[ntbirdman]

This was mentioned in another post but deserves its own discussion thread:

Going to Extremes: Contrasting Rates of Diversification in a Recent Radiation of New World Passerine Birds
F. Keith Barker, Kevin J. Burns, John Klicka, Scott M. Lanyon and Irby J. Lovette

abstract:
http://sysbio.oxfordjournals.org.pro...ys094.abstract

Proposes splitting this group into 15 families, some of them wholly new:

Superfamily Emberizoidea

Family Calcariidae Ridgway, 1901; Genera: Calcarius, Plectrophenax, Rhynchophanes

Family Rhodinocichlidae Ridgway, 1902; Genera: Rhodinocichla

Family Emberizidae Vigors, 1825; Genera: Emberiza, Latoucheornis, Melophus, Miliaria

Family Passerellidae Cabanis and Heine, 1850-51; Genera: Aimophila, Ammodramus, Amphispiza, Arremon, Arremonops, Atlapetes, Calamospiza, Chlorospingus, Chondestes, Junco, Lysurus, Melospiza, Melozone, Oreothraupis, Oriturus, Passerculus, Passerella, Pezopetes, Pipilo, Pooecetes, Pselliophorus, Spizella, Torreornis, Xenospiza, Zonotrichia

Family Spindalidae (new family); Type genus: Spindalis; Diagnosis: Containing a single genus, this family is diagnosed by the generic characters of Spindalis (Jardine and Selby 1837); Genus: Spindalis

Family Nesospingidae (new family); Type genus: Nesospingus; Diagnosis: Containing a single genus, this family is diagnosed by the generic characters of Nesospingus (Sclater 1885); Genus: Nesospingus

Family Phaenicophilidae Sclater, 1886; Diagnosis: This family was originally erected for the genus Phaenicophilus alone. However, the two species in this genus share an olive back, wings and tail, gray underparts, and a broken white eye ring with both Xenoligea and Microligea; Genera: Phaenicophilus, Xenoligea, Microligea

Family Zeledoniidae Ridgway, 1907; Genus: Zeledonia

Family Teretistridae Baird, 1864; Genus: Teretistris

Family Parulidae Wetmore et al., 1947; Genera: Seiurus, Helmitheros, Parkesia, Vermivora, Mniotilta, Protonotaria, Limnothlypis, Oreothlypis, Leucopeza, Oporornis, Geothlypis, Catharopeza, Setophaga, Myiothlypis, Basileuterus, Cardellina, Myioborus (see Lovette et al. 2010; Chesser et al. 2011).

Family Icteriidae Baird, 1858; Genus: Icteria

Family Icteridae Vigors, 1825; Genera: Agelaioides, Agelaius, Agelasticus, Amblycercus, Amblyramphus, Cacicus, Chrysomus, Curaeus, Dives, Dolichonyx, Euphagus, Gnorimopsar, Gymnomystax, Hypopyrrhus, Icteria, Icterus, Lampropsar, Macroagelaius, Molothrus, Nesopsar, Ocyalus, Oreopsar, Psarocolius (including Gymnostinops), Pseudoleistes, Quiscalus, Sturnella (including Leistes), Xanthocephalus

Family Calyptophilidae Ridgway, 1907; Genus: Calyptophilus

Family Mitrospingidae (new family); Type genus: Mitrospingus; Diagnosis: We know of no morphological characters that unite these three genera of South and southern Central America. In lieu of such characters, we list 13 unreversed molecular synapomorphies of the group, from four different genes. These changes include (numbered by their position in each gene alignment) CYTB: A627C, T798C, C801T, A1074G; ND2: C27T, T195C, C231T, C372A, G637A, C710T, T968C; ACO1-I9: C977T; RAG1: A1987G. Cladistically, we define this family as the descendants of the common ancestor of Mitrospingus cassinii and Lamprospiza melanoleuca; Genera: Lamprospiza, Mitrospingus, Orthogonys

Family Cardinalidae Ridgway, 1901; Genera: Amaurospiza, Cardinalis, Caryothraustes, Chlorothraupis, Cyanocompsa, Cyanoloxia, Granatellus, Guiraca, Habia, Passerina, Periporphyrus, Pheucticus, Piranga, Rhodothraupis, Spiza

Family Thraupidae Cabanis, 1847; Genera: Acanthidops, Anisognathus, Bangsia, Buthraupis, Calochaetes, Camarhynchus, Catamblyrhynchus, Catamenia, Certhidea, Charitospiza, Chlorochrysa, Chlorophanes, Chlorornis, Chrysothlypis, Cissopis, Cnemoscopus, Coereba, Compsothraupis, Conirostrum, Conothraupis, Coryphaspiza, Coryphospingus, Creurgops, Cyanerpes, Cyanicterus, Cypsnagra, Dacnis (including Pseudodacnis), Delothraupis, Diglossa (including Diglossopis), Diuca, Dolospingus, Donacospiza, Dubusia, Emberizoides, Embernagra, Eucometis, Euneornis, Geospiza, Gubernatrix, Haplospiza, Hemispingus, Hemithraupis, Heterospingus, Idiopsar, Incaspiza, Iridophanes, Iridosornis, Lanio, Lophospingus, Loxigilla, Loxipasser, Melanodera, Melanospiza, Melopyrrha, Nemosia, Neothraupis, Nephelornis, Nesospiza, Orchesticus, Oreomanes, Oryzoborus, Parkerthraustes, Paroaria, Phrygilus, Piezorhina, Pinaroloxias, Pipraeidea, Poospiza, Porphyrospiza, Pyrrhocoma, Ramphocelus, Rhodospingus, Rowettia, Saltator, Saltatricula, Schistochlamys, Sericossypha, Sicalis, Sporophila, Stephanophorus, Tachyphonus, Tangara, Tersina, Thlypopsis, Thraupis, Tiaris, Trichothraupis, Urothraupis, Volatinia, Wetmorethraupis, Xenodacnis, Xenospingus

Here is their taxonomic justification:

APPENDIX 1. A TAXONOMY OF NEW WORLD NINE-PRIMARIED OSCINES The New World nine-primaried oscines are traditionally classified in five families, with nothing to associate them with one another except adjacency within the linear order of passerines. Sibley and Monroe (1990), in an attempt to determine rank by genetic divergence, classified these five families as tribes within a single subfamily, the Emberizinae, within the family Fringillidae. As discussed in the main text, there is substantial evidence from molecular data for the existence of five clades corresponding to the traditional families. However, a number of genera appear to be more distantly related, raising the question of how those groups should be classified, and how the relatedness of all of them to one another should be reflected taxonomically. The sister group to this clade, comprising the chaffinches, goldfinches, honeycreepers, and allies, is currently recognized as a single family, the Fringillidae, by most taxonomies (American Ornithologists' Union 1998, Dickinson 2003). One possible treatment for the group under consideration would be to rank it as a family (the Emberizidae), and to rank lineages within it as subfamilies; however, to do so would overturn more than a century of taxonomic practice. Instead, we have chosen to minimize changes to higher-level avian classification and to continue to rank the lineages within this group as families. In addition, as a further effort to maintain stability, we have chosen to continue to recognize the five core lineages (Emberizidae, Cardinalidae, Thraupidae, Parulidae, and Icteridae) as families in accordance with universal practice (excepting Sibley and Monroe 1990). Perhaps unfortunately, given the constraint of naming only monophyletic groups, recognizing these five families requires that we recognize eleven additional families within this larger radiation. One of these—the Calcariidae—which includes the genera Calcarius and Plectrophenax, in addition to the recently resurrected genus Rhynchophanes, has already been recognized by the AOU (American Ornithologists' Union 1998). Another five—the Rhodinocichlidae, Zeledoniidae, Teretistridae, Icteriidae and Calyptophilidae—have also been previously recognized (Bock 1994) for their corresponding genera. Another family-level name, Phaenicophilidae, has previously been applied to that genus alone (Sclater 1886), but we suggest expanding its definition to include Xenoligea and Microligea. In addition to these previously-named groups, we propose the following new family names (see descriptions below): Spindalidae (genus Spindalis), Nesospingidae (genus Nesospingus), and Mitrospingidae (genera Mitrospingus, Lamprospiza and Orthogonys). These steps would triple the number of families in this diverse radiation of birds. More importantly, this would give formal recognition to the deep evolutionary history preserved in these unique lineages of birds, especially those in Caribbean habitats, some of which are threatened by habitat fragmentation and loss.

Finally, in addition to the five traditional core families and the nine families listed above, we recommend recognition of a separate family name for the New World sparrows. As discussed previously, although our data strongly support monophyly of the Old World buntings and New World sparrows, they also demonstrate significant conflict over the placement of these groups relative to one another, with mtDNA strongly supporting a monophyletic grouping of the two, and ACO1-I9 and combined analyses favoring their separation: analyses including additional gene regions will be used to address this conflict in more detail. For the sake of future taxonomic stability, and to recognize real biological differences between the two groups, we propose their separation and recognition of both at the family level. Aside from the obvious biogeographic difference (one group is exclusively New World, and the other is Old World with a single species breeding marginally in Alaska), these groups also differ in the frequency of sexual dichromatism (buntings are generally sexually dichromatic, sparrows monochromatic), and have often been seen as more closely allied to the genera Plectrophenax and Calcarius (Paynter 1970, Patten and Fugate 1998). Restriction of the family name Emberizidae to the Old World buntings (genus Emberiza and allies) requires application of another name to the sparrows. Several names for this group are already available, but the oldest is the Passerellidae (Cabanis and Heine 1850; see Bock 1994).

Here we recognize the primarily New World group of nine-primaried oscines that form a clade sister to the finch family Fringillidae as the superfamily Emberizoidea. We list the families in our currently preferred order, with the genera (as recognized in the taxonomy of Dickinson 2003, with emendations from AOU 1998 and Lovette et al. 2010) assigned to each listed (alphabetically with the exception of the Parulidae; ordering within other families will depend on better-sampled analyses of relationships in each, combined with standard sequencing conventions), although we recognize that substantial revision of generic limits will be necessary in the near future. Type designations and diagnoses are given for three new families, and a diagnosis is given for a fourth family not previously used for two of the genera we place within it.

[ntbirdman]

Discussion of the merits of this kind of family-level splitting probably belongs back in this current thread: http://www.birdforum.net/showthread.php?t=246207

I'll just say its good to finally see this work published, and finally see where a lot of these enigmatic taxa things fall out, regardless of taxonomic rank. I'll see if I can post a tree of the relationships later.

[Richard Klim]

Wow. Thanks for the summary, Nick. Very timely/topical!

[njlarsen]

Family Icteriidae Baird, 1858; Genus: Icteria

Family Icteridae Vigors, 1825; Genera:

Do I remember wrongly, or is there a rule somewhere in the Code that prohibits names that differ by this little to both be in use? (or is that only for species level taxa?)

Niels

[fugl]

Originally Posted by njlarsen

Do I remember wrongly, or is there a rule somewhere in the Code that prohibits names that differ by this little to both be in use?

If not, there certainly should be! For one thing, how is a mere mortal to remember which is which?

[Richard Klim]

Originally Posted by njlarsen

Do I remember wrongly, or is there a rule somewhere in the Code that prohibits names that differ by this little to both be in use? (or is that only for species level taxa?)

It seems to be acceptable for family-group names (Article 55.4), but not for species-group names (Article 58.14) - hopefully someone more knowledgeable can clarify/correct...

www.nhm.ac.uk/hosted-sites/iczn/code

[Snapdragyn]

Do they list any divergence dates? I'm curious to what degree this may be a matter of 'we can't overturn tradition by lumping, so we'll do it by splitting all over the place instead', vs. 'these clades diverged so long ago we can't see lumping them & thus obscuring such ancient lineages'.

[andrew147]

Snapdragyn, I think this bit of their "taxonomic justification" may provide the answer.

"One possible treatment for the group under consideration would be to rank it as a family (the Emberizidae), and to rank lineages within it as subfamilies; however, to do so would overturn more than a century of taxonomic practice".

Can't believe they missed Saltatoridae and Seiuridae

[Snapdragyn]

Oh, I did see that - hence my comment that they seem to have no problem overturning tradition by splitting, just not by lumping (even when the latter would require less overall change to 'traditional' classification). I'm just wondering if there's an additional justification not mentioned in the excerpt above.

[Mysticete]

Some of those oddball taxa have been frequently debated in the past, and have jumped families numerous times (Zeledonia, Yellow-breasted Chat).

I know as an American birder, I don't see any proposal that would result in loss of Parulidae and other familiar groups going over so well.

The original version of this (which I saw at AOU) had Yellow-breasted Chat as sister to Icteridae, and they just lumped it into that family. Would certainly make things easier :P

[James Jobling]

The family names Icteridae and Icteriidae are not homonyms and are valid under the ICZN (art. 55.4) (altho' to calm down those who don't know their Crax from their Crex the case could be referred to the Commission for a ruling, e.g. to replace Icteriidae with another name). Similarly, the genera Icterus and Icteria are not homonyms (art.56.2). Specific names are now more rigorously controlled (e.g. nigricinctus and nigrocinctus are deemed homonyms) (art. 58).

[alytothrix]

There are some viable taxonomic treatments in between the extremes of lumping everything into Emberizidae or accepting all of the authors' splits.

For example, Cardinalidae and Thraupidae form a highly supported (100% bootstrap) clade with "Mitrospingidae". Why not lump them into a single family? So many genera have been shuffled between these families already that it seems pointless to continue to treat them as distinct, and it would avoid creating a new family based solely on molecular "apomorphies". Calyptophilus could be wedged in there too, although the support for that is much weaker.

Similarly, Icteria and the Icteriidae form a highly supported clade (again 100% bootstrap) as Mysticete mentioned. Although the chat doesn't much look like a typical blackbird, icterids are already relatively heterogenous and I could see lumping them being a viable option. This would also prevent the hassles of counting "I"s.

[alytothrix]

Originally Posted by Snapdragyn

Do they list any divergence dates? I'm curious to what degree this may be a matter of 'we can't overturn tradition by lumping, so we'll do it by splitting all over the place instead', vs. 'these clades diverged so long ago we can't see lumping them & thus obscuring such ancient lineages'.

The divergence time analysis suggests all of their family-level clades split rapidly between 9-14 million years ago. This compares with a Fringillidae-Emberizoidea split of ~21 million years (this is a prior, rather than a result, of their analysis). These are definitely not ancient lineages.

[Paul Clapham]

"One possible treatment for the group under consideration would be to rank it as a family (the Emberizidae), and to rank lineages within it as subfamilies; however, to do so would overturn more than a century of taxonomic practice".

Hardly a "century" -- the AOU had a giant Emberizidae family (with subfamilies Parulinae, Coerebinae, Thraupinae, Cardinalinae, Emberizinae, and Icterinae) which they only split in 1998 with the 7th edition.

[Mysticete]

My understanding however that the big Emerizidae "Lump" it itself fairly recent (80's?), so there is probably less historic inertia for keeping them together.

At this point, any proposal to "merge" Cardinalidae and Thraupidae is probably too late. That proposal should have come before Piranga and kind was moved. Merging the two families (or merging a bunch of them) would reverse a good chunk of taxonomic changes SACC and NACC have completed in the last decade. Note that most of these results would have already been known by AOU committee members. I in fact saw the talk that outlined most of these taxonomic changes by the authors something like 6 years ago at the Wyoming AOU meeting.

[Mysticete]

Also, from reading the paper, I think Icteriidae was kept out because, despite sampling a but load of genes, Where Icteriidae fell out varied considerably based on the genes used and the method of analysis. I suspect Icteriidae was created because of that uncertainty. In the AOU talk from 6 years ago that I saw, the initial idea was to "lump" Icteriidae into Icteridae.

[Snapdragyn]

Originally Posted by alytothrix

The divergence time analysis suggests all of their family-level clades split rapidly between 9-14 million years ago. This compares with a Fringillidae-Emberizoidea split of ~21 million years (this is a prior, rather than a result, of their analysis). These are definitely not ancient lineages.

Oh good grief. These are piddling divergences, on par w/ some tribes! Well, that answers what to do for my personal list then - all one family!

[alytothrix]

It certainly isn't "too late" for lumping the Cardinalidae and Thraupidae. My understanding is that a sister relation between the two was not strongly supported at the time of the previous moves, while it is now. Previous classifications also didn't involve creating a new family based on molecular characters alone, and I hope doing so would be a concern to committee members.

I think the variable position of Icteria (and indeed all of the new "families" in the parulid-icterid clade) is a concern, but the evidence from the concatenated dataset is as strong as that for many previous changes. I personally don't mind monotypic families, but for those who dislike the extensive splitting of this study this is certainly somewhere to look.

[MichaelRetter]

Icteria is listed above in both Icteriidae and Icteridae. Which does the paper propose?

[Richard Klim]

Originally Posted by MichaelRetter

Icteria is listed above in both Icteriidae and Icteridae. Which does the paper propose?

It's listed in both in the advance access paper, but presumably mistakenly in the latter case.

[MJB]

Originally Posted by alytothrix

The divergence time analysis suggests all of their family-level clades split rapidly between 9-14 million years ago. This compares with a Fringillidae-Emberizoidea split of ~21 million years (this is a prior, rather than a result, of their analysis). These are definitely not ancient lineages.

To the sizeable minority who align with the calculations of Bishop Ussher (and others of that time), these lineages certainly are unacceptably ancient!
MJB

[fbarker]

Originally Posted by Richard Klim

It's listed in both in the advance access paper, but presumably mistakenly in the latter case.

Yeah, Icteria was moved to Icteriidae due to lack of resolution as sister group ti Icteridae, but the lead author failed to delete it from Icteridae. It will be fixed in proof.

Cheers

[pianoman]

I can't think of a good reason for splitting Old World Buntings and New World Sparrows. Buntings differ from each other at least as much as any of them differ from NW sparrows. So a Yellowhammer and Rock Bunting would be in the same family, but Rock Bunting and White-crowned Sparrow in different families?

[njlarsen]

fbarker, welcome to birdforum!

Niels

[fbarker]

Originally Posted by pianoman

I can't think of a good reason for splitting Old World Buntings and New World Sparrows. Buntings differ from each other at least as much as any of them differ from NW sparrows. So a Yellowhammer and Rock Bunting would be in the same family, but Rock Bunting and White-crowned Sparrow in different families?

That argument hinges on a "degree of difference" criterion in determining taxonomic rank. Given that ranking of taxa is an art rather than a science, opinions on the validity of that approach will differ. However, evidence is mounting that buntings and sparrows are not sister taxa, and I don't think you can argue that yellowhammers and rock buntings have less in common than white-crowned sparrows and red-legged honeycreepers. The only way to keep them in the same family would be to bung all the Emberizoidea (warblers, tanagers, sparrows, buntings, blackbirds, etc.) into one family (Emberizidae), as others on this thread have suggested. I wouldn't be surprised to see some taxonomies taking that route, in which case all of the families listed above become subfamilies (still family-level groups governed by the Code).

However, taxonomies are tools that gain at least some of their usefulness from stability. From that perspective, it makes sense to keep recognizing the five families New World ornithologists have been using for many decades, and rearrange the rest to fit, as done here.