalist.org/flags/535033
For context, dandelion taxonomy has been discussed extensively on the forum: https://forum.inaturalist.org/t/overlooked-dandelion-diversity-in-bc-and-everywhere-in-north-america/3808
For Taraxacum taxonomy in general, there are two interpretations. The one used by most European taxonomists is sections containing many (micro)species. Others and most North Americans (in theory) use broadly defined macrospecies instead of the sections. If applied consistently, I think(?) it would be relatively easy to switch back and forth between these since it would just involve renaming the taxa as sections/species/subspecies. The problem is that the macrospecies model has never been consistently applied anywhere (e.g. no one has laid out a species>subspecies/varieties system).
I think the idea that 99% of introduced dandelions can be reduced to T. officinale and T. erythrospermum comes from the 2006 Flora of North America: http://floranorthamerica.org/Taraxacum
This taxonomy competes with POWO and in practice is oversimplified because other sections such as section Hamata have been identified in North America since then. FNA says that T. officinale is in section Ruderalia (now sect. Taraxacum), so I don't think T. officinale can be interpreted to include e.g. sect. Hamata even following FNA.
Since POWO accepts microspecies, it's inconsistent and confusing for iNat to try to follow the macrospecies model only in North America especially when European observers (e.g. @joshstyles and @snedergaard) are using microspecies. We should choose one or the other.