Issue 17694

Decide which bird taxonomy to follow

17694
Reporter: mdoering
Assignee: mdoering
Type: Task
Summary: Decide which bird taxonomy to follow
Priority: Major
Status: Open
Created: 2015-07-16 10:53:14.794
Updated: 2016-02-25 21:11:36.008
        
Description: There are several major competing bird classifications in use, most of which we have as dwc archives. We (or our users) should decide which taxonomy GBIF should follow. Unfortunately not all bird names exist in one source, mostly because of the many changes incl recombinations happening in Aves.

The current Catalog of Life follows ITIS for all bird names within Aves:
http://www.catalogueoflife.org/col/details/database/id/115
http://www.catalogueoflife.org/col/browse/tree/id/67847d320573ff18b0384e22afe97ed8

The ITIS version we have has 33 accepted orders, 193 families and 9.924 species: http://www.gbif.org/species/107233360

"Our" CoL has 33 accepted orders, 225 families and 10.306 species:
http://www.gbif.org/species/141411231


Alternatively we have the Clemens checklist from Cornell:
http://www.birds.cornell.edu/clementschecklist/
http://www.gbif.org/dataset/47f16512-bf31-410f-b272-d151c996b2f6
with 34 orders, 234 families.
10,404 species, 20,768 subspecies

Or the IOC World Bird List from the International Ornithologists Union:
http://www.worldbirdnames.org/
http://www.gbif.org/dataset/c696e5ee-9088-4d11-bdae-ab88daffab78
with 40 orders, 238 families.
10,567 extant species and 149 extinct species.
Includes subspecies


ITIS recognizes 40 bird orders:
http://www.itis.gov/servlet/SingleRpt/SingleRpt?search_topic=TSN&search_value=174371


Is there a clear main bird taxonomy out there that most birders follow?]]>
    


Author: mdoering@gbif.org
Comment: [~trobertson@gbif.org], [~Donald Hobern], [~rdmpage] any ideas how we should settle this? Just keep following CoL or ask the community?
Created: 2015-07-16 10:54:15.759
Updated: 2015-07-16 10:54:15.759


Author: rdmpage
Created: 2015-07-16 12:01:36.021
Updated: 2015-07-16 12:03:22.052
        
[~mdoering@gbif.org]  I guess I'd do one of two things.

Given that Denis LePage has studied bird classifications in some detail http://dx.doi.org/10.3897/zookeys.420.7089 I'd ask him (but he's likely to say we need to support multiple classifications, which doesn't really help).

Alternatively, we could ask the community by querying GBIF. Given that a huge fraction of GBIF's data comes from eBird, what do they use? Even better, is it possible to to the following query:

Given the set of all bird names in each classification, which classification shares the greatest number of name with those submitted as verbatim names in bird occurrences? That's your best classification.
    


Author: donald hobern
Comment: It would be good to check what plans ITIS and Sp2K actually have - there is clearly a tension between the US-centric approach of ITIS and the broader IOC  and Birdlife/HBW trends.  The Birdlife/HBW work on a new checklist (non-passeriformes published last year, other volume next year) will underpin bird conservation globally and may be the best long-term basis.  However it is not yet complete.  Melanie Heath of Birdlife might be a good contact on that.
Created: 2015-07-17 09:07:41.08
Updated: 2015-07-17 09:07:41.08


Author: rdmpage
Created: 2015-07-20 11:10:53.846
Updated: 2015-07-20 11:12:00.202
        
[~mdoering@gbif.org] [~Donald Hobern]
Another consideration is that the global phylogeny of birds http://birdtree.org (from Walter Jetz's group) follow Birdlife, see http://birdtree.org/taxonomy/ Given that this global phylogenyis likely to figure in a lot of biodiversity analyses, it would be nice if there was an easy mapping to whatever classification GBIF adopts.

Yet another consideration is subspecies, which birds are full of (and,of course, a taxon may happily bounce between species and sub-species rank).

One would think that, in theory, if LePage has done all the work in AviBase, mapping these classifications, it should be possible to pick one for GBIF to use, and derive all the synonym mappings automatically so that any user with a bird name from any of the major classifications would be able to recover the bird they had in mind. As far as I can tell AviBase isn't actually available to download, but would that be a conversation worth having?
    


Author: mdoering@gbif.org
Created: 2015-07-22 23:28:54.291
Updated: 2015-07-22 23:28:54.291
        
[~rdmpage] ebird follows the Clements list which is maintained by them: http://help.ebird.org/customer/portal/articles/1006825-the-ebird-taxonomy

Clements and IOC both seem to me the most complete ones recognizing more taxa than ITIS. Would those not be a better choice than CoL? Especially Clements if it is the ebird taxonomy?
    


Author: rdmpage
Comment: [~mdoering@gbif.org] I'd be happy with that choice. Makes sense to match the eBird classification as closely as possible, given that to a first approximation GBIF is a database of birds ;) I like that it has subspecies as well - not that I'm a fan of subspecies as such, but this helps capture a lot of taxa that would otherwise be missed if we ignore that rank.
Created: 2015-07-23 01:00:35.092
Updated: 2015-07-23 01:00:35.092


Author: danstoner
Created: 2016-02-25 21:11:36.008
Updated: 2016-02-25 21:11:36.008
        
"Would those not be a better choice than CoL?"

What is necessary to help CoL become as good as these other choices?

It seems that there is a lot of "work" being done that does not contribute to CoL improvement.