search

gbif / backbone-feedback

2 stars 0 forks source link

Incorrect taxonomy for some plants #472

Open ekoblogist opened 5 years ago

ekoblogist commented 5 years ago

There are some accepted names that were erroneously ascribed as synonyms to other accepted names (sometimes they are taxonomically or geographicaly very distant). Here is the list of some of them:

https://www.gbif.org/species/3090717 Amphoricarpos neumayeri (Vis.) Vis. is not synonym of Echinops spinosissimus subsp. neumayeri (Vis.) Kožuharov (corrected in CoL)

https://www.gbif.org/species/3748573 Cheilanthes fragrans (L.) Sw. is not synonym of Dryopteris fragrans (L.) Schott

https://www.gbif.org/species/7127713 Cyclamen repandum Sm. is not synonym of Cyclamen hederifolium subsp. hederifolium Aiton (corrected in CoL)

https://www.gbif.org/species/2706239 Festuca pratensis Huds. is not synonym of Festuca rubra L.

https://www.gbif.org/species/3172300 Ligustrum vulgare L. is not synonym of Syringa vulgaris L. (corrected in CoL)

https://www.gbif.org/species/7451761 Orchis tridentata Scop. and https://www.gbif.org/species/6315349 Orchis tridentata subsp. commutata (Tod.) Nyman are not synonyms of Platanthera clavellata (Michx.) Luer (corrected in CoL)

https://www.gbif.org/species/3166703 Phyteuma pseudorbiculare Pant. is not synonym of Phyteuma orbiculare L.

https://www.gbif.org/species/3021922 Prunus cerasus L. and Prunus cerasus var. caproniana L. are not synonyms of Prunus avium (L.) L. (corrected in CoL)

https://www.gbif.org/species/2926634 Rosmarinus officinalis L. is not synonym of Salvia rosmarinus Schleiden

https://www.gbif.org/species/5583101 Salix grandifolia Ser. is not synonym of Salix laggeri Wimm.

https://www.gbif.org/species/3045290 Thlaspi montanum L. is not synonym of Noccaea fendleri (A.Gray) Holub

I wanted to know what is the practice in GBIF with such backbone errors, i.e., how long does it usualy take to correct such nomenclatural or taxonomical errors. I saw that some of the issues are open for more than a year now, but on the other hand some names that had incorrect taxonomy in april this year, are now corrected, e.g.:

Hippocrepis ciliata (https://www.gbif.org/species/5353690) Luzula multiflora (https://www.gbif.org/species/2700934) Polygonatum latifolium (https://www.gbif.org/species/7920718)

Considering that majority of mistakes come from CoL, you can see from the list that half of the names have already been corrected in CoL.

What if they are not correct in the CoL, but there is correct usage of the names in for example The Plant List database? When do you switch?

yroskov commented 4 years ago

False accusations of the CoL. The only comment "Phyteuma pseudorbiculare Pant. is not a synonym of Phyteuma orbiculare L." relates to CoL data, and reflects different opinions. Other comments may refer to GBIF back bone (misinterpretation of CoL data?).

ekoblogist commented 4 years ago

My appologies to CoL.

When I stated:

Considering that majority of mistakes come from CoL, you can see from the list that half of the names have already been corrected in CoL

I just made an assumption that errors were coming from CoL, because GBiF reffered to CoL as a source.

Now, I am checking previous versions of Annual Checklists of CoL and there is no trace of such errors, so I agree that this is probably GBIF backbone issue.

However, my question still stands:

I wanted to know what is the practice in GBIF with such backbone errors, i.e., how long does it usualy take to correct such nomenclatural or taxonomical errors. I saw that some of the issues are open for more than a year now, but on the other hand some names that had incorrect taxonomy in april this year, are now corrected, e.g.: Hippocrepis ciliata (https://www.gbif.org/species/5353690) Luzula multiflora (https://www.gbif.org/species/2700934) Polygonatum latifolium (https://www.gbif.org/species/7920718)

Anyway, thank you @yroskov, for this is the first reply to this issue ever since I posted it (for almost a year now).

yroskov commented 4 years ago

Dear Vladimir,

My apology, I am responsible for original Catalogue of Life (CoL) content at www.catalogueoflife.org/annual-checklist/2019 only and cannot explain how GBIF interprets our data in their portal.

It looks like presentation of CoL data in GBIF portal and references/credits to CoL are confusing and cause misunderstanding among GBIF users. We’ll discuss further how to improve the situation.

If you’ll have questions on CoL practices and procedures, I am happy to tell you our stories via my email yroskov@illinois.edumailto:yroskov@illinois.edu or Skype.

Yours, Yuri

Dr Yury Roskov Catalogue of Life Executive Editor

(Office 2020 in Evers Lab bldg)

Illinois Natural History Survey Prairie Research Institute University of Illinois at Urbana-Champaign 1816 South Oak Street (MC-652) Champaign, IL 61820-6960, USA

Email: yroskov@illinois.edu Phone: +1 217 300 0595

www.catalogueoflife.org; www.sp2000.org

From: Vladimir Stupar notifications@github.com Sent: Friday, November 1, 2019 10:45 AM To: gbif/portal-feedback portal-feedback@noreply.github.com Cc: Roskov, Yury yroskov@illinois.edu; Mention mention@noreply.github.com Subject: Re: [gbif/portal-feedback] Incorrect taxonomy for some plants (#1677)

My appologies to CoL.

When I stated:

Considering that majority of mistakes come from CoL, you can see from the list that half of the names have already been corrected in CoL

I just made an assumption that errors were coming from CoL, because GBiF reffered to CoL as a source.

Now, I am checking previous versions of Annual Checklists of CoL and there is no trace of such errors, so I agree that this is probably GBIF backbone issue.

However, my question still stands:

I wanted to know what is the practice in GBIF with such backbone errors, i.e., how long does it usualy take to correct such nomenclatural or taxonomical errors. I saw that some of the issues are open for more than a year now, but on the other hand some names that had incorrect taxonomy in april this year, are now corrected, e.g.: Hippocrepis ciliata (https://www.gbif.org/species/5353690) Luzula multiflora (https://www.gbif.org/species/2700934) Polygonatum latifolium (https://www.gbif.org/species/7920718)

Anyway, thank you @yroskovhttps://github.com/yroskov, for this is the first reply to this issue ever since I posted it (for almost a year now).

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/gbif/backbone-feedback/issues/472?email_source=notifications&email_token=AG2AELD6FAPEV6PQVGYYJT3QRRFG5A5CNFSM4GIBY5I2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEC3JSQA#issuecomment-548837696, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AG2AELF4TZBE6QGHCFXSF73QRRFG5ANCNFSM4GIBY5IQ.

ekoblogist commented 4 years ago

Dear Yuri,

Thanks for the explanation.

We are trying to make the checklist of vascular flora for Bosnia and Herzegovina, and while we are collecting and georefferencing available floristic data (literature and collections supported with recent field work), since we don't have enough stuff, we decided not to deal with nomenclature, synonymy and systematics on our own, but to resolve it using one of the online taxonomic databases.

Due to several different reasons we've chosen GBIF, but now I am not so sure any more.

Anyway, thank you for your kind offer to explain your procedures, I will most certainly contact you (I already have several questions in mind).

Sorry for my English, and thank you very much,

All the best, Vladimir Stupar

On Fri, Nov 1, 2019 at 7:07 PM yroskov notifications@github.com wrote:

Dear Vladimir,

My apology, I am responsible for original Catalogue of Life (CoL) content at www.catalogueoflife.org/annual-checklist/2019 only and cannot explain how GBIF interprets our data in their portal.

It looks like presentation of CoL data in GBIF portal and references/credits to CoL are confusing and cause misunderstanding among GBIF users. We’ll discuss further how to improve the situation.

If you’ll have questions on CoL practices and procedures, I am happy to tell you our stories via my email yroskov@illinois.edu<mailto: yroskov@illinois.edu> or Skype.

Yours, Yuri

Dr Yury Roskov Catalogue of Life Executive Editor

(Office 2020 in Evers Lab bldg)

Illinois Natural History Survey Prairie Research Institute University of Illinois at Urbana-Champaign 1816 South Oak Street (MC-652) Champaign, IL 61820-6960, USA

Email: yroskov@illinois.edu Phone: +1 217 300 0595

www.catalogueoflife.org; www.sp2000.org

From: Vladimir Stupar notifications@github.com Sent: Friday, November 1, 2019 10:45 AM To: gbif/portal-feedback portal-feedback@noreply.github.com Cc: Roskov, Yury yroskov@illinois.edu; Mention < mention@noreply.github.com> Subject: Re: [gbif/portal-feedback] Incorrect taxonomy for some plants (#1677)

My appologies to CoL.

When I stated:

Considering that majority of mistakes come from CoL, you can see from the list that half of the names have already been corrected in CoL

I just made an assumption that errors were coming from CoL, because GBiF reffered to CoL as a source.

Now, I am checking previous versions of Annual Checklists of CoL and there is no trace of such errors, so I agree that this is probably GBIF backbone issue.

However, my question still stands:

I wanted to know what is the practice in GBIF with such backbone errors, i.e., how long does it usualy take to correct such nomenclatural or taxonomical errors. I saw that some of the issues are open for more than a year now, but on the other hand some names that had incorrect taxonomy in april this year, are now corrected, e.g.: Hippocrepis ciliata (https://www.gbif.org/species/5353690) Luzula multiflora (https://www.gbif.org/species/2700934) Polygonatum latifolium (https://www.gbif.org/species/7920718)

Anyway, thank you @yroskovhttps://github.com/yroskov, for this is the first reply to this issue ever since I posted it (for almost a year now).

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub< https://github.com/gbif/backbone-feedback/issues/472?email_source=notifications&email_token=AG2AELD6FAPEV6PQVGYYJT3QRRFG5A5CNFSM4GIBY5I2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEC3JSQA#issuecomment-548837696>, or unsubscribe< https://github.com/notifications/unsubscribe-auth/AG2AELF4TZBE6QGHCFXSF73QRRFG5ANCNFSM4GIBY5IQ>.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/gbif/backbone-feedback/issues/472?email_source=notifications&email_token=AITWACFEH2GG22AUWYTBPBLQRRV5HA5CNFSM4GIBY5I2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEC3W6MQ#issuecomment-548892466, or unsubscribe https://github.com/notifications/unsubscribe-auth/AITWACCLMPCPY5NHH7R3XDTQRRV5HANCNFSM4GIBY5IQ .

MattBlissett commented 4 years ago

Hi Vladimir,

First, our (GBIF's) apologies for not replying to this issue sooner, and thank you for taking the time to document these problems with the GBIF backbone.

We aim to fix most issues with the backbone each time it is released, which is roughly every 6 months. I don't know why these ones were missed. The next release of the backbone will probably be in January 2020.

The backbone is produced by automatically combining many datasets (listed here). The primary purpose is to organize the occurrence data in GBIF (so people searching on GBIF.org can find relevant data), it's not necessarily the best source for a new checklist. I've referred this to a colleague, as I'm not a taxonomist.

yroskov commented 4 years ago

The main difference between GBIF and CoL:

ekoblogist commented 4 years ago

Thank you, Matt, for your answer.

I admit that back when we started our work on the B&H flora checklist (with occurrence data), I didn't fully understand the purpose nor differences between GBIF and CoL, as well as other databases, such as ThePlantList.org or Euro+Med Plantbase, however we had to choose one.

Here in Europe, botanists have a consensus to use Euro+Med Plantbase as a taxonomic reference, but its taxonomical concept is very much different from the one in CoL or GBIF. However, their database didn't have an API for the use of their data. ThePlantList.org was the second choice, but again no API.

So then I started to experiment with CoL and GBIF APIs and found the latter more intuitive and usable for our purpose. Regarding the taxonomy, I figured that the majority of data are sourced back to CoL, and as much as GBIFs primary purpose is organizing and providing species occurrences, this is not possible without correct taxonomy (just take a look at datasheet for Syringa vulgaris which incorporates data for Ligustrum vulgare because the letter is erroneously treated as a synonym of the former), so I went with GBIF.

Also, in some, not frequent cases, GBIF offered correct solutions which included other databases. Back then I didn't go very much into detail, but now it seems that the majority of such cases fall into the category of different spelling of the name (Salix eleagnos vs. S. elaeagnos or Avenula blavii vs. A. blaui).

So now, I see I made a mistake, and will probably go with the CoL original data for our database.

I have a list of other "errors" and possible misconceptions of names in GBIF backbone that I've "caught" during our database development, so I can send it to you if you find it useful.

ManonGros commented 4 years ago

Thank you @ekoblogist! Yes please send us the issues you found, it is very useful. The backbone should follow the CoL so we shouldn't have the issues you reported. We will be investigating what is happening. For more information on how the backbone is generated and how much of it is based on the catalogue of life, you can check this recent blogpost on the topic: https://data-blog.gbif.org/post/gbif-backbone-taxonomy/

ManonGros commented 3 years ago

Issues that have been solved in the new backbone:

ManonGros commented 3 years ago

Leaving the issue open because 4 issues remain.

ManonGros commented 2 years ago

According to this paper: https://doi.org/10.12705/661.7, Rosmarinus officinalis L. is a synonym Salvia rosmarinus Schleiden

I couldn't find any literature to check if Phyteuma pseudorbiculare Pant. is or isn't a synonym of Phyteuma orbiculare L. Same issue to check if Salix grandifolia Ser. is or isn't a synonym of Salix laggeri Wimm. Will close the issue now, unless someone can provide some references.

ekoblogist commented 2 years ago

According to Kew's Plants of the World Online which uses the World Checklist of Selected Plant Families as a backbone Phyteuma pseudorbiculare Pant. is synonym of P. orbiculare L., while Salix grandifolia Ser. is synonym of S. appendiculata Vill. and not S. laggeri Wimm. According to them Salix laggeri Wimm. is accepted species from S. grandifolia complex.