Hi Everyone

I've not followed this discussion too closely, but I thought some more background information would be helpful (please excuse me if you already have this). Some of this is covered by Brian's email below, which I realised after I wrote most of this.

BirdLife recently did an extensive revision of their list, based on a published method (see attached). The revision work was lead by Nigel Collar, working largely with Lincoln Fishpool (at least for most African taxa) and with Peter Boesman doing the vocal analyses but taking recommendations from a large number of people as to what is worth investigating. My understanding is that large numbers of museum specimens were consulted, with actual measurements of the features generally made in the case of morphology (see below "effect size for wing 3.32"), and qualitative assessments made of colour variation in cases of plumage. Actual measures made off sonograms are used for vocalisation comparisons, although are not relevant here, specifically.

My understanding is that the list revision was done to provide consistency to the way species are recognised. Much of the current taxonomies is based on early assessments made by people with limited experience and small sample sizes, with little rigour or consistency. When we first open bird books we assume that the 'species' delimited therein are correct and we get 'comfortable' in these ideas. When any change is recommended we automatically assume it is wrong and what was before is right. I myself did this too. However, I think it was well worth a new perspective, and if nothing else it has made me look more closely at a lot of birds, for which I'm grateful.

Obviously revising the entire checklist was a monumental task. However, this list is not seen as an end point, but rather a work in progress, with further investigation/work to follow. Given the large number of revision made I think it is understandable that each one was not published (it would be impractical and the revision would have taken 50 years instead of 5), but the data is there available for anyone to access. What I can say, however, is that I was initially very sceptical myself, but I have checked many of these new proposal myself by looking at skins in museums, and the more I look the more impressed I am.

Coming to this individual case of the barbets, the following motivation is given for the split, specifically regarding cineraeiceps (available online, this from HBW website (free section):

Taxonomy:
Hitherto considered conspecific with G. bonapartei owing to zone of intergradation in C DRCongo (where populations formerly separated as race intermedius), but this not an obstacle here to its separation, based on larger size (effect size for wing 3.32) (2); proportionately rather longer nasal tufts (but these not measured; ns[1]); pale iris (3); whiter, narrower streaks on forecrown (1); blacker tail (1); broad hybrid zone (1). Monotypic.

The number is parentheses is the score given to that criterion, from 0 to 4, with 4 being very different and 0 not being different at all. Thus it is larger (score = 2; this is an objective measure with an effect size of 3.32, see the original paper for further details), has longer nasal tufts (score = 1), has a pale versus dark eye (score = 3), has narrow streaks on the forecrown (score = 1), and a blacker tail (score = 1) than the nominate subspecies. I don't remember how the hybrid zone scoring works, but that gets a score of 1 in this case; I think the fact that they hybridize but still have remained stable, identifiable taxa over much of their ranges suggests at least that there is assortative mating and thus there is some support that they are 'biological' species. Add these scores up and you get the total score, which is 9 in this case. If I remember correctly, from the early paper it was determined that in very well studied groups from Europe, the threshold for recognising a split was 7 points, so anything above 6 is split. Obviously there is some subjectivity to this in terms of how different colours are scored (the measurement data are much more objective), so one could argue that eye colour should be only a 2, but this still leaves us with a score of 8. One could also argue that nasal tuft length should be excluded because it was not measured. This brings it down to 7. So this is a close one, but based on their transparent assessment of a number of characters that (in this case) differ subtly, BirdLife has split the barbet. Obviously when a bird is split new English names are required. I must admit that I'm not particularly fond of many of the English names, and I think keeping the names as similar as possible would make sense. So I would have gone for Western Grey-throated Barbet and Eastern Grey-throated Barbet, but that is another debate.

Coming back to the method used, there are two major issues I see with it, and that is (i) genetic data are not considered and (ii) there is a problem with the way they have compared taxa involving more than two subspecies, which is mainly a statistical problem.

In the prior case I think this is a shortfall in only a limited number of situations. Most taxa that could be split are sister taxa; in other words, each other's closest relatives. I believe using percentage genetic difference is fairly meaningless in delimiting species, so it then comes down to morphological, plumage and vocal differences. In some cases taxa are not each other's closest relatives, in which case I believe they must be split. One such case is the Common Fiscal complex, which is generally split into Northern Fiscal and Southern Fiscal. Based on genetic data they must be split, because Souza's Shrike sits between them. However, for a significant proportion of species there is no genetic data and so to keep the system comparable for all birds, genetic information has been excluded.

The second problem is perhaps best illustrated with the White-headed Barbet group, which consists of five subspecies split, now, into three species. The problem I have is that none of the three newly recognised species have any unique characters; all characters are shared with at least one other taxon, which means, really, that they should score 0 on all characters. But the taxa were compared in a pair-wise manner, which is scientifically incorrect.

Coming back to some points made, in this case I don't believe there are any elements of 'self promotion' and having names attached to species. I have been involved in several splits and the BirdLife team were always very careful not to try to take credit for the work. Huambo Cisticola from Angola is a case in point, and they offered to withhold the split from the list until I had published a paper that I have been working on but is still not ready for publication.

Also, I would prefer to call it the BirdLife list (as Brian did), as the work was done by BirdLife, essentially. HBW is involved, but this is the BirdLife list.

A further point to keep in mind regarding describing new species, there is a difference between that and what BirdLife has done. BirdLife has not described any new taxa - they all have scientific names. All BL has done is to simply raise taxa to species level, when they were regarded as subspecies on other world lists. Many of the taxa were probably even described as species when they were first given names, but later lumped. This is actually the case for this barbet, which was described as Gymnobucco cinereiceps Sharpe, 1891. I can bet you there was no detailed study lumping cinereiceps with bonapartei, just an opinion of someone who compiled a regional or world list, which others followed without question. So cinereiceps was described as a species initially, before we came to know it in the field guides as a subspecies of bonapartei. Really there is nothing new about it at all, and it has already been described as a new species but lumped possibly without good justification. BirdLife is perhaps just rectifying what was a good species and described as such, which got lumped without any proper study.

That's my two cents worth.

Kind regards
Michael

On 2019/01/05 14:24, Adam Scott Kennedy adamscottkennedy@gmail.com [kenyabirdsnet] wrote:

 

Hi Brian et al.,

I'm not entirely sure if you were asking for this but with these new developments on HBW, I've found the citation to be located at the very bottom of the page. Find attached for your attention just in case.

I agree largely with what Brian has said here and I also wonder about the formality of such 'studies' (if there were any) and what are the origins of such splits being discussed in the first place; museums, field observations, coffee shop chit-chat? With the studies that Alex Kirschel has done o, the extensive field research is combined with mDNA to produce a worthwhile reason

I am hopeful that the primary motivation is species conservation although I think there are also elements of 'self promotion' and getting one's name attached to a new species, especially given at least one of the parties implicated here and their general 'type'. As Brian states; 

"all this is a farce and makes mockery of the set of procedures that are in place to prevent cavalier presentation of new species, it could be just as easily you or I or anyone for that matter following this same route."

Indeed it is a bit farcical but no doubt Messers del Hoyo, Kirwan and Collar will have a perfectly reasonable explanation. Would anyone like to request a comment from them on this matter?

On another note, watch out for a Yellow-spotted Barbet split in due course as I recently received an RFI from Nigel Collar in this regard.

Cheers

Adam

On Sat, 5 Jan 2019 at 07:45, Brian Finch birdfinch@gmail.com [kenyabirdsnet] <kenyabirdsnet-noreply@yahoogroups.com> wrote:

 

Hi Neil and all,
As promised I have attached the description (passed to me by Nigel
Hunter yesterday), of the Grey-headed Barbet Gymnobucco cinereiceps,
presented as a distinct species by BirdLife as opposed to
Grey-throated Barbet G. bonapartei.

I support the Tobias principal that a bird that is morphologically
different should get recognition, and receive appropriate protection
for the form. I also think that BirdLife are doing a sterling job at
highlighting little recognised forms that deserve our attention.

Lynx Publications are a private commercial publishing company with
their offices in Spain. They merely act as Editors/Publishers for
BirdLife, and of course as such have no rights to launch new species
onto the planet, and are just printing what they are told to.

If we now look at this revolutionary new Barbet, that occurs in East
Africa, we must look at the parameters that have been set that lead to
the acceptance by both scientific and interested communities of a
formally described new species. Is there any scientific-backed
evidence, and structural evidence, comparison of bill structure and
size, tarsal length, wing and tail lengths, egg, immature plumages and
moult, vocalisations, nesting habits, and on and on….. Is there a
formal write up of the species which we can refer to in the reference
to see in detail what the proposition for regarding this as a new
species is, is there an author credited as to the research on the
study…. Was there a study?

All of these relate to supportive evidence, most of these are
mandatory when writing up a new species to science, and to all of
these the answer is negative, there is no paper, no study, no author,
no genetic evidence etc. No-one in the past hundred years has ever had
a species accepted as novel merely based on a few feather details, and
minor size variation.

Looking further into the paragraph, items such as different lengths of
nasal tufts, have not even been compared and measured, there is a
broad zone over overlap which is brushed aside, all this is a farce
and makes mockery of the set of procedures that are in place to
prevent cavalier presentation of new species, it could be just as
easily you or I or anyone for that matter following this same route.
No new species should ever be admitted without peer reviewage prior to
formal publication.

We are very close to publishing the new Kenya Checklist, is anything
going to change for Grey-throated Barbet…. I think not. Are any of the
world checklist authorities going to accept this on the evidence
provided, I hope not.

There are very many other new species that they have also “launched”
that are in all likelihood forms deserving specific identity and
hopefully all will be followed up with convincing material, but for
this Barbet alone it is all too wishy-washy and unconvincing and there
lies the problem.

All this may seem harsh, and as I say BirdLife have my support, but
with this they are transgressing the ethical boundaries of species
representation, great for the world listers but is it the road to
complete chaos? With all the necessary evidence for this form provided
and following the correct guidelines, it may well be that they are
right after all, but at present they are not.

Feel free to comment and give your opinions,

Best to All
Brian

On 1/3/19, Brian Finch <birdfinch@gmail.com> wrote:
> I don't know as it is the first I heard if it. In theory there should
> be a cited reference with details of the paper, I will see and let you
> know but do not have HBW Alive I will ask Nigel when I see him
> tomorrow,
> Happy New Year
> Brian
>
> On 1/3/19, Neil and Liz Baker <tzbirdatlas@yahoo.co.uk> wrote:
>> Brian
>> Is there a peer-reviewed paper for this barbet split?
>> thanks
>> Neil
>> Neil BakerTanzania Bird AtlasP.O. Box 396, Iringa, Tanzania Mobile: +255
>> 753-513603 http://tanzaniabirdatlas.netSubscribe to:
>> tanzaniabirds-subscribe@yahoogroups.com
>>

--

-- 
------------------------------------------------
Michael Mills
BIRDS ANGOLA
www.birdsangola.org