Mammal Taxon Curators
Hi @bobby23 @jwidness, @jakob, @maxallen et al. - thanks for your help curating Mammal taxa and your patience as I've been getting the bugs out of some new Taxon Curator features.
Mammals on iNat are now relatively 'clean' in that what's on iNat still more or less follows IUCN, but the associated Taxon Frameworks are now mapped to MDD. All the IUCN taxon ranges and conservation statuses for mammals are imported and up to date. And I think we all agree that if we want to have the mammal taxonomy on iNaturalist evolve (by moving towards MDD or elsewhere) we should do a thorough curation job to ensure that we don't loose distribution data or otherwise corrupt/introduce misspecification into the taxonomy. Luckily, I think with the new Taxon Framework tools we can do that. The purpose of this post is to:
1) start a conversation about what taxonomy we should be aiming for. I know @jwidness and others are pushing for MDD while @jakob and others are pushing alternatives. Before we start curating taxa to move it in one direction or another, we need to agree on what taxonomy we are all aiming towards. I think having a Google Hangout meeting would be a nice place to have this discussion. Would someone be interested in taking the lead on organizing such a discussion (e.g. starting a doodle poll https://doodle.com/create)?
2) more training about how to use the new Taxon Framework tools and other tools for good curating housekeeping. I wrote a post on how to work with Taxon Framework Relationships: https://www.inaturalist.org/projects/inaturalist-mammal-working-group/journal/21192-how-to-work-with-taxon-framework-relationships @jwidness also made an awesome guide for curating Taxon Ranges here:
https://docs.google.com/document/d/1LyavoWDxsZ5ObwjDDMbThoLZ3ZZf4_ciC2h19Vebyt0/edit
So we can start eliminating me as an unwilling bottle neck here, I added @bobby23 as a taxon curator for 'Rodents', @jwidness as a taxon curator for Shrews/Moles and @jakob as a taxon curator for bats (@maxallen I also just made you a curator - I suggest you get started with reading the curator guide and spend some time on mammal ssp which aren't covered by the taxon framework to get the hang of things). As for bobby, jakob and jane - this is just to learn how to use these tools at this point. Even though you now have permissions, please don't commit any taxon changes or otherwise alter the taxonomy until we have a chance to discuss (1 above).
For now, it would be great if you could read my post section about Marmota kastschenkoi and how curation was required to change the Taxon Framework Relationship from 'non-internal' to a 'many-to-many' with Marmota baibacina. You now have permissions to do these for your respective groups (rodents etc.). Please give it a try with a few of the non-internal Rodents, or non-internal Bats, or non-internal Shrews/Moles.
As I mention in the post, non-internal mappings should be rare. In most cases they need to be replaced with other deviations (one-to-one, many-to-one etc.) and in the rare cases where the non-internal mappings are legit, please explain your reasoning (e.g. for Galagoides kumbirensis). Please try to 'resolve' at least one or two of the non-internal mappings in your group by replacing them with more complex mappings or by adding an explanation for why non-internal is justified. Please let me know if you have any trouble or encounter any bugs. But also please be careful as all the kinks aren't worked out here and things can require alot of cleanup if the Taxon Framework Relationships get too screwed up. And a reminder - please don't actually change the taxa (e.g. taxon changes) until we have a chance to discuss (1) above.
Also thanks for being guinea pigs here. These tools are by no means perfect, but the best way to improve them and document them is to try them out. Many thanks for bearing with all the clunkyness.
Lastly, if any other folks than the ones mentioned here are interested in getting into this please give a shout. But remember, being a good curator/taxon curator on iNat has more to do with the time and interest in doing lots of tedious manipulation of records than it does actually being the world's expert in anything and its definitely not for everyone. Please make sure you really want to dig into these tedious and clunky interfaces before you commit to getting involved.
Thanks!
Scott
Comments
I deleted the non-internal relationship for Suncus niger and added it as a many-to-one match with Suncus montanus. Does it look ok?
that looks great jwidness - thanks for giving it a try and for adding that info!
@loarie what do you think about displaying the notes for each relationship on the taxon framework relationships page?
Then we could see which non-internal ones have already been addressed.
It would also be super helpful to have some sort of indicator on the taxon page that it is already entered as a deviation from the accepted authority.
@loarie another question: when you say "All the IUCN taxon ranges and conservation statuses for mammals are imported and up to date" does that mean that there shouldn't be any species that have outdated IUCN maps? (Like Gazella cuvieri in the taxon range editing guide.)
jwidness, I think I could add at least the first few sentences from the notes to each the taxon framework relationship panel on the taxon framework relationship search page. I don't think we can put in all the notes since they're meant to accomodate images and lots of text etc. - would that work?
bouteloua, right now all that information is on the taxonomy_details page that's linked to from the taxon page. What is it more precisely that you'd like to show on the taxon page?
I'm still wrapping my head around this new tool, but maybe—
On the taxon page, taxonomy tab, next to the link to the "Taxonomy Details" page, display a letter for the relationship status, matching the layout of the number of taxon changes below it?
-"D" for "deviation" (green background)
-"M" for "match" (grey bg)
Possibly "U" for "unknown"/blank (pink bg) ? What's an example case for relationship status="unknown"? Only display a pink "alert" (unknown/blank) if the taxon is covered by a framework — most vascular plants would currently have a pink U, but for most insects, nothing would be displayed there)?
@loarie yeah, a snippet would be fine
jwidness, correct. I updated all the mammal ranges and statuses with IUCN. It was mostly easy because mammal taxonomy currently mostly matches IUCN. These are the exceptions I'm aware of. In these cases I did not add/update in the IUCN range/status because we have a different concept on iNat than IUCN:
{iucn: ["Thylamys cinderella"], inat: ["Thylamys sponsorius", "Thylamys cinderella"]},
{iucn: ["Sminthopsis fuliginosus"], inat: ["Sminthopsis fuliginosus", "Sminthopsis aitkeni"]},
{iucn: ["Sorex antinorii"], inat: ["Sorex arunchi", "Sorex antinorii"]},
{iucn: ["Crocidura poensis"], inat: ["Crocidura fingui", "Crocidura poensis"]},
{iucn: ["Ictidomys mexicanus"], inat: ["Ictidomys mexicanus", "Ictidomys parvidens"]},
{iucn: ["Sciurus vulgaris"], inat: ["Sciurus meridionalis","Sciurus vulgaris"]},
{iucn: ["Glaucomys sabrinus"], inat: ["Glaucomys sabrinus","Glaucomys oregonensis"]},
{iucn: ["Otospermophilus beecheyi"], inat: ["Otospermophilus beecheyi","Otospermophilus atricapillus"]},
{iucn: ["Thomomys umbrinus"], inat: ["Thomomys umbrinus", "Thomomys sheldoni", "Thomomys atrovarius", "Thomomys nayarensis"]},
{iucn: ["Loxodonta africana"], inat: ["Loxodonta cyclotis","Loxodonta africana"]},
{iucn: ["Dipodomys phillipsii"], inat: ["Dipodomys phillipsii", "Dipodomys ornatus"]},
{iucn: ["Urocitellus brunneus","Urocitellus endemicus"], inat: ["Urocitellus brunneus"]},
{iucn: ["Hylobates muelleri","Hylobates abbotti","Hylobates funereus"], inat: ["Hylobates muelleri"]},
{iucn: ["Marmosops caucae", "Marmosops impavidus"], inat: ["Marmosops impavidus"]},
{iucn: ["Cebus brunneus", "Cebus castaneus"], inat: ["Cebus olivaceus"]},
{iucn: ["Cebus albifrons","Cebus cuscinus","Cebus aequatorialis","Cebus cesarae","Cebus malitiosus","Cebus versicolor"], inat: ["Cebus albifrons"]},
{iucn: ["Molossus bondae", "Molossus currentium"], inat: ["Molossus currentium"]},
{iucn: ["Plecotus kolombatovici"], inat: ["Plecotus kolombatovici", "Plecotus gaisleri"]},
{iucn: ["Rhinolophus pusillus"], inat: ["Rhinolophus monoceros", "Rhinolophus pusillus"]},
Also I ignored the range for Ochotona pallasi since the IUCN range seems to be out of sync with the their own concept and replaced it with a manually fixed version.
The trick is going to be how to preserve taxon ranges if/as we move in the direction of MDD. Since doing so will require splitting/lumping a lot of IUCN ranges
bouteloua - that sounds good. If there's no taxon framework relationship for a taxon then I'd recommend its blank. I can add a 'U' for unknown, but any taxon framework relationship with relationship unknown has some problem associated with it. I hope there aren't places in the UI where taxon framework relationships with unknown's can be created but there could be if something funky happens
Oh... I had manually edited a few IUCN range maps based on non-IUCN sources that got blown away by the refresh. For example, I re-did Odocoileus hemonius (and added subspecies maps). You can see how the subspecies maps no longer match the species, e.g. the cut-outs in the southwest, and the range expansion in Texas.
I had also extended the range for Heliosciurus undulatus because a number of sources (including iNat observation data) put the boundary further south.
There's no way to restore those, right? Or even just to get a list of ones that I had edited?
hm - shoot I didn't realize that. We need a better way of tracking 'deviations' from IUCN. Unfortunately the database isn't currently tracking created_by/updated_by for taxon ranges so we have no way of knowing which ones particular people added/updated. In theory one could dig into the backups and pull out the ranges but its a big job and we'd have to know what ranges you changed. Do you have data on your side for what you added? - more philosophically, and admittedly not having thought about this that much, I'm not sure how I feel about changing the IUCN ranges when one exists and its not a concept mapping issue (e.g. sensu lato/sensu stricto type of thing). Range maps are supposed to be abstractions so trying to get them 'perfect' is a slippery slope with no clear stopping point. But curious to hear what others think
Sorry for the delayed response! I am happy to be involved with the management of iNaturalist's taxonomic tree.
I have yet to do anything with the Taxon Framework for rodents, but I added a deviation for Cephalopoda and it worked just fine. (I added the non-external order Idiosepiida based on a recently published review in Hanlon R, Vecchione M, and Allcock L. (2018). Octopus, Squid, and Cuttlefish: A Visual, Scientific Guide to the Oceans’ Most Advanced Invertebrates. 10.7208/chicago/9780226459738.001.0001.) Everything worked out fine. I plan on providing a more thorough look through the deviations listed under Rodentia this weekend to practice using the Framework tool.
I agree with @bouteloua that a visual indicator on the taxon page would be nice for deviations from our authorities. It seems tucked away under "Taxonomy Details". However, I don't think this is a pressing issue. What's unclear to me is if a listed 'deviation' is implied to always be intentional, or evidence of an oversight to be corrected. There are intentional deviations in Cephalopoda, for example, and it would be helpful if there was some way to clarify this for observers looking at the Taxonomy Details.
I'm pretty sure I still have all my kml files, it's just a matter of figuring out which ones were edited. I have a ton of files, but most were either direct updates from IUCN or subspecies maps (e.g. waterbuck, kob, hartebeest). A few maps had other issues, like only the introduced part of the range was displaying. I think there are only maybe a handful of actual edited maps.
I agree that we'll never have perfect range maps, but the ones I edited were for species where I had personally found that others were having trouble with IDs due to "out of range" observations, and the edits I made followed reliable outside sources (and were cited in the upload).
I'll hold off on re-posting my maps, and if anyone would prefer to keep them all set to IUCN, that's ok with me.
I think your maps are fine, @jwidness, and some of which are preferable to what the IUCN has published right now.
No map is a perfect representation of a species' true distribution, but as long as the maps accurately reflect the taxon we are trying to convey, then I see no problem with deviating from the maps that IUCN has published.
i don't have strong feelings about replacing IUCN maps - but we should come up with a system for tracking that. FWIW I didn't meddle with ssp (range maps or taxonomy) since they're not really covered by IUCN or MDD
@jwidness - the snippets you requested are up
I also added some taxon framework data to taxon changes as requested by https://www.inaturalist.org/taxon_changes as suggested by bouteloua
Thanks, the snippets looks great.
I thought @bouteloua was requesting something like in the picture below?
That too. :) The other requested changes to taxon changes were via email.
e.g. https://www.inaturalist.org/taxon_changes/46699 output taxon says "Relationship: Match"
I've noticed a few taxon changes happening, but I'm still not entirely clear on whether there was a consensus for how to proceed. Can we discuss what the plan is?
my understanding was that @jakob volunteered to organize a meeting to discuss what our target for curation should be (e.g. MDD, IUCN, Willson Reeder etc.). Is that correct jakob that we should await such a meeting invite or otherwise?
For what its worth, my personal preference is that we should be curating towards MDD but that we should aim not to loose atlases and range maps so we should only move away from IUCN and towards MDD once curators have manually split the IUCN range maps etc as jwidness has done a great job at doing. But I'd be happy to discuss this more (preferably in a voice meeting setting) if there's not broad consensus for that approach
I admittedly have gotten antsy and started doing (very light) work on rodents and playing around with Framework tools, but I have not made any change without reviewing the source material beforehand (as oppose to blindingly adopting what the MDD or IUCN propose), and I have not committed any significant changes that affect atlases or taxon ranges. Reasons being that: (1) I am still not well and using QGIS is too much for me right now and (2) I understand how it's important to preserve the KMLs.
@jwidness made this taxon split https://www.inaturalist.org/taxon_splits/55662 which when committed would 'resolve' this deviation https://www.inaturalist.org/taxon_framework_relationships/109509 (ie move from IUCN towards MDD)
This fits my 'personal preference' I described in my prev comment above. But I thought @jakob was maybe not on board with this. But I'm not in favor of holding up jwidness's work if jakob is too busy/not super engaged. @jakob - any objections to committing taxon swaps like this one that move from IUCN to MDD (when work on ranges and atlases is put in as jwidness did here)?
To clarify: this is a split that IUCN also supports, so it's not really moving away from IUCN. The problem is they've only updated some of the pages associated with this split, for example https://www.iucnredlist.org/species/40646/92572335. But it's clear IUCN and MDD both support C. wolfi as distinct from C. pogonias.
ah got it - this is one of those annoying situations where IUCN partially splits something (presumably because they're rolling it out over a prolonged period) in a way that is not internally consistent. I'll commit then
Hi guys, please go ahead whatever suits iNat best. I simply don't find the time to substantially engage in this and related discussions.
@loarie @bobby23 do we currently have a plan of action? I'm not doing anything in Eulipotyphla because that's not really my area of interest, but we sort of started in on Carnivora. Are we going back to that or are we each doing a separate part of the tree? How are we deciding on deviations? etc. etc.
I think how we manage Mammalia could be a good model for how we handle Frameworked taxa going forward, but I think a little more discussion is needed (some good discussion is going on in the Forum: https://forum.inaturalist.org/t/questions-on-taxon-curation/4134).
But I at least would like to see Mammalia opened up a bit more, instead of dividing curators across multiple taxa.
Add a Comment