Proposal (940) to South American Classification Committee

Recognize Pseudocolaptes johnsoni as a separate species from P. lawrencii

Pseudocolaptes johnsoni was described as a new species by Lönnberg and Rendahl in 1922 based on a specimen reportedly from Baeza, Ecuador (ca. 1800 m on the Eastern slopes of the Andes). Subsequently, Zimmer (1936) found four specimens that matched perfectly Lönnberg & Rendahl’s description of P. johnsoni but collected on the Western slopes of the Andes. The match was so striking that Zimmer (1936) not only concluded that johnsoni inhabits the western slopes but also cast doubts about the provenance of the type specimen, as all other specimens reported from that region are clearly boissonneautii. Zimmer (1936) also made clear that the distinctive plumage of johnsoni cannot be confused with the juvenal plumage of boissonneautii. Zimmer ended up considering johnsoni as a subspecies of P. lawrencii instead, but without providing any evidence other than presumed “closer affinities”.

Based on plumage differences and elevational preferences Robbins & Ridgely (1990) suggested that johnsoni deserves species status, a treatment followed by few lists (Ridgely & Tudor 1994, Ridgely & Greenfield 2001, del Hoyo & Collar 2016). A previous SACC proposal based on this evidence was rejected because of lack of additional evidence such as vocalizations: https://www.museum.lsu.edu/~Remsen/SACCprop28.htm

A more recent proposal was based on results of playback experiments showing that the song of johnsoni does not elicit any response in lawrencii individuals, suggesting significant vocal differences potentially producing premating isolation (Freeman & Montgomery 2017):

https://www.museum.lsu.edu/~Remsen/SACCprop754.htm

The analyses of Spencer (2011) and Boesman (2016) show the differences in song characteristics between johnsoni and lawrencii. If the samples are representative of each taxon, they show well-marked differences:

“johnsoni: Song is a high-pitched rattled series of notes slowing into stuttering and ending (always) with a characteristic high-pitched down-slurred note.”

“lawrencii: Song is a number of well-spaced staccato notes (always present unlike johnsoni) followed by a trill, which usually first ascends in pitch and then slightly descends while slowing down in pace.” (Boesman 2016)

New Information:

Forcina et al. (2021) revisited the issue of the species status of lawrencii by reanalyzing DNA sequences from a previous study (Derryberry et al. 2011). They claimed that researchers so far have “neglected” the DNA evidence that indicates that johnsoni is a separate species. They found that the single sample of johnsoni showed levels of divergence comparable to those between lawrencii and boissonneautii. In other words, the three lineages diverged nearly simultaneously about 2 million years ago (using mitochondrial clocks; compare to ~ 4 Ma in Derryberry et al. 2011). The analysis could not resolve whether johnsoni is closer to lawrencii or to boissonneautii. They discuss all the evidence accumulated so far, including morphology, vocalizations, and habitat, and conclude that johnsoni should be elevated to species status.

Maximum likelihood analysis (Forcina et al. 2021 fig 1)

Bayesian analysis (Forcina et al. 2021 fig 1)

Discussion:

The data needed for evaluating the species status of johnsoni has accumulated slowly over decades, a little at a time, mostly due to the rarity of the species. But all pieces of evidence suggest that johnsoni is a distinctive species. It is clearly diagnosable by plumage; it is not a cryptic species by any means. I don’t understand why Forcina et al. claim that this is a case of “cryptic diversity.”

Photographs of the type specimen in the Stockholm museum are now publicly available here:

https://www.nrm.se/forskningochsamlingar/zoologi/samlingar/ryggradsdjur/typsamlingfaglar/tattingar/furnariidae/nrm569749.3562.html

And the photographs confirm the original description and, in my opinion, the match between the type specimen and the AMNH series of 4 birds from W Ecuador.

Photographs available online of live birds may be a bit confusing because of the artificial variation in color produced by different light conditions and digital adjustments, and the presence of birds in juvenal plumage, but here are what I consider good representatives of adult johnsoni (very similar the AMNH specimens):

https://macaulaylibrary.org/asset/178142751?_gl=1*1pz6d24*_ga*NjE4NTUxOTA5LjE2NDQ5NzU5MDk.*_ga_QR4NVXZ8BM*MTY0NzQwNzI3NC4yNS4xLjE2NDc0MDczMjkuNQ..#_ga=2.131747151.1345091137.1647401651-618551909.1644975909

https://macaulaylibrary.org/asset/41029281?_gl=1*30cct8*_ga*NjE4NTUxOTA5LjE2NDQ5NzU5MDk.*_ga_QR4NVXZ8BM*MTY0NzQ1Mjg5NS4yNi4xLjE2NDc0NTMwMTQuNDQ.#_ga=2.201353681.1345091137.1647401651-618551909.1644975909

In both the holotype and the photos, note the deep rufous tones on back and flanks (versus brown in the other two taxa), no light streaks on the mantle (vs. streaked in boissonneautii), whitish “cheeks” (vs. buffy in lawrencii), dark lower throat (not forming a gular band that is continuous with the light cheeks like in the other two taxa), inconspicuous superciliary stripe (versus thin but well-demarcated stripe in the other two taxa), blackish breast with white rhomboid spots and rest of the belly rufous (versus predominating buffy spots that coalesce to form a light-colored central belly in the other two taxa).

The song is clearly distinctive and not recognized by lawrencii in playback experiments (Freeman & Montgomery 2017). For both plumage and vocalizations, a more detailed analysis of geographic variation across the Colombian Andes would have been desirable but I think that the evidence is compelling.

Finally, the genetic evidence adds a bit of additional information. But contrary to what the title in Forcina et al. suggests, it doesn’t “untangle” anything. Genetic data for just three individuals do not tell much about species limits. A random sample of three individuals from a single large and old population can produce a tree similar to the one found by Forcina et al. The species delimitation algorithm based on the Poisson Tree Processes (PTP) used by Forcina et al. is basically a mathematical formalization of a genetic divergence threshold/yardstick criterion, in itself a rather weak species delimitation criterion. The method classifies the nodes of the tree into speciation events and coalescent (intraspecific) events based on levels of divergence. It thus formalizes the observation that the three Pseudocolaptes are rather divergent.

The recovered topology doesn’t help much either. A strongly supported sister relationship between johnsoni and boissonneautii would have clearly made the case for separating johnsoni from lawrencii, but there were conflicting topologies across methods, and clade support was nil. In sum, the genetic data is consistent with the species status of johnsoni, but the evidence is rather weak.

In any case, I think that plumage, songs, and genes together provide sufficient evidence suggesting that johnsoni deserves species-level status and there is not a hint of evidence that suggest that johnsoni is conspecific with lawrencii or with boissonneautii.

Recommendation:

I recommend the treatment of johnsoni as a separate species.

Boesman, P. 2016. Notes on the vocalizations of Buffy Tuftedcheek (Pseudocolaptes lawrencii). HBW Alive Ornithological Note 87. In: Handbook of the Birds of the World Alive. Lynx Edicions, Barcelona.

Del Hoyo, J., & N. J. Collar 2016. Illustrated checklist of the birds of the world, Volume 2: Passerines. Lynx Edicions, Barcelona, Spain, & BirdLife International, Cambridge, UK.

Forcina, G., P. Boesman, & M. J. Jowers 2021. Cryptic diversity in a Neotropical avian species complex untangled by neglected genetic evidence. Studies on Neotropical Fauna and Environment: 1-8. https://doi.org/10.1080/01650521.2021.1915674

Freeman, B. G., G. A. Montgomery. 2017. Using song playback experiments to measure species recognition between geographically isolated populations: a comparison with acoustic trait analyses. Auk. 134(4):857–870.

Lönnberg, E., & H. Rendahl 1922 A contribution to the ornithology of Ecuador. Arkiv för Zoologi 14(25): 1-87.

Ridgely, R. S. & P. J. Greenfield. 2001. The birds of Ecuador. Vol. I. Status, distribution, and taxonomy. Cornell University Press, Ithaca, New York.

Ridgely, R. S. & G. Tudor. 1994. The bird of South America. Vol. II. The suboscine passerines. University of Texas Press, Austin, Texas.

Robbins, M. B., & Ridgely, R. S. 1990. The avifauna of an upper tropical cloud forest in southwestern Ecuador. Proc. Acad. Nat. Sci. Philadelphia 142: 59-71.

Spencer, A. 2011. Variation in tuftedcheek vocalizations. https://www.xeno-canto.org/article/99

Zimmer, J. 1936. Studies of Peruvian birds, no. 21. Notes on the genera Pseudocolaptes, Hyloctistes, Hylocryptus, Thripadectes, and Xenops. American Museum Novitates 862: 1-25.

Santiago Claramunt, March 2022

Note from Remsen on English names: After internal discussion with SACC members, it seems best to stick with Pacific Tuftedcheek, despite the marine connotation, because this is the name used at the onset of the suggestions to treat this as a separate species, and because several other bird species in the region are also referred to as Pacific Something, with the understanding that this refers to the Pacific side of the Andes. Therefore, we think that no formal proposal is needed.

_______________________________________________________________________________________

Comments from Robbins: “YES for recognizing Pseudocolaptes johnsoni as a species. As we pointed out in our 1990 paper, we suspected that johnsoni deserved species status based on plumage and elevational differences. In the original SACC proposal, I voted against recognizing it as a species because of the lack of vocal data. However, given that Spencer (2011) and Boesman (2016) have established vocal differences, I now vote to elevate it to species status. As pointed out by Santiago, I consider the genetic data equivocal.”

Comments from Stiles: “YES. As noted by Santiago, genetics, distributions, plumage, and vocalizations all support recognition of johnsoni as a species (the genetic evidence is weakest, but nothing suggests that johnsoni is NOT a distinct species).”

Comments from Areta: “YES. Vocal and plumage data support the split. The playback experiments also support this, even when these lack rigor. Genetic data is conflicting, but show a relative deep divergence between lawrencii, boissonneautii and johnsoni. Harvey et al. (2020) recovered a sister relationship (with low support) of boissonneautii and johnsoni with lawrencii sister to them:”

Comments from Lane: “YES. Between the distinctive vocalizations and the likelihood that P. lawrencii and P. johnsoni are not sisters, it seems the only reasonable move to separate the two at the species level.”

Comments from Pacheco: “YES. Accumulated evidence including genetics, morphology, vocalizations, and habitat support the recommendation of this proposal.”

Comments from Bonaccorso: “YES. All the evidence points towards species status. There are distinctive plumage differences between P. lawrencii, P. johnsoni, and P. boissonneautii (still, this level of plumage differences alone does not grant more than subspecies status in other groups). The difference in song is clear, and the genetic differentiation (although from just one specimen per species) is deep. Even if we cannot know the sequence of splits between all three species, which seem to have happened very fast in evolutionary time, the available evidence points to three lineages that have been evolving independently from each other for about 2-4 million years. Also, I know that allopatry is frequently used against granting species status, but I love seeing a clear geographic break between P. lawrencii and P. johnsoni.”

Comments from Remsen: “YES, barely. I think burden-of-proof is now on treating them all as conspecific, but not entirely because of the reasons in the proposal. Although all the relatively new evidence points towards recognition of johnsoni as a species, my rationale derives from the phylogenetic hypothesis in Harvey et al. (2020), as shown in Nacho’s comments above. Assuming that additional taxon-sampling would not affect the topology (see further comments below) and assuming that the weakly supported sister relationship of johnsoni and boissonneautii has a fair chance of being correct, then treatment of johnsoni as a subspecies of lawrencii would be incorrect, and treatment of it as a subspecies of boissonneautii is also incorrect because these two taxa are close to being elevationally parapatric in the Western Andes without any physical barriers between them --- those two have to be treated as separate species. Add the fragmentary vocal data to this, and in my opinion, johnsoni has to be treated as a species until shown otherwise.

“However, the negative wording and implications in the Forcina et al. paper merit a critical response here lest those who don’t know the details take them at face value. First, the title of the paper is an in-your-face, unjustified swipe at Derryberry et al. (2011). The title proclaims that we ‘neglected genetic evidence’, and that Forcina et al. have come to the rescue to “untangle it”. Their abstract also proclaims:

‘We assemble already available mitochondrial and nuclear DNA sequences to assess their taxonomy and to provide appointed committees with specific proof [italics inserted by me] to ascertain the number of Pseudocolaptes species.’

“Where to start? First, Derryberry et al.’s paper was a conceptual paper in which any taxonomic comments would have been inappropriate and edited out by reviewers, and so this paper is not to blame for “neglecting” the issue of johnsoni. Even so, there would have been good reasons to “neglect” the genetic data. They were based on an analysis of 6 neutral loci of only 3 individuals, including only 1 of the 8 subspecies of P. boissonneautii. Further, the unsampled nominate subspecies would have to be included in any analysis to anchor any taxonomic conclusion; nominate boissonneautii is also the taxon geographically closest to the range of johnsoni, and thus its absence is an additional reason to be cautious about the interpretation of the existing results. It would have been just as unwise for SACC or NACC to have reached any resolution on species rank for johnsoni on such weak data as it would have been for Derryberry et al. to have commented on it. No amount of ex-post-facto reanalysis by Forcina et al. of DNA sequences from those three samples (the only new data presented in Forcina et al.) will fix that crippling problem. Rather than “untangling” the “neglected” data, their oversimplification of the problem is a potential setback. As for “proof”, the only thing they “proved” is that they don’t understand the problems of using genetic data in a complex system, e.g. highly polytypic P. boissonneautii is not a panmictic gene pool or that failure to include the nominate subspecies in an analysis prevents any firm taxonomic conclusions.

“Forcina et al.’s Abstract stated:

‘However, even when multiple lines of evidence lean toward lumping or splitting of species, some taxonomic committees refuse to acknowledge their validity until convincing genetic evidence is produced and integrated with other sources of data.’

“This attack on SACC and NACC requires a direct response. Contrary to Forcina et al.’s (2021) hyperbolic claim to have restored sanity to taxonomy with resounding evidence, the data are borderline. The plumage differences, for example, tell us only that they are diagnosable taxa, not necessarily species under the BSC. Many subspecies that are more distinctive in plumage than these three taxa intergrade at contact zones and thus show that plumage differences themselves are seldom isolating mechanism. As for the vocalizations, the N is small, the playbacks are not a controlled experiment, and variation within boissonneautii and lawrencii was noted by Spencer (2011) as adding some complexity to interpretations of inter-taxon vocal differences:

“With such a small sample size it's not possible for me to tell if this is a systematic difference or just individual variation, but with what I can see it appears that populations of Streaked Tuftedcheek vary as much vocally as Buffy do. The break appears to be somewhere in Colombia, as at least some of the recordings from Colombia are of the lower pitched song types, and songs I have heard in Ecuador and Peru have also been of this type.”

“As for “refuse to acknowledge”, this reflects a misunderstanding of the process by which those committees operate. To make a change in the classification requires a proposal, which then has to be approved by a 2/3 majority. SACC rejected an earlier proposal before the online summaries of vocal information by Spencer and Boesman. The “johnsoni problem” has been on our publicly available list of proposals to be done for at least a decade. As noted above, the genetic data of Derryberry et al. and Forcina et al. is far from “proof” of species rank for johnsoni, yet I pressured Santiago to write up a SACC proposal. Does this constitute being labeled as “refuse to acknowledge”?

“Forcina et al. also stated:

“Derryberry and coauthors did produce genetic data from an individual of the contentious P. johnsoni but did not use such information in their phylogeny nor for any comparative analyses. In spite of the overt differences in phenotypes, the SACC (South American Classification Committee) along with the NACC (North American Classification and Nomenclature Committee) of the AOS (American Ornithological Society) and, as a consequence, the eBird and Clements checklists still treat P. johnsoni and P. lawrencii as conspecifics.”

“First, the taxon johnsoni has been treated as a subspecies of lawrencii for a century or so, including by Dickinson (2003), which is clearly indicated as the starting point for the first SACC classification. And as Forcina et al. noted later in their paper, a previous SACC proposal in 2003 rejected treatment of johnsoni as a species because it was largely based on anecdotal information. Second, “overt differences in phenotype” are not sufficient grounds for treatment as a separate species under any BSC-based classification, so the relevance of that point is uncertain. Further, although lacking modern analyses (see caveats in Remsen 2003 HBW Furnariidae account), the 8 subspecies of P. boissonneautii also purportedly have “overt phenotypic differences” among them --- otherwise, they would not be treated as subspecies.

“Forcina et al. then stated:

‘In this study, we assemble available genetic data as a completion of multiple lines of evidence to provide taxonomic committees with all the tools needed for making a sound decision about the number and identity of Pseudocolaptes species, thus eliminating the confusion arising from conflicting taxonomic treatments and assisting future conservation actions.’

“This over-confidence is unwarranted. The data are still weak, with gaping holes and small N. My vote on this is a tenuous YES only because continued inclusion of johnsoni as a subspecies of lawrencii cannot be justified by the data in Harvey et al. (2020), which was itself “neglected” by Forcina et al. (published online in July 2021). Also, I have confidence in the experience of Spencer and Boseman in evaluating species-level vocal differences, although I worry that until all 8 subspecies of boissonneautii are included in vocal comparisons, we may be premature in assessing vocal variation as well.

“Finally, Forcina et al. put SACC, NACC, and Derryberry et al. on the defensive because their non-recognition of johnsoni as a species is implied to be an obstacle to conservation. However, the Chocó region has been recognized as a region of high endemism for over a century (since Frank Chapman’s monographs) and as a biodiversity hotspot since the term was invented decades ago. Whether johnsoni is recognized as a species does not affect those conclusions, much less act as some pivot point, and any conservation measures that affect this region will also affect johnsoni. Finally, in a seminal analysis of Neotropical regions of endemism (Cracraft 1985 paper in the “Eisenmann” AOU Monograph volume), subspecies as well as species were used to define those regions of endemism, as was done also in Chapman’s monographs; thus, the taxonomic rank does not prevent johnsoni from being included in analyses of endemism, nor does the current SACC-NACC treatment prevent conservation measures from treating johnsoni as a species given that BirdLife International has treated it as a species since 2016.”

Comments from Jaramillo: “YES – the vocal data help here.”