Proposal (1023) to South American Classification Committee

Recognize Trichothraupis griseonota as a separate species from T. melanops

Our current SACC note reads:

"14f. The Andean population of Trichothraupis melanops was described as a new species, Trichothraupis griseonota, by Cavarzere et al. (2024). SACC proposal badly needed."

Cavarzere et al. (2024) described the Andean-foothill populations of T. melanops as a new taxon at the species level under the name T. griseonota, following the PSC (but arguing that the same interpretation would be possible under the BSC). They based their conclusions on the examination (and measurement) of 314 specimens from the Atlantic Forest population, and 52 from the Andean-foothill population. They also examined 520 sound recordings of the Atlantic Forest population, and 19 from the Andean-foothill population.

The current distribution of both taxa is more extensive than shown in their specimen map (Figure 5, below; e.g., there are overlooked documented photographic, sound, and published specimen records in NW Argentina, photographic records in the tall, riverine forests in eastern Formosa in NE Argentina, and numerous records along the Paraguay river).

Earlier, Zimmer (1947; note number 51, and not 52 as cited in Cavarzere et al.) referred to geographic variation in Trichothraupis stating that

"There is a slight possibility that distinctions might be found in Peruvian specimens as compared with Paraguayan and Brazilian material, but the range of variation in the species is so great (126 specimens examined) that little is to be expected. Two adult males from Bolivia (Province of Sara) have a greater posterior extension of the black area on the sides of the face than any other males at hand from any locality, but there is considerable variation within this extreme limit. These two males also have the longer upper tail-coverts decidedly blackish, with narrow olive tips, but this character is present in a small number of Brazilian and Argentine birds, also, and is of doubtful taxonomic value.”

Prior to this, Hellmayr (1936) stated in a footnote:

"The supposed distinction of a northern form (auricapilla), which we at one time advocated, has not been corroborated by additional material since. Birds from Espirito Santo and Rio de Janeiro seem to be precisely like others taken in southern Brazil and Paraguay. Bolivian and Peruvian specimens merely differ by very slightly paler under parts, but the divergency appears to me too insignificant to justify its recognition in nomenclature, although there is obviously a wide gap between the eastern and the Andean ranges of the species."

Therefore both, Zimmer and Hellmayr discussed slight plumage variation in T. melanops.

The holotype of griseonota is "MACN 8979a, Male. BOLIVIA, Buenavista, Santa Cruz, 30 July 1916, J. Steinbach leg., 450 m". So, how is it possible that the holotype is not shown in the paper itself or in the supplementary material? It is not too much asking to show photographs of a holotype in the 21st century!

Cavarzere et al. (2024) provide the following diagnosis:

"Diagnosis. Four fixed plumage traits and one morphometric trait diagnose Trichothraupis griseonota from its sister species, T. melanops. The first and most noteworthy is the black facial mask. In the new species, it includes the auricular region (Fig. 1; Fig. S3), whereas in T. melanops this mask is only a narrow line behind the eye, not reaching the auriculars. In a few Atlantic specimens (AMNH 774505, LSUMZ 59405, MZUSP 27851) there is some black in the auriculars, but in those cases, it is mixed with green, producing a mottled appearance, unlike the homogeneous black in T. griseonota. A second diagnostic character is the paler underparts of adult males and females. This is especially evident in the undertail coverts, which are a cream color to buff-yellow in T. griseonota, versus cinnamon in T. melanops, and in the chest, in which both species have a more orange tone than in the belly, but less distinctly so in the new species, where the color of the chest is more buff-yellow, versus buff in T. melanops (Fig. 1). Adult females of T. griseonota are also somewhat paler in the underparts than in T. melanops, but the difference is not as pronounced as in males, and sexual dimorphism of plumage color in the underparts is more evident in T. griseonota than in T. melanops. The third plumage characteristic is the color of the back of adult males, which is greyer in T. griseonota, versus being more greenish-olive in T. melanops (Fig. 1; Fig. S3). The last character in which the two species differ is the breast coloration of females, which is subtly but consistently darker in Atlantic populations (Fig. 2). Trichothraupis griseonota also has significantly shorter tarsi compared to T. melanops (Fig. 4)."

Looking at these photographs we see minimal plumage differences showing very slight differences in dorsal coloration, even more subtle differences in ventral coloration, and a noticeable (yet variable) difference in the amount of black in the goggle.

"Morphometrics. The Andean population has significantly shorter tarsi, on average, than the Atlantic population, but all traits show measurement overlap. The mean measurements of both populations are shown in Table 1. Principal Component Analysis indicates that morphometrics segregate the two populations, although incompletely (Fig. 3). Within this context, the two-factor comparison of sex and populations only showed significant differences for culmen (p<0.010) and wing (p=0.000) measurements between sexes, while tarsi measurements (p=0.000) were significantly shorter among Andean populations (Table 1). There were no significant differences in the interactions between them (Bonferroni post hoc test p>0.02). Therefore, the Andean population is morphologically distinguished based on plumage color patterns and tarsus length (Table 2)."

The PCA show wide overlap between the taxa (i.e., no segregation between taxa when sorted by sex), and univariate analyses (for which see their Table 1) do not show well-marked differences (although the authors claim a difference in tarsus length, the effect size for such a comparison is small). Thus, morphometry show at most subtle differences between the taxa.

Despite the large number of available recordings, the vocal analyses included only two recordings of songs of the Andean-foothill population:

"Vocalizations. Of the recordings of the Andean population, only two were loudsongs, and only one consisted of a reasonable-quality recording to allow for the creation of spectrograms. The lack of Andean data precluded us from quantitatively analyzing loudsongs. Qualitative inspections failed to find visual differences between contact calls (Fig. S1). The Atlantic loudsong is clearly composed of a series of upslurred and downslurred notes, forming a “W” shape. It seems to share a two-note homologous pattern with the Andean loudsong, but the number of loudsong emissions within the recording did not allow for further evaluation (Fig. S2)."

"Contact calls are visually identical in spectrograms and sound quite similar for both allopatric populations, but for most passerines, when sister clades are morphologically and genetically distinct, supposedly homologous loudsongs also typically differ quantitatively, including for Neotropical oscine and suboscine species (Isler et al. 1998; Bocalini & Silveira 2016). Therefore, we also recommend analyzing more loudsongs as more recordings of T. griseonota are made available. "

The calls of both populations are identical (our own recordings of both populations from Argentina further reaffirm that), and there was not enough vocal material to adequately characterize songs (which can be quite complex in T. melanops, therefore not an easy task to accomplish).

The genetic data used by Cavarzere et al. (2024) to support their species-level taxon comes from Trujillo-Arias et al. (2018). Most notably, their discussion bears on this issue:

"the divergence between these rainforests in a sample of three birds varied from 1.3 My (Amazona pretrei/A. tucumana, Rocha et al., 2014) to about 0.72 My and 0.15 My (T. melanops and Pipraeidea melanonota, this study and Lavinia, 2016, respectively). However, it should be noted that although the re-current connections between these biomes could have led to periodic re-establishment of gene flow between regions after the initial separation, the historical gene flow between these rainforests has not been high enough to preclude divergence.

“Finally, a definitive analysis of the phenotypic variation of T. melanops is necessary to determine the taxonomic status of its populations. Our results indicate that the Andean and Atlantic populations of T. melanops are genetically isolated (i.e., migration between populations M < 1 individuals per generation), and therefore, its current classification as a monotypic species might not be adequate to reflect its evolutionary history. "

Of course, these two being allopatric populations, one would not expect much reciprocal genetic influx. This being said, the genetic differentiation between the Atlantic Forest and Andean-foothill populations is very low.

Recommendation: We recommend a NO vote. The reduced levels of phenotypic and genetic differentiation coupled with a lack of vocal differences in calls (and unknown, but apparently trivial levels of vocal differentiation in song; pers. obs.) indicate to us that griseonota can be a good, mildly differentiated subspecies of T. melanops but not a separate species from it.

References

Cavarzere, V., Costa, T.V.V., Cabanne, G.S., Trujillo-Arias, N., Marcondes, R.S. & L.F. Silveira. (2024) A new species of tanager (Aves: Thraupidae) from the Eastern slopes of the Andes. Zootaxa 5468: 541–556.

Hellmayr, C.E. (1936) Catalogue of birds of the Americas, part IX, Tersinidae, Thraupidae. Field Museum of Natural History Publications in Zoology 13:1–458.

Trujillo-Arias, N., Calderón, L., Santos, F.R., Miyaki, C.Y., Aleixo, A., Witt, C.C., Tubaro, P.L. & Cabanne, G.S. (2018) Forest corridors between the central Andes and the southern Atlantic Forest enabled dispersal and peripatric diversification without niche divergence in a passerine. Molecular Phylogenetics and Evolution 128: 221–232.

Zimmer, J.T. (1947) Studies of Peruvian birds. No. 51. The genera Chlorothraupis, Creurgops, Eucometis, Trichothraupis, Nemosia, Hemithraupis, and Thlypopsis, with additional notes on Piranga. American Museum Novitates 1345.

Juan I. Areta and Mark Pearman, July 2024

Comments from Remsen: “NO. The authors did a great job of quantifying the differences between the two populations, especially the newly discovered diagnostic plumage traits of the two populations. So, the data strongly justify describing the Andean population as a new taxon … but not as a species under the BSC but rather as a subspecies. This designation calls attention to their separate evolutionary history by formal recognition through taxonomy of this level of biodiversity. What is needed now, in my view, is a study to show that these two taxa have diverged to the point of “no return” in terms of divergence in the comparative framework of Thraupidae divergence through a formal analysis of song differences. Strictly on the basis of genetic differences and time-since-separation, the data suggest a better fit with subspecies-level divergence (although I will spare us my usual diatribe on using genetic divergence as THE arbiter of species vs. subspecies assignment.”

Comments from Stiles: “NO. To me, the differences in plumage between griseonota and melanops are sufficient for recognizing the former as a good subspecies, but fall short of favoring its status as a separate species, and the N = 1 of its loudsong likewise makes support for the split less convincing. The genetic information is also borderline and given the records (including specimens?) from Argentina and Paraguay that narrow appreciably the gap in distributions of the two, any such specimens should at least be subjected to genetic analyses.”

Comments from Bonaccorso: “NO. As Van and Gary pointed out, overall, the data support a subspecies, not a species. The plumage differences are subtle and the morphological data overlapping. Also, according to the phylogenetic analysis of Trujillo-Arias et al. (2018), the genetic differences are minor, and the two taxa are not reciprocally monophyletic. I don´t mean that reciprocal monophyly would be necessary, but it would make a stronger case for species status. Finally, the contact calls and loudsongs don´t seem very different (although I am not an expert), and not including data from individuals close to the potential contact zone is problematic.”

Comments from Robbins: “For the reasons Areta and Pearman state in their proposal, I vote "NO" for treating griseonota as a species.”

Comments from Del-Rio: “NO. Because of low genetic differentiation and slight morphological differences, griseonota should be treated as a subspecies.”

Comments from Lane: “A solid NO here. This taxon, while worthy of a name, is hardly distinctive enough to warrant splitting from T. melanops by BSC standards. This is why PSC species are a whole different ball of wax from those that most taxonomists would recognize as distinct.”

Comments solicited from Kevin Burns: “Cavarzere et al. (2024) is an exciting and important paper because it reveals previously unreported differences between Andean and Atlantic Forest populations of Trichothraupis melanops. It’s both surprising and fascinating that these differences remained underappreciated until now. In addition, the genetic study of Trujillo-Arias et al (2018) provides a nice complement. It is a thorough and rigorous study that includes phylogenies, population genetic analyses, and tests of demographic models. After reading both of these papers, I’m left feeling this is a borderline case, but see below.\

“The model testing of Trujillo-Arias et al (2018) inferred a scenario of peripatric divergence, with dispersal from the Atlantic Forest to the Andean Forest. This is also illustrated in their phylogeny, where there is not reciprocal monophyly between the two sets of populations. Instead, the Andean populations form a clade embedded within the Atlantic Forest populations. This is what you would expect in a case of recent peripatric speciation. The timing of the splitting event is estimated as 0.728 million years ago (mid-Pleistocene). On the surface, this seems low compared to other avian species splits. However, I went back to the data from Burns and Racicot (2009), and I found similar levels of divergence between sister species. In Burns and Racicot, we looked at phylogeny and biogeography of subfamily Tachyphoninae (Tachyphonus, Ramphocelus, Lanio, Eucometis, Trichothraupis, and relatives) using cyt b. Therefore, our dating is based on similar methods and markers as the Trujillo-Arias et al. study. Several recognized species splits in Burns and Racicot are of a similar age or even more recent than the Trichothraupis split. These include Lanio aurantius and L. leucothorax (0.594 mya), Ramphocelus carbo and R. melanogaster (0.515 mya), and Ramphocelus dimidiatus and R. nigrogularis (0.856 mya). Based on these data, the level of divergence seen between the two proposed species of Trichothraupis cannot be dismissed outright as “too low”.

“For the morphological data, there is quite a bit of overlap in the PCA plots of Cavarzere et al, but the two proposed species do generally fall out in different portions of the graphs. The authors also describe a significant difference in tarsus length, but I’m not sure what to make of this. Overall, I don’t think the morphology adds much to make the case for a species split.

“For the plumage, the two forms differ in ventral and dorsal color, which is obvious in the photographs. I wish we had quantitative (spectrophotometric) data comparing plumage color of all individuals so we could see any potential overlap or lack of overlap between the two forms. Also, an examination of the color differences using an avian visual model might be helpful when thinking about differences and how they might play out in courtship. Nevertheless, I am more convinced by the differences in the “goggles” – specifically the extent of melanin on the face mask. There is clear discrete difference in the facemask, with the Andean birds having black extending onto the auricular. The authors also note that the crest/crown has black framing in some of the Andean birds.

“Based on the genetics and plumage differences, these two forms are easily two species under the phylogenetic species concept.

“So what about the biological species concept? Well we need to think more about reproductive isolating mechanisms and try to predict what they would do if they came into contact. This is the spatial dimension problem that the BSC consistently faces for geographically isolated populations. We know that vocalizations are very important in songbirds for reproductive isolation. The authors note that both species have similar call notes, so that doesn't provide evidence for a split. For songs, there aren’t enough recordings of the Andean birds for comparison to Atlantic forest birds. The fact that the Andean birds apparently sing less (something also mentioned in the Islers’ tanager book) is intriguing. However, based on available information, vocalizations aren’t helpful in arguing these are two biological species.

“Since the genetics shows enough time to evolve reproductive isolation (as seen in related sister pairs) and the morphology and vocalization data are inconclusive, I think it comes down to the significance of those plumage differences. For a Yes vote on biological species, I think you would have to argue that the plumage difference are important in courtship. For a No vote on biological species, I think you would have to argue that those differences are inconsequential for courtship. I don’t think there is much, if anything, known about Trichothraupis courtship, but personally I haven’t been able to observe them much in the field. Overall, the differences in plumage seem slight compared to that seen in other sister pairs in this subfamily. Therefore, if I were asked to vote on this, I would probably vote No on biological species status for these two proposed species, although they would qualify under other species definitions.

“Regardless of species status, I think the Cavarzere et al. (2024) paper is an important one for understanding diversification and also demonstrates what discoveries await in museum collections!

Burns, K.J. and Racicot, R.A., 2009. Molecular phylogenetics of a clade of lowland tanagers: implications for avian participation in the Great American Interchange. The Auk, 126(3), pp.635-648.

Trujillo-Arias, N., Calderón, L., Santos, F.R., Miyaki, C.Y., Aleixo, A., Witt, C.C., Tubaro, P.L. and Cabanne, G.S., 2018. Forest corridors between the central Andes and the southern Atlantic Forest enabled dispersal and peripatric diversification without niche divergence in a passerine. Molecular Phylogenetics and Evolution, 128, pp.221-232.”