Proposal (980) to South American Classification Committee

Treat Celeus grammicus as conspecific with Celeus undatus

Background: The treatment of Celeus grammicus as a separate species from Celeus undatus has long been considered questionable (Short 1972), because of their very similar plumage morphology. Field workers have known for decades that the vocalizations of these two taxa are extremely similar, if not inseparable (cf. Macaulay Library, xeno-canto). Benz and Robbins (2011) provided the first genetic perspective, establishing that they were indeed sister taxa and there was extremely low genetic differentiation between them. Because of limited sampling across the entire distribution of the two taxa, they refrained from suggesting that the two be treated as conspecific.

New Information: Sampaio et al. (2018) obtained much more complete sampling (n=56 individuals from 6 of the 7 subspecies) and a more in-depth genetic assay of the two, concluding that there was a lack of reciprocal monophyly in all phylogenetic analyses. They recommended that they be treated as a single species.

Recommendation: Plumage, vocalizations and genetic data unequivocally indicate that grammicus should be considered conspecific with undatus. Undatus (Linnaeus 1766) has priority over grammicus (Natterer and Malherbe 1845).

References:

Benz, B.W. and M.B. Robbins. 2011. Molecular phylogenetics, vocalizations and species limits in Celeus woodpeckers (Aves: Picidae). Molecular Phylogenetics and Evolution 56:29-44.

Sampaio, L., A. Aleixo, H. Schneider, I. Sampaio, J. Araripe, P.S.D. Rego. 2018. Molecular and plumage analyses indicate incomplete separation of two woodpeckers (Aves, Picidae). Zool. Scr. 47:418-427. https://doi.org/10.1111/zsc.12287

Short, L.L. 1972. Relationships among the four species of the subspecies Celeus elegans (Aves, Picidae). Am. Mus. Novitates 2487:1-26.

Mark Robbins, August 2023

Comments from Zimmer: “YES, for all of the reasons stated in the Proposal. This would constitute a welcome change for me – I’ve never been able to separate these two “species” vocally, and there are areas on the South Bank of the Amazon where North Bank undatus occurs to the exclusion of widespread South Bank grammicus, making you obligated, when working in a new location, to track down every vocalizing C. undatus/grammicus for visual confirmation of identity.”

Comments from Lane: “YES. I think the evidence makes this a pretty clear case, and agree with Mark's comments about vocal similarity.”

Comments from Areta: “YES. All lines of evidence support the conspecificity of grammicus and undatus, with the later having precedence.”

Comments from Remsen: “NO. Whether they are “reciprocally monophyletic” based on 3 mtDNA and 3 nDNA markers is, to me, irrelevant. In contrast to many studies that proclaim “reciprocal monophyly”, at least Sampaio et al. (2018) had a large N (= 57) with good geographic sampling. Nonetheless, reciprocal monophyly with respect to this minute portion of the genome, assumed to be of no biological importance, is not relevant to species limits, in my opinion. What counts under the BSC is whether there is unrestricted gene flow between the two. If these two are parapatric, as is evidently the case, with no phenotypic signs of hybridization, then there must be a barrier to free gene flow. That they cannot be distinguished vocally is indeed concerning, because almost all woodpecker contact zones in which there is no or restricted gene flow are accompanied by vocal differences, and those in which there are no known vocal differences (e.g., North American Colaptes auratus/cafer). But not all. In the North American Sphyrapicus varius complex, in which there are no known vocal differences, gene flow between the component species has been shown in several studies to be restricted … contrary to my personal expectations from lack of vocal differences. Evidently, plumage differences are sufficient to prevent free interbreeding, although hybridization is moderately extensive. I don’t have a good grasp of voice in the various Picumnus species complexes, but my understanding is that vocal differences are not known.

“Even Lester Short, who used almost any sign of gene flow to treat woodpecker (and other) taxa as conspecific, treated grammicus and undatus as separate species. He (Short 1982, Woodpeckers of the World, Delaware Museum of Natural History) based this on apparent parapatry in the Rio Negro region and close approach without signs of interbreeding in two other areas. This parapatry is also noted above by Kevin in his comments.

“As outlined by Short (1982) and Sampaio et al. (2018), these two species have diagnostic plumage characters that allow visual discrimination wherever they occur. No intermediate specimens have been reported, as far as I can determine, so current data suggest no gene flow despite ample geographic opportunity. As uncomfortable as it is for my tidy world view of vocalizations marking species boundaries, the data do not uphold this view is this case, at least as far as what is known. This is a recent speciation event, only ca. 50,000 years old by the Sampaio et al. estimates, with little time for coalescence at all loci. Assuming the diagnostic plumage characters have a genetic basis, then indeed the two species are “reciprocally monophyletic” with respect to whatever gene complexes control those phenotypic features. Absence of gene flow between two populations is prima facie evidence for species rank in any species concept in which gene flow or lack of it is the foundation. From the taxonomic standpoint, parapatry without gene flow means “no further data needed” for assigning taxonomic rank, whether or not we understand the underlying mechanism for genetic isolation.”

Additional comments from Zimmer: “Upon further reflection, I would like to change my vote on this one to NO. I found Van’s comments on this one compelling, particularly his chosen example of the North American Sphyrapicus varius complex, wherein varius, nuchalis and ruber are vocally nearly indistinguishable to my ears (and this applies to call notes, complex interactive calls, and territorial “shrieking” calls), readily respond to audio playback of one another’s voices, and, are known to hybridize to a limited extent in contact zones, yet maintain their phenotypic integrity, and are treated by us a constituting separate species under the BSC. Yet another parallel, might be found in Piculus, where some vocalizations of flavigula, leucolaemus and chrysochloros are virtually indistinguishable from one another, and yet, they seem to have no problems maintaining their genetic isolation in sympatry (I would also note that lack of vocal divergence in this group may be obscuring the presence of cryptic species within both P. flavigula and P. chrysochloros, but that’s a topic for another time.). As stated in my earlier comments, I can’t separate C. grammicus and C. undatus by their vocalizations, and I kind of doubt that they can either, since I’ve had each respond to playback of the other on numerous occasions, but, as with the sapsucker example, it appears that vocalizations are not the mechanism by which these particular birds are maintaining their reproductive isolation. As Van noted, these taxa are parapatric in the Rio Negro region, as well as on the south bank of the Amazon in different areas, and there is no indication that I’ve seen of hybridization – it’s either one form or the other in any given spot, but not both, so the birds seem to be treating one another as separate species.”

Comments from Claramunt: “YES. Sampaio et al. (2017( clearly showed that this pair of woodpeckers lack gene tree reciprocal monophyly, and are not strictly diagnosable because some characters vary clinally, others are just polymorphic, and vocalizations are indistinguishable. The photos of the specimens clearly show intermediate specimens that show trait mosaicism (Figure 3). All this suggests that gene flow is ongoing, and this complex constitutes a single lineage and a single biological species.”