Proposal (289) to South American Classification Committee

Lump the genera Arremon, Buarremon, and Lysurus in an expanded genus Arremon

Effect on SACC list: Three currently recognized genera would be lumped into a single genus.

Background and New information: Relationships among genera of Neotropical Emberizine finches and sparrows have traditionally been inferred based on superficial analyses of phenotypic variation, but are starting to be assessed in a more rigorous fashion as molecular phylogenetic studies start to accumulate. In particular, the affinities of birds in the genus Buarremon have been somewhat uncertain. The three species currently recognized in the genus (one of which is extralimital for the SACC) were merged for quite a while in the genus Atlapetes, whose members are now known not be particularly closely related to them. Hackett (1992) indicated that species of Buarremon are more closely allied to members of the genus Lysurus than to Atlapetes, a result corroborated recently by an ongoing study on the phylogeny of all the nine-primaried oscines, whose preliminary analyses clearly show that Buarremon belongs in a strongly supported clade with Lysurus and another similar genus, Arremon (J. Klicka et al., unpublished data).

A recently completed molecular phylogenetic study (Cadena et al. 2007; pdf available at: http://evolvert.uniandes.edu.co/Site/Publicaciones_files/mpe2007-1.pdf) analyzed the relationships among taxa in these three genera based on sequences of four mitochondrial (ND2, cytb, ATPase 6, ATPase 8) and two nuclear (MUSK, ACO1) genes. Mitochondrial data provided strong support for a clade formed by the Buarremon torquatus complex and the genus Arremon, indicating that Buarremon as currently defined is not monophyletic. A clade formed by B. brunneinucha and B. virenticeps consistently appeared to be sister to the genus Lysurus, but this was not strongly supported. Variation in one of the nuclear genes was entirely consistent with these results, but analyses of sequences of the other nuclear gene placed Arremon outside a clade formed by Buarremon and Lysurus, albeit with relatively weak support.

When mitochondrial and nuclear data were analyzed together, strong support (1.0 Bayesian posterior probability and 80% maximum-likelihood bootstrap) was obtained for the B. torquatus – Arremon clade. In these combined analyses, the B. brunneinucha-B. virenticeps clade was recovered as sister to Lysurus, but support for this relationship remained relatively low (although it closely approached significance in a Bayesian analysis, with a posterior probability of 0.94). It is noteworthy that a clade formed by all taxa placed in Buarremon in our current classification was not recovered in any single tree in this study, even considering the 120,000 trees that were sampled in Bayesian MCMC searches. From a Bayesian perspective, this means that given the data and the models employed in analyses, the probability that Buarremon is monophyletic is zero.

Analysis: Molecular data indicate strongly that as currently defined, Buarremon is not a monophyletic genus because B. torquatus is the sister taxon to the genus Arremon, which in retrospect is not entirely surprising considering that some taxa currently placed in Arremon (e.g. some A. taciturnus) seem like "miniature" versions of B. torquatus given their high similarity in plumage. In turn, B. brunneinucha and B. virenticeps seem to form a clade with Lysurus, but support for this relationship is not very compelling. In any event, the problem of the non-monophyly of Buarremon needs to be addressed. I believe that the best alternative is to merge the three genera into a single genus for two reasons. First, birds currently placed in Buarremon, Arremon, and Lysurus are rather similar in overall plumage, behavior, microhabitat, and vocalizations, so lumping them does not result in a highly heterogeneous genus. Second, because the sister relationship between Lysurus and B. brunneinucha - B. virenticeps is not very strongly supported, it seems that recognizing two genera (one for each of these clades) would not be the best option because support for the monophyly of one of the clades is weak. Because it has priority, the name for the expanded genus would be Arremon. For a more detailed discussion of the issues involved, please refer to Cadena et al. (2007).

Recommendation: I would suggest voting YES to lump Buarremon, Arremon and Lysurus.

References

Cadena, C. D., J. Klicka & R. E. Ricklefs. Evolutionary differentiation in the Neotropical montane region: molecular phylogenetics and phylogeography of Buarremon brush-finches (Aves, Emberizidae). Molecular Phylogenetics and Evolution, in press.

Hackett (1992) - See SACC Literature Cited.

C. Daniel Cadena, June 2007

______________________________________________________________________________________________________

Comments from Remsen: "YES. As Daniel noted above, an expanded Arremon is the only classification that is consistent with current phylogenetic data. Whether current Buarremon is paraphyletic may not be 100% certain from Cadena et al., but the monophyly of Buarremon +Arremon + Lysurus seems indisputable."

Comments from Nores: "YES. Al mostrar que Buarremon no es monofilético resulta aceptable que la mejor manera de resolver la cuestión es unificar los géneros en Arremon que es el más antiguo. Queda por resolver si A. virenticeps y A. crassirostris deben ser incluidos en la SACC list."

Comments from Stotz: "YES. Makes sense in terms of morphology. Once broad Atlapetes bit the dust it seemed clear that there would be further refinements of the taxonomy of the Brush-finches and allies."

Comments from Jaramillo: "YES - Logic and data tell me it is so, although my heart is fighting it. I guess it is fighting the issue of having what seems like such a nice and neat genus, Arremon, become somewhat more heterogeneous. But don't let my sentiments be misunderstood, the data appears solid, and as mentioned in the proposal it makes sense for various reasons (voice for example, habitat etc.)."