Internal Reference Taxonomies: Amphibian Pilot

Generally, iNaturalist’s taxonomic policies are to follow one or more taxonomic sources (e.g. Amphibians of the World and SSAR). The role of iNaturalist curators is to keep the Live Taxonomy on iNaturalist.org up to date with this External Source Taxonomy.

A problem with this approach is that the iNaturalist community has been reluctant to completely buy into any single External Source Taxonomy, and articulating priority among multiple sources (especially if they are not global) can be difficult. As an alternative, I propose a system where the iNaturalist community agrees upon an Internal Reference Taxonomy which acts as a middleman between the External Source Taxonomy and the Live Taxonomy. We’ll use amphibians as a case study to pilot this alternative.

The Internal Reference Taxonomy relies on one External Source Taxonomy as a foundation. In this case, that external source is Amphibians of the World. However, the Internal Reference Taxonomy can differ from the external source. The process for deciding how the Internal Reference Taxonomy differs from the foundation external source and any other external sources will be to reach consensus among the iNaturalist curator community. Philosophically, we strive to minimize discrepancies between the foundation external source and the internal reference so any discrepancies must be made explicit and justified with reasoning. Philosophically, we also give priority to local external references for taxa that are locally endemic (e.g. if a species is endemic to the US, then deviating from the foundation external source Amphibians of the World to accommodate a US local source such as SSAR may be more justified). Beyond discussing and reaching consensus on the make-up of the Internal Reference Taxonomy, curators will also strive to keep iNaturalist’s Live Taxonomy in sync with this internal reference.

I (@loarie) will take on the work of ‘amphibian czar’ for this pilot to:
A. update the Internal Reference Taxonomy with any agreed upon changes agreed upon here
and
B. in the event that consensus can’t be reached, act as the tie-breaker

This pilot is an experiment. The costs are that it creates a lot of work for said ‘czar’ and may lead to more taxonomic arguments among curators and the community. The benefits are that it may make for a better maintained taxonomy on iNaturalist with more buy-in from the community. If the benefits outweigh the costs, it might be worth expanding this approach to other taxa or potentially making some functionality changes to better facilitate this approach. If it doesn’t, we’ll revert back to the previous approach.

To kick things off, I’ve made this read-only Internal Reference Taxonomy for Amphibians in a Google Doc. It only considers extant species and explicitly differs from a snapshot of Amphibian Species of the World made on August 8, 2017 as follows:

1. Lumps Aneides flavipunctatus, Aneides iecanus, and Aneides niger as Aneides flavipunctatus
Reasoning: Follows SSAR for these USA endemics

2. Lumps Desmognathus marmoratus, Desmognathus aureatus, and Desmognathus melanius as Desmognathus marmoratus
Reasoning: Follows SSAR for these USA endemics

3. Lumps Desmognathus auriculatus and Desmognathus valentinei as Desmognathus auriculatus
Reasoning: Follows SSAR for these USA endemics

4. Lumps Pseudotriton montanus and Pseudotriton diastictus as Pseudotriton montanus
Reasoning: Follows SSAR for these USA endemics

5. Lumps Trachycephalus typhonius, Trachycephalus macrotis, Trachycephalus quadrangulum, and Trachycephalus spilomma as Trachycephalus typhonius
Reasoning: iNaturalist has a lot of observations of Trachycephalus typhonius so the split will be disruptive. Splitting may be premature since Mexican frogs (Trachycephalus “spilomma”) weren’t formally treated in the paper that coined the split Amphibian Species of the World follows this paper.

6. Lumps Hyla and Dryophytes as Hyla
Reasoning: SSAR hasn’t yet adopted this split, and it will be disruptive on iNaturalist

7. Lumps Pseudacris and Hyliola as Pseudacris
Reasoning: SSAR hasn’t yet adopted this split, and it will be disruptive on iNaturalist

8. Lumps Eurycea quadridigitata, Eurycea hillisi, Eurycea paludicola, and Eurycea sphagnicola as Eurycea quadridigitata
Reasoning: Follows SSAR for these USA endemics

9. Lumps Eurycea spelaea, Eurycea nerea, and Eurycea braggi as Eurycea spelaea
Reasoning: Follows SSAR for these USA endemics

10. Splits Pelophylax bergeri and Pelophylax lessonae from Pelophylax lessonae
Reasoning: While Amphibian Species of the World treats Pelophylax bergeri as a subspecies of Pelophylax lessonae, there are observations of both on iNaturalist so @danieleseglie should probably be consulted before lumping these taxa.

11. Includes Pristimantis bounides, Pristimantis humboldti, and Pristimantis puipui
@coreyjlange added these three newly discovered Peruvian endemics described here that aren’t yet in Amphibian Species of the World but they are legitimate.

While Amphibian Species of the World is a global External Source Taxonomy, SSAR is a local external source for the United States and Canada only. I've also described how our Internal Reference Taxonomy differs from this local external source (using a snapshot of SSAR on August 8, 2017) as follows:

1. Splits Rhinella marina and Rhinella horribilis from Rhinella marina
Reasoning: Amphibian Species of the World has adopted this split and the Mexican community e.g. @coatzin began using it. Because this species which ranges far beyond the US and Canada, it seems reasonable that references other than SSAR should be considered.

2. Includes Anaxyrus williamsi
Reasoning: @aambos observed and @coreyjlange identified this newly described US endemic that Amphibian Species of the World has included. So it seems reasonable to depart from SSAR and include this species.

Next steps

As part of this pilot, curators should now strive to make sure that the amphibian Live Taxonomy on iNat matches the Internal Reference Taxonomy. That means that active taxa not on that list should be swapped accordingly. I've updated the Curator Guide Policies section on Amphibians to reference this post, and please direct any inquiries about iNaturalist’s taxonomic approach for amphibians here.

Similarly, if anyone would like to propose changes to the Internal Reference Taxonomy, please add a comment to this post with your proposal and reasoning. I envision that we’ll have a discussion that loops in the appropriate local expertise on the site and reaches a consensus. I’m hopeful that we can build a polite and open-minded culture towards reaching taxonomic consensus among the community of curators.

Once a change to the Internal Reference Taxonomy is made, curators can help by making the necessary taxonomic change to update the Live Taxonomy. There's been some confusion about how to properly make taxonomic changes on iNaturalist, so hopefully this will be an opportunity to describe the proper steps in more detail necessary to keep everything in sync.

Here's the top identifier figure mentioned in this comment below


Posted by loarie loarie, August 09, 2017 01:40 AM

Comments

Thumb

Hmm, interesting. This is a nice read. I do think this is a good idea, that's all I have to say on it

Posted by cliygh-and-mia 12 months ago (Flag)
Thumb

Hi! I think is a good idea the use of an Internal Reference Taxonomy.
Concerning amphibian taxonomy in Italy, for example, we recently experience some issues on the use of a single External Source Taxonomy (ASW)...
Indeed, the Societas Herpetologica Italica (the national authority for Amphibians and Reptiles in Italy: http://www-3.unipv.it/webshi/) follows the more stable taxonomy of Amphibiaweb (http://www.amphibiaweb.org/taxonomy/index.html)... and also the taxonomy used by the Italian committee of the IUCN (http://www.iucn.it/index.php) match more the amphibiaweb taxonomy than that of the ASW.
E.g., Bombina pachypus is considered a species on Amphibiaweb (http://www.amphibiaweb.org/cgi/amphib_query?where-genus=Bombina&where-species=pachypus) and on IUCN.it (http://www.iucn.it/scheda.php?id=-291037272), wihle only a subspecies on ASW (as B. variegata pachypus: http://research.amnh.org/vz/herpetology/amphibia/Amphibia/Anura/Bombinatoridae/Bombina/Bombina-variegata).
E.g., The species Lissotriton meridionalis (ASW: http://research.amnh.org/vz/herpetology/amphibia/Amphibia/Caudata/Salamandridae/Pleurodelinae/Lissotriton/Lissotriton-meridionalis) is not commonly utilized in Europe, but is considered a subspecies, Lissotriton vulgaris meridionalis (http://www.amphibiaweb.org/cgi/amphib_query?where-genus=Lissotriton&where-species=vulgaris , http://www.iucn.it/scheda.php?id=-1405096456)

So, in principle, I agree with the idea to give priority to local external references for taxa that are locally endemic (but some problem will arise if two different local external references don't agree...)
Considering my experience in Italy, for the endemic or almost-endemic species, to follow Amphibiaweb or IUCN.it should be better (more conservative and more accepted by the Italian scientist community).

So, having said that, I propose the following changes to the Internal Reference Taxonomy:
1. Follow ASW and consider Pelophylax bergeri and only a subspecies of Pelophylax lessonae, P. lessonae bergeri
Reasoning: Also the IUCN.it treats Pelophylax bergeri as a synonymous of Pelophylax lessonae
2. Consider B. variagata pachypus as a full species: B. pachypus
Reasoning: Bombina pachypus as generally accepted as a full species in Italy (As stated above)
3. Consider Lissotrion meridionalis only a subspecies of Lissotriton vulgaris (L. vulgaris meridionalis)
Reasoning: For Lissotrion vulgaris complex there is no global consensus on taxonomy; in Europe is more accepted the taxonomy of amphibiaweb (As stated above)
4. Consider Lissotrion graecus only a subspecies of Lissotriton vulgaris (L. vulgaris graecus)
Reasoning: For Lissotrion vulgaris complex there is no global consensus on taxonomy; in Europe is more accepted the taxonomy of amphibiaweb (As stated above)

Best!
D.

Posted by danieleseglie 12 months ago (Flag)
Thumb

interesting. i will reserve judgement. i don't want the plant taxonomy shaken up personally, because it bounces all over the place. But it isn't always internally consistent, because it can't be because taxonomy isn't consistent.

Posted by charlie 12 months ago (Flag)
Thumb

Wow, quite the undertaking @loarie! But the need is warranted. I've been discussing this specific topic as it pertains to A. boreas / A.boreas halophilus with a few other iNatter's since SSAR no longer recognizes the subspecies.

Posted by zabbey 12 months ago (Flag)
Thumb

I completely agree with @danieleseglie remarks! :-)

Posted by finrod 12 months ago (Flag)
Thumb

Thanks for the general feedback folks. And @danieleseglie thanks for the specific proposals. Lets use your proposed Bombina variegata split as an example for how things would proceed (and an opportunity to lecture folks about the proper way to make taxonomic changes to keep all existing content in sync using the hopelessly complex and confusing iNaturalist taxon change tools).

You're proposing to deviate from Amphibian Species of the World to instead follow IUCN & Amphibiaweb in elevating the Bombina variegata variegata and Bombina variegata pachypus to species status - which can simultaneously be expressed as splitting Bombina variegata (sensu lato) into Bombina variegata (sensu stricto) and Bombina pachypus

Assuming everyone is onboard with this change, I've made 3 draft taxon changes that would accomplish it. In order I'd do the following:
Step 1: Commit Swap 24092 which would inactivate Taxon 480215 while activating and moving content to Taxon 556621
Step 2: Commit Swap 24093 which would inactivate Taxon 479255 while activating and moving content to Taxon 24500
Step 3: Commit Split 24094 which would inactivate Taxon 24496 and distribute content among Taxon 556621 and Taxon 24500 based on their respective atlases.
(This assumes the targets of the taxon changes exist. In this example, I had to create Taxon 556621 but I was able to repurpose Taxon 24500 which was still around from an earlier Taxon Swap I had made in 2015 (looking back, that one should have been a merge rather than a swap, oops).

Yes, I know this is complicated, but these 3 taxon changes are necessary to have all the existing iNat content move to its proper location and keep everything in sync.

Once these taxon changes are made, I (acting as 'amphibian taxonomy czar') would then remove Bombina variegata (sensu lato) [Taxon 24496] from the Internal Reference Taxonomy and add Bombina variegata (sensu stricto) [Taxon 556621] and Bombina pachypus [Taxon 24500].

Does this make sense to everyone? Any objections to implementing this Bombina variegata split? If not, then I'll commit the taxon changes and update the Internal Reference Taxonomy

Posted by loarie 12 months ago (Flag)
Thumb

Also as suggested by @danieleseglie:

Here's 3 draft taxon changes to lump Pelophylax bergeri and Pelophylax lessonae (sensu stricto) as Pelophylax lessonae (sensu lato)
http://www.inaturalist.org/taxon_changes?change_group=Pelophylax+lessonae+merge to bring the Internal Reference Taxonomy in line with Amphibian Species of the World for Pelophylax lessonae.

And here's 3 draft taxon changes to lump Lissotriton vulgaris (sensu stricto), Lissotriton graecus and Lissotriton meridionalis as Lissotriton vulgaris (sensu lato)
http://www.inaturalist.org/taxon_changes?change_group=Lissotriton+vulgaris+lump
to bring the Internal Reference Taxonomy in line with Amphibiaweb for Lissotriton vulgaris.

Combined with the Bombina variegata split described earlier, making these changes would bring things in line with danieleseglie and @finrod's proposals. Everyone on board with these changes?

Posted by loarie 12 months ago (Flag)
Thumb

Perfect! It's all clear to me. I agree with the changes.
Thanks a lot!

Posted by danieleseglie 11 months ago (Flag)
Thumb

@loarie thanks, great! I owe you a beer next time you'll come to Italy (there's a craft brewery near where Daniele lives, and their logo is a toad!) or maybe some wine :-)

Posted by finrod 11 months ago (Flag)
Thumb

I think this is a great way to handle the complexities of the iNaturalist taxonomy. In a perfect world, there'd be a global taxonomic authority that iNaturalist could follow 1:1 but that's not the case for most taxonomic groups and even when there is there may be a very good argument to make small (or large) deviations. The method you propose really is just a better way to document the decisions that curators are already making which I think is a great idea.

Posted by mikeburrell 11 months ago (Flag)
Thumb

Saying "hi" to follow this discussion ;-)

What about adding a few columns to the Google doc where a bunch of things could be documented, eg source taxonomy, reference(s), taxon from which it was split?

Posted by jakob 11 months ago (Flag)
Thumb

Another question, if we do this, can we finally find a way to add a few 'subgenus' type units for really common species groups that can't be determined from photos most of the time? It seems like if we have one 'taxon coordinator' for each group we would be able to pull that off and it keeps coming up again and again. I want carex groupings (though i am not sure i'm willing to offer to be the carex overlord, scary!) and there's been interest in that for spiders as well, as discussed in Google Groups. I know a lot of people are also interested in updating the Mimulus/Erythranthe taxonomy ( @Naomibot is the obvious expert in that realm but probably not on iNat enough to recruit to do taxonomy curation?) and there are plenty of other examples. Just 'hands off my cornus' please! :)

C

Posted by charlie 11 months ago (Flag)
Thumb

@jakob, good idea, I added columns for source and notes to the internal reference taxonomy.

@charlie, the amphibian internal reference taxonomy has columns for family and order which in the case of amphibians are all the nodes in between the root (in this case Class Amphibians) and species. I'd assume if this approach were extended to other taxa that intermediate nodes (family, subgenus etc.) could also be explicitly laid out in the internal reference taxonomy as I've done here with amphibians.

Lastly, looks like ASW has added Pristimantis latro, a newly described Brazilian endemic. Any objections to adding it to our Internal Reference Taxonomy?

Posted by loarie 11 months ago (Flag)
Thumb

One other thing that could be added to the internal reference taxonomy is some sort of taxonomic order value, so that it can be sorted in taxonomic order. This could be multiple fields (i.e. in the case of the amphibian example, an "Order_number", "Family_number", "Genus_number", and "Species_number") or just a single "taxonomic_order" field. I would think that maintaining the order (sorting) values might be a tad easier if it there were multiple fields, so a big taxonomic shuffle wouldn't necessarily mean new numbers for everything each time, just the appropriate level. So, for instance if you moved the order Anura you could just renumber the "Order_number" and all of the family, genus, and species numbers within Anura wouldn't necessarily change.

This could hopefully be used at some point in the future when features on iNaturalist allow taxonomic sorting.

Posted by mikeburrell 11 months ago (Flag)
Thumb

I am not an expert, but rather depend on a friend of mine who is the expert for Louisiana herps, Brad Glorioso. He suggested to me that all Dwarf salamnder in Louisiana with the exception of a couple of parishes are now E. paludicola. His words: Actually paludicola is east of the MS River as well. Only extreme Eastern FL Parishes have quadridigitata.

here is the observation that freshened my memory on it and prompted me to check with Brad: https://www.inaturalist.org/observations/8793224

here is his page he keeps up-to-date : http://www.louisianaherps.com/dwarf-salamander-eurycea-qu.html

anyone disagree? should it (or can it) be updated on inat?

Posted by royaltyler 8 months ago (Flag)
Thumb

Hi royaltyler. This is one of the places where we're explictly not following ASW. Read "8. Lumps Eurycea quadridigitata, Eurycea hillisi, Eurycea paludicola, and Eurycea sphagnicola as Eurycea quadridigitata. Reasoning: Follows SSAR for these USA endemics" Above.
I don't have strong feelings about whether or not to follow this split or not. But if we we were to follow the split it would be important to have range data on Eurycea quadridigitata (sensu stricto) and the 4 other child taxa. I can't get ahold of Wray et al. 2017, in theory its on ResearchGate but the PDF is loading for me. But if anyone has access to that would be good to see if there's range data available in the paper.

Posted by loarie 8 months ago (Flag)
Thumb

royaltyler -

I don't personally have strong feelings about splitting Eurycea quadridigitata (sensu lato) into Eurycea hillisi, Eurycea paludicola, Eurycea sphagnicola & Eurycea quadridigitata and altering the concept of E. chamberlaini (which will bring iNat inline with ASW, but out of sync with SSAR).

It might be nice to form a working group from iNat's top amphibian identifiers who can help decide on whether or not to embrace taxonomic changes like these. Are there ~5 people from the following top IDers who would like to serve in this role with me for amphibians?
@herpguy @lucareptile @ritt @pintail @sandboa @wild-about-texas @danieleseglie @amarzee @coatzin @amplex4love @jakob @sullivanribbit @pagophila @cliygh-and-mia @tonyg @john8 @toby @hydaticus @kucycads @jplarry @sindic @kevin0226 @holzheuser

Duties would involve deciding when to keep iNat in sync with ASW and when to deviate from it (see above).

I checked iNat's current amphibian taxonomy with Amphibian Species of the world and found the following new discrepancies:

1. ASW splits Smilisca manisorum off from Smilisca baudini following http://www.mesoamericanherpetology.com/uploads/3/4/7/9/34798824/mccranie_paper_sept_2017.pdf I suggest we hold off on this split since its a commonly observed species on iNat and McCranie doesn't include range maps.

2. ASW makes the following swaps
Ameerega andina -> Paruwrobates andinus
Ameerega erythromos -> Paruwrobates erythromos
Anomaloglossus astralogaster -> Ectopoglossus astralogaster
Anomaloglossus atopoglossus -> Ectopoglossus atopoglossus
Anomaloglossus confusus -> Ectopoglossus confusus
Anomaloglossus isthminus -> Ectopoglossus isthminus
Anomaloglossus lacrimosus -> Ectopoglossus lacrimosus
Colostethus argyrogaster -> Leucostethus argyrogaster
Colostethus fugax -> Leucostethus fugax
Glyphoglossus capsus -> Glyphoglossus capsa
Grandisonia brevis -> Hypogeophis brevis
Hyloxalus whymperi -> Paruwrobates whymperi
Sallywalkerana diplosticta -> Walkerana diplosticta
Sallywalkerana leptodactyla -> Walkerana leptodactyla
Sallywalkerana phrynoderma -> Walkerana phrynoderma
Eupsophus nahuelbutensis -> Eupsophus roseus
Eupsophus altor -> Eupsophus migueli
Eupsophus contulmoensis -> Eupsophus roseus
Nanorana bourreti -> Nanorana yunnanensis

3. ASW has the following new species
Spinomantis beckei
Pristimantis latro
Pristimantis nimbus
Phrynopus lapidoides
Phrynopus unchog
Bufotes baturae
Bufotes shaartusiensis
Nymphargus caucanus
Sachatamia electrops
Ectopoglossus absconditus
Ectopoglossus saxatilis
Limnonectes conspicillatus
Limnonectes kong
Limnonectes mocquardi
Nanorana phrynoides
Sphaerotheca maskeyi
Charadrahyla esperancensis
Ptychohyla zoque
Phyllodytes amadoi
Scinax onca
Megophrys insularis
Megophrys katabhako
Megophrys lishuiensis
Megophrys sanu
Oreophryne brunnea
Kalophrynus kiewi
Nasikabatrachus bhupathi
Odontophrynus juquinha
Pithecopus araguaius
Abavorana nazgul
Rana luanchuanensis
Kurixalus lenquanensis
Rhacophorus lishuiensis
Hynobius mikawaensis
Tylototriton anhuiensis
Hypogeophis pti

If we can get a working group of ~5 people from our top IDers, it would be nice to just do an efficient vote on whether to make or hold off on these changes now as well as occasional updates (every 3 months or so) moving forward. Sound good?

Posted by loarie 8 months ago (Flag)
Thumb

Hello @loarie

I wish I had seen this post earlier. I read from the top and there is something that bothers me a bit regarding this point.

6. Lumps Hyla and Dryophytes as Hyla
Reasoning: SSAR hasn’t yet adopted this split, and it will be disruptive on iNaturalist

Indeed SSAR hasn't recognised Dryophytes yet, but the IUCN and AmphibiaWeb have (thank you @cliygh-and-mia for making the same comment 3 months ago), and as most Dryophytes are in North America, it doesn't bother people very much as the majority of iNat users are located there (or not in NE Asia at least). By working in North East Asia I'm in a very different position. Hyla crossed the Berring Bridge 60ish mya while Dryophytes did so about 16 mya. I don't have the exact divergence date in mind but I don't think it makes sense to anyone to pool Dryophytes and Hyla here, especially that there is literature supporting the split (Duellman 2016) and it has been recognised by other international leading groups.
On the other hand, there is one paper suggesting that Dryophytes suweonensis may be the same as D. immaculatus, based on a single individual, and the two species are pooled on iNat while it still is a debated question, the range of the two species is separated by the Yellow Sea and the calls are different (although unpublished).

I feel uncomfortable with the unbalanced reaction. Such a thing would not have happened for US or Eruoperan species, where indeed there may be more people active, as there is a clear lack of data.

So I agree and quote @cliygh-and-mia: "So, in principle, I agree with the idea to give priority to local external references for taxa that are locally endemic (but some problem will arise if two different local external references don't agree...)

I also like the recommendation from @jakob about adding a few columns to the Google doc where a bunch of things could be documented.

Regarding your question on a working group from iNat's top amphibian identifiers, I would be happy to help as much as I can.

Thank you for all the developments!

Amael

Posted by amarzee 8 months ago (Flag)
Thumb

Hi!
Forming a working group for me is a great idea! I would be happy to help as much as I can (mainly on European species).
It would be great have one or two people per continent...

Concerning the specific discrepancies iNat-ASW, I agree with your suggestion and I have no reasons to to reject the proposed swaps and new species (But I'm not a specialist on these taxa)
Best!
D.

Posted by danieleseglie 8 months ago (Flag)
Thumb

I can help some as well because I believe this could be a good thing and clear up a lot of the confusion. That being said have a few bones to pick with what has been clumped so far, mainly regarding US caudates. I must agree with @royaltyler on the dwarf split but I also believe that lumping of D. valentinei and D. auriculatus should not be done. This goes the same for the lumping of the E. spelaea complex. My reasoning behind this is because looking at the ASW it lists the papers published describing the species under the "comments" section. The reasoning is the same for the other species that I believe should stay lumped like the D. marmoratus complex. If you read the comments section on those species it will state that the species has yet to be formally described. I understand the inclusion of these species would call for more data and I believe I have the valentinei paper saved, or at least a range map. I know there was an E. spelaea complex map created I'll just need to search for that paper again. The use of amphibiaweb.org as a general reference is a good suggestion by @danieleseglie not just for Italian species. Overall though I think this is a great idea and like previously stated taxonomy in general is a hassle to deal with so I appreciate the effort.

Posted by kevin0226 8 months ago (Flag)
Thumb

Hi folks, I wanted to give a quick update on the Internal Reference Taxonomy approach we’ve been piloting here. Unfortunately, I'm about to embark on a long 6 comment recap - apologies for all the text.

In a nutshell, the approach one solution for how to track explicit discrepancies between an external source taxonomy (ie Amphibian Species of the World aka ASW) and the taxonomy we intended to have live on iNaturalist. For example of what I mean by an explicit disagreement: on iNat but unlike ASW we are choosing to consider Bombina pachypus as distinct from Bombina variegata.

I piloted maintaining an Internal Reference Taxonomy as a separate Google Doc so we could write down a statement about our taxonomic intent and maintain control over when this changes via discussions here. ASW changes all the time. Similarly the Live Taxonomy on iNaturalist changes all the time as iNaturalist curators make changes. So maintaining an independent doc was the only way we could control our intent for what we wanted the taxonomy to be.

We’ve since made a change that allows the a taxonomic clade to be locked down and only edited by one or more ‘taxon curators’. You can read more here where I’ve been piloting this approach with the Ode community (Odes are simpler because for the time being we’re not intentially deviating from the External Source Taxononmy). I’ve locked down Amphibians from Class down to species meaning that I, as the sole taxon curator at the moment, am the only one who can make changes to that part of the Amphibian tree.

One clear advantage of having the iNaturalist taxonomy no longer changing all the time is that we no longer need an Internal Reference Taxonomy to control this process. All we need to do is compare the iNaturalist amphibian taxonomy with our External Source Taxonomy (ASW) taking into account any explicit discrepancies that we are intentionally choosing to make. Even though ASW continues to change all the time, since our iNat taxonomy and our Explicit Discrepancies are now controlled via discussions here, we can, at any point, compare iNat to ASW and choose to either explicitly update the iNat taxonomy or add explicitly add to our list of discrepancies.

Here’s a simple Ruby script that pulls data from the public iNaturalist API and scrapes the ASW website. The script uses our hard-coded discrepancies to compare the two. You can think of these discrepancies as a key for translating between ASW (asw) and iNaturalist (inat). Some are just ways of dealing with apparent typos in ASW (e.g. "Microhyla mihintaleiWijayathilaka,"). Others are one-to-one name changes - for example, ASW uses Dryophytes andersonii while iNat uses Hyla andersonii. Others are more complicated splits or lumps - for example, ASW has three species Aneides iecanus, Aneides niger and Aneides flavipunctatus (sensu stricto) whereas iNaturalist only has one Aneides flavipunctatus (sensu lato).

Posted by loarie 8 months ago (Flag)
Thumb

Here's the 41 explicit discrepancies:


discrepancies = [
{asw: ["Ichthophis cardamomensis"], inat: ["Ichthyophis cardamomensis"]},
{asw: ["Ichthophis catlocensis"], inat: ["Ichthyophis catlocensis"]},
{asw: ["Ichthophis chaloensis"], inat: ["Ichthyophis chaloensis"]},
{asw: ["Microhyla mihintaleiWijayathilaka,"], inat: ["Microhyla mihintalei"]},
{asw: ["Chlacorana crassiovis"], inat: ["Chalcorana crassiovis"]},
{asw: [], inat: ["Dendropsophus nekronastes"]},
{asw: [], inat: ["Hylodes caete"]},
{asw: [], inat: ["Megophrys koui"]},
{asw: ["Aneides iecanus","Aneides niger","Aneides flavipunctatus"], inat: ["Aneides flavipunctatus"]},
{asw: ["Desmognathus aureatus","Desmognathus melanius","Desmognathus marmoratus"], inat: ["Desmognathus marmoratus"]},
{asw: ["Desmognathus auriculatus","Desmognathus valentinei"], inat: ["Desmognathus auriculatus"]},
{asw: ["Pseudotriton diastictus","Pseudotriton montanus"], inat: ["Pseudotriton montanus"]},
{asw: ["Trachycephalus typhonius","Trachycephalus macrotis","Trachycephalus quadrangulum","Trachycephalus vermiculatus"], inat: ["Trachycephalus typhonius"]},
{asw: ["Dryophytes andersonii"], inat: ["Hyla andersonii"]},
{asw: ["Dryophytes arboricola"], inat: ["Hyla arboricola"]},
{asw: ["Dryophytes arenicolor"], inat: ["Hyla arenicolor"]} ,
{asw: ["Dryophytes avivoca"], inat: ["Hyla avivoca"]},
{asw: ["Dryophytes bocourti"], inat: ["Hyla bocourti"]},
{asw: ["Dryophytes chrysoscelis"], inat: ["Hyla chrysoscelis"]},
{asw: ["Dryophytes cinereus"], inat: ["Hyla cinerea"]},
{asw: ["Dryophytes euphorbiaceus"], inat: ["Hyla euphorbiacea"]},
{asw: ["Dryophytes eximius"], inat: ["Hyla eximia"]},
{asw: ["Dryophytes femoralis"], inat: ["Hyla femoralis"]},
{asw: ["Dryophytes gratiosus"], inat: ["Hyla gratiosa"]},
{asw: ["Dryophytes immaculatus"], inat: ["Hyla immaculata"]},
{asw: ["Dryophytes japonicus"], inat: ["Hyla japonica"]},
{asw: ["Dryophytes plicatus"], inat: ["Hyla plicata"]},
{asw: ["Dryophytes squirellus"], inat: ["Hyla squirella"]},
{asw: ["Dryophytes versicolor"], inat: ["Hyla versicolor"]},
{asw: ["Dryophytes walkeri"], inat: ["Hyla walkeri"]},
{asw: ["Dryophytes wrightorum"], inat: ["Hyla wrightorum"]},
{asw: ["Hyliola cadaverina"], inat: ["Pseudacris cadaverina"]},
{asw: ["Hyliola hypochondriaca"], inat: ["Pseudacris hypochondriaca"]},
{asw: ["Hyliola regilla"], inat: ["Pseudacris regilla"]},
{asw: ["Hyliola sierra"], inat: ["Pseudacris sierra"]},
{asw: ["Eurycea quadridigitata","Eurycea hillisi","Eurycea paludicola","Eurycea sphagnicola"], inat: ["Eurycea quadridigitata"]},
{asw: ["Eurycea spelaea", "Eurycea nerea", "Eurycea braggi"], inat: ["Eurycea spelaea"]},
{asw: ["Bombina variegata"], inat: ["Bombina pachypus","Bombina variegata"]},
{asw: ["Lissotriton graecus","Lissotriton vulgaris","Lissotriton meridionalis"], inat: ["Lissotriton vulgaris"]},
{asw: ["Smilisca manisorum","Smilisca baudin"], inat: ["Smilisca baudin"]},
{asw: ["Megophrus koui"], inat: ["Megophrys koui"]}
]

These discrepancies aside, iNaturalist should map to ASW. By this I mean if we run the above script to compare iNaturalist with ASW and there are any discrepancies not accounted for in our Explicit discrepancies, we should either update these discrepancies or update iNaturalist.
For example, imagine we no longer wanted to deviate from ASW via our “Aneides flavipunctatus merge” or “Bombina variegata split” coded in the discrepancies above. If we removed these and ran the script the output would read:

These are species in the iNat, not in ASW…
Bombina pachypus
These are species in ASW, not in the iNat…
Aneides iecanus
Aneides niger
Posted by loarie 8 months ago (Flag)
Thumb

To bring iNaturalist in line with ASW, we must address both sets of species. Beginning with the species not in ASW, we must first determine where to move observations and other data associated with Bombina pachypus before inactivating it. In most cases the species to be inactivated is now considered a synonym of some other valid species (in this case Bombina variegata) and the solution would be to swap the species to be removed into this target species. In some cases, however, it's not clear that the species to be removed is a synonym of any other valid species. In these cases it’s best to swap the species to be removed into some coarser taxon such as the genus Bombina.

The one caveat to keep in mind is that if swapping the species to be removed into the target taxon would expand the concept of the target and the target has a taxon range or an atlas, it’s probably better to do a merge. For example, Bombina variegata (sensu stricto) has a range and an atlas that excludes the Italian peninsula (pink below). If Bombina pachypus is swapped into Bombina variegata (sensu stricto) the taxon range and the atlas of the target would have to be altered in order to make it accommodate this expanded concept into the peninsula Bombina variegata (sensu lato). Sometimes it’s easier to merge Bombina pachypus and Bombina variegata (sensu stricto) into a new taxon for Bombina variegata (sensu lato) with a separate atlas and taxon range rather than remembering to alter Bombina variegata (sensu stricto). But it’s really up to you.


One could also argue that other aspects of the concept Bombina variegata (sensu stricto) would need to be altered if Bombina pachypus were swapped into it. For example, IUCN considers Bombina variegata (sensu stricto) Least Concern and Bombina pachypus Endangered. Technically, Bombina variegata (sensu lato) would have to be reassessed by IUCN so it really should have no conservation status, but considering it Least Concern isn’t a huge deal.

Posted by loarie 8 months ago (Flag)
Thumb

Now lets consider addressing the species not in ASW. In this example they are Aneides iecanus and Aneides niger. Here the obvious action would be to create new taxa for Aneides iecanus and Aneides niger. But similar to the concept change caveat mentioned above, we must be wary that creating new taxa may unintentionally confuse the concepts of existing taxa. For example, Aneides flavipunctatus (sensu lato) includes salamanders not just from coastal California north of the San Francisco Bay (blue) but also salamanders from the Santa Cruz mountains south of San Francisco (pink) and salamanders inland around the Mount Shasta region (orange).

If we add Aneides niger as a species (salamanders Santa Cruz mountains south of San Francisco) this we are implying a change in the concept of Aneides flavipunctatus from senu lato to sensu stricto. Many bits of data associated with Aneides flavipunctatus (sensu lato) may no longer be suitable. For example, default taxon photos, taxon ranges, atlases, conservation statuses. Most importantly, observations of Aneides flavipunctatus from the Santa Cruz mountains must be re-identified as Aneides niger. For these reasons, it’s almost always easier and cleaner to split Aneides flavipunctatus (sensu lato) into Aneides iecanus and Aneides niger and a new Aneides flavipunctatus (sensu stricto) rather than trying to futz with Aneides flavipunctatus (sensu lato) to reinvent it as Aneides flavipunctatus (sensu stricto).

Posted by loarie 8 months ago (Flag)
Thumb

A problem is that it’s often not clear whether a ‘new species’ like Aneides niger impacts the concepts of any other species like Aneides flavipunctatus. If the clade is really poorly known, it might not be possible/worthwhile to stress out trying to proactively understand if adding a new taxon will impact other taxa in the clade so it's best to just add the new taxon.

In either case, often what happens is that only later someone realizes that creating a new taxon has impacted the meaning of another species by flagging the impacted species. When this happens, it’s usually a good time to reactively split the taxa. For example, at the time of this writing iNaturalist has a taxon for Boa constrictor that implies sensu lato. It has observations, an atlas and a taxon range that extend from Northern Mexico down to Argentina. However, I noticed that the taxon Boa imperator also exists with a range from Northern Mexico south to the Andes.

The presence of Boa imperator as an active taxon would imply that Boa constrictor is in fact sensu stricto, but the observations, range, and atlas tell another story. This is an example of a confused, ambiguous, incompatible concepts. The solution (assuming the intent was to go with “the Boa constrictor split”) would be to retroactively replace the Boa constrictor (sensu lato) taxon with one for Boa constrictor (sensu stricto).

But in these cases, rather than trying to shoehorn the taxon for Boa constrictor (sensu lato) into one for Boa constrictor (sensu stricto) by altering the former taxon by manually re-identify observations, altering the atlas, taxon range, conservation status, default taxon photos etc. the proper approach would be to create a new taxon for Boa constrictor (sensu stricto) with the appropriate atlas, taxon range etc. and to retroactively split Boa constrictor (sensu lato) into Boa constrictor (sensu stricto) and Boa imperator.

Posted by loarie 8 months ago (Flag)
Thumb

All these above mechanics of how to manage the amphibian taxonomy aside, we clearly need a way of deciding which explicit exceptions from ASW we need to make. More generally, there are actually a few decisions here that we need to make (or at least be aware that we’ve made them):

What taxonomic reference to use?
Currently, we’re curating from the Order Amphibia down to species. We aren’t actively curating subspecies, extinct taxa, or hybrids. We’re not including any non-standard nodes (e.g. subfamilies). Our external taxonomic reference is ASW with 41 explicit discrepancies mentioned above. Adding/removing additional explicit discrepancies need to be agreed upon.

How often do we update iNat to ASW?
I propose every three months proactively and reactively if someone complains (‘I observed species X which is in ASW why is it not in iNaturalist?’)

Who are the taxon curators and for how long?
The only people who can change the taxonomy of Amphibia on iNaturalist are now the Amphibia taxon curators. The role of the taxon curators is to help facilitate reaching consensus on any decisions affecting the taxonomy of Amphibians on iNaturalist (as I’m doing here) and also doing the tedious work of altering the taxonomy (as I just described above). We can have multiple taxon curators (ideally 1 to 2 more) and I’m proposing we serve for the duration of a year (starting now, but officially through 2018). Chime in if you want to be an ode taxon curator - but please do so if you have time to help me with the roles mentioned above.

Note on dispute resolution process...
In an ideal world, we all agree on these decisions. But I'd prefer we have a protocol in place for what to do if there's a disagreement that can't be resolved. My proposal is to set up an Amphibian taxonomic working group from top-identifiers similar to the one proposed for Odes here.

A working group consisting of the top 5 global amphib IDers + the top 3 IDers from each continent would result in a working group composed of the following people:
-=Global=-
herpguy, ritt, lucareptile, sandboa, danieleseglie
-=Europe=-
danieleseglie, leibele, jakob
-=Africa=-
jakob, alexanderr, mantellaman90
-=Oceania=-
adammyates,reiner,herpguy
-=North America=-
herpguy, ritt, lucareptile
-=Asia=-
amarzee,pintail,herpguy
-=South America=-
herpguy, diogoprov, iuvel

I've pasted a figure that shows these top-IDers in global context at the top of this post (and linked to here).

I propose that this working group is in effect for 2018. If there's any decision that needs to be made thats controversial, we post a link here to an online survey for voting with a deadline and this working group votes.

Does this sound like it could work?

Posted by loarie 8 months ago (Flag)
Thumb

Hi Scott, 2 points in continuation of our discussion of the Odonate group:
- I suggest to have the global top IDers (here herpguy, ritt, lucareptile, sandboa, danieleseglie) _plus_ those from the regions. When selecting from the latter, I suggest to skip those global top IDers (eg in Europe danieleseglie as he's already included in the global top IDers), but selecting the user(s) who are next in the list. This would result in 3 IDers per region.
- I also feel somewhat uncomfortable to see myself both in the Eurpe and Africa group as I'm not a particular amphibian expert in either region. I'd rather have experts such as sindic and hcuohc in Europe or markscherz, wernerconradie and devinedmonds in Africa fill my place here.

Posted by jakob 8 months ago (Flag)
Thumb

Hi Scott! I completely agree with your proposal!
Concerning the working group composition, I agree with the "meritocracy based on community activity" selection, and also with that @jakob has just suggested (global top IDers + those from the regions).
Just some considerations on the additional taxonomic experts and how to most credibly select them...
Yes, having taxonomic experts might produce better outcomes but if they aren't very active on iNat (although they are already users) probably they don't have time to devote to the working group...
Could it be sufficient that working group members (at their own discretion) consult an external expert where necessary?
Otherwise, if it is opted to add "hand selected" taxonomic experts, I'd take the option 2 proposed by @loarie (First order of business by the working group is to appoint 2 additional members).
Best!
D.

Posted by danieleseglie 8 months ago (Flag)
Thumb

Hello Scott!
I like the idea, and I'm happy to be helping, but I don't think I can say that I know amphibians in the whole Asia. Maybe a subdivision is needed? Though I'm happy to help if you think I am able to.
Best,
Amael

Posted by amarzee 8 months ago (Flag)
Thumb

@amarzee imho Russia, Mid East and Central Asia countries could be merged with Europe to make a sort of "extended" Western Palearctic zone, while China, Japan, Korea with South Asia to make a "restricted" (but still huge) Asia. Also North Africa could be put in the Palearctic "iNat zone".

Posted by finrod 8 months ago (Flag)
Thumb

@finrod I like the idea, that seems helpful. Most species north of the Yangtzae River are not present much further South, so there is quite a bit of variation in what would be "restricted Asia", but it's still easier.
Amael

Posted by amarzee 7 months ago (Flag)
Thumb

The idea of selecting people from a bunch of major geographic regions came up to balance the bias of a largely North American user community and not so much to have these as specialists only assigned to the respective regions.

Posted by jakob 7 months ago (Flag)
Thumb

thanks for everyone's (ongoing) help getting Amphibian taxonomy in line. Here's some of the fruits of our labor: https://www.inaturalist.org/blog/13377-tallying-observed-vertebrate-species-with-complete-taxa

Posted by loarie 7 months ago (Flag)
Thumb

From @sandboa: "we need to add the two new species E. paludicola and E. sphagnicola to the this species group. They are both accepted by AMNH and SSAR."

In contrast to the journal post above, SSAR appears to now accept Eurycea hillisi, E. paludicola, and E. sphagnicola based on Wray 2017.

Edit to add: It's unclear to me whether taxon curators get notified when locked taxa are flagged, or what the correct protocol is in your hybrid curator/taxon curator system. Please let me know whether I should bother @ tagging you Scott and/or commenting on these journal posts (or Max for birds). Thanks!

Posted by bouteloua 4 months ago (Flag)
Thumb

Hi Folks, I pulled in an update from Amphibian Species of the World so we're now up to date aside from the 11 'explicit deviations' we're making. Also I created a project with its own journal where we can continue these discussions https://www.inaturalist.org/projects/amphibian-taxonomic-working-group/journal/17227-amphibian-species-of-the-world-update seems more appropriate than my personal journal here. Please follow the project if you want to be looped into Amphibian taxonomy discussions

Posted by loarie about 1 month ago (Flag)

Add a comment

Sign In or Sign Up to add comments

Is this inappropriate, spam, or offensive? Add a flag