[Daniel Philippe]

3 species ? Did you know that ? I did not and saw these birds only on their wintering grounds, so today I do not know what I saw:

24th International Ornithological Congress, Hamburg, Germany, 13-19 August 2006
ABSTRACT

Saitoh T1, Nishiumi I2, Alström P3, Olsson U4, Ueda K5
(1) Dept. of Life Science, Rikkyo University, 3-37-1, Nishiikebukuro, Toshima-ku, Tokyo 171-8501, Japan, takema@sc4.sonet.ne.jp
(2) Dept. Zoology, National Science Museum, Hyakunin-cho, Shinjuku-ku, Tokyo 169-0073, Japan, nishiumi@kahaku.go.jp
(3) Swedish Museum of Natural History, Box 50007 SE-104 05 Stockholm and Dept. of Systematic Zoology, Evolutionary Biology Centre, Uppsala University, Norbyvagen 18D, SE752 36, Uppsala, Sweden, per.alstrom@ebc.uu.se
(4) Dept. of Zoology, University of Göteborg, Box 463, 405 30 Göteborg, Sweden, urban.olsson@zool.gu.se
(5) Lab. of Animal Ecology, Rikkyo University 3-37-1, Nishi-ikebukuro, Toshima-ku, Tokyo 171-8501, Japan, keisuke@rikkyo.ac.jp

Deep phylogeographical divergences among Far Eastern populations of the widespread Arctic Warbler
In contrast to the numerous phylogeographical studies of North American and European birds, very few have been undertaken in far eastern Asia. The Arctic Warbler (Phylloscopus borealis), a migratory passerine with a wide breeding range from Norway to Alaska, is unsettled taxonomically. Three to seven subspecies are usually recognized: borealis from Scandinavia to the Chukotskiy Peninsula, sometimes split into talovka, transbaicalicus and borealis; hylebata in Ussuriland and Sakhalin, although the Ussuri population is sometimes lumped with borealis and the Sakhalin with Japanese xanthodryas); examinandus in Kamchatka, also often lumped with xanthodryas); xanthodryas in Japan; and kennicotti in Alaska. These subspecies are poorly differentiated morphologically, but some have distinctive songs. We analyzed variation in 1011 base-pairs of cytochrome b (cyt b) gene sequence of mitochondrial DNA from breeding populations in Siberia, Alaska, Kamchatka, Sakhalin and Japan (Hokkaido, Honshu, Shikoku and Kyushu). We recovered four clades: clade A from Kamchatka and Hokkaido; clade B from Siberia and Alaska; clade C from Sakhalin; and clade D from Honshu, Shikoku and Kyushu. Genetic distances between the four clades ranged between 2.3% and 5.8%. Assuming substitutional rates 2% per million years in the cyt b gene, the divergence between clades A and D was estimated at approximately three million years, far greater than those among typical subspecies. In contrast, samples of borealis and kennicotti were very similar genetically. We conclude that examinandus and xanthodryas might advisedly be treated as distinct species, because they have been separated from each other and from borealis for a considerable time, probably throughout the Pleistocene.

J Ornithol 147 suppl (2006)


Daniel

[Markus Lagerqvist]

Interesting! Does anyone have the full paper, would be nice to read!

[James Lowther]

looks like the abstract for a forthcoming presentation, therefore no full paper as yet??

[njlarsen]

A google search didn't turn up anything ...

Niels

[Frenchy]

And maybe next week someone else will repeat the study and come up with completely different results. Don't reach for the checklist just yet!

[l_raty]

Only two clades are apparent in the bar-code data from the database on http://www.barcodinglife.org/views/idrequest.php, but they have no birds from Japan or Sakhalin, so this is not too surprising.

The first clade includes birds from Norway, central Russia, Mongolia, and one bird from the Madagan area (N coast of the sea of Okhotsk). This would presumably fit "clade B".
The second clade includes a bird from Kamchatka and a second bird from Madagan. Based on the bird from Kamchatka, this could fit "clade A".
(A number of additional samples labeled simply "Russia" fall in both clades. Five samples from S Korea are also parted over the two clades, but I presume these were almost certainly migrants - in any case, my references do not include S Korea in the breeding range of the species. Intriguingly, the three Korean samples that fall in the first clade [and thus should be borealis or kennicotti] are all labelled explicitly "Ph. b. xanthodryas", while the two that fall in the second clade [and thus should probably be examinandus = xanthodryas sensu lato] are left without subspecific ID.)

The genetic distance could certainly be compatible with a species limit, although "distance-alone" arguments to split Phylloscopus spp. are always a bit weakened by the present treatment of Greenish Warbler, I feel. (The western and eastern clades of this species appear [just a bit] more divergent than the Arctic Warbler clades, but these are said to intergrade smoothly and for this reason are virtually never split [see http://www.actazool.org/downloadpdf.asp?id=5099 though]. Thus whether this distance is sufficient to ensure reproductive isolation in this genus may be questionable.)

Whether we are on secure grounds with the application of the names should also probably be checked before taking things any further. (Although perhaps it has already been done...?) Hylebatus, xanthodryas and examinandus are all based on migrant/wintering birds - with taxa similar enough that they are often synonymized, this is probably a situation in which a mistake could easily have happened at some point. If hylebatus really applies to the birds of Ussuriland, it would also be particularly important to have these sampled, because this name would have priority over xanthodryas and examinandus. (BWP includes this population in xanthodryas, suggesting the name hylebatus is best not used because it is based on migrants from China (and thus presumably doubtfully identifiable). But so is xanthodryas, too - in fact, both names were introduced by the same author (Swinhoe) and have the same type locality: Amoy (=Xiamen), on the Taiwan strait, hence I'm unclear why one was deemed better than the other. Examinandus is based on birds from Bali.)

[Daniel Philippe]

Thanks for shedding some light on this secretive little brown job, Laurent, even if for me the bar-code data in the database do not appear.

From the IOC abstract, I do not feel the case of the “arctic” is similar to the “greenish” though. I mean the “arctic” does not look like a ring species as birds from northern Europe seem genetically similar to those from Alaska. The significant genetic distances pertain to the isolated populations (Kamchatka & Japan).

On the other hand if the names of the different taxa are mixed up, we may misunderstand their respective migration routes then. However do we have a clue whether distinct populations have different migratory timing (they are seen in South-east Asia from August to April) ?

[l_raty]

Originally Posted by Daniel Philippe

From the IOC abstract, I do not feel the case of the “arctic” is similar to the “greenish” though. I mean the “arctic” does not look like a ring species as birds from northern Europe seem genetically similar to those of Alaska. The significant genetic distances pertain to the isolated populations (Kamchatka & Japan).

The Japanese population is geographically isolated, but that of Kamchatka is arguably not - or at least not completely.
Also, the Ussuriland populations (apparently not sampled by Saitoh et al.) is regularly grouped taxonomically with that of Sakhalin (e.g., hylebatus as described in the abstract), or even treated as part of a broader group also including the populations of Japan and Kamchatka (e.g., xanthodryas as recognized in BWP), thus this population could at least conceivably be part of one of the differentiated haplogroups. This population is presumably not geographically isolated either.

I'm not saying the case is very similar to that of greenish warblers as a whole, of course - indeed, this does not look like a ring species, and if Ticehurst's hypothesis about the creation of the greenish warbler population ring is correct, the processes that led to the two situations may have been very different.
But I think it still remains that, with greenish warblers, we have an example of a closely-related species in which similar genetic distances exist between populations, while gene flow remains (locally) strong enough, and the variation in other characters remains smooth enough, that almost nobody regards them as different species.
In this context, I just don't feel fully comfortable with splitting based on data that would not fully address the possibility that such contact zones could exist between haplogroups.

L -

[Richard Klim]

Originally Posted by Frenchy

And maybe next week someone else will repeat the study and come up with completely different results. Don't reach for the checklist just yet!

A posting on the OrientalBirding mailing list today reminded me of the recent publication of this paper, also suggesting the recognition of Phylloscopus xanthodryas, as 'Pacific Warbler':

• Reeves, Drovetski & Fadeev 2008. Mitochondrial DNA data imply a stepping-stone colonization of Beringia by arctic warbler Phylloscopus borealis. J Avian Biol 39(5).

[Thanks to Max Berlijn for original notification.]

Richard

[l_raty]

Originally Posted by Richard Klim

A posting on the OrientalBirding mailing list today reminded me of the recent publication of this paper, also suggesting the recognition of Phylloscopus xanthodryas 'Pacific Warbler':

• Reeves, Drovetski & Fadeev 2008. Mitochondrial DNA data imply a stepping-stone colonization of Beringia by arctic warbler Phylloscopus borealis. J Avian Biol 39(5).

I'm a bit disturbed by the fact that the results of this study apparently contradict strongly those of the Saitoh et al. study originally cited by Daniel.
Saitoh et al.: "We recovered four clades: clade A from Kamchatka and Hokkaido; clade B from Siberia and Alaska; clade C from Sakhalin; and clade D from Honshu, Shikoku and Kyushu."
Reeves et al.: "[...] it revealed two divergent (3.8% ML-corrected divergence) clades strongly supported by bootstrap (100%). One of these clades corresponded to birds sampled on Sakhalin Island and Kamchatka Peninsula."

According to Saitoh et al., the birds of Sakhalin would be quite divergent from those of Kamchatka. In the results of Reeves et al., they are clearly not... (Sakhalin and Kamchatka share identical haplotypes.)

Which study should we believe?

[Richard Klim]

As is often the case, the lack of direct correspondence between clades and traditional subspecies, and differing views on subspecies distributions, makes this a confusing subject.

Saitoh et al consider that examinandus traditionally refers to the Kamchatka population. They conclude that xanthodryas and examinandus should each possibly be treated as species distinct from borealis based on the genetic distance between their clades A (Kamchatka & Hokkaido – P (b) examinandus) and D (Honshu, Shikoku & Kyushu – P (b) xanthodryas). Although not clear from the abstract, presumably their suggested P (b) examinandus must also include clade C (Sakhalin) given its geographical relationship, despite a genetic distance of at least 2.3%.

This arrangement does not fully align with Clement 2006 (HBW11) which restricts the range of 'examinandus' to
S Kurils & Japan, and leaves the Kamchatka population within xanthodryas (sensu strictu). Given that xanthodryas and examinandus were collected in Amoy (Xiamen) and Bali respectively, it is unclear on what basis these differing associations with supposed breeding ranges were made. It certainly doesn’t help to resolve the conflict between Kamchatka vs Kurils/Japan in the case of examinandus.

A further potential confusion is the probably incorrect inclusion of the Chukotka population within xanthodryas (sensu lato) by Clement 2006 & H&M3 (contra BWP, Baker 1997, Clements 2007 & Saitoh et al).

It’s unfortunate that Reeves et al did not include any specimens from the range of Saitoh et al’s clade D (ie their P (b) xanthodryas). As far as I can see, P xanthodryas [Reeves et al] = P (b) examinandus [Saitoh et al].

Richard

[l_raty]

Originally Posted by Richard Klim

Although not clear from the abstract, presumably their suggested P (b) examinandus must also include clade C (Sakhalin) given its geographical relationship, despite a genetic distance of at least 2.3%.

If so, then the sequence in which they listed their clades is quite surprising - usually sister clades are placed next to each other. (They will always unavoidably appear next to each other in a tree, thus if you label groups on a tree, you don't even have to think about it - this simply comes naturally.) Here they listed their Sakhalin clade between their Siberian/Alaskan and their S Japan clade - this really does not suggest that it was sister to their Kamchatka clade.
Geographically, Sakhalin is also arguably close to the continent...

Plus, there is no reciprocal monophyly between the Kamchatka and Sakhalin birds in the result of Reeves et al., and the divergence levels within this group are smaller than those within the Siberian clade. It would be quite surprising that the cyt-b could suggest reciprocal monophyly, with a 2.3% divergence, for the same birds.

[Richard Klim]

Originally Posted by l_raty

If so, then the sequence in which they listed their clades is quite surprising - usually sister clades are placed next to each other. (They will always unavoidably appear next to each other in a tree, thus if you label groups on a tree, you don't even have to think about it - this simply comes naturally.) Here they listed their Sakhalin clade between their Siberian/Alaskan and their S Japan clade - this really does not suggest that it was sister to their Kamchatka clade.
Geographically, Sakhalin is also arguably close to the continent...

Plus, there is no reciprocal monophyly between the Kamchatka and Sakhalin birds in the result of Reeves et al., and the divergence levels within this group are smaller than those within the Siberian clade. It would be quite surprising that the cyt-b could suggest reciprocal monophyly, with a 2.3% divergence, for the same birds.

I was just trying to make sense of the fact that Saitoh et al have identified 4 clades but probably only 3 potential species.

It is clear that clades A, B & D correspond to their potential species examinandus, borealis & xanthodryas respectively.

If clade C is assumed to be included within P (b) xanthodryas rather than within P (b) examinandus as I suggested, then we are left with the Hokkaido population (clade A, examinandus) directly between two populations of xanthodryas (clade C on Sakhalin and clade D on Honshu etc), which seems rather unlikely...???

Richard

[PS: Sorry Laurent, I must be getting too tired! Presumably you're suggesting that clade C (Sakhalin) is included within mainland P (b) borealis, which does indeed make more sense.
I was attaching too much importance to the fact that BWP, H&M3, Clement 2006 & Clements 2007 all included the Sakhalin population within the traditional definition of xanthodryas (sl).]

[Scott Atkinson]

To all:

Have been following the Pacific Warbler thread with great interest and just wanted to chime in with a (hopefully) slightly new perspective. First, you may be interested in Mark Brazil's treatment of the Arctic Warbler in his Birds of East Asia released this year. He singles out examinandrus of the Kamchatka Peninsula and n. Kurile Islands for commentary, noting clear difference in song from other forms. He also cites plumage differences, which have been documented for a longer period of time in both Russian and English-language source material.

I spent very considerable time in the RFE (Kamchatka, s. Sakhalin, southern mainland) in the 1990s and though on extended business travel, birded a lot. I encountered examinandrus, xanthodryas and hylebata respectively, as well as nominate on mainland in migration. Kennicottii I encountered once on migration through s. Sakhalin. Xanthodryas on s. Sakhalin was chiefly a migrant, but the songs I do recall were reminiscent of mainland birds.

My experience for examinandrus corraborated Brazil's entry. Examinandrus had a very distinctly different song from other forms--a complex warble with variation in pitch, longer and more varied than songs of other forms. Tone quality seemed like other forms, as did calls. Around Petropavlosk-Kamchatskii at s.e. Kamchatka, this was a very common bird, and one could hear it singing from spring until as late as mid-August, along Stone Birch-laden hillsides, even right into town.

What is unclear to me are the range limits of this taxon. Beyond Kamchatka and the n. Kuriles, Brazil mentions Chukotka also for it, which doesn't square with earlier Russian-language material, yet I believe a recent Russia entry is the source of his mention of Chukotka. I believe Dementev's 5-volume series was the first to outline range of examinandrus, but some Russian authors don't mention it at all, apparently lumping with xanthodryas. Seems it would be worthwhile to determine if there are zones of integradation with other forms in the n. Kuriles and n. Kamchatka; although I have plenty of earlier Russian-language sources, I don't have knowledge of more recent material on this matter. I would appreciate it if anyone aware of recent material would bring it to my attention.

Taken with the fascinating clade study, it would seem that at least examinandrus is distinct from other P. borealis forms.

Best regards,

Scott Atkinson

[Richard Klim]

Originally Posted by Scott Atkinson

...you may be interested in Mark Brazil's treatment of the Arctic Warbler in his Birds of East Asia released this year. He singles out examinandrus of the Kamchatka Peninsula and n. Kurile Islands for commentary, noting clear difference in song from other forms.

As far as I can see, Brazil 2009 only describes differences in the song of xanthodryas (sensu lato) cf borealis & kennicotti, but doesn't mention any differences in the song of 'examinandus' per se.

Richard

[petemorris]

I have just come back from Japan where we noted several individuals in the montane areas of Honshu. Their songs are clearly different from Eurasian birds. In limited time available (I was leading a tour and groups have limited interest and patience for such activities) I tried some very basic playback experiments on a series of five individuals. All five showed no response to pre-recorded songs from Schulze's WP collection, but all five showed strong responses to playback of another individual of their population recorded that day. Only that individual showed any interest in the Schulze recording, and that was after it had already been wound up by playback. My understanding from Japanese friends is that the birds on Hokkaido (which I have never seen or heard) sing very differently to those on Honshu.

Cheers

Pete

[Richard Klim]

283 Wild Bird Songs of Japan (Hideo Ueda, 1998, 3 CDs) provides an interesting direct comparison of the vocalisations of xanthodryas and borealis:

• xanthodryas (recorded in Shizuoka Prefecture, Honshu) has a simple song "chori-chori, chori-chori, chori-chori", and "ju, ju" call notes;
  
• borealis (described as migrants recorded in Aomori Prefecture, N Honshu) has a different "chori-chi, chori-chi, chori-chi" song, and harsh "chichi" call notes.

Richard

[recbirds]

A few thoughts from some other recordings.

From Veprintsev's LPs "Birds of the Soviet Union:A Sound Guide" there are recordings attributed to borealis and xanthodryas. The borealis recordings are from Sayan Mts, Tuva (sound like European birds as presented on the Schulze CDs, ie. very rapid song) and from Kril'on, Sakhalin (sound like Ueda's borealis, ie. "chori-chi, chori-chi"). The xanthodryas recording is from Petropavlosk-Kamchatskiy, Kamchatka and sounds similar to the Sakhalin recording, but is slightly higher-pitched sounding 'thinner'. This suggests that the birds called borealis on Ueda's CDs are not borealis (in the sense of Western birds) and that Japanese breeding birds (? xanthodryas) are not the same as those from further north (ie. Sakhalin and Kamchatka = ? examinandus). BTW recordings from Alaska (kennicotti) have the melodic "chori-chori" song but the pace is in between that of the rapid Western borealis and slow Japanese 'xanthodryas'.


Shaun Peters

http://www.aviandiscography.webs.com

[johnallcock]

Here in Hong Kong, Arctic Warbler is a reasonably common passage migrant in both spring and autumn. Reading the above it seems that any (or even all?) of the discussed taxa could occur in this area.

Are there any features which can be reliably used to separate the various taxa either in the field or in the hand? Song differences are not useful (the species doesn't regularly sing here) but how about calls? Wing formula? Any other features?

[Richard Klim]

Originally Posted by johnallcock

Are there any features which can be reliably used to separate the various taxa either in the field or in the hand?

Brazil 2009 (Birds of East Asia):

• borealis has browner upperparts, pale brown underparts, faintly yellow-washed flanks, and long, slender yellow-white supercilium; slightly finer bill than xanthodryas.
  
• kennicotti as nominate but bill finer, shorter and less broad-based, slightly brighter green upperparts and brighter yellow underparts.
  
• xanthodryas brighter green above, has broader, yellower wingbars and the yellowest underparts (especially flanks), whiter on belly.
  
• examinandus slightly larger than xanthodryas with heavier bill and less yellowish underparts. Those in NE Hokkaido and Sakhalin are slightly smaller than in Kamchatka.

Baker 1997 (Warblers of Europe, Asia and North Africa):

• kennicotti like nominate, but has finer and less broad-based bill and yellower underparts.
  
• xanthodryas greener above than nominate, and averages slightly more yellow and greyish below. The supercilium and cheeks tend to be yellower. Broader bill base. In fresh plumage it is one of the easier subspecies to identify, but in worn plumage, as with other forms, separation becomes very difficult, though larger size may help to identify birds in the hand.

[See post #17 for possible differences in call.]

All in all, I suspect that reliable field identification away from the breeding grounds could be rather tricky...

Richard

[Richard Klim]

Alström, Saitoh, Williams, Nishiumi, Shigeta, Ueda, Irestedt, Björklund & Olsson 2011. The Arctic Warbler Phylloscopus borealis – three anciently separated cryptic species revealed. Ibis 153(2): 395-410.
http://onlinelibrary.wiley.com/doi/1...116.x/abstract
http://onlinelibrary.wiley.com/doi/1...116.x/suppinfo

(pdf will presumably soon be posted at SMNH.)

[See also www.birdforum.net/showthread.php?t=182939]

Richard

[Mysticete]

I keep kicking myself for not really knowing the distribution of these forms last summer. I knew about Japanese Leaf Warbler potentially being split off, but not that the Hokkaido represented a different species :(

[Richard Klim]

Birdwatch Listcheck article by David Callahan:
http://www.birdwatch.co.uk/categorie...c=147&item=741

[Xenospiza]

Originally Posted by Mysticete

I keep kicking myself for not really knowing the distribution of these forms last summer. I knew about Japanese Leaf Warbler potentially being split off, but not that the Hokkaido represented a different species :(

Your the second person I know who's suffered an armchair dip with these while being aware!

[Richard Klim]

Splits of Phylloscopus examinandus Kamchatka Leaf Warbler & P xanthodryas Japanese Leaf Warbler accepted for v2.8 (Draft), following Alström et al 2011.
www.worldbirdnames.org/updates-spp.html