[Daniel Philippe]

Still Spizella arborea, but it seems that it is closer to Passerella than to the other Spizella. See for instance Carson & al. 2003 (A phylogenetic analysis of the emberizid sparrows based on three mitochondrial genes).

Is there a new genus name for this species ?

Daniel

[AlexC]

Can you cite any other references supporting this? From what I've read, reliance on mtDNA to determine close sister taxa is iffy at best.

[Daniel Philippe]

These ref are in the Carson & al. article:

Mayer and Short (1970). I do not have the paper.

Zink and Dittmann (1993): http://elibrary.unm.edu/sora/Wilson/...0399-p0413.pdf

Dodge and al. (1995): http://elibrary.unm.edu/sora/Wilson/...0641-p0654.pdf

Patten and Fugate (1998): http://elibrary.unm.edu/sora/Auk/v11...0412-p0424.pdf

Daniel

[l_raty]

Originally Posted by Daniel Philippe

Still Spizella arborea, but it seems that it is closer to Passerella than to the other Spizella. See for instance Carson & al. 2003 (A phylogenetic analysis of the emberizid sparrows based on three mitochondrial genes).

Is there a new genus name for this species ?

Not as far as I know. (The only synonym of Spizella I know is Spinites Cabanis, 1851, but this was proposed as a replacement name, hence is attached to the same type species.) Spizella arborea could of course be transferred to Passerella, but so far their sister-group relationship is indeed "iffy at best", so there would be no guarantee that the move remains viable in the long term.

American sparrows and relatives (New World Emberizini) are strongly oversplit at the generic level, IMHO, particularly in comparison to their putative Old World sister group. (There is about as much variability in Emberiza alone as in the 25-or-so recognised genera of NW Emberizini.)
And several recent proposals, instead of reducing this problem, will worsen it (from both sides) if they become accepted.

L -

[l_raty]

Originally Posted by AlexC

From what I've read, reliance on mtDNA to determine close sister taxa is iffy at best.

Reliance on a single gene is always a bit iffy for closely related taxa - or, as a more general rule, to resolve relationships in a part of the tree where internodal distances are short. (In fact, some potential problems, such as getting a gene tree different from the real taxon tree due to a retention of ancestral polymorphism followed by stochastic lineage sorting, may even be more likely to occur with a nuclear gene.)

But anyway, in Carson & Spicer's analyses, Spizella arborea does not appear closely related to any other Spizella spp at all - it is part of a group that also includes Passerella, Junco and Zonotrichia, and that is itself related to two other groups, one made of Pipilo and Aimophila, the other including some Ammodramus (but not the type of this genus), Pooecetes, Amphispiza belli (but not A. bilineata), Passerculus and Melospiza. All the other Spizella spp (including S. pusilla, the type of the genus) are in a still more distant group.
So you might indeed use this argument to cast doubt over an assumed sister-taxon relationship between Spizella arborea and the Fox Sparrow; but it would remain more than unlikely that its sister taxon is to be found among other Spizella.

L -

[Xenospiza]

Originally Posted by l_raty

American sparrows and relatives (New World Emberizini) are strongly oversplit at the generic level, IMHO, particularly in comparison to their putative Old World sister group. (There is about as much variability in Emberiza alone as in the 25-or-so recognised genera of NW Emberizini.)
And several recent proposals, instead of reducing this problem, will worsen it (from both sides) if they become accepted.

And the few monotypic genera that are recognized within the Old World Emberizini (Melophus, Miliaria, Latoucheornis) don't hold up to scrutiny either. However, the African buntings (including House/Striolated Bunting) may well be put into a separate genus (is Fringillaria available?)
(Lapland Bunting (Longspur), Snow Bunting and a few American relatives are no Buntings or American Sparrows, but a separate (old?) group within the huge Emberizidae).

[l_raty]

Originally Posted by Xenospiza

And the few monotypic genera that are recognized within the Old World Emberizini (Melophus, Miliaria, Latoucheornis) don't hold up to scrutiny either.

Latoucheornis is interesting. The single species included in this genus was described in the genus Junco - the male being strikingly similar to a male Slate-colored Junco. But their similarities do not stop there. The genetic distance separating Junco and Latoucheornis from their respective closest, more "typical emberizine" relative (Zonotrichia in the first case; Emberiza (Cristemberiza) elegans in the second - the main ref is http://dx.doi.org/10.1016/j.ympev.2007.12.007) is almost exactly the same. If there is a biological reason to keep one of them in a separate genus, it almost certainly also applies to the other.

The main difference is that in the case of Latoucheornis, a generic lump results in transferring one single species to Emberiza - this is the traditional catch-all genus, so we don't feel "hurt" by the move. Junco has priority over the generic names associated to all of the species that are parts of its group, so a lump would result in a sudden broadening of a genus that is perceived as "special", and that in addition is identical to the vernacular name of a more restricted group...

However, the African buntings (including House/Striolated Bunting) may well be put into a separate genus (is Fringillaria available?)

Think so. (With E. capensis as its type.) But even then, Emberiza would still be much more variable than any New World Emberizini genus. And the basal position of this group rests on a single gene (ODC)...

L -

[Gentoo]

Originally Posted by Xenospiza

And the few monotypic genera that are recognized within the Old World Emberizini (Melophus, Miliaria, Latoucheornis) don't hold up to scrutiny either. However, the African buntings (including House/Striolated Bunting) may well be put into a separate genus (is Fringillaria available?)
(Lapland Bunting (Longspur), Snow Bunting and a few American relatives are no Buntings or American Sparrows, but a separate (old?) group within the huge Emberizidae).

Do you mean birds like the Lark Bunting, Dickcisell and others?

[ntbirdman]

Originally Posted by Gentoo

Do you mean birds like the Lark Bunting, Dickcisell and others?

Plectrophenax and Calcarius (Snow Bunting, McKay's Bunting, and the four Longspurs) form a clade basal to all of the new world nine-primaried oscine radiation. See Jonsson and Fjeldsa's Oscine Supertree paper (PM me if you want a pdf copy) and the papers they cite.

~ Nick

[Xenospiza]

The Lark Bunting is (not surprisingly) an “American Sparrow” (I'm always surprised how this prairie species mirrors the steppe-dwelling Black Lark in its colors both as male and female), but the Dickcissel is a Cardinal (related to the Indigo Bunting etc.)!

It is not yet certain that the "Longspurs and Snow Buntings" are really basal to the "nine-primary assemblage" – they may also be a separate group (which would be a family if you think the warblers, cardinals, tanagers and orioles are... I guess lumping them all in one big family as the AOU once did is a lot smarter).

[Gentoo]

Originally Posted by Xenospiza

The Lark Bunting is (not surprisingly) an “American Sparrow” (I'm always surprised how this prairie species mirrors the steppe-dwelling Black Lark in its colors both as male and female), but the Dickcissel is a Cardinal (related to the Indigo Bunting etc.)!

It is not yet certain that the "Longspurs and Snow Buntings" are really basal to the "nine-primary assemblage" – they may also be a separate group (which would be a family if you think the warblers, cardinals, tanagers and orioles are... I guess lumping them all in one big family as the AOU once did is a lot smarter).

Thanks, I never kept track of what they did with the Dickcisell once they resplit that catch-all mess of a family.

The Lark Bunting is an interesting bird to say the least.

[ntbirdman]

Originally Posted by Xenospiza

It is not yet certain that the "Longspurs and Snow Buntings" are really basal to the "nine-primary assemblage" – they may also be a separate group (which would be a family if you think the warblers, cardinals, tanagers and orioles are... I guess lumping them all in one big family as the AOU once did is a lot smarter).

Help me out here, why are people resistant to more families rather than less? Why is lumping a whole bunch of things under one name smarter than recognizing all of these diverse groups? I hear that sort of opinion a lot (for any taxonomic level) and I guess I don't really understand it.

Thanks,
Nick

[njlarsen]

Originally Posted by ntbirdman

Help me out here, why are people resistant to more families rather than less? Why is lumping a whole bunch of things under one name smarter than recognizing all of these diverse groups? I hear that sort of opinion a lot (for any taxonomic level) and I guess I don't really understand it.

Thanks,
Nick

I probably should not be the one writing an answer to this because I am not sure I agree with all the arguments for keeping things together, but I can retell what I read in a recent paper on thrushes: There was arguments in a DNA analysis for several rather deep clades within the thrushes currently in genus Turdus. Basal to them all was one species, which happened to be the type species for Turdus. What that meant was, that if the genus was split, the large genus of Turdus would then become a one-member genus, with e.g., European Blackbird belonging to Merula if I remember correctly; this was seen as a too radical change by the authors (again if I remember correctly). Additionally, several of the groupings within Turdus was very close to each others, with much larger interspecies distances in other groups: the former groups were therefore seen as superspecies, because if you looked at those as genera, it would necessitate a splitting of the clades into several genera each. I am personally not sure that the mere fact that Turdus would be reduced from a lot of species to only one would be a strong enough argument for status quo; the latter argument against necessarily allowing superspecies to become a genus seems a much stronger one.

I hope I did not hopelessly mangle this up, my copy of the Turdus paper is in my office, not here at home.

One observation that can be made in all of this is that the average number of species per genus is about three in the hummingbirds, while the genus Turdus contains sixty-odd species (about 66). This raise the suspicion that not all orders have been treated equally so far (and probably never will be ).

Cheers
Niels

[Gentoo]

Originally Posted by njlarsen

I probably should not be the one writing an answer to this because I am not sure I agree with all the arguments for keeping things together, but I can retell what I read in a recent paper on thrushes: There was arguments in a DNA analysis for several rather deep clades within the thrushes currently in genus Turdus. Basal to them all was one species, which happened to be the type species for Turdus. What that meant was, that if the genus was split, the large genus of Turdus would then become a one-member genus, with e.g., European Blackbird belonging to Merula if I remember correctly; this was seen as a too radical change by the authors (again if I remember correctly). Additionally, several of the groupings within Turdus was very close to each others, with much larger interspecies distances in other groups: the former groups were therefore seen as superspecies, because if you looked at those as genera, it would necessitate a splitting of the clades into several genera each. I am personally not sure that the mere fact that Turdus would be reduced from a lot of species to only one would be a strong enough argument for status quo; the latter argument against necessarily allowing superspecies to become a genus seems a much stronger one.

I hope I did not hopelessly mangle this up, my copy of the Turdus paper is in my office, not here at home.

One observation that can be made in all of this is that the average number of species per genus is about three in the hummingbirds, while the genus Turdus contains sixty-odd species (about 66). This raise the suspicion that not all orders have been treated equally so far (and probably never will be ).

Cheers
Niels

This make sense to me. Didn't a similar thing happen with finches? Early on, most were in the genus Fringilla until they were all split later, even into two subfamilies. with now only three species remaining in that genus, or four if they split the Chaffinch like I heard they were going to.

[l_raty]

Originally Posted by Gentoo

Didn't a similar thing happen with finches? Early on, most were in the genus Fringilla until they were all split later, even into two subfamilies. with now only three species remaining in that genus, or four if they split the Chaffinch like I heard they were going to.

Actually, the largest finch genus in Linnaeus' 1758 work was Loxia...

(But, granted, Fringilla was close.)

L -

[Gentoo]

Originally Posted by l_raty

Actually, the largest finch genus in Linnaeus' 1758 work was Loxia...

(But, granted, Fringilla was close.)

L -

I think I remember that now. Weren't a lot of the Carduelis (acanthis and a few others) placed in that genus as well?