Proposal (587) to South American Classification Committee
Split Gymnopithys leucaspis into two species
Effect on SACC: This proposal, if passed, would split Gymnopithys leucaspis into two species, a trans-Andean G. bicolor and a cis-Andean G. leucaspis.
Background: Gymnopithys bicolor is a humid lowland forest antbird distributed both east and west of the Andes. Nine subspecies are currently recognized: five in Central America and western Colombia/Ecuador (the bicolor group, hereafter bicolor) and four in the northwest Amazon basin (the leucaspis group, hereafter leucaspis). The status quo classification followed by SACC (tagged as “proposal badly needed”) lumps bicolor and leucaspis following the rationale outlined by Zimmer (1937a). However, various authors have followed the alternate treatment of splitting bicolor and leucaspis as separate species (Willis 1967, Hilty & Brown 1986, Sibley & Monroe 1990). There is now sufficient data describing patterns of genetic, vocal, and plumage variation within this complex for SACC to vote on these alternatives.
Genetic data: Hackett (1993) found significant (~5%) allozyme divergence between trans-Andean bicolor and Amazonian leucaspis, though refrained from making a taxonomic recommendation and suggested analyzing these populations with more sensitive molecular markers. More recently, Brumfield et al. (2007) included single samples of both leucaspis and bicolor (as well as single samples of the other three recognized Gymnopithys species) in a broader phylogeny of ant-following antbirds. This study used both mitochondrial (cyt b, ND2, ND3) and nuclear (f5) markers, and found strong support for a sister relationship between leucaspis and G. rufigula, a congeneric species with an allopatric Amazonian distribution (G. rufigula is a Guianan Shield species, leucaspis present in northwest Amazonia), with trans-Andean bicolor sister to the combined leucaspis and G. rufigula group (see snapshot of their tree below).
The relevant portion of the maximum-likelihood tree presented by Brumfield et al (2007). Bayesian (before slash) and bootstrap (after slash) support values are given. This tree was inferred from the concatenated data matrix of three mitochondrial and one nuclear gene sequences.
Vocal data: Statistical differences in antbird vocalizations have been used to justify splitting of antbird species. However, there is no published analysis of vocalizations within G. leucaspis (as currently defined, including bicolor). The species account in Handbook of Birds of the World (Zimmer and Isler 2003) provides a detailed verbal description of leucaspis and bicolor loudsongs, stating that bicolor song “starts with long, upslurred whistles that shorten rapidly and gain in intensity, followed by shorter notes that drop in pitch and intensity before becoming harsh” while that of leucaspis “begins with upslurred whistles at an even pitch that shorten into rather abrupt notes dropping in frequency and intensity, then lengthen and increase again in intensity, finally decreasing in intensity and becoming harsh.” Additionally, the loudsong of bicolor is reported to be ten notes, compared to 20 for leucaspis, though the HBW account (Zimmer and Isler 2003) also notes that loudsongs are “quite variable in length”.
Plumage data: For antbirds, plumage is rather divergent within Gymnopithys: G. rufigula is entirely brown with patches of cinnamon, and G. salvini and lunulata are sexually dichromatic, with gray males and brown females. In contrast, plumage variation in bicolor and leucaspis is relatively slight, with subspecific plumage variation in head/side coloration and overall darkness. Nevertheless, there appears to be diagnostic plumage differences between these two groups: the bicolor group has two plumage traits – a black subocular area and blue-gray plumage behind the eye – that the leucaspis group lacks.
Taxonomic possibilities: There are two possible treatments at this time.
1. Maintain the status quo, leaving all taxa within both bicolor and leucaspis groups in a broadly defined G. leucaspis.
2. Split G. bicolor from G. leucaspis.
Recommendation: I suggest that current evidence supports splitting bicolor from leucaspis. The strongest data supporting this split is Brumfield et al.’s (2007) finding that the Amazonian leucaspis is sister to Amazonian G. rufigula and not trans-Andean bicolor. As the species status of G. rufigula has not been questioned, these genetic relationships strongly argue bicolor and leucaspis should be treated as different species.
Vocal and plumage data supporting this split are less conclusive. Loudsongs may differ (Zimmer and Isler 2003), but have not yet been subjected to quantitative analysis or behavioral playback experiments. Plumage is likewise similar between bicolor and leucaspis, though there are diagnostic differences in multiple plumage patches, providing weak support for the proposed split.
In sum, genetic divergence and the sister relationship of leucaspis with G. rufigula support splitting bicolor from leucaspis. This split is weakly supported by vocal and plumage divergence. This proposed treatment is also consistent with the commonly found pattern of divergence between cis- and trans-Andean populations of widely distributed lowland forest taxa.
Vernacular Names: If passed, this proposal would require new English names for bicolor and leucaspis. Ridgeley and Greenfield (2001) suggested “White-cheeked Antbird” for leucaspis and retaining “Bicolored Antbird” for bicolor. This treatment emphasizes the most prominent plumage difference between the two taxa, the white “cheek” of leucaspis. These English names therefore seem appropriate, though the committee could also consider alternatives.
Brumfield, R. T., J. G. Tello, Z. Cheviron, M. D. Carling, N. Crochet, and K. V. Rosenberg. 2007. Phylogenetic conservatism and antiquity of a tropical specialization: army-ant-following in the typical antbirds (Thamnophilidae). Molecular Phylogenetics and Evolution 45:1-13.
Zimmer, K. and M. Isler. 2003. Family Thamnophilidae (typical antbirds). Pages 448-681 in J. del Hoyo, A. Elliot, and D. A. Christie, editors. Handbook of the Birds of the World. Vol. 8. Broadbills to Tapaculos. Lynx Edicions, Barcelona.
Other papers are cited in the SACC bibliography
Ben Freeman, September 2013
Comments from Stiles: “YES, although a more comprehensive analysis of vocalizations, including use of playbacks, would be nice. Because both species are common, vocal birds, surely enough recordings exist in Cornell and Xenocanto to perform such an analysis. However, the genetic data seem convincing and the slight but diagnostic plumage differences seem on a par with several other recent splits in the Thamnophilidae.”
Comments from Pacheco: “YES. O somatório de evidências, sobretudo o mais recente estudo de Brumfield et al. (2007), dão suporte apropriado à adoção da proposta aqui apresentada.”
Comments from Zimmer: “YES. I think that the only reason that the proposed split “is weakly supported by vocal and plumage divergence” is that there has been no quantitative vocal analysis of the two taxa. I suspect that such an analysis would show that loudsongs of leucaspis and bicolor differ at least as does either of them from rufigula. Vocal differentiation between allospecies in the obligate ant-following genera (particularly within this species-group of Gymnopithys and the entire genus of Rhegmatorhina) appears to be pretty conservative from an evolutionary standpoint – perhaps there is some conferred advantage in signal recognition between different species that are all in the mode of constantly searching for the same moving target (ant swarms), and having to coexist at swarms while competing for prey. At any rate, the genetic data revealing the sister relationship between leucaspis and rufigula pretty much dictates either splitting off bicolor, or lumping all three groups (rufigula, leucaspis, bicolor) into one very bizarre (given the striking plumage differences between rufigula and the other two) species, something that I don’t think anyone is advocating.”
Comments from Jaramillo: “YES – leucaspis sister to rufigula I find odd and disconcerting although vocally these three forms are very similar. Still, the genetic separation between leucaspis and bicolor is strong, and although it would be great to have a more formal analysis of voice, what is available suggests diagnosable differences in voice, although slight. But rather than lumping rufigula into this species, I think the option of separating bicolor and leucaspis is the better one.”
Comments from Robbins: “YES, for splitting Gymnopithys leucaspis into two species, as genetic data show that Amazonian leucaspis is more similar to the morphologically distinct G. rufigula than leucaspis is to trans-Andean bicolor.”
Comments from Pérez-Emán: “A tentative YES based on Brumfield et al. (2007) data, as the alternative possibility to lump rufigula, bicolor and leucaspis is not in agreement with plumage differences between rufigula and bicolor/leucaspis.”
Comments solicited from Mort Isler: “Although I have no new vocal analysis to add, I agree with Kevin Zimmer's remarks in his response to Proposal 587 that it is highly likely that quantitative vocal analysis of Gymnopithys taxa will show that ‘loudsongs of leucaspis and bicolor differ at least as does either of them from rufigula.’ In this case, for reasons described in the proposal, it seems reasonable to make the split without thorough vocal analysis.”
Comments from Remsen: “YES, with my thoughts exactly as stated by Jorge.
“As for English names, there will be those who go ballistic that we don’t come up with new names for both daughter species. As much as I appreciate that logic and urge that we do so in most cases, there are extenuating circumstances. First, in this case, the two are not sister species – that’s the rationale for the split – so any sort of compound name is completely out unless rufigula also gets one, which would concoct and create a novel name for a long-established one. In this case, “White-cheeked Antbird” has a long history, going back to at least Ridgely’s (1976) Panama book and used by Sibley & Monroe (1990), so we already have a good name with a historical track record for leucaspis.”