Proposal (#219) to South American Classification Committee

Split Thryothorus

Effect on South American Checklist: This proposal is to substitute three generic names (two old, one new) for species currently in Thryothorus, as well as slightly modifying the species order within the former genus.

Background: Prior to Hellmayr (1934), wrens in the genus Thryothorus were typically classified in three genera, Thryothorus, Pheugopedius, and Thryophilus (see especially Baird, 1874). Subsequent workers have followed Hellmayr's example, with little additional comment.

New Information: Mann et al. (2006) report analyses of mitochondrial sequence data from nearly all species of Thryothorus. This study clearly shows these species falling into four well-supported clades, two corresponding to Thryothorus and Pheugopedius as previously recognized, and the other two corresponding to Thryophilus. Support for monophyly of the latter two clades was very poor, and the authors chose to restrict Thryophilus to one of these clades, and to erect a new generic name (Cantorchilus) for the other. Their study also included nuclear DNA sequence from a subset of taxa, but these data were nearly uninformative with regard to relationships among the mitochondrial clades, except in separating Thryothorus ludovicianus (the type of the genus) from all the remaining species, necessitating revision of genus definitions.

Proposed Sequence: Based on these data, the following sequence is proposed for South American "Thryothorus":

Pheugopedius:

spadix

fasciatoventris

euophrys

eisenmanni

mystacalis

genibarbis

coraya

rutilus

sclateri

Thryophilus :

rufalbus

nicefori

Cantorchilus:

leucopogon

nigricapillus

superciliaris

leucotis

guarayanus

longirostris (type species for Cantorchilus)

griseus (incertae sedis)

The placement of T. griseus is speculative, based on an offhand comment of Hellmayr. The species P. spadix, P. eisenmanni, and T. nicefori were not included in the study, but each shows clear affinities with other species in the two genera.

Literature Cited:

Baird, S.F. 1874. Review of American birds in the Museum of the Smithsonian Institution, Part 1. Smithsonian Misc. Coll. 12: 484 pp.

Hellmayr, C.E. 1934. Catalogue of birds of the Americas and the adjacent Islands. The Field Museum of Natural History, Chicago, IL.

Mann, N.I., et al. 2006. Molecular data delineate four genera of "Thryothorus" wrens. Mol. Phylogenet. Evol. 40: 750-759. http://www.tc.umn.edu/~barke042/pdfs/Mann.et.al06.pdf

F. Keith Barker, June 2006

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

Comments from Cadena: "YES, but somewhat tentatively because one could argue that the genetic data are not very solid. The only well-supported node in the dataset that is relevant for this proposal is the one showing that Thryothorus ludovicianus is sister to Thryomanes and not to other Thryothorus, which justifies the need for a revision of nomenclature unless one favors merging Thryomanes with Thryothorus and argues that support for other relationships is not strong enough to make any further changes. Indeed, the relationships of the clades that Mann et al. propose to call different genera are not well established, and it is actually quite remarkable that none of the relevant nodes have posterior probabilities greater than 50% (nevermind 95%). Looking at the very short internodes connecting branches in their Figure 2, I would predict that a likelihood-based test (e.g. Shimodaira-Hasegawa) would not reject the hypothesis that "Pheugopedius", "Thryophilus ", and "Cantorchilus" form a monophyletic group in favor of the hypotheses that they are a para- or polyphyletic assemblage. However, this is probably true of most relationships in that large clade including many additional genera, whose diversification appears to have been very rapid; this will be a very difficult phylogeny to resolve with good support. The question is whether one wants to recognize one, two, or three genera for the "southern" birds. As Mann et al. point out, one would be most conservative, and I would be willing to reconsider my vote to take this position if committee members have cogent arguments to support this idea. Stability at this point in time would probably not be a very good argument because the generic allocation of all these birds is going to change anyway (i.e. none of them will remain in Thryothorus), and I doubt that subsequent studies will prove these form a monophyletic group. Even if they were to form a clade, one could still maintain them in three different genera, which could be used as an argument for stability in the long run.

"In sum, although not strongly supported by the results of the phylogenetic analyses, for now I agree with the decision to rank the four clades of "Thryothorus" as different genera. This seems well justified on the basis of genetic distances, which show they are all quite distinct from each other and are divergent to the same level that well-established genera differ from each other. I realize it is a bit of a shame that the different "southern" genera proposed by Mann et al. do not appear to have morphological or behavioral synapomorphies that would aid in their recognition (birders will have a hard time remembering which species go in which genus), although it appears to me that the songs of Thryophilus (rufalbus and nicefori, which I am familiar with) stand out as rather unique when compared to those of other "Thryothorus" I know."

Comments from Jaramillo: "YES - This is a solid paper, clarifying a large part of wren systematics. I don't see any reason not to support this division of Thryothorus."

Comments from Thomas Donegan: "A point which may be of interest: I exchanged some emails with Keith Barker as to whether the authors of the above paper had considered applying to ICZN to change the type species of the genus Thryothorus. The crux of the need for re-classification is that the type species T. ludovicianus s not closely related to other current Thryothorus taxa. If (i) T. ludovicianus were held congeneric with Thryomanes or given a new genus and (ii) Thryothorus were assigned a new type species by ICZN, then Thryothorus would arguably be capable of being retained for all the other taxa. Keith Barker, however, pointed out that until Hellmayr's (relatively recent in zoological time) 1930s treatment, Thryothorus was not used for many of the Neotropical species whilst two of the three proposed new names were so used. Thryothorus could therefore be thought not sufficiently strongly entrenched for many of these taxa to warrant an ICZN application."

Comments from Stiles: "YES, but like Daniel, rather tentatively. The separation of Thryothorus (ludovicianus only) from the rest of the species was already indicated in a previous study and is clearly mandated; the three groups of species in our area are about equally distinct (and roughly equally distinct from other genera of wrens, according to a previous paper) such that each could be given generic status - although support for the nodes involved leaves much to be desired. The whole group might prove to be paraphyletic, but (I'm guessing) this would probably not affect the monophyly of the genera as defined here. Since they are all going to depart from Thryothorus in any case, I suspect that looking forward, the four genera proposed will probably require the least changing in the future."

Comments from Robbins: "NO. The Mann et al. paper is puzzling on a couple of issues. First, it is a bit hard to swallow that the authors make a statement "Where possible systematic work should be based on material vouchered in publicly accessible collections (with refs); however, the majority of Thryothorus samples included in this study were aliquots of blood." when a number of the taxa where they used only unvouchered blood samples **are** represented with tissues and vouchers, and have been for > 15 years, at museums that readily loan such material. If you are going to make such a statement, then you should practice what you preach!

"The current data strongly support two conclusions: 1) the name Thryothorus should be restricted to ludovicianus and its apparent sister taxon albinucha. 2) the "southern" wrens form a monophyletic clade. However, given the weak support, I believe it is premature (and the authors state: "Statistically, we cannot distinguish among three possible relationships for these three remaining three clades" and note how various genera bounce around based on inclusion of genes and analyses; compare figs. 1-3) to split these into three different genera. The prudent course is to place all into the genus Pheugopedius until data from additional genes and perhaps vocal and morphological data can resolve the relationships.

Additional comments from Stiles: "After reading Mark's comments and rereading the Mann et al. manuscript, I wish to change my vote to a qualified, partial YES - namely, to place all the "southern Thryothorus" in Pheugopedius for now. The monophyly of this group as a whole seems well established, but that of the three clades proposed by Mann et al. is less so, as is the membership of each clade. Support for the nodes in question leaves something to be desired, so I prefer take a more conservative approach, go with the one well-established conclusion and await less equivocal evidence, hopefully including better morphological and vocal data, before splitting Pheugopedius into three genera."

Comments from Pacheco: "YES. A abrangência taxonômica do estudo é bastante boa, ficando apenas quatro espécies de fora e o suporte estatístico dos nós genéricos são altos. Pessoalmente reconheço apesar dos comentários dos colegas que os resultados obtidos pelo estudo apresentam uma hipótese mais adequada para a recuperação do padrão da história evolutiva dos até então denominados "Thryothorus. Antevejo, ainda, que mais estudos serão necessários para resolver as relações interespecíficas dentro do complexo de táxons leucotis-longirostris.

Comments from Zimmer: "I would join with Mark and Gary in a partial, qualified "YES". In other words, I would favor the separation and placement of all other species except ludovicianus into Pheugopedius (a move that seems to be well supported by the molecular evidence), but not the further separation of the various clades into other genera. Even given that further generic-level splitting may be inevitable, I would prefer to wait until a more thorough analysis, particularly one that examines vocal characters, is available to refine the placement of the various species in the respective new genera."

Comments from Remsen: "YES, in the broad sense of recognizing two other genera within Pheugopedius. [I've gone back and forth on this several times.] Daniel has just reminded me that the genetic support for monophyly of broadly defined Pheugopedius is insufficient. However, the three genera proposed by Mann et al. are monophyletic, not only in cyt b analysis (their Fig. 2), but also when FGB-14 sequences are combined with cyt b data, albeit with weak support (their Fig. 3). The combined analysis includes fewer taxa, and the branch with their single representative of Thryophilus (sinaloa) cannot be placed confidently. Mann et al. noted that there is no morphological diagnosis for Cantorchilus, which in itself is not so bad, but when you compare what would become Cantorchilus modestus and Thryophilus sinaloa to each other, the morphological and plumage differences are slight and below the subjective level typically associated with generic limits. But the same could be said for Thryothorus ludovicianus vs. these same species (the presumed reason for the original lump that produced the broadly defined and polyphyletic Thryothorus. So, to avoid maintaining a clearly polyphyletic Thryothorus, we are forced to chose between two "ugly" choices: (1) recognize three separate genera that are almost certainly monophyletic but morphologically "below" the traditional level of generic distinctiveness, as well as leaving griseus either Incertae Sedis or placed in Thryophilus based on Hellmayr, or (2) recognize a broadly defined Pheugopedius for which there is no real support for its monophyly. I'll go with the former as an admittedly unsatisfactory but hopefully temporary measure rather than risk perpetuating another taxon for which there is no support for its monophyly."

Comments from Stotz: "NO. I agree with Mark. The best course to me looks like split off Thryothorus and place all the SA taxa in Pheugopedius."