Proposal (618) to South American Classification Committee

Recognize newly described Myrmotherula oreni and split Myrmotherula iheringi into two species

Effect on South American CL: If adopted, this proposal would add a newly described species of Myrmotherula to the list and elevate to full species status one subspecies of M. iheringi (heteroptera).

Background: The Ihering' s Antwren is endemic to southwestern Amazonia, where it occurs in Bolivia, Brazil, and Peru on both sides of the Madeira River (Zimmer and Isler 2003). Currently, two parapatric subspecies are recognized in polytypic M iheringi based primarily on differences in female plumage patterns (Todd, 1927): M. i. iheringi (distributed in the Madeira-Tapajós interfluvium); and M. i. heteroptera (west of the Madeira and south of the Amazon rivers). More recently, the morphological diagnoses of the subspecies of M. iheringi have been questioned (Zimmer and Isler 2003), thus underscoring the need of a taxonomic revision of the M. iheringi complex.

New information: As a result of a multi-character taxonomic revision of the M iheringi complex, Miranda et al. (2013) recovered phylogenies with three main, reciprocally monophyletic clades also diagnosed by distinct loudsong note structures and female plumage patterns, which correspond to a monophyletic M. i. iheringi and a paraphyletic M. i. heteroptera.

Analysis/Recommendation: The analyses included 44 skins, 62 tape-recordings, and 14 tissues belonging to all taxa and from the entire range of Myrmotherula iheringi. All Bayesian and Maximum Likelihood trees obtained recovered the same topology in which three main lineages not corresponding to current subspecific limits within M. iheringi were found (hereafter named as clades A, B, and C). Clade A includes specimens inhabiting the Madeira–Tapajós interfluvium, where the type locality of nominate M. iheringi is located; clade B includes birds from the southern part of the Inambari area of endemism, which have been treated up to now as M. i. heteroptera; finally, clade C groups specimens occurring west of the Madeira river in the northern part of the Inambari area of endemism (Silva et al. 2005) and to which the name heteroptera applies (type locality Hyutanaã on the Rio Purus). Thus, all phylogenies obtained recovered a paraphyletic M. i. heteroptera as indicated by current taxonomy (Zimmer and Isler 2003), because clades C (to which the name heteroptera truly applies) and B do not share a most recent common ancestor; instead, clade B is sister to clade A, to which the name iheringi applies, hence highlighting the need to name clade B as a new taxon (oreni). Clade C is especially well differentiated molecularly and vocally from its closest relatives, hence supporting an independent species status for heteroptera, which we recommend be known henceforth as the Purus Antwren.

Literature Cited

Miranda, L. S.; A. Aleixo; B. M. Whitney; L. F. Silveira; E. Guilherme; M. P. D. Santos, and , M. P. C. Schneider (2013) Molecular systematics and taxonomic revision of the Ihering's Antwren complex (Myrmotherula iheringi: Thamnophilidae), with description of a new species from southwestern Amazonia. Pp. 268-271. in del Hoyo, J.; A. Elliott; J. Sargatal and D. A. Christie. (Org.). Handbook of the Birds of the World, Special Volume: New Species and Global Index. Barcelona.

Silva, J.M.C., S. B. Rylands, G. A. B. Fonseca. (2005). The fate of the Amazonian areas of endemism. Conservation Biology 19:689-694.

Todd, W. E. C. (1927). New gnateaters and antbirds from tropical South America, with revision of the genus Myrmeciza and its allies. Proceedings of the Biological Society of Washington 40: 149- 178.

Zimmer, K. J. and M. L. Isler (2003). Family Thamnophilidae (Typical Antbirds). Pp. 448-681 in: del Hoyo, J., A. Elliott, and D. A. Christie (orgs.). Handbook of the Birds of the World, Volume 8: Broadbills to Tapaculos. Lynx Edicions, Barcelona.

Leonardo S. Miranda and Alexandre Aleixo, December 2013

______________________________________________________________________________

Comments from Remsen: “YES, but barely, and tentatively, pending a plea to Mort Isler for an evaluation of the songs. The plumage analysis shows that there is indeed an overlooked, diagnosable, undescribed taxon, to which now the name oreni applies. The question is … subspecies or species? The genetic data are of no real use, and I consider the “paraphyly” incorrectly applied. All we have is an mtDNA gene tree, not necessarily a “species tree” (and the words “gene tree” do not appear in the text). Further, the N of individuals (14 for three taxa) makes statements concerning “reciprocal monophyly” potentially premature, i.e. if the next individual sampled has an allele shared with one of the other taxa, then the “reciprocal monophyly” disappears. [On my “wish list” is for someone out there to publish something on this, with intense sampling throughout the ranges of parapatric/allopatric taxa, including potential contact zones, to analyze, via subsampling, what the minimum N is to determine true reciprocal monophyly, at least for a few real-world examples so we can see what we’re up against.] Further, if I am reading Fig. 3 correctly, the nearest samples of heteroptera and oreni are roughly 1400 km apart, and so there are major assumptions in terms of the weak geographic sampling representing the populations as a whole, particularly since the samples of heteroptera are as far from the potential contact area as one can get. Even the samples of oreni and iheringi (s.s.) are roughly 500 km apart, and none of the localities is anywhere near the area towards the headwaters, where detection of any gene flow would seem most likely. As for genetic distance, the 1.8% difference between oreni and iheringi (s.s.) falls within the “subspecies level” of genetic distance even for mobile temperate latitude birds if we were to apply some sort of bar-coder standards. The bamboo specialization by oreni is suggestive, but not a certain species-level character because some species (e.g. Epinecrophylla ornata) are only regionally specialized on bamboo.

“That leaves voice, specifically song, and very specifically, note shape. To me those shapes do indeed differ dramatically and strongly suggest species rank. I do not see any other differences unless further analyses would find that the difference in duration in the sonograms between heteroptera and the other two holds up. By the Isler-Isler-Whitney point of reference of a minimum of 3 vocal characters, oreni and heteroptera may fall short, depending on what is considered a character … and so I am actively seeking Mort Isler’s guidance on this and may change my vote.”

Comments solicited from Mort Isler: “I do not think the committee should even ask the question of whether note shape differences alone are a sufficient basis to recommend that these populations be considered species. Before considering this as a question, I think there is an obligation on the part of authors to determine whether vocal characters other than loudsong note shape distinguish populations. This should include quantitative measurements of loudsongs and, equally important, examination of other vocal types (i.e., calls). Although the committee has on occasion accepted species status for thamnophilids without a complete vocal examination (e.g., Gymnopithys species), obvious differences between such populations in their songs greatly exceeded note shape. I think it puts an undue burden on the community to assess species status based on note shape alone.

“Even within the confines of note shape, variability within populations should have been examined.  With this as a concern, I examined note shape in a half a dozen recordings of each population and found substantial within population variation. From my quick perusal, I think that this would not be a problem in distinguishing note shape between heteroptera (whose notes ascend in frequency) and the other two populations (whose notes primarily descend in frequency). However, I did find recordings in my small samples of nominate iheringi and oreni that were much closer in note shape than the two examples that were published.  On request, I can send members of the committee spectrograms to provide an idea of how similar they can be.  I am not saying that they cannot be distinguished, but their similarity demands that all available loudsongs recordings from iheringi and oreni be examined in "blind tests" to determine whether a qualitative difference in note shape is diagnostic.

“In conclusion, I do not believe that the vocal analysis presented in Miranda et al. provides an adequate basis for their recommendations.  A thorough vocal analysis is needed to support recommendations for species status.”

Additional comments from Remsen: “I change my vote to NO, for the reasons outlined by Mort above.”

Comments from Stiles: “NO. Although I was initially inclined towards supporting the three-way species split, Isler´s comments have convinced me that the evidence as it stands is insufficient. Given the genetic data, evidence for splitting off heteroptera is strongest, but that for splitting oreni from iheringi is much more tenuous; I also note that a photo or two illustrating the plumages of all three forms is important to include as these are relatively subtle and the painting does not show clearly the differences adduced in the diagnosis of oreni. In sum, more extensive genetic, morphometric and vocal data over the entire distributions of all three forms, especially where their distributions approach most closely, could make a solid case but as it stands, the necessary evidence for splitting iheringi into three species is lacking.”

Comments from Nores: “NO. I agree with Morton Isler that note shape differences alone are not a sufficient basis to recommend that these populations be considered species. As mentioned by Gary, more extensive genetic, morphometric and vocal data over the entire distributions of all three forms, especially where their distributions approach most closely, is necessary for splitting iheringi into three species.”

Comments from Zimmer: “NO, particularly given the comments by Mort Isler. While conceding that species recognition based on a single vocal character is not ideal, I am actually not philosophically opposed to the idea of species recognition based solely on note shape. The Isler et al three vocal characters “guideline” for species recognition in antbirds is just that, a “guideline”, not a rule. I know of at least one case where two parapatrically distributed populations are behaving as good biological species (no signs of intergradation, all pairs in the contact zone assortatively mated, neither population responds to playback of the other, etc), yet a vocal analysis revealed only one diagnosable character that differed, albeit significantly, between the two populations. My point is that one character may be enough, if the difference in that character is dramatic. Note shape differences are reflected in qualitative differences in how those notes sound – if note shape is dramatically different, then the resulting quality of the sounds produced can be dramatically different. A good example is provided by notes that show a high level of frequency modulation versus those that don’t. The spectrogram etchings of such notes appear very different, and the notes sound qualitatively very different, even to the human ear. So, for me, note shape differences as the sole vocal character are not an automatic deal-breaker. However, in this case, there appears not to have been a thorough vocal analysis to identify possible additional characters. More troubling are Mort’s comments regarding intra-population variation in note shape, and the sense that this variation was not only overlooked, but that the most divergent examples were cherry-picked to make the case, without the use of blind tests to confirm diagnosability for the greater sample. The authors may well be correct in their three species conclusions, but I don’t think the published evidence proves their case. Mort’s call for a quantitative vocal analysis, more comprehensive sampling, and blind tests to impartially judge qualitative characters such as note shape are well-reasoned, and may well provide the proof justifying the proposed taxonomic changes.”

Comments from Robbins: “NO. Given Mort’s comments on treating oreni and heteroptera as species.”

Comments from Pacheco: “NO. A partir das incertezas na análise e na distribuição da amostragem vocal, evocadas pelo Mort, debatidas em detalhes por Van e Kevin, sobretudo com relação à variabilidade dentro das populações, forçoso é compartilhar o voto com os demais colegas do Comitê.”

Comments from Jaramillo: “NO – based on Isler’s comments, as I understand it the vocal data sample is not large enough to take into consideration individual variation, furthermore a wider analysis of various aspects of vocalizations in the group are necessary to confirm the results of the paper.”