Species limits in Thamnophilus ruficapillus

Proposal (968) to South American Classification Committee

Species limits in Thamnophilus ruficapillus

Our current SACC notes read:

"7c. Short (1975) suggested that the Andean subspecies (marcapatae and jaczewskii) might warrant recognition as a separate species from Thamnophilus ruficapillus, but see Ridgely & Tudor (1994).

7f. Plumage and vocal characters strongly suggested that Thamnophilus ruficapillus and T. torquatus should be placed next to the T. doliatus group in linear sequences (Ridgely and Tudor 1994), and this change in sequence was made by Zimmer & Isler (2003), who considered them to form a superspecies. Genetic data (Brumfield & Edwards 2007) strongly support this, with T. ruficapillus and T. torquatus are sister species and closely related to T. doliatus. SACC proposal passed to change linear sequence of species."

The Rufous-capped Antshrike T. ruficapillus is polytypic, with the following 5 subspecies – and overall distribution:

Atlantic Forest and Parana/del Plata basin:

• T. r. ruficapillus Vieillot, 1816 – E Paraguay and SE Brazil (S from E Minas Gerais and Espírito Santo) S to NE Argentina (Misiones, Corrientes, Entre Ríos, NE Buenos Aires) and Uruguay.

Andes (from N to S)

• T. r. jaczewskii Domaniewski, 1925 – Andes of N Peru (C Cajamarca, Amazonas S of R Marañón, NW San Martín).

• T. r. marcapatae Hellmayr, 1912 – E slope of Andes of S Peru (E Cuzco, Puno).

• T. r. subfasciatus P. L. Sclater & Salvin, 1876 – E slope of Andes of NW Bolivia (La Paz, W Cochabamba)

• T. r. cochabambae (Chapman, 1921) – E slope of Andes of C Bolivia (E Cochabamba and SW Santa Cruz) S to NW Argentina (Jujuy, Salta, Tucumán).

Previous treatments

Short 1975 (pages 262-263) wrote:

"?Thamnophilus ruficapillus

Rufous-capped Antshrike

Taxonomy. Polytypic. Forms a superspecies with Andean marcapatae (disjunct populations of latter in northern and in southeastern Peru). Thamnophilus marcapatae is darker, grayer, and less pale rufescent than ruficapillus, and females are very dark (rufous) below, not whitish-buff as in ruficapillus.

Ecology. Found in forest undergrowth and in canopy. Nonmigratory.

Distribution and variation. The superspecies is endemic in central South America. Thamnophilus ruficapillus occurs from northern Bolivia S along the Andes to Tucumán, and, disjunctly, from Espirito Santo and eastern Paraguay S to northern Buenos Aires and Uruguay. It occurs outside of the chaco, but is very likely to reach the western fringes of the chaco in Salta or western Paraguay, and, in the E, it possibly may reach the W bank of the Parana River in the Santa Fe chaco. The eastern nominate ruficapillus is more rufous and less gray than the western races. Of the two western forms, cochabambae occurs from central Bolivia S, and is the one likely to occur in the western chaco fringes. It is lighter gray and less fully barred below than is west-central Bolivian subfasciatus."

Ridgely & Tudor (1994) remarked on the vocal similarities, highlighted the distinctiveness of marcapatae and suggested that subfasciatus was intermediate between the northernmost Andean taxa and cochabambae, proposing that variation was clinal:

Pajarografo Sólido:Users:javierareta:Desktop:R&T 1994 Thamnophilus ruficapillus 1-2.png

Pajarografo Sólido:Users:javierareta:Desktop:R&T 1994 Thamnophilus ruficapillus 2-2.png

The BirdLife split:

Del Hoyo & Collar (2016), see also del Hoyo et al. (2017), split the three north Bolivian/Peruvian taxa (subfasciatus, marcapatae and jaczewskii) as the Northern Rufous-capped Antshrike T. subfasciatus with the following argument:

"Belongs to the “T. doliatus group” (which see). Hitherto treated as conspecific with T. ruficapillus, but differs in its denser barring on breast, this extending to belly and onto throat, in male (3); rufous vs whitish below in female (3); shorter tail (effect size 2.79, score 2). Three subspecies recognized."

Genetics

A series of genetic studies, using different combinations of taxa, are consistent among them and point to relatively deep divergences among taxa:

• Brumfield & Edwards (2007) found that a sample of cochabambae and one of torquatus were sister (3.6% divergent in ND2, 5.2% in Cyt b).

• Kerr et al. (2009) found a deep (ca. 4%) divergence in COI between cochabambae and ruficapillus with two samples per taxon:

Pajarografo Sólido:Users:javierareta:Desktop:Kerr et al 2009 T. ruficapillus Fig 2.png

More recently, with better geographic sampling but just one sample per taxon Harvey et al. (2020) found T. torquatus as sister to nominate ruficapillus, whereas cochabambae was sister to marcapatae and jaczewskii (lacking samples of subfasciatus):

Importantly, note that according to this tree T. ruficapillus would be paraphyletic whether we follow the traditional arrangement or whether we decide to split it along BirdLife lines. With all this in mind, one could solve the lack of monophyly of T. ruficapillus either by moving cochabambae as a taxon within subfasciatus or by following a more drastic course and performing a three-way split, separating cochabambae (monotypic) and subfasciatus (polytypic) from a monotypic ruficapillus. However, (a) Harvey et al. (2020) had no genetic samples of nominate subfasciatus, (b) there is just one sample per taxon, and (c) no rigorous vocal studies have been conducted.

Vocalizations

The vocalizations of two undisputed species, T. torquatus and T. ruficapillus, are quite similar. However, no formal quantitative studies have been published to date for this complex. Mark and I have recorded cochabambae and ruficapillus extensively, in the hope of being able to sit down and analyze the data properly. Although we still have to do that, our preliminary analyses (Areta & Pearman, unpublished) indicate that there are some differences, but the differences are far from obvious: birds in the Andes of NW Argentina (cochabambae) deliver faster paced songs than lowland birds in NE Argentina (ruficapillus). This is an interesting study case to apply the Isler et al. (1998) criteria, that has been instrumental to most taxonomic changes in Thamnophilidae during the last 25 years.

Plumage

In terms of plumage, the subfasciatus group is the most distinctive, given its extensive and profuse ventral barring, grey flanks and back, and grey face (https://macaulaylibrary.org/asset/204867361), whereas cochabambae has a greyish face (https://macaulaylibrary.org/asset/205684661) that distinguishes it from the browner-faced, yet very similar ruficapillus (https://macaulaylibrary.org/asset/68731871). Overall, despite variation, cochabambae has less extensive ventral barring and is paler dorsally than ruficapillus. Ridgely & Tudor (1994) (see above) suggested that plumage varied clinally in Andean taxa.

Brief analysis

Given the apparent vocal conservatism in the ruficapillus/torquatus clade, we think that a full comparative study is necessary to correctly sort out the species limits here. More genetic samples of all taxa (especially including subfasciatus) would also strengthen the case. Although the plumage differences (the main features used to describe the taxa) seem well established, a more thorough characterization of geographic variation would also be welcome, especially to assess whether variation is clinal across the Andean taxa cochabambae-subfasciatus-marcapatae-jaczewskii. The BirdLife treatment was based just on plumage and tail length: features that can hardly be used as solid characters in antbird species-level taxonomy, given the number of well-studied widespread (biological) species with striking plumage and morphological differences that exhibit clinal variation.

Voting options

A) Treat T. ruficapillus as a single, polytypic species, including the 5 traditional subspecies --- The traditional treatment.

B) Split T. ruficapillus (with ssp. ruficapillus and cochabambae) from T. subfasciatus (with ssp. subfasciatus, marcapatae and jaczewskii) --- The BirdLife treatment.

C) Split T. ruficapillus (monotypic), and T. subfasciatus (with ssp. subfasciatus, marcapatae, jaczewskii and cochabambae) --- A different 2-species option that achieves monophyly for both: there would be one Western (Andes) species and one Eastern (Atlantic Forest and Parana/del Plata river basin) species

D) Split T. ruficapillus (monotypic), T. cochabambae (monotypic) and T. subfasciatus (with ssp. subfasciatus, marcapatae and jaczewskii) --- A 3-species option that has not been suggested before in the literature to our knowledge.

Recommendation

We recommend a YES to option A and NO to options B, C, and D. More studies, especially including genetic samples of nominate subfasciatus and rigorous, large-scale comparative vocal analyses are needed to sort this complex out adequately. Collectively, the data indicate that the current single-species treatment seems erroneous, but we lack enough data to make fully informed improvements to our classification.

Additional references

del Hoyo, J., and N. J. Collar (2016). HBW and BirdLife International Illustrated Checklist of the Birds of the World. Volume 2. Lynx Edicions, Barcelona, Spain.

del Hoyo, J., N. Collar, and G. M. Kirwan (2017). Northern Rufous-capped Antshrike (Thamnophilus subfasciatus). In Handbook of the Birds of the World Alive (J. del Hoyo, A. Elliott, J. Sargatal, D. A. Christie, and E. de Juana, Editors). Lynx Edicions, Barcelona. Retrieved from Handbook of the Birds of the World Alive: https://birdsoftheworld.org/hbw/species/rucant4/1.0

Kerr KCR, Lijtmaer DA, Barreira AS, Hebert PDN, Tubaro PL. (2009) Probing evolutionary patterns in Neotropical birds through DNA barcodes. PLoS ONE 4(2): e4379. doi:10.1371/journal.pone.0004379

Juan I. Areta & Mark Pearman, February 2023

Comments from Remsen: “YES to A and NO to B, C, and D, as recommended in the proposal. I strongly agree with the theme of the analysis: this is a complex situation that really needs a thorough study of voice and potential contact zones, with multiple genetic samples from strategic locations. Making splits based on differences in tail length and plumage in antbirds is just silly, and using genetic distance is problematic, for conceptual and practical reasons (that I’ve elaborated on in many previous proposals; further, the reported distances in the taxa sampled are routine in Neotropical birds for taxa ranked as subspecies for lack of vocal differences or evidence of free gene flow in contact zones).”

Comments from Stiles: “YES to A- maintain a polytypic T. ruficapillus pending more conclusive data supporting a split.”

Comments from Del-Rio: “YES to A, and NO to B, C and D. Waiting on an updated phylogenetic tree to clarify whether ruficapillus would be polyphyletic, for now, keep the status quo.”

Comments from Zimmer: “YES to Option A (continue to treat T. ruficapillus as a single, polytypic species, including the 5 traditionally recognized subspecies.), and NO to options B, C and D. Although the incomplete genetic data, and striking differences in both male and female plumage of some of the Andean taxa versus nominate ruficapillus are certainly suggestive of more than one species-level taxon being involved, I agree completely with Nacho and Mark that more genetic, vocal, and morphological/plumage analyses are needed from throughout the Andean distribution of taxa to make sense of the variation, and, to rule out clinality in morphological and vocal characters. This case actually reminds me a lot of the situation with Thamnophilus caerulescens, which Mort Isler and I grappled with while working on the Thamnophilid chapter for Volume 8 of HBW. As is the case with ruficapillus, caerulescens is a polytypic species (I think Peters recognized 12 ssp., but Mort and I only recognized 8.), with a broad, and somewhat unusual distribution that includes a chain of multiple Andean taxa, as well as Atlantic Forest representatives ranging from NE Argentina, SE Paraguay, Uruguay, and SE Brazil (and even to NE Brazil in Ceará and Pernambuco, so even more extensive than the distribution of ruficapillus). As with ruficapillus-group, the caerulescens-group encompasses a considerable amount of plumage variation in both male and female plumages, with males of some subspecies-pairs and females of others being exceptionally dichromatic. Also, as in the ruficapillus-group, vocalizations of the caerulescens-group are remarkably consistent across such a broad range, particularly given the extent of plumage variation within the group. Our conclusion, when researching the antbird chapter for HBW, was that, in spite of the apparent discontinuities in diagnostic plumage characters separating many of the described subspecies, that it would take a broadly sampled quantitative vocal analysis to demonstrate any diagnostic vocal characters, and to eliminate the possibility that perceived distinctions in pace of the loudsongs (which was the only vocal difference between populations that we could discern from a qualitative “ear test”) between some populations did not differ clinally. A couple of years later, (2005) the Islers and Robb Brumfield published just such a study of caerulescens, albeit one dealing only with the Andean taxa, and not with any of the Atlantic Forest populations, and found that vocalizations did vary clinally across named populations. In fact, as I recall, the only diagnostic vocal character they could find was pace of loudsong (maybe also number of notes?), and even that single character differed only between populations at either end of the cline, with no vocal characters allowing diagnosis of any two adjacent populations. I thought of this as soon as I read Nacho’s comments in the Proposal regarding the preliminary results of the vocal study that he and Mark have been doing on ruficapillus and cochabambae – ‘some differences…far from obvious…cochabambae delivering faster-paced songs than ruficapillus.’

“All of this having been said, I still suspect that a similar, thoroughly sampled genetic, vocal and plumage analysis of the various eastern (NE Brazil south to Argentina, Paraguay and Uruguay) taxa in the caerulescens-group would support recognition of at least one species distinct from the Andean populations. Similarly, I would not be surprised if such a detailed analysis across all populations of the ruficapillus-complex would justify some splitting of taxa too, particularly since there is a vocally similar sister species (T. torquatus) whose range basically fills the gap between the range of nominate ruficapillus and that of the other subspecies. But the point is, that it will take robust analyses such as the molecular and vocal studies of caerulescens by Brumfield (2007) and Isler et al (2005), not simplified, point-based scoring systems of plumage or tail length, to resolve the situation in my opinion.”

Comments from LANE: A) YES. NO to the remaining options. Until it is clear with which group cochabambae falls, and what (if any) characters vocal analyses can unearth, I think we should not make changes that could require reversals once studies are performed. One issue I think is overlooked in this proposal is the status of T. torquatus with respect to the rest of T. ruficapillus. In the proposal, they are called "undisputed species" but I wonder? I will point out that T. torquatus and T. ruficapillus (nominate group) have extremely limited overlap, and with birds that look like this

at sites where both species reportedly occur makes me wonder just how “undisputed” the species status of T. torquatus really is? Other than the black cap of the male, I don’t see or hear a whole lot that distinguish the two… and is that black cap as important a feature in the birds’ eyes as it seems to be for us? My point is: I am not entirely convinced that we should be leaving T. torquatus out of this complex simply because it looks so distinctive to human eyes. It may be best inserted into T. ruficapillus, and then the paraphyly of the latter is no longer! According to Wagner Nogueira, it seems that the present-day contact of the two taxa is thanks to recent anthropogenic habitat alteration, so a broad region of contact and overlap is not to be expected, and the above evidence of potential interbreeding could be a precursor of a larger future swarm.”

Comments from Rafael Lima: “I agree with Daniel that it may be best to lump T. torquatus with the T. ruficapillus complex, at least provisionally while thorough genetic and acoustic trait analyses are not available. Although this possibility has not been considered in the original proposal, it would equally eliminate paraphyly.

“I have recently cursorily examined dozens of specimens of almost all taxa of the complex at the AMNH, LSUMNS, and MZUSP, and my impression is that if two biological species were to be recognized in this complex for the time being, based on morphology alone, these species should be the eastern taxa (ruficapillus + torquatus) comprising one species and the Andean taxa (subfasciatus + marcapatae + jaczewskii + cochabambae) comprising the other species. Subspecies cochabambae is difficult to place in one of the two morphologically-delimited species, but I think this arrangement makes much more sense than treating the Andean taxa plus ruficapillus as one species and torquatus as another. This arrangement would also result in two monophyletic species taxa. Nevertheless, I think that based on current evidence (including the many photos on WikiAves suggesting that torquatus and ruficapillus hybridize freely where they come into contact), the best option is to treat all of them as a single biological species until a thorough study of geographic variation is available.”

Comments from Robbins: “YES to A, no to B, C, D for reasons outlined in the proposal and based on comments by other committee members.”

Comments from Claramunt: “Option A. The situation is clearly complex and requires a detailed analysis of the evidence. Nothing new has been published. NO to B,C,D.”

Comments from Bonaccorso: “A, YES. No to B, C, D. The problem with the last three proposals is not only the lack of vocal data (as the proposal indicates), but there is also a lack of good geographic and genetic sampling across the ranges of all taxa involved in the potential alternatives.”