Proposal (1055) to South American Classification Committee

 

 

Revise species limits in Onychorhynchus coronatus.

 

 

Effect on SACC: This could split widespread Onychorhynchus coronatus into as many as six species.

 

Background: Our current Note is as follows:

 

2. Ridgway (1907), Cory & Hellmayr (1927), and Pinto (1944) considered the four subspecies groups in Onychorhynchus coronatus as separate species: mexicanus of Middle America and northwestern Colombia, occidentalis of western Ecuador and northwestern Peru, coronatus of Amazonia, and swainsoni of southeastern Brazil. Meyer de Schauensee (1966, 1970) treated them all as conspecific without providing justification, and this was followed by Traylor (1977<?>, 1979b), AOU (1998), Sibley & Monroe (1990), Fitzpatrick (2004), Ridgely & Tudor (1994), who provided rationale for their continued treatment as conspecific, and Dickinson & Christidis (2014), but this was not followed by Wetmore (1972), who considered the evidence insufficient for the broad treatment. Ridgely & Greenfield (2001) and Hilty (2003) returned to the classification of Cory & Hellmayr (1927). Collar et al. (1992) considered occidentalis as a separate species. See Whittingham & Williams (2000) for analysis and discussion of morphological characters. Del Hoyo & Collar (2016) recognized four species: O. mexicanus of Middle America and n. South America; O. occidentalis of the Tumbesian region; O. coronatus of Amazonia; and O. swainsoni of the Atlantic Forest.  Reyes et al. (2023) presented data relevant to recognition of as many as six separate species based mainly on deep divergence in mtDNA. SACC proposal badly needed.

 

Birds of the World/Clements has instituted a 2-way split: https://ebird.org/species/royfly1

 

Ridgely & Tudor (1994) thought that perhaps four species could be recognized, but that treating them all as conspecific was the best course given similarities I behavior and voice, as far as was known at the time.  (They mistakenly cited AOU 1983 as having recognized two species.)

 

Six taxa are recognized in the complex (Dickinson & Christidis 2014):

 

• mexicanus (s. Mexico to Panama)

• fraterculus (n. Colombia and nw. Venezuela

• occidentalis (Tumbesian region)

• coronatus (Guianan Shield and n. and e. Amazonian Brazil)

• castelnaui (w. Amazonia)

• swainsoni (se. Brazil)

 

Here are Hilary Burn’s illustrations from del Hoyo and Collar (2016), which illustrate the differences in plumage, primarily in degree of markings on breast and general plumage tone.  Note that swainsoni and geographically distant occidentalis are the palest and least spotted of the group, and I suspect this is what influenced Meyer de Schauensee’s reasoning for treating them all as conspecific.

 

 

In Harvey et al.’s (2020) massive phylogenetic analysis of the suboscines using genomic (UCE) data, 4 subspecies were included.  Swainsoni was “basal” to the other taxa, with an estimated divergence time of ca. 6 MYA; nominate coronatus was the sister to mexicanus + occidentalis, with divergence estimate of ca. 3 and ca. 1 MYA, respectively:

 

 

Just eye-balling node depth in adjacent clades in the Harvey et al. tree indicates that 6 MYA divergence is more typical for taxa treated at the species level, whereas 3 and 1 MYA are more typical for taxa traditionally treated as conspecific.  This is no substitute for a formal analysis but is meant only to provide a crude comparison of node depths.  Feel free to make your own comparisons, of course.

 

New information: Reyes et al. (2023; note that the “et al.” includes our Luciano and Elisa) produced the first genetic analysis of the complex.  They used a single mtDNA marker: NADH.  They sampled 40 individuals, including individuals of all taxa in the complex.  They had only 1 swainsoni but had at least 5 for the other 5 taxa. 

 

 

The topology of the tree was consistent with that of Harvey et al. (2020), including the placement of the taxon with arguably the most divergent plumage, occidentalis, within the northern mexicanus group.  Their geographic sampling was reasonably thorough except for the absence of specimens from the populations attributed to coronatus from Brazil south of the Amazon and e. Bolivia:

 

 

The mean estimate of the divergence time between swainsoni and the rest was 6.1 MYA, remarkably similar to that of Harvey et al. (2020).  The focus of the paper was on historical biogeography, not taxonomy, so the paper did not produce any firm taxonomic recommendations.  Nevertheless, their analysis produced evidence for 6 independently evolving mtDNA lineages that they suggested could be treated as species.  Caution is required because the geographic and numerical sampling is possibly insufficient to be certain that the lineages are monophyletic, at least in southwestern Amazonia, where large areas of coronatus distribution in potential contact with castelnaui have not been sampled. Their concluding paragraph is as follows:

 

“Furthermore, this study helped to reveal independently evolving lineages that might have to be treated as separate species with different conservation concerns. Complementary studies that include nuclear DNA, morphology, niche differentiation, and vocalizations with thorough sampling throughout the Onychorhynchus distribution are needed to fully resolve the evolutionary relationships and delimit species within this genus.”

 

Potential changes in species limits are also implied in the section titled Cryptic Diversity.  I have to point out that the diversity cannot be described as “cryptic” if it has been known for more than a century with at least 4 taxa treated as separate species for the first half of the 20th century, and the only taxon not described before 1860 was fraterculus Bangs 1902.  In fact, Reyes et al. make this very point in this section.

 

As for vocalizations in this group, analyses are handicapped by how few samples there are.  Royal Flycatchers vocalize infrequently, and dawn songs are poorly known.  Thus, It is no surprise to me that Peter Boesman did not include this species as one of his 400+ “Ornithological Notes” (for HBW/BLI) despite this complex being a prime candidate for a preliminary analysis.

 

Kirwan et al. (2024a, b) summarized what is known about vocalizations in this group, with links to recordings.  In Birds of the World, they are treated as two species: O. swainsoni for se. Brazil and O. coronatus for everything else.  Kirwan et al. (2024b) outlined the differences used by del Hoyo & Collar’s (2016) adoption of the Tobias et al. scoring system and said this about voice: “Also distinctive in call, which is significantly shorter (effect size 4.9; score 2) and higher-pitched (effect size also 4.9; score 2) than that of other taxa.”  [It sounds lower-pitched to me.] Here are the recordings cited by Kirwan et al. (2024b):

 

https://search.macaulaylibrary.org/catalog?taxonCode=royfly5&mediaType=audio&sort=rating_rank_desc

 

Kirwan et al. (2024a) described the vocalizations of the coronatus group as follows:

 

“Song. A rarely heard series of whistles, which apparently varies geographically. There are however hardly any recordings available, thus the following are merely examples of specific cases:

 

·       Northern group: audio 2 In Costa Rica and Panama, a series of rather sharp downslurred whistles preceded by a short introductory note, whit..eeeuw...eeeuw...eeeuw ..., uttered at a rate of about one whistle per second. In the literature, Song in this region is described as a long series of higher, sharper notes with a most peculiar intonation (8) or (in Mexico) a descending, slowing series of plaintive whistles, usually 5‒8, whi' peeu peeu peeu peeu peeu ..., or wh' wheeu wheeu ... (65).

·       Amazonian group: In Brazil, a series of long melodious whistles, starting with a loud flat-pitched introductory note and followed by a series of lower-pitched disyllabic mellow whistles wheeee-pihuuw-pihuuuw-pihuuw. Another variant is structurally similar, but disyllabic whistles are more modulated wheeeee-priririuuw-priririuuw... audio Also described as a squeaky PEE'u occasionally followed by a lower, musical PEE'u-brrrr (66).

·       Pacific group: Possibly a homologous vocalizations is described as a squeaky whi-CHEW in a series in a display (66). No recordings of Song are available.

·        

Primary call. audio 3 A short disyllabic nasal keeeyup, repeated many times at intervals of ca. 2 seconds. On sonogram, note has a characteristic shape, initially reaching a flat-pitched top around 3 kHz, after which frequency drops sharply with a hiccup around 2 kHz, for a total duration of about 0.20‒0.25 second. Also described as a low-pitched sur-líp, sometimes [repeated] over and over (67); a loud, mellow, hollow-sounding whistle, usually two-syllabled: keeeyup or keee-yew (8); a squeaky to hollow, plaintive whee-uk or see-yuk (65); and as a loud, plaintive squeak: PEE'yuk (42).”

 

Three recordings of the primary call are presented here, one from mexicanus, one from castelnaui, one from coronatus: https://birdsoftheworld.org/bow/species/royfly1/cur/sounds#vocal.  These indeed sound very different from swainsoni, and sound very similar to each other.  No primary call of occidentalis was presented.

 

Discussion

 

This complex has been on everyone’s radar “forever” in terms of species limits, especially because classifications in the early 1900s treated them as for species, and Meyer de Schauensee provided no rationale for the lump.  One could argue on that basis alone that a return to the taxonomy of Ridgway-Cory-Hellmayr-Pinto is warranted, and thus place burden-of-proof on treatment as a single species.

 

Based on qualitative perusal of sonograms of what is called the primary call, one could easily justify treating swainsoni as a separate species, as done by Birds of the World.  Based on that same qualitative perusal of N=1 primary calls of mexicanus, coronatus, and castelnaui, the primary calls sound similar to me, and their sonograms have that distinctive appearance noted by Kirwan et al.  Calls of occidentalis and fraterculus were not presented, but perusal of xeno-canto suggests that occidentalis (N=1; https://xeno-canto.org/264700) and fraterculus (N=1; https://xeno-canto.org/664102) are also qualitatively similar.  However, none of this is a substitute for a quantitative analysis, so that concerns me.

 

It is not the job of SACC members to do original quantitative analyses, and this is what would be ideal for a decision on species limits in Onychorhynchus.  The qualitative difference between swainsoni and the others could be considered sufficient, in my opinion, for placing burden-of-proof on a single species treatment.  This is also consistent with the genetic data of Harvey et al. (2020) and Reyes et al. (2023). However, in my opinion, further splits would require quantitative analyses of the vocalizations.  Studies of contact zones would be the best of all, but occidentalis is isolated, and I’m not sure if contact zones exist between mexicanus and fraterculus, fraterculus and coronatus, fraterculus and castelnaui, or coronatus and castelnaui.

 

For voting purposes let’s break this down as follows, with YES/NO votes on the following

 

A. Retain traditionally defined broad O. coronatus.  If YES, then B, C, and D are automatically NO.

 

B. Recognize two species (as in HBW and Birds of the World): O. swainsoni and O. coronatus, which would include all other taxa.

 

C. Recognize four species: O. mexicanus, O. occidentalis, O. swainsoni, and O. coronatus (return to the classification of Cory & Hellmayr and others, following del Hoyo and Collar 2016):

 

D. Recognize six species (O. mexicanus, O. occidentalis, O. swainsoni, and O. coronatus) for each of the lineages elucidated by Reyes et al. (2023).

 

 

Pending input from those more knowledgeable than I, my recommendations are as follows: A. No – the original lump was never justified; B- Yes – based on published sonograms of primary call note and a genetic distance more consistent with species rank in related lineages; C. No – this would require additional, qualitative research on vocalizations in my opinion; D. No – likewise, and I oppose using mtDNA lineages for taxonomy, particularly because of the well-known potential problem of gene trees conflicting with species trees..

 

Literature Cited:

 

KIRWAN, G. M., R. SAMPLE, B. SHACKELFORD, R. KANNAN, AND P. F. D. BOESMAN.  2024a.  Atlantic Royal Flycatcher (Onychorhynchus swainsoni), version 1.1. In Birds of the World (T. S. Schulenberg and B. K. Keeney, Editors). Cornell Lab of Ornithology, Ithaca, NY, USA. https://doi.org/10.2173/bow.royfly5.01.1

KIRWAN, G. M., R. SAMPLE, B. SHACKELFORD, R. KANNAN, AND P. F. D. BOESMAN.  2024b.  Atlantic Royal Flycatcher (Onychorhynchus swainsoni), version 1.1. In Birds of the World (T. S. Schulenberg and B. K. Keeney, Editors). Cornell Lab of Ornithology, Ithaca, NY, USA. https://doi.org/10.2173/bow.royfly5.01.1

REYES, P., J. M. BATES, L. N. NAKA, M. J. MILLER, I. CABALLERO, C. GONZALEZ-QUEVEDO, J. L. PARRA, H. F. RIVERA-GUTIERREZ, E. BONACCORSO, AND J. G. TELLO.  2023.  Phylogenetic relationships and biogeography of the ancient genus Onychorhynchus (Aves: Onychorhynchidae) suggest cryptic Amazonian diversity.  J. Avian Biology 2023: e03159.

WHITTINGHAM, M. J., AND R. S. R. WILLIAMS.  2000. Notes on morphological differences exhibited by Royal Flycatcher Onychorhynchus coronatus taxa.  Cotinga 13: 14-16.

 

 

Note on English names: BOW uses “Atlantic Royal Flycatcher” for swainsoni and “Tropical Royal Flycatcher” for everything else.  If Option B passes, then I think we can save ourselves plenty of work and simply adopt the current BOW names Tropical Royal-Flycatcher and Atlantic Royal-Flycatcher.  If you object and would like a separate proposal, and are also willing to write that proposal, speak out.  There will be the usual howls concerning use of names of oceans as names of terrestrial species, but I think it is widely understood what the implication of such names are for land birds.  If Option C or D passes, then we will have to do a separate SACC proposal on English names depending on which taxonomy we adopt.  See the illustration above for names used by BLI for a 4-way split.

 

 

Van Remsen, May 2025

 

 

 

 

Voting Chart: https://www.museum.lsu.edu/~Remsen/SACCPropChart1044+.htm