Proposal (633) to South American Classification Committee
Modify linear sequence of genera and species in Emberizidae
Effect on SACC: This would modify the linear sequence of genera and species to reflect the recently published phylogeny of the family by Klicka et al. (2014).
Background: The current sequence is listed below. It is based largely on historical momentum and reflects the transfer of many genera to the Thraupidae and the transfer of Chlorospingus to the family.
New information: Klicka et al. (2014) produced the first comprehensive phylogeny for the family, which they further restricted to New World genera under the name Passerellidae. They found that the Emberiza buntings and relatives are not the sister group to the New World sparrows (but that should be the topic of a separate proposal, likely to go to NACC first). The Klicka et al. phylogeny has exceptionally strong taxon sampling (at least for mtDNA) and gene sampling (nuDNA sequenced for subset of critical taxa), so I do not think we’ll be seeing anything more detailed any time soon. The support values for most nodes in the tree are strong:
Using our standard sequencing conventions, the following sequence reflects Klicka et al.’s Fig. 1. I included all the taxa restricted to North and Middle America (shaded gray) for completeness, and I used indentation and skipped lines to help mimic the branching pattern. The one area of semi-exception to the conventional rule is in Atlapetes. If you look at the tree, you can see that there are many nodes with little or no support, and even so, Fig. 1 is based entirely on mtDNA and thus may have gene tree/species tree problems. In fact, the *BEAST species tree in Fig. 2c shows a large polytomy among the taxa sampled except for the northern group, of which we only have A. albinucha. The nuDNA-only tree (Fig. 2b) shows a similar pattern. Therefore, the proposed sequence, clearly tentative, is a mix of incorporating the few solid nodes with geography, which is clearly the best predictor of relationships in the genus, not color (as I predicted in an Auk paper back in 1995 or so). The exception to geography as a predictor of relationships would be if the Tepui species personatus is actually more closely related to a cluster of narrowly distributed species in central/southern Peru, which would be very interesting (and which has parallels, I think, in Myioborus and Myiothlypis).
Note that Chlorospingus flavovirens is not in Fig. 1 - -this is because Klicka et al. (2014) found that not only is it not a Chlorospingus but it is a member of the Thraupidae, not Emberizidae/Passerellidae. Until we have a genus for flavovirens, and until we know where in the Thraupidae it goes, I suggest we leave it at the end of Chlorospingus, with appropriate footnote.
One of the many interesting points illuminated by the Klicka et al. (2014) that would be reflected in the new sequence is that Chlorospingus is indeed embedded in the family, but sister to everything except the enigmatic Oreothraupis. Another is that many of us might think of our “northern” sparrows as a monophyletic group, but they are not. Also, the “brush-finches”, Atlapetes and former Buarremon, are even more distantly related than was recognized. (Extralimitally, note that the Pselliophorus is really just an Atlapetes, and that Ammodramus and Melozone are polyphyletic.)
[Chlorospingus flavovirens – retained here temporarily pending additional publication]
Arremon costaricensis SS
Arremon basilicus SS
Arremon perijanus SS
Arremon atricapillus SS
Arremon phaeopleurus SS
Arremon phygas SS
Arremon assimilis SS
Arremon torquatus SS
Arremon aurantiirostris SS
Arremon abeillei SS
Arremon schlegeli SS
Arremon taciturnus SS
Arremon franciscanus * SS
Arremon semitorquatus SS
Arremon flavirostris SS
Junco hyemalis SS
Junco phaeonotus SS
Junco vulcani SS
“Ammodramus”=Ammospiza s nelsoni
Pipilo maculatus SS
Pipilo erythrophthalmus SS
Pipilo ocai SS
Recommendation: The current sequence does not reflect the most recent phylogenetic data and has to be changed. I encourage inspection of the tree and the sequence to see whether tweaks are needed, especially in Atlapetes, which has severe problems with respect to placement of A. personatus and A. rufigenis if the sequence is to combine geography and phylogeny.
Van Remsen and John Klicka, May 2014
Comments from Robbins: “YES, to the sequence change based on the new molecular data.”
Comments from Stiles: “YES, to bring the sequence into line with the phylogeny. Do we need a separate proposal regarding splitting Passerellidae (New World) from the Old World Emberizidae sensu stricto?
Comment from Remsen in response to Stiles: Yes we definitely need that proposal. I was hoping that NACC would take the lead on this, but if anyone here wants to go forward with one, do it.”
Comments from Nores: “? Although I agree that the current sequence does not reflect the most recent phylogenetic data and has to be changed, the new sequences (Arremon and Atlapetes), in my opinion, need tweaks.”
Comments from Jaramillo: “YES – This paper is a good one! Very well done. Apart from the various considerations we are dealing with here, I was quite surprised to see were Torreornis fell out. Lots of unexpected issues here. There may be some generic changes to come, I am not sure which species carries the name Ammodramus for example.”
Comments from Pacheco: “YES. Although the current sequence does not reflect the most recent phylogenetic data, this suggestion to new linear sequence is an improvement.”
Comments from Zimmer: “YES. Not perfect, but the new linear sequence does reflect the most recent molecular data, and represents a step forward.”