Change the sequence of hummingbird genera

Proposal (267) to South American Classification Committee

Change the sequence of hummingbird genera

This proposal would modify substantially the sequence of genera in the family Trochilidae.

Recent phylogenetic studies in the Trochilidae have radically revised our understanding of the relationships of many genera (and the composition of not a few). It is evident from both DNA hybridization (Bleiweiss et al. 1997) and DNA sequencing (Altshuler et al. 2004, McGuire et al. 2007?) that several major clades exist - and that all sequences of genera in current use scatter members of most of these clades about in a seemingly haphazard manner. The high degree of consistency among these studies strongly suggests that the arrangement proposed therein is quite robust. The latest and most exhaustive study - both from the number of genes sequenced and the number of taxa studied - is that of McGuire et al. (2007?), which also involves a detailed examination of how best to treat data from a number of genes when different ones produce trees that differ in details. I therefore propose that this study serve as the basis for ordering the genera of hummingbirds. Because not all genera were sequenced, I have attempted to place some unstudied genera (marked with an asterisk) into the sequence in accord with previous estimations of relationship, where these seem clearly established and consistent with the genetic data. Several genera that have not yet been sequenced and for which I am not aware of solid data regarding relationships are listed as "incertae sedis" at the end - presumably as genetic studies include them, their places will be clarified. Given that our current sequence is a thoroughly mixed bag, with members of clades often scattered among unrelated genera, the sequence proposed here should come much closer to producing a stable and reliable framework for organizing the family. However, I should note that the last dozen or so genera at the end represent a hodgepodge at this point: results indicate that members of currently accepted genera are often not closest relatives and various genera are para- or polyphyletic; only much denser taxon sampling will resolve the melee, so I have presented a sequence of genera as traditionally defined, while recognizing that it will surely suffer changes in the future; at the least, the proposed sequence places this assemblage of problem genera together. I separate by a blank line the major clades recognized in these studies.

Florisuga

Topaza

Eutoxeres

Ramphodon*

Glaucis

Threnetes

Anopetia*

Phaethornis

Doryfera

Schistes

Augastes*

Colibri

Androdon

Heliactin

Heliothryx

Polytmus

Avocettula*

Chrysolampis

Anthracothorax

Heliangelus

Sephanoides

Discosura

Lophornis

Phlogophilus

Adelomyia

Aglaiocercus

Sappho*

Oreotrochilus

Opisthoprora

Ramphomicron

Chalcostigma

Oxypogon

Oreonympha

Metallura

Haplophaedia

Eriocnemis

Aglaeactis

Coeligena

Lafresnaya

Ensifera

Pterophanes

Boissonneaua

Ocreatus

Urochroa

Urosticte

Sternoclyta*

Hylonympha*

Heliodoxa

Patagona

Heliomaster

Myrtis

Eulidia*

Rhodopis

Thaumastura*

Chaetocercus

Myrmia

Microstilbon

Calliphlox

Chlorostilbon

Chlorestes

Klais

Phaeochroa

Campylopterus

Eupetomena

Chalybura

Thalurania

Aphantochroa

Taphrospilus

Leucochloris

Leucippus

Amazilia

Chrysuronia

Goethalsia

Goldmania

Lepidopyga

Damophila

Hylocharis

Incertae sedis: Anthocephala, Clytolaema, Loddigesia, Polyonymus, Stephanoxis, Taphrolesbia (I would welcome comments from the committee on the placement of any of these genera, as I have little or no personal experience with any of them.)

As noted above, from Leucochloris or Leucippus on, things get messy and at least the larger genera will likely have to be redefined, split or sunk once a majority of species has been sequenced. This would appear to be the most actively/recently speciating assemblage in this clade, and morphological differences are slight on the whole. Whatever its uncertainties, the present sequence of genera comes much closer to providing a phylogenetic framework than does our current SACC List or the treatments by Peters (1945), Meyer de Schauensee (1966) or Schuchmann (1999), hence I strongly advocate a YES on this proposal.

References

Altshuler, D. L., R. Dudley & J. A. McGuire. 2004. Resolution of a paradox: hummingbird flight at high elevations does not come without a cost. PNAS USA 101:17731-17746.

Bleiweiss, R., J.A.W. Kirsch & J. C. Matheus. 1997. DNA hybridization evidence for the principal lineages of hummingbirds (Aves: Trochilidae). Mol. Biol. Evol. 14:325-343.

McGuire et al. 2007 (?). Phylogenetic systematics of hummingbirds: partitioned Bayesian and Maximum likelihood Analysis of multilocus sequence data and an unranked phylogenetic taxonomy for the Trochilidae. Systematic Biology: in press.

Meyer de Schauensee 1966

Peters 1945

Schuchmann 1999: HBW.

F. Gary Stiles, May 2007

_______________________________________________________________________________________________

Comments from Cadena: "YES. Of the missing genera, I know at least Anthocephala has already been sequenced by collaborators of McGuire et al. Why don't we ask them about the position of this genus (and others)? They may want to reserve this information for future publications, but if not, I would prefer an informed decision rather than the incertae sedis placement, even if the analyses have not been published. Van?"

Comments from Pacheco: "YES. A sequência baseada nos recentes estudos filogenéticos (sobretudo McGuire et. al.) representa um avanço significativo ao conhecimento das relações entre os gêneros de beija-flores e deve ser implementado, contra a sequência meramente convencional."

Comments from Zimmer: "YES. This is clearly a work in progress, but better to make some headway now rather than clinging to the old mess while waiting for perfect resolution. McGuire et al (2007) provides a good foundation for change."

Comments from Nores: "YES. La evidencia molecular es muy consistente y resulta fundamental en estos casos donde las diferencias morfológicas y comportamentales muestran en general poca evidencia sobre relaciones filogenéticas."

Comments from Jaramillo: "YES - Obviously this is a temporary order, but a much improved order from what we have now. The various molecular datasets are congruent, and the new order makes sense, it seems like a great improvement."