Proposal (1067) to South American Classification Committee

 

 

Treat Thamnophilus cearensis as a separate species from Thamnophilus caerulescens

 

 

Background:

The Variable Antshrike (Thamnophilus caerulescens) currently includes eight subspecies distributed from Peru to eastern Brazil. These taxa have traditionally been treated as conspecific due to general plumage similarity and apparent clinal variation among some populations (Zimmer and Isler 2003).

 

Previous studies focusing on the western subspecies aspersiventer, paraguayensis, and dinellii indicated weak reproductive isolation (Brumfield 2005, Isler et al. 2005, Marcondes et al. 2020). A range-wide genetic analysis, however, found that the geographically isolated cearensis, from northeastern Brazil, is deeply divergent from all other taxa (Bolívar-Leguizamón et al. 2020). Nonetheless, both this and other large-scale studies (e.g., Marcondes et al. 2020 for plumage) suffered from sparse geographic sampling, preventing a thorough taxonomic assessment. Moreover, vocal variation has never been comprehensively studied across the species’ range, despite the widely perceived importance of vocalizations in antbird taxonomy.

 

New Information:

Lima et al. (2025) conducted an integrative analysis of genome-wide nuclear data, plumage, morphology, vocalizations, and behavioral responses to song playbacks across the full range of the Variable Antshrike, using much denser geographic sampling than any previous study.

 

They identified nine phenotypically distinct populations, corresponding to seven of the eight currently recognized subspecies plus two newly described ones. The following figure illustrates the geographic ranges of the nine taxa recognized by Lima et al. (2025):

 

 

 

 

Crucially, Lima et al. (2025) confirmed that cearensis is noticeably divergent genetically from all other taxa (FST on the order of 0.24 to 0.38 between cearensis and other taxa) and also that it has a remarkably distinct song, similar only to that of the geographically distant melanchrous. In field playback experiments, cearensis individuals responded to song recordings of its own taxon and melanchrous while ignoring others.

 

Recommendation:

The combination of distinct song, discrimination in playback experiments, and deep genomic divergence indicates that cearensis has likely evolved substantial reproductive isolation from all other populations of the T. caerulescens complex. While song-based behavioral isolation between cearensis and melanchrous appears weak, the two show the highest pairwise FST values, suggesting that postzygotic genetic incompatibilities may be strongest between them.

 

Although direct tests of postzygotic isolation are unavailable, genetic divergence provides a reasonable proxy. The number of Dobzhansky–Muller incompatibilities is expected to increase exponentially with genetic distance between two populations (Orr and Turelli 2001, reviewed in Coyne and Orr 2004). Consistent with this expectation, Lima et al. (2025) found that parapatric subspecies with lower genetic differentiation tend to form broader hybrid zones than those with deeper genetic divergence (see their Fig. 12). Thus, the high genetic differentiation between cearensis and melanchrous (FST = 0.38) likely indicates meaningful postzygotic barriers. Inferring reproductive isolation from this pattern is no more speculative than inferring behavioral isolation from territorial responses in playback experiments.

 

Based on this evidence, I recommend a YES vote to treat Thamnophilus cearensis as a separate species from Thamnophilus caerulescens.

 

References:

Bolivar-Leguizamón, S. D., Silveira, L. F., Derryberry, E. P., Brumfield, R. T., & Bravo, G. A. (2020). Phylogeography of the Variable Antshrike (Thamnophilus caerulescens), a South American passerine distributed along multiple environmental gradients. Molecular Phylogenetics and Evolution, 148, 106810.

Brumfield, R. T. (2005). Mitochondrial variation in Bolivian populations of the Variable Antshrike (Thamnophilus caerulescens). The Auk, 122(2), 414–432.

Coyne, J. A., & Orr, H. A. (2004). Speciation. Sinauer Associates.

Isler, M. L., Isler, P. R., & Brumfield, R. T. (2005). Clinal variation in vocalizations of an antbird (Thamnophilidae) and implications for defining species limits. The Auk, 122(2), 433–444.

Lima, R. D., Bolívar-Leguizamón, S. D., Bocalini, F., Marcondes, R. S., Brumfield, R. T., & Silveira, L. F. (2025). Geographic variation, population genetic structure, and taxonomy of the Variable Antshrike (Thamnophilus caerulescens). Zoological Journal of the Linnean Society, 205(2), zlaf129.

Marcondes, R. S., Stryjewski, K. F., & Brumfield, R. T. (2020). Testing the simple and complex versions of Gloger’s rule in the Variable Antshrike (Thamnophilus caerulescens, Thamnophilidae). The Auk, 137(3), ukaa026.

Orr, H. A., & Turelli, M. (2001). The evolution of postzygotic isolation: accumulating Dobzhansky‐Muller incompatibilities. Evolution, 55(6), 1085-1094.

Zimmer, K. J. & Isler, M. L.  (2003). Family Thamnophilidae (Typical Antbirds). In: del J Hoyo, A Elliot, DA Christie (eds), Handbook of the Birds of the World. Barcelona: Lynx Edicions, 448–681. 

 

 

Rafael D. Lima, November 2025

 

 

Note from Remsen on English names: Cory & Hellmayr (1924) called this taxon “Ceará Ant Shrike”.  I suggest we simply adopt that sensible name (as Ceara Antshrike); if anybody has any better ideas, speak out – and do a separate proposal for the English name.

 

 

 

Vote tracking chart:

https://www.museum.lsu.edu/~Remsen/SACCPropChart1044+.htm

 

 

Comments from Remsen: “YES. Vocal differences with playback experiments make this one an easy decision, in my opinion.  That cearensis is the most isolated taxon, in the complex, geographically, also makes sense.”

 

Comments from Gustavo Bravo (guest voter): “YES. Given the vocal, morphological, and genetic evidence, it is appropriate to recognize Thamnophilus cearensis as a separate species.”

 

Comments from Zimmer: “YES. Way back when the Islers were actively looking into geographic variation in the T. caerulescens-complex (more focused on the western array of taxa), and in the lead-up to Mort and I doing the antbird chapter for HBW, I suggested to Mort that the taxon in the complex to be most focused on was cearensis, which was not only the most isolated taxon, but which, to my ears, was vocally distinct from all other Brazilian taxa in the complex, and which showed the most divergence in female plumage as well.  I even did some informal, one-way playback trials in the Serra de Baturité, employing only songs of nominate caerulescens and gilvigaster, and found, like Lima et al (2025), that cearensis did not respond to playback of either caerulescens or gilvigaster vocalizations, but did respond aggressively to playback of pre-recorded vocalizations of other cearensis.  I never did reciprocal playback trials of cearensis songs to either caerulescens or gilvigaster populations, nor did I do any trials involving any of the western taxa in the caerulescens-complex.  So, nothing as rigorous or comprehensive in scope as what Lima et al (2025) have published, but this is just to say that all of my own fieldwork with this complex is in line with their conclusions.  Assuming this passes, I would also be strongly in favor of Van’s suggested English name of “Ceara Antshrike” for T. cearensis.”

 

Comments from Naka: “YES. I agree with the new treatment, which was long overdue.”

 

Comments from Stiles: “YES - evidence from genetics, phenotype, and distribution justify this split.

 

Comments from Robbins: “YES, given the new data.”

 

Comments from Lane: “YES. It looks like all datasets agree with splitting these two groups.”

 

Comments from Areta: “A fascinating system and a great study by Lima et al. 2025. I am however mostly not satisfied with the taxonomic decisions of the work. It is of course impossible to tackle all the fronts of this widespread species at once, and while some results are clear, others are not (not to me at least).

 

“cearensis as a species: plumage differences vs the closest taxon (newly described bahiae) are very minor and limited to females. The only "Isler-Whitney" difference between the geographic closer cearensis and bahiae is on note shape, which falls short of the 3 such characters that set the standard level of differentiation in order to be considered different species in the family by the same Isler-Whitney team. The playback experiments (Table S3) are suggestive, but overall inconclusive: nearly 50% of the caerulescens tests did not discriminate, cearensis could not discriminate against melanchrous, and the number of playback stimuli indicates rampant problems of pseudoreplication by lack of adequate number of stimuli. More rigorous playback experiments are needed. Of course, one can choose to ignore methodological problems and take an "overall" message, but I don´t think that the overall message is that clear either. The Fst between the newly described bahiae and cearensis is on par with those between other populations (especially comparisons involving the other geographically extreme taxon melanchrous) which we are not considering as a species-level candidate. The deepest differentiation between melanchrous and cearensis (the two more distant and disconnected taxa) is quite high, but they are bridged by intermediate populations. Finally, do the phylogenetic trees recover bahiae and cearensis as reciprocally monophyletic? It seems from Figure 11, that cearensis is embedded within bahiae. I might be alone in this world, but I vote NO to the split of cearensis. I am comfortable with considering it as an allopatric population of the widespread caerulescens, being similar vocally and in plumage to its closest (yet allopatric) neighbour, bahiae. It is clear that different populations show different degrees of isolation and hybridization, and cearensis is the most extreme one but not yet commensurate to species level differentiation by Thamnophilus standards.

 

“The two new subspecies described have diagnostic females, but not diagnostic males, which cannot be distinguished from neighbouring populations. Given the impressive number of specimens analyzed, I would have expected heavy doses of photographs instead of drawings. Not even the holotypes were pictured! I am not enchanted by the descriptions.

 

“atricapillus (new taxon #1): This taxon seems hard to justify, despite its apparent bold and easily diagnosable female aspect, in which "They are closest in appearance to females of caerulescens and gilvigaster, from which they differ by having a black crown" (Lima et al. 2025: 16). First of all, this subspecies would have a minute range (or even no proper range), and furthermore Lima et al. 2025 (16) stated that "we documented plumage intergradation between caerulescens and new taxon #1 in central Brazil (Fig. 7). Although the contact zone appears narrow (∼31 km), phenotypic intergradation spans the entire range of new taxon #1 (Fig. 7). ". So, if phenotypic intergradation spans the entire range, then I don´t see how we can delimit a range for this taxon whose full range is an intergradation zone between itself (an intergrade) and another taxon. Confusing. In looking at the specimen table (Table S1) I find 20 females "f_new1" phenotype individuals and 52 males "m_new1_or_caerulescens" phenotype individuals. Although the diagnosis indicates the taxa from which males are diagnosable, it does not mention those from which they are not. If the phenotypic intergradation spans the entire range of the new taxon, then it seems like this taxon does not have an independent existence and is therefore questionable and fuzzy. I vote NO to the recognition of atricapillus, mostly because of how confusingly the situation was described.

 

“bahiae (new taxon #2): the females are extremely similar to those of cearensis, and males are indistinguishable from the latter. Females of bahiae "are closest in appearance to females of cearensis, from which they differ by being more brownish and less olivaceous on the belly, breast, back, and wing coverts" (Lima et al 2025: 16). I can accept this taxon, on account of the difference of females from the neighbouring caerulescens.”

 

Additional comments from Remsen: “After reading Nacho’s much more thorough comments on this, I feel bad for not having seen those points myself, and change my vote to NO.  Also, this has no bearing on my view of the quality of the paper itself – in general, the quality of the paper is independent from the taxonomic conclusions from it.”

 

Comments from Claramunt: “Reluctant YES. I truly commend Rafael and co-authors for such a thorough study with great sample sizes and an integrative approach. But there is no clear picture emerging from the data and analyses regarding the species status of cearensis. The genetic analyses presented in this study are not totally convincing. It is mostly based on methods borrowed from numerical taxonomy (PCA, k-means clustering). I would like to see more model-based phylogenetic and population genetic analyses of this complex to determine potential relationships and levels of gene flow. PCA on discrete data such as SNPs always tends to exaggerate differences among group. The phylogenetic network in Figure 11 is the closest to a phylogenetic analysis and the long branch leading to cearensis suggests that it might be the most basal branch, but without outgroup, Nacho’s interpretation of cearensis deriving from bahiae is a plausible alternative. Bolívar-Leguizamón et al. produced a mtDNA tree in which cearensis is indeed the most distantly related lineage in the complex.

 

“According to the taxonomic summary, females of cearensis“ differ diagnostically in plumage from all other taxa” but I could not find the evidence for this statement. The cited Fig. 1 shows only illustrations of idealized individuals in which females of cearensis and bahiae look indistinguishable. The diagnosis of bahiae indicates that they are diagnostically browner, less olivaceous, than cearensis, but the term “diagnostic” is used very liberally in this study with most “diagnosable taxa” having wide “transition” zones and polymorphisms that are incompatible with strict diagnostic status.

 

“However, all things considered, the phylogenetic data plus partial diagnosability in plumage and song mildly suggest that cearensis may be better treated as a full species.”

 

.