Talk:Rheum (plant)

Taxonomy

Hello to all interested Rheum taxonomy editors. I have been squatting here at the Rheum pages for a bit. I originally wrote the following in response to Plantsurfer's edits, who feels it should be posted here:

The TPL version 1.1 you linked to is built up of two insecurely melded lists, the unavailable preliminary WCPF Polygonaceae database and the TROPICOS database. In the case of the WCPF only the synonymy for R. rhaponticum, R. x. hybridum and R. rhabarbarum was truly worked out; the other names were just added from the Index Kewensis to the database without judging the validity of the species. The TROPICOS data was imported in 2012; if you look at the taxa validated and synonymised, compared to those labelled 'unresolved' then it is clear that the 'unresolved' category is for taxa which do not occur within the remit of the projects of the MOBOT; i.e. Nepal, Pakistan, China. In other words, species which have a distribution outside of those three countries are 'unresolved'. TPL version 1.0 differs from 1.1 in a way which I assume means it was downloaded from TROPICOS before the FoC was completed (2003). Furthermore, taxa for the Flora Iranica (some possibly dubious) and the Middle East are not present in either database.

Earlier, the species list was made by Peter Coxhead & referenced to POWO. I removed that and started modifying, as the PoWO is many ways worse than the TPL. Most annoyingly it is unsourced, but judging by the pattern of synonymity they are following the Checklist of (former) USSR 1995. This was itself citing Borodina in 1989 Plants of Central Asia, who built her very lumpy species interpretation from a number of obscure floras from 1930s to 1960s Mongolia, Uzbekistan, Tibet and Xinjiang. These are not in agreement with each other, different taxa are being synonymised left and right, so Borodina basically lumps all these taxa into one superspecies. The USSR checklist peels a number of species off that complex again. In the 2003 FoC, the author is the same Borodina (along with Bao Bojian), but after 14 years she has changed her mind and split up all of her earlier synonymies. Besides this problem with PoWO, I also noted a number of other problems, like missing or double taxa.

As these database generated lists are untrustworthy I am hence going with the 2017 plant species checklist of China as source for validity in that country, and the other regional floras with the newest having highest priority, which is how synonymy in taxonomy is actually supposed to be done, no? I can happily explain why I feel certain individual taxa should be included.

I want to remove the TPL reference altogether. It just confuses people and isn't useful.

Regarding the list of synonyms Plantsurfer removed (emodi, undulatum), I didn't add them, but do think they should remain, as they are still in common use. The point of this page is to be informative, and as such the section is handy. Personally, because the misapplication in usage of the name R. rhabarbarum, I have found researching R. undulatum is the best way to study the taxon unambiguously.

Cheers, Leo86.83.56.115 (talk) 11:24, 5 April 2019 (UTC)

I have no objection to the inclusion of synonyms where they are necessary for understanding the topic, provided they are declared as such and not as additional species, but it is not the job of this article to provide a list of all species names, let alone all synonyms. The species list could therefore justifiably be shortened, particularly with a view to removal of synonyms.

You do not say why you regard R. rhabarbarum as misapplied. The name dates from 1753, while R. undulatum dates from 1762, so the prior name should have precedence. Anyway, if it is merely your view, then it amounts to WP:OR and has no place in the article. If there is a reliably-sourced consensus view that R. rhabarbarum should be rejected in favour of R, undulatum, then that is a different matter, and in view of the notability of the species that case should be brought to the attention of readers. Otherwise it should remain. Your objection to TPL as a source is close to being a WP:POV issue. TPL is a respectable source, hosted by two authoritative botanical institutions in the English-speaking world. Its content derives from 2012/13 and is not being constantly fiddled with and updated. As such, it represents a benchmark view that can be referenced without fear that positions will have changed since last month or are likely to be out of date by next week. The purpose of this encyclopedia is to summarize the consensus understanding of a topic that can be backed up by reliable secondary sources. It is not intended as a vehicle for unresolved minutiae being discussed currently at a research level. Plantsurfer 13:38, 5 April 2019 (UTC)

@Plantsurfer: while I respect your work re plants on Wikipedia, I must very strongly disagree re TPL. It is absolutely not a reliable source as regards its extracts from Tropicos, as is very well known (this was discussed at WT:PLANTS not long after TPL emerged; I think Plantdrew knows more). All 'scraping' databases are only reliable if they extract information correctly. Tropicos is concerned with "nomenclatural, bibliographic, and specimen data" as its home page says; it does not attempt to provide an single coherent taxonomic view, unlike, say, WCSP. So when TPL sources data from Tropicos or combines its data with data from other sources, you must investigate Tropicos to check what it actually says. Tropicos is reliable as to the existence of a name and as to what literature it appears in, but should not be used as the source that says the name is "accepted", because it doesn't. This is not WP:POV or WP:OR, it's just checking the reliability of sources, as we are required to do.

As per its purpose, Tropicos does not always give a single "accepted" name (or indeed any "accepted" name). Thus if you look at the entry for Reynoutria japonica Houtt., then under the "Accepted Names" tab, there are two names matched to the three sources that accept them. Tropicos itself does not accept either of them. We could use TPL to argue for Reynoutria japonica (if there were no better sources, which there are) because TPL's entry is derived from WCSP (see [1]), which does accept names. If TPL's entry had been derived from Tropicos, we would need to investigate further and see who actually accepted the name, and use that source for "acceptance".

There are more trivial problems, too, such as the same name appearing more than once because the authority doesn't exactly match (the separate entries at [2] and [3] are an example of not matching WCSP and Tropicos names).

Further, TPL is now obsolete. It won't be updated again, and it and WCSP are effectively being replaced by PoWO. So even where its information was correct at the time it appeared, it can now be out-of-date. Peter coxhead (talk) 14:34, 5 April 2019 (UTC)

Hi there. Peter got here before I was done writing. I understand you are unfamiliar with the "minutiae" of the subject, as far as validity of Rheum species are minutiae, and are being cautious. Let me give an example or two. Certain taxa such as R. palaestinum or R. maximowiczii are listed as 'unresolved' at TPL, simply because they are not native to China or Pakistan because the data was imported from Tropicos which doesn't try to compile the complete synonymy for the genus Rheum, just those species dealt with in their projects, not because they are synonyms or aren't recognised by relevant references; i.e. the floras of Israel or the former USSR. Regarding the R. undulatum thing, until recently in Europe most floras list R. x. hybridum as R. rhabarbarum, i.e. misapplication. In the US and the USSR, the floras list R. x. hybridum as 'R. rhaponticum. That things changed happened fairly recently. This has been covered at Wikipedia talk:WikiProject Plants twice before. Cheers, Leo. 86.83.56.115 (talk) 14:53, 5 April 2019 (UTC)

Commenting since I was pinged. I don't have much to add. TPL's list of Rheum species is a prime example of their failure to properly handle Tropicos data. TPL has two entries for Rheum undulatum, one from Tropicos, and one from WCSP, with identical authorities. I don't know why those weren't merged; I haven't noticed a TPL failure like that before. PoWo is Kew's successor to TPL; Kew rushed it out unilaterally and I've heard they no longer have employees tasked with developing it further. World Flora Online is the successor to TPL supported by multiple institutions (including Kew and MoBot). World Flora Online is still a work in progress; target date for completion is "by 2020", although I'm not sure that will be met. Plantdrew (talk) 15:43, 5 April 2019 (UTC)

This is great. Saturday morning rhubarb taxonomy. More to bolster my argument. Plantsurfer, you state I am being subjective in my inclusion of taxa in the species list, fair enough, but I am backing up my point of view with good taxonomic works, it shouldn't matter that some are Russian or Chinese. But take R. palaestinum, you imply this taxon is a synonym of something and shouldn't be included in Wikipedia as it is labelled as 'unresolved' in an out-of-date and badly constructed online list (albeit by "two authoritative botanical institutions in the English-speaking world", that is the 'appeal to authority' argument). Yet it is in the Red List of Israel, the government has introduced legislation about it and the Jerusalem Botanic Garden is right now running a propagation experiment with the species. And yet the world should be deprived of its existence as a species on this website? I have pointed out the problem in TPL -the data comes from Tropicos, which has all names, species or synonym, which represent taxa outside of the remit of MOBOT projects, labelled as 'unresolved' -it's really that simple. One could also say you have a POV, but one that is sillier than mine. So ha.

Let me give another example to illustrate why PoWO is not trustworthy. The last taxon I added to this list -which I will freely admit is a work in progress, was R. hissaricum: this appears to be rare endemic in a region of Tajikistan known for rare endemics, some now likely extinct in the wild (Tulipa subpraestans, the Nurek Dam possibly flooded a number of important collection locales). It is considered a valid species in all the Soviet/Russian literature I have come across, even in Borodina's very lumpy 1989 work, and Czerepanov's 1995 work. It it listed as 'unresolved' at TPL and Tropicos, fine, we have discussed why this is, but in PoWO it is simply missing from the database -not a synonym, not unresolved, just missing. This is very strange, because judging by the species and synonymies recognised it looks like Czerepanov's work was followed at PoWO (but no ref., so...), so it seems someone just omitted to fill this in or something? So yeah, I would rather judiciously rely on Czerepanov here until something better turns up.

My friend, 6 or so weeks ago, similarly to you, I started with the PoWO list, but as I started on the redlinked species and read through the relevant works, it became apparent that I needed to abandon it. I know what I'm doing, (usually) serious student of botany here, thank you. Please go over the literature yourself, I am certain you will join team R. hissaricum!

I have been using my own cache of Soviet and Russian books, but I recently found this great site, at which I think you can download all the relevant literature: [[4]].

And now off to the garden to putter around with new seeds of rhubarb from interesting places, instead of doing my taxes. Yay. — Preceding unsigned comment added by 86.83.56.115 (talk) 10:05, 6 April 2019 (UTC)