Proposal (644) to South American Classification Committee

Revise the classification of the Phoenicopteridae

The flamingos are a well-defined and small family without major controversies in taxonomy. One open question had been whether to afford species status to American Flamingo Phoenicopterus ruber versus Old World roseus, the SACC voted to accept species status for ruber in proposal 274.

Two other open questions have been (1) whether the genus Phoenicoparrus should be retained, and (2) the proper linear sequence in this small family. Phoenicoparrus is defined by a “deep keeled” bill structure quite different from Phoenicopterus and a lack of a hind toe, among other features. The importance of these features in designating a separate genus has been questioned, although seemingly no one has questioned the sister relationship between jamesi and andinus. Sibley & Monroe (1990) merged Phoenicoparrus into Phoenicopterus based on small genetic distances among all flamingos as measured by DNA-DNA hybridization (Sibley & Ahlquist 1989).

New information: Torres et al. 2014 (http://www.biomedcentral.com/1471-2148/14/36) sampled all of the extant flamingo species, including both ruber and roseus. Data for 12 nuclear loci, and 2 mitochondrial loci were extracted for analysis. They generated a robust phylogeny with high level of bootstrap support. They also found that the division between American and Greater flamingoes is relatively old, as old as between James’s and Andean, so they are well differentiated, supporting the SACC decision to separate American Flamingo as a species.

Phoenicoparrus: They found a deep separation between the deep-keeled and shallow keeled-flamingos, supporting the division of the family into at least two genera. Although outside of our region of interest, they decided to merge the Lesser Flamingo (P. minor) into Phoenicoparrus rather than retaining three genera in the family.

Proposal A. Merge Phoenicoparrus into Phoenicopterus

As noted above, Sibley and Monroe (1990) merged these two genera based on a short genetic distance between them, and presumably also because the overall morphology of the two groups is very similar other than differences in bill structure and presence/absence of the hallux.

Proposal B. Modify linear sequence of species

Our current linear sequence is:

PHOENICOPTERIDAE (FLAMINGOS)

Phoenicopterus ruber American Flamingo

Phoenicopterus chilensis Chilean Flamingo

Phoenicoparrus andinus Andean Flamingo

Phoenicoparrus jamesi James's Flamingo

This needs only 1 minor tweak to conform to our rules for sequencing. Because Chilean Flamingo is an earlier offshoot in Phoenicopterus, it should precede American. So the tweak would be:

PHOENICOPTERIDAE (FLAMINGOS)

Phoenicopterus chilensis Chilean Flamingo

Phoenicopterus ruber American Flamingo

Phoenicoparrus andinus Andean Flamingo

Phoenicoparrus jamesi James's Flamingo

Recommendations:

A- Merge Phoenicoparrus into Phoenicopterus – I recommend a NO.

B- Tweak the linear sequence of flamingos to conform to the convention of least-diverse branch first. I recommend a YES.

Literature Cited

TORRES, C. R., L. M. OGAWAL, M. A.F. GILLINGHAM, B. FERRARI and M. VAN TUINEN 2014. A multi-locus inference of the evolutionary diversification of extant flamingos (Phoenicopteridae). BMC Evolutionary Biology 2014, 14:36 

SIBLEY, C. G., AND J. E. AHLQUIST. 1990. Phylogeny and classification of birds. Yale University Press, New Haven, Connecticut.

SIBLEY, C. G., AND B. L. MONROE, JR. 1990. Distribution and taxonomy of birds of the World. Yale University Press, New Haven, Connecticut.

Alvaro Jaramillo, September 2014

________________________________________________________________________________________________________

Comments from Stiles: “A. NO, in part; the only thing I question is including minor in Phoenicoparrus but as that is extralimital for us, we need not worry about it!”

Comments from Remsen: “A. YES, emphatically! The paper estimates the divergence between Phoenicopterus and Phoenicoparrus as follows: “The deep- and shallow-keeled clades diverged in either the Pliocene or earliest Pleistocene (1.7-3.9 mya).” This is very recent for separation of genera. For example, in the Furnariidae, the first groups that we (or anyone previously by traditional criteria) consider as genera are at least 4 million years old, and most are in the 8-18 mya range. Further, I don’t know these birds very well, but at least superficially they appear to be a single, conservative morphotype differing primarily in bill structure (due to feeding differences). There is no character MORE PLASTIC in morphology than bill shape – it has long been dismissed as a character on which to base genera. Just look at Anas, Aythya, Calidris, etc., in waterbirds (much less something like Hemignathus, which is a young group).” B. YES. Trivial but necessary tweak to fit the phylogeny.”

Comments from Pacheco: “A. YES. I agree with Van’s arguments. There are two groups, but the level of divergence and morphological differences are not suitable for treatment in two genera. B. YES.”

Comments from Zimmer: “Part (A): YES. Van’s points regarding both the estimated time of divergence between Phoenicopterus and Phoenicoparrus (very recent for generic separation) and the evolutionary plasticity of bill morphology are well taken. Part (B): YES on the necessary phylogenetic housekeeping.”

New comments from Stiles: A. YES. First, after reading your comments on the flamingo proposal, I am willing to vote against recognizing Phoenicoparrus - I hadn't caught the recency of the split, which does indeed indicate to me that the evolution of the more specialized feeding apparatus of the Phoenicoparrus types probably involved selection for divergence in feeding methods vs. Phoenicopterus types, probably due to sympatry - reflecting selection to reduce competition and facilitate coexistence in a very special, probably limited habitat, saline lagoons.”

Comments from Cadena: “A. NO. If we were to start classifying these birds from scratch, then I would likely agree with Van in that they should all probably be included in a single genus. However, there is a tradition in recognizing two separate genera in our baseline list and the two-genus treatment is perfectly consistent with the phylogeny, so there is no need to change. I also agree with Van in that genera should be more than simply clades (i.e., they should have a long history of divergence from other clades and have meaningful phenotypic/ecological differences), but I think stability trumps all this: we should only change when strictly necessary. Regarding bills being highly plastic, yes, but here there is no indication that phenotypic similarity does not reflect homology because as far as I understand the case, the (admittedly minor) phenotypic differentiation is consistent with the phylogeny.

B. YES.”

Comments from Areta: “A-NO. I agree with Van in that the bill is an extremely plastic feature, as has been shown repeatedly. However, in this case bill-shape and presence/absence of hallux are coincident with degree of genetic differentiation, indicating that there is some important and consistent differentiation between Phoenicopterus and Phoenicoparrus. The fact that bills are plastic does not necessarily mean that in all instances the differences have to do with plasticity only. In terms of information contents, I think it is more informative to keep two genera than to merge everything into a single one. I am not convinced that an absolute genetic yardstick can be used to split genera, although I appreciate deeper splits supporting generic-level taxa. Instead, some measure of evolutionary change seems more appropriate to make the subjective judgment of generic species limits. Alternatively, analyses of heterogeneity can be performed to find units that may later be named genera.”