Thanks. Isn’t it better to swap Caracara plancus ss into C.p.plancus, as this maintains the IDs as Southern Crested Caracara, rather than them all becoming the less precise C.plancus sl?
1 Like
I think its a gray area. I agree with you that the benefit would be maintaining the C.plancus s.s. precision, but the cost is the additional swap (incurring replacing IDs and changing obs ancestries) and also since ssp are controversial there’s a cost/risk there. I know alot of people don’t like to ID to ssp and suspect many might not like to have their species level IDs automatically replaced with ssp level IDs when the species name is still valid (though I agree with you that technically that would be needed to maintain precision). I think you could argue it either way, but IMO the costs outweigh the benefits
1 Like
I don’t think this has been touched on yet but how are we doing splits for species that are regularly nomadic. I see the Larus canus split being particularly tricky because in places like Eastern US, L. canus and L. brachyrhynchus are both equally likely to be your bird. I even hear there’s a few cases that L. canus kamtschatschensis are present every now and then. And because that subspecies breeds in eastern Russia, they likely being reported as L. brachyrhynchus along the US Pacific coastline. So we can atlas the two species according their regular ranges but there’s going to be the conflict of misapplied ids.
1 Like
IDs of the split input taxon will either be replaced with one of the output taxa if the obs uniquely falls in one atlas, or the common ancestor if the obs is outside of all atlaess or overlaps multiple atlases. FOr the Larus canus split https://www.inaturalist.org/taxon_changes/99241 it would be best if the places where the atlaeses don’t overlap focused on places where one species is overwhelmingly likely over the other. For places where both occur (e.g. Eastern US) I’d expect those to be outside of all atlases and thus IDs will be replaced with genus Larus.
Good point about vagrants though. In places where species X is overwhelmingly likely there’s still the chance of a weird vagrant of species Y occurring there and thus a split replacing the ID of species Y with species X but I wouldn’t sweat this too much. You can always manually comment on/mention/ID on that obs to get the community to alter IDs if its not practical to design IDs to accomodate it.
2 Likes
Including the Long-requested and observed Dinopium psarodes x bengahalense.
Maybe another dumb question but will genus level replacement only occur to observations previously identified at species level? Because if an observation in eastern US is already identifed to L. canus canus and so forth, it wouldn’t make sense to bump it up to genus id.
1 Like
I’ve created atlases for both species. Basically, I added only places where either species is expected. For example, it is almost certainly correct to believe all western US sightings are Larus brachyrhynchus but since they neither breed or winter there, I left those states out of the atlas because they’re vagrants. The atlases should be aligned to the atlas of L. canus (sensu lato), just divided between New World and Old World.
1 Like
@loarie are all the subspecies splits supposed to be in the system? I’m looking at Bubo virginianus, which is supposed to have some subspecies splits to accommodate the “new” taxa B. v. algistus, B. v. pinorum, and B. v. deserti, but I don’t see any splits set up – though the inactive taxa are there.
1 Like
@ jwidness yes ssp splits should be there. If there not there its because they’ve were mischaracterized as ‘wholly new’ in one of the steps above. Please list them out with ‘split notation’ (filling in the XXX for Bubo virginianus and adding new ones) and I’ll make draft splits e.g.
Bubo virginianus XXX → B. v. algistus, B. v. pinorum, and B. v. deserti
As for our splits needing atlases - thanks for all the work on these - I went throug the above any for many people just weren’t crossing them off, so we’re further along than I thought. I updated the post above and am repasting the list here with just the remaining ones that still need to be done.
Splits needing atlases
split#2 - Zosterops senegalensis
split#10 - Colluricincla megarhyncha
split#14 - Alophoixus ochraceus - using IUCN maps
split#15 - Pyriglena leuconota
split#29 - Synallaxis stictothorax - using IUCN maps
split#31 - Cynanthus latirostris - using IUCN maps
split#32 - Euphonia affinis
split#33 - Zosterops abyssinicus
split#38 - Ficedula narcissina - using IUCN maps
split#40 - Forpus xanthopterygius
split#41 - Gallinago paraguaiae
split#42 - Grallaria rufula
split#43 - Herpsilochmus rufimarginatus
split#67 - Sylvia cantillans
We’re getting close! thanks for all your help
4 Likes
I’ve taken care of atlases for splits #2, #15, #33, #41, #43.
I will note that there’s overlap between the species split#2 and split#41, and I think it’d be best if we just let the community reidentify countries of overlap because trying to atlas the erratic borders of Tanzania and Argentina states/provinces would cause more problems than solve.
3 Likes
Slight update: Ninox fusca will have to be split now (as there is now an observation of it).
3 Likes
This topic has gone pretty static. I’ve dealt with one of the two remaining splits still needing atlases; split #40. The only one still needing work is split #67 and I started that, but the species’ seem to exhibit high levels of movement in both breeding and wintering grounds so even if we get them atlases, we’re looking at a high potential of mislabeling observations.
@loarie I’m not sure what else is needed to get done but what timeframe are we looking at for getting these splits committed? Do you think we can get a majority of them done by new year’s?
Is there anything else that needs to be done (save for #67)? I have a bit more time now so might be able to help.
Nope, literally everything is done.
1 Like
#67 should be a two-way split only. Sylvia subalpina already exists on iNat, though will need updating to Curruca. @loarie. I’ll have a go at atlases, there is a lot of overlap but some areas only have one species
1 Like
Sorry to be a pester, but we’re starting the month of February and to the best of my knowledge, we have not moved forward in implementing the new changes. Do we have a timeline on when this will happen?
6 Likes
I have gone through the splits and done all the outstanding ‘non-splits’ ie raising subspecies to species, so I think all the splits are now complete. However, I have deleted rather than updated the ranges of the parent species. I also added English names to all splits.
1 Like
Have also activated all the new species, and added English names
As new subspecies are starting to be added haphazardly, have made a start on the subspecies. Have activated the completely new ones (ie species previously monotypic), and some of the others where there are no subspecies IDs in the range of the new subspecies. Have also inactivated a couple of recently added subspecies that might need a split. Not sure how best to deal with splits as detailed ranges are hard to obtain for subspecies
I noticed when the owl taxon splits were committed, some IDs didn’t update properly (they reverted to genus instead of species). These observations:
https://www.inaturalist.org/observations/20303444
https://www.inaturalist.org/observations/78827428
https://www.inaturalist.org/observations/9836154
have large accuracy values, so maybe that’s why (even though there’s still only 1 species within the circle)? These observations:
https://www.inaturalist.org/observations/39863321
https://www.inaturalist.org/observations/25965379
https://www.inaturalist.org/observations/25965409
https://www.inaturalist.org/observations/5706608
https://www.inaturalist.org/observations/9998597
https://www.inaturalist.org/observations/110724077
however, have much smaller accuracy values and I don’t know why most IDs went to genus, especially since the observations around them weren’t affected.
I believe I found all the affected observations, but there could be more.