Proposal (#291) to South American Classification Committee

Change linear sequence in Tangara

Effect on SACC: All that this would do is change the linear order of species in Tangara to reflect published data.

Background and New information: The current linear sequence is essentially unchanged from Storer's sequence in the Peters checklist and is maintained largely by historical momentum rather than explicit rationale:

Current SACC sequence =

Tangara inornata
Tangara palmeri
Tangara mexicana
Tangara chilensis
Tangara fastuosa
Tangara seledon
Tangara cyanocephala
Tangara desmaresti
Tangara cyanoventris
Tangara johannae
Tangara schrankii
Tangara florida
Tangara arthus
Tangara icterocephala
Tangara xanthocephala
Tangara chrysotis
Tangara parzudakii
Tangara xanthogastra
Tangara punctata
Tangara guttata
Tangara varia
Tangara rufigula
Tangara gyrola
Tangara lavinia
Tangara cayana
Tangara vitriolina
Tangara meyerdeschauenseei
Tangara peruviana
Tangara preciosa
Tangara rufigenis
Tangara ruficervix
Tangara labradorides
Tangara cyanotis
Tangara cyanicollis
Tangara larvata
Tangara nigrocincta
Tangara nigroviridis
Tangara vassorii
Tangara heinei
Tangara phillipsi
Tangara viridicollis
Tangara argyrofenges
Tangara cyanoptera
Tangara velia
Tangara callophrys

Burns & Naoki (2004) sequenced about 1500 bps of two mitochondrial genes, cytb and ND2, for 43 of 49 species of Tangara. Their hypothesis of relationships within Tangara (their Fig. 2) contains about 23 nodes with 97-100% Bayesian support that involved more than 1 species. To incorporate that information on sister relationships into our classification, we have produced the following sequence, using the conventions of basal taxa listed first and allotaxa listed NW to SE. Otherwise, we then incorporated the remaning topology of their Fig. 2, but maintained historical stability wherever possible.

Tangara ruficervix
Tangara cyanoptera
Tangara viridicollis
Tangara heinei
Tangara argyrofenges
Tangara phillipsi [not sampled]
Tangara palmeri
Tangara peruviana [not sampled but almost certainly sister to peruviana]
Tangara preciosa [not sampled in original study but unpublished Burns data support this placement]
Tangara meyerdeschauenseei
Tangara cayana
Tangara vitriolina
Tangara nigrocincta
Tangara larvata
Tangara cyanicollis
Tangara varia
Tangara rufigula
Tangara guttata
Tangara xanthogastra
Tangara punctata
Tangara vassorii
Tangara nigroviridis
Tangara labradorides
Tangara cyanotis
Tangara inornata
Tangara mexicana
Tangara chilensis
Tangara velia
Tangara callophrys
Tangara seledon
Tangara fastuosa
Tangara cyanocephala
Tangara desmaresti
Tangara cyanoventris [not sampled in original study but unpublished Burns data support this placement]
Tangara lavinia
Tangara gyrola
Tangara rufigenis [not sampled; placed here tentatively following Isler & Isler 1989]
Tangara chrysotis
Tangara xanthocephala
Tangara parzudakii
Tangara schrankii
Tangara johannae
Tangara arthus
Tangara florida
Tangara icterocephala

Some of this sequence will likely require tweaking with additional analyses, although Burns's unpublished data with more taxa and sequences strongly supports the suggested sequence.

Recommendation: Yes. Regardless of any future tweaking, the above sequence reflects published data, whereas the traditional one does not.

References (I have pdf if you need it):

BURNS, K. J., AND K. NAOKI. 2004. Molecular phylogenetics and biogeography of Neotropical tanagers in the genus Tangara. Molecular Phylogenetics and Evolution 32: 838-854.

Van Remsen & Kevin Burns, June 2007

============================================================================

Comments from Zimmer: "YES. Even if modifications are needed, this represents a distinct step forward."

Comments from Nores: "YES, aunque mirando el árbol de Burns y Naoki (2004) no me resulta tan claro como fue finalmente logrado el ordenamiento. De todos modos parece que es un buen avance comparado con la tradicional secuencia."

Comments from Stotz: "NO. Given that Kevin Burns was involved in creating this proposal, perhaps it is a bit hard to fathom that I am voting against it because of further data from Kevin.  I feel this is similar to the Hemispingus case, where more data is indicating that there are broader issues that remain to be dealt with.  In this case Tangara is not monophyletic, and given the limited information that is contained within the order of taxa in a large genus like Tangara, I would prefer to wait until the more complete dataset is available."

Comments from Stiles: "YES. The proposal best reflects the phylogenetic evidence available. I may be myopic, but I fail to understand Doug's objection here. The branch to Tangara in the Burns & Naoki paper has 100% Bayesian support in both trees, and it leads only to Tangara (no putative Tangara species on other branches, no non-Tangara on the Tangara branch; there are evidently several major clades within Tangara, but this does not affect the monophyly of the genus)."

Comments from Jaramillo: "YES - Even if refinements are needed in the future, the new order will be easier to work with than our present order. This applies to events such as the need to divide Tangara to maintain monophyly, or the addition of other species into Tangara ... whatever may arise with further sampling of tanagers (reading between the lines of Doug Stotz's comments)."

Comments from Kevin Burns: "Doug is correct that our new data show that six species of Thraupis are imbedded within Tangara.  Thus, Tangara is not monophyletic.  These data are strongly supported with nuclear and mtDNA data.  So, when our results are published, these species will need to be moved within the sequence.  But note that Al Jaramillo is correct in his comments in that this wouldn't result in rearrangement of all the Tangara.  These Thraupis species would simply be inserted as a group within the sequence.  (most likely between larvata and guttata in the above arrangement).

"Hemispingus is a different story.  In that case, the nonmonophyly of Hemispingus is the result of multiple species being inserted at multiple places within Hemispingus."