Convergence of IOC, Clements and BirdLife Taxonomies? (1 Viewer)

[Paul Clapham]

There are some truly messy groups. For example Asian Koel is split by BirdLife, but not by IOC or Clements. One of the subspecies is not recognized by IOC as it is synonomized with another subspecies. BirdLife however, recognize both the subspecies and the synonomized subspecies, but places them in separate species. How can you create a group for that unless you extend a list to include unrecognized subspecies, and update range info to facilitate recognition of the faux subspecies?

One of my goals when I started this project is that I would have one database for sightings and many taxonomies to interpret them. For example if (in some imagined world) I went to Kalidupa Island in 2019 and saw a scops owl, I would attach that sighting specifically to a kalidupae taxon and it would be filtered up to Moluccan Scops Owl in my IOC list and to Sulawesi Scops-Owl in my Clements list. And then when I looked at my lists later, after the 2022 updates had been applied I would see it as Moluccan in both lists. Without me having to go back and change my sighting.

But those groups you're talking about, I think they correspond to what Denis LePage calls "taxonomic concepts". If you look at his Avibase site you can get an idea of what they are. For my kalidupae example he has five different "concepts" which describe its relationships; see Otus manadensis kalidupae (Sulawesi Scops-Owl (kalidupae)) - Avibase to look at them.

 

[Jon.Bryant]

Going down the rabbit hole again but… the paper fell short of stating fraud, possibly for legal reasons. I was not trying to make it sound nasty, but I did think the language in the paper was quite pointed and personal.

Was a subsequent paper done showing non-assortative mating? Was this subsequent to the NACC decision? The paper I read states evidence of non-assortative mating but didn’t seem to present new survey data, but perhaps there was a reference I missed.

Again I was not challenging the NACC decision, which in fact is also adopted by IOC, Clements and BirdLife. The Dutch Taxonomic authority seem to be the only outlier. But then again I doubt many American’s would give two hoots about this regional authority.

What I was trying to say was ‘can we as citizen scientist resolve some taxonomic arguments by observation, photography or sound recording - if we were pointed in the right direction with clear instruction by taxonomic comment on what the missing parts of the jigsaw are.’

As I say, I went down the rabbit hole with Thayer’s Gull as I personally did not like the paper or the comments - perhaps a cultural issue, as I am a stuffy old brit.

Thayer’s was probably the wrong example, but I had literally read a back issue of Dutch Birding on taxonomy of Thayer’s Gull yesterday evening.

I still think that it would be nice if taxonomic authorities better articulated their decisions. The best I have seen so far are the BirdLife Taxonomic notes, although some notes are (or at least were in version 6) out of date, and the taxonomic reports by the Oriental Bird Club.

 

[Jon.Bryant]

But those groups you're talking about, I think they correspond to what Denis LePage calls "taxonomic concepts".

Not sure about this. My goal is pretty much the same as yours. But let’s say for arguement sake you go to Papua New Guinea and see an Owlet Nightjar. Using the BirdLife taxonomy and Phil Gregory’s book (based on the same taxonomy) you ID it as Archibold’s. You then decide to switch your list to IOC (or with the latest update Clements) and the record become Mountain Owlet-nightjar. But IOC and Clements treat Mountain Owlet-nightjar as monotypic, so if you get bored with IOC or Clements and switch back to BirdLife there is no way to track back to Archibald’s rather than Mountan Owlet-nightjar. In this case IOC and Clements think that Mountain Owlet-nightjar shows plumage variations, whereas BirdLife treat one of these variations (not subspecies) as warranting species recognition.

There are also plenty of examples where the various taxonomies allocate subspecies to different species, but all recognise species on both sides of the divide. To resolve this surely you need groups such as Palla’s Reed Bunting (Common) and Pallas’s Reed Bunting (Pallas’s Group) which seems a bit contrived.

 

[Paul Clapham]

In my system I have Aegotheles albertisi archboldi as a taxon. In BLI it's marked as a species, but in IOC and Clements it's marked as a synonym of Mountain Owlet-nightjar. So a sighting of that taxon would appear as Archbold's in BLI and it would appear as Mountain in IOC and Clements.

 

[Jon.Bryant]

So a sighting of that taxon would appear as Archbold's in BLI and it would appear as Mountain in IOC and Clements

But it f your base was IOC and you started off recording it as Mountain Owlet-nightjar, how could this be mapped to Archibold’s - surely you need to have Mountain (Mountain) and Mountain (Archibald’s) as groups in IOC? Presumably you are using some unique identifier behind the less prescriptive synonym?

 

[Paul Clapham]

I don't have a base taxonomy. I have a collection of taxons which can be used to build taxonomy trees for anybody's checklist. One of them is Birds/Aves which goes at the top of all of the trees, for example. Yes, every taxon has a base name which is just the name used by the taxonomy where I first created it, but it can have a different name in every place I use it subsequently.

So I have a couple of taxons which represent Mountain and Archbold's O-N. In the BLI tree they are sister species, below the taxon for the genus Aegotheles. Whereas in Clements and IOC the Mountain taxon is still a species below Aegotheles but the Archbold's taxon is below Mountain as a synonym.

I also have species groups (like Mallard/Mexican Duck), subspecific groups like the ones Clements uses, and subspecies which can be used to hold the taxonomy trees together.

 

[awiner]

I don't have a base taxonomy. I have a collection of taxons which can be used to build taxonomy trees for anybody's checklist. One of them is Birds/Aves which goes at the top of all of the trees, for example. Yes, every taxon has a base name which is just the name used by the taxonomy where I first created it, but it can have a different name in every place I use it subsequently.

So I have a couple of taxons which represent Mountain and Archbold's O-N. In the BLI tree they are sister species, below the taxon for the genus Aegotheles. Whereas in Clements and IOC the Mountain taxon is still a species below Aegotheles but the Archbold's taxon is below Mountain as a synonym.

I also have species groups (like Mallard/Mexican Duck), subspecific groups like the ones Clements uses, and subspecies which can be used to hold the taxonomy trees together.

For what it's worth, I have a similar problem in Scythebill, which lets people enter in either Clements/eBird or IOC and automatically map to the other. (But not Birdlife, which hasn't seemed worth supporting and only gets requested occasionally.)

I implemented it by treating eBird/Clements as the "base" taxonomy - mostly because (A) I started with that and (B) its groups are darn useful for simplifying mappings. Subspecies that are only supported in IOC have a special, fiddly implementation. I'd describe it all as "good enough". It would've been just a bit better - but a lot more work - to define my own collection of taxa that then get mapped to both IOC and eBird/Clements taxa.