Proposal (921) to South American Classification Committee

Treat Trogon rufus (Black-throated Trogon) as consisting of five species, including one newly described

Effect on SACC list: This proposal would treat our Trogon rufus as consisting of five species, one of which is newly described.

Background: Our current Note reads as follows:

7c. Dickens et al. (2021) found evidence that T. rufus should be treated as five separate species, including one newly described: Trogon muriciensis of the Atlantic Forest patches of northeastern Brazil; they recommended elevating the subspecies tenellus, cupreicauda, and chrysochloros to species rank. SACC proposal badly needed.

Trogon rufus (Black-throated Trogon) is a polytypic species with one of the largest distributions of any trogon, with taxa treated as (nine described) subspecies occurring in three disjunct regions: (1) from Honduras to the Chocó of northwestern South America, (2) Amazonia; and (3) the Atlantic Forest from Alagoas through Brazil to eastern Paraguay and Misiones, Argentina. They have all been treated as conspecific from Cory (1919; albeit as T. curucui due to early name confusion), Pinto (1937), and Peters (1945) through the present.

Here's the map (from Dickens et al. 2021), which will be useful in evaluating the proposal:

New information: Jeremy Kenneth Dickens and colleagues examined 906 specimens at 17 different museums, including all taxa and all available type specimens, and gathered spectrophotometric data, patterning, bare parts coloration, and standard morphometric data on subsets of these specimens. They quantified vocal characters from 273 songs from throughout the distribution and including all named taxa. They also analyzed genetic samples (ND2 and Cyt-b) from 29 specimens from throughout the distribution and all taxa. In other words, Dickens, Britton, Bravo, and Silveira conducted an amazing study in terms of sample size, geographic coverage, and critical data. Although the genetic sampling included only mtDNA, I don’t think data from additional genes would have made a difference in terms of determining species limits. The analyses are really excellent, and I encourage everyone to check out the great graphics (my favorite is Fig. 3a on male uppertail covert hue, although the color illustrations by Eduardo Brettas of the taxa and their critical features is tough to beat). If only we had papers of this quality and depth, from sophisticated analyses to classical taxonomy. The only weakness, given the vocal differences, is the absence of playback trials, although to do that properly would require fieldwork in multiple regions in the Neotropics.

A detailed synopsis of all these data would take up a lot of space here. Check out the details for yourselves, but here’s what stands out to me: the voices of all their proposed species are qualitatively and quantitatively different; in contrast, no such major differences are found among the Amazonian taxa that they recommend be treated as subspecies of T. rufus sensu stricto. For three of their proposed species-level taxa, 100% of the recordings were correctly classified to species using linear discriminant analysis for three of their proposed species-level taxa (chrysochloros, cupreicauda, tenellus).

As for the genetic data (ND2 1041 bp, cytb1011 bp), the results show prefect congruence between geographic samples and relationships, and genetic distances within geographic clusters and within taxa are small. The cis-Andean taxa are all weakly differentiated, with time calibrations suggesting divergence times among species at ca. 3-4 million years. The big break is between cis-Andean and trans-Andean taxa, with a divergence time estimated at ca. 5 million years ago. The single sample of cupreicauda is fairly divergent from the 9 samples of tenellus.

(There are lots of little nuggets within this 42 page paper that I could itemize, but I will stop at just one because it has direct relevance to species limits. Atlantic forest chrysochloros is almost exclusively insectivorous, in contrast to the other omnivorous taxa, and it has a more heavily serrated bill for grasping large arthropods; they are also known to be regular followers of monkeys, army ants, and coatis, evidently more so than other trogons, so this all fits. Actually, I can’t resist adding a second one because as the authors note, it is important to keep in mind when assessing plumage in trogons: they found evidence that there may be an environmental influence on iridescence, which changes with elevation in chrysochloros.)

A summary of their recommended species classification is as follows. See the paper for detailed diagnosis of each, including coloration, pattern, eyering color, and song features. These sections also contain detailed descriptions of each taxon, detailed synonymies, and useful notes on type specimens

• Trogon tenellus Cabanis, 1862: Central America to extreme NW Colombia (dpto. Chocó)

• Trogon cupreicauda (Chapman, 1914: N Colombia south on Pacific slope to NW Ecuador

• Trogon rufus Gmelin, 1788 (including nominate rufus of Guianan Shield, T. r. sulphureus of western Amazonia, and T. r. amazonicus of eastern Amazonia)

• Trogon muriciensis sp. nov.: Alagoas Forest region; known only from type locality at Estação Ecologica de Murici.

• Trogon chrysochloros Pelzeln, 1856: Atlantic Forest region

Here is a screen shot of the outstanding color plate by Eduardo Brettas that illustrates the taxa:

The quality of evidence for species rank varies among the 4 newly recognized or new species, so I think we should subdivide the proposal as follows below. I suggest listening to the recordings of the various taxa on xeno-canto: https://www.xeno-canto.org/explore?query=trogon+rufus&pg=1

A. Treat tenellus as a separate species from T. rufus

Because tenellus is entirely trans-Andean, it is not parapatric with any subspecies of T. rufus, and thus the decision on taxon rank must rely on comparative methods. The vocal differences from other taxa are given as follows:

“Song: Diagnosed from neighbouring T. cupreicauda by fewer notes per phrase, longer note durations and generally higher note frequencies, particularly the introductory note high frequency. Note frequencies, particularly the introductory note high frequency, are higher than for T. rufus subspecies. Fewer notes per phrase, slower pace and longer durations of notes and pause following introductory note than T. chrysochloros.”

In terms of phenotype, It has a breast band, unlike T. r. sulphureus or T. r. amazonicus, from which it differs in several other less conspicuous details (p. 26, and see the color plate (Fig. 6, p. 14, and above). The Diagnosis states that its pale (blue-gray to white) eyering that distinguishes is from everything else, including parapatric cupreicauda, but chrysochloros also is listed as having the same color eyering.

B. Treat cupreicauda as a separate species from. T. tenellus

This is the other trans-Andean taxon, and it is sister to tenellus as one would predict on biogeographic grounds. If someone does not endorse tenellus as separate from the rufus group, then also endorsing cupreicauda as a separate species from either of those would be unlikely and would require special explanations. The vocal differences from other taxa are given as follows:

“Song: Compared to T. tenellus, the song has more notes per phrase, shorter note durations and generally lower note frequencies. It also has more notes per phrase, shorter note durations but a longer pause after introductory note, and generally higher note frequencies, especially for the introductory note, than in T. rufus subspecies. Compared to T. chrysochloros, the song has a slower pace, longer pause following the introductory note, generally longer note durations and generally lower note frequencies.”

In their Factor Analysis (Fig. 4A), cupreicauda shows no overlap with the other taxa on the axis heavily weighted by pace (slow) and note length (short).

In terms of phenotype, the most striking feature to me is that the tail color is closer to distant chrysochloros than to any other taxon, and in fact, it differs the most in this feature from parapatric tenellus. The yellow eyering distinguishes it from all other taxa. See p. 28 for a listing and discussion of other color differences.

C. Recognize T. muriciensis as a species.

As rightfully emphasized by Dickens et al., little comparative material was available:

“Diagnosis: We had little material available for the diagnosis of the new species Trogon muriciensis, particularly regarding external morphology, so caution must be taken until more information is collected. For comparison of plumage coloration and barred patterning, only the holotype was available. For morphometric traits, in addition to the holotype, we had measurements from the paratype and a ringed individual. For other discrete traits, we had photos from online depositories, in addition to those of the holotype (Supporting Information, Fig. S8) and ringed individual. For the song, we had slightly more material, with recordings from five separate individuals (including the holotype).”

The vocal differences from other taxa are given as follows:

“Song: Compared to T. chrysochloros, the song of T. muriciensis has fewer notes per phrase, slower pace, longer note durations, longer pause following introductory note and generally lower note frequencies. It is similar to T. r. rufus but with generally more notes per phrase, higher introductory note frequencies and higher loudsong note low frequencies. Compared to T. r. sulphureus, it has wider bandwidth frequencies and generally more notes per phrase, whilst against T. r. amazonicus, it has faster pace, shorter note durations and a higher frequency introductory note. In relation to T. tenellus, it has a greater number of notes per phrase, shorter pause after the introductory note a generally lower introductory note high frequency, and generally lower peak and high loudsong note frequencies. It differs from T. cupreicauda by having fewer notes per phrase, longer note durations but a shorter pause after the introductory note. The bandwidth frequencies of the introductory and loudsong notes are generally wider than all other taxa, except T. chrysochloros.”

In terms of phenotype (with their caveats concerning N), there is no single diagnostic character than I can see, but rather a combination of differences not shared with any other taxon. See p. 30 for an enumeration of the ways it differs from adjacent chrysochloros and members of the nominate group. Genetically (Fig. 5), the single sample is sister to all 9 samples of chrysochloros; however, all nine are distant, i.e. from São Paulo south (no samples from closer Bahia to RJ), so that result would be expected on the basis of isolation by distance alone.

D. Treat chrysochloros as a separate species from T. rufus

The vocal differences from other taxa are given as follows:

“Song: More notes per phrase, faster pace, shorter note durations and pause following introductory note, as well as higher note frequencies and wider introductory note bandwidth than T. rufus subspecies. The greater number of notes per phrase, faster pace and shorter durations are also diagnostic against T. tenellus. Compared to T. cupreicauda, the pace is faster, the pause duration shorter and frequencies usually higher.”

In their Factor Analyses (Fig. 4A, B [beware that B is mis-labeled as Trans-Andean]), chrysochloros shows no overlap with the other taxa on the axes heavily weighted by pace, pauses between notes, and number of notes.

As for phenotype, this taxon has a relatively smaller bill that is more highly serrated than any other taxon. The density of barring on the wing panel and undertail coverts is diagnosably higher than for any other taxon. The Diagnosis says that it can be diagnosed from all other taxa (except evidently muriciensis) by its blue-gray to white eyering, but I think this is an error – see tenellus. See p. 25 and the plates for additional differences.

Discussion:

There is no doubt that all taxa are diagnosable at some phenotypic level. But at what rank? Dickens et al. noted introgression at the phenotypic level between T. r. rufus and T. r. amazonicus, T. r. amazonicus and T. r. sulphureus, and T. r. rufus and T. r. sulphureus; thus, they treated them as subspecies of the same species, which is the logical treatment. Contact zones are valuable test cases. Those three contact zones provide a defensible standard for seeing which phenotypic characters matter and which ones don’t in terms of barriers to gene flow. In the vocal analyses, these three taxa mostly overlap in every feature analyzed, so song, as we have known empirically for 70+ years, matters. The characters that do not seem to matter in terms of barriers to gene flow, at least in this group of trogons, are tail color (which varies dramatically among the three), presence of subterminal tail band, presence of breast band (no surprise here, because there is intraspecific variation in this in other trogons), width of undertail barring, eye-ring color (ranges from blue to yellow even within amazonicus) and any mensural characters. In other words, almost every plumage and morphological characters they measured is irrelevant as a barrier to gene flow, and so an important message of this paper is that all these characters may be irrelevant when considering species limits in trogons. (And I wish they had measured the implications of this for the highly flawed Tobias-BLI 7 point scoring scheme in terms of taxon ranking, but this would have added a tangent that would likely have increased the paper length by a page.)

The only other known contact zone is the one between tenellus and cupreicauda at the Panama-Colombia border, where there are no documented cases of introgression. Therefore, with the usual caveats, I think we can take the differences between tenellus and cupreicauda as potential isolating mechanisms, at least if they differ from those between taxa in the T. rufus group. These “if/then” comparative extrapolations come with obvious solutions, but in the absence of alternatives, at least provide defensible rationale for extrapolation to allopatric taxa in terms of assessing potential barriers to gene flow, for better or worse. However, none of the phenotypic characters other than voice show any difference from those shown by Dickens et al. to not be potential isolating mechanisms in the T. rufus group.

That leaves vocalizations as the best proxy for estimating gene flow or lack of it.

I personally am not the person to assess what the vocal differences mean in a trogon framework -- -way too rusty. Are the reported differences comparable to species level differences in other trogon groups? I will leave that up to those of you who have extensive recent comparative experience with trogon voices. To me, when I listen to the recordings, all I hear are the features in common, which to me are considerable; but then again, many for-sure trogon species sound moderately similar. Is the absence of playback trials crippling given what I perceive as subtle differences? Ridgely & Greenfield (ergo also Mark Robbins) in the taxonomy volume of Birds of Ecuador were consistently alert to vocal differences between bird taxa west and east of the Andes in that country; however, they considered the differences between voices cupreicauda and sulphureus to be only “slight”, which is slightly worrisome to me. So, I look forward to your comments.

A. Treat tenellus as a separate species from T. rufus. I hesitate to make a recommendation on this one other than noting my subjective feeling that I trust the authors of this paper on this one because of the depth to which they have gone in these analyses. Certainly, tenellus occupies a nearly unique multivariate space in the Factor Analysis of plumage characters in both sexes (Fig. 2), more distinctive than any other taxon other than chrysochloros. However, in terms of vocal characters, tenellus overlaps nearly completely with the T. rufus group despite differing in average ways from them.

B. Treat cupreicauda as a separate species from. T. tenellus. Because there is no phenotypic evidence of gene flow between presumably parapatric cupreicauda and tenellus, I regard this alone as a sufficient criterion for species rank between those two. See the Discussion on p. 32 --- the authors were keenly aware of the significance of this. With cupreicauda vocalizations not overlapping with those of the other taxa in multivariate space, we have additional indirect evidence for species rank. I’m also impressed with the phenotypic differences between cupreicauda and tenellus, which are arguably greater than between any two adjacent taxa. The genetic distance between the two appears at least as larger as that between any two adjacent taxa in the tree; however, there were no genetic samples from Colombia, much less NW Colombia, closer to the contact zone than the sample of cupreicauda from Provincia Esmeraldas, and perhaps sampling within that ca. 700 km gap might produce a different result.

C. Recognize T. muriciensis as a species. With a tiny N and no truly diagnostic characters known, this one unfortunately represents the weakest case for species rank, as the authors noted.

D. Treat chrysochloros as a separate species from T. rufus. With its distinctive plumage characters and vocalizations, the evidence for this split is strong in my opinion.

Recommendations:

A. Treat tenellus as a separate species from T. rufus. I’m ambivalent on this one, which is awkward, because the evidence is nearly mandatory for treating its sister taxon cupreicauda as a separate species, which would make Trogon rufus a paraphyletic taxon if tenellus were included in T. rufus but not cupreicauda. On the other hand, at the population-species level I don’t think monophyly, especially when only two labile mtDNA loci were sampled, is a valid requirement for species rank (as I have argued several times previously). With every passing month, it seems that new data reveal ancient hybridization among species that would make perilous the use of any single gene tree as representing the “true” history.

B. Treat cupreicauda as a separate species from T. tenellus. YES. I think these have to be treated as separate species given the parapatry with no sign of introgression.

C. Recognize T. muriciensis as a species. Ambivalent. The endangered status of this one should not, in my opinion, influence the taxonomic decision; otherwise, this undermines the credibility of the scientific process. Certainly this should be recognized as a separate subspecies, minimally, and I look forward to others’ comments on this.

D. Treat chrysochloros as a separate species from T. rufus. YES on this one for reasons given above.

E. English names: Dickens et al. proposed the following:

• Trogon tenellus = Graceful Black-throated Trogon

• Trogon cupreicauda = Kerr’s Black-throated Trogon

• Trogon rufus = Amazonian Black-throated Trogon

• Trogon muriciensis = Alagoas Black-throated Trogon

• Trogon chrysochloros = Southern Black-throated Trogon

• Trogon chrysoc

A lot of you don’t like long compound names. In fact, Tom’s blood pressure just skyrocketed when he read this. I understand that view, but in this case I like the compound names because: (1) it keeps intact the link to Black-throated Trogon, thus making it clear within a long list of trogon names which species are members of the T. rufus superspecies; (2) maintaining Black-throated in the name retains a somewhat useful character in separating them from other trogons within their range (I think); (3) they are marginally distinguishable anyway by plumage; and (4) if one drops the “Black-throated”, then we would probably have to invent new names for two of the species because “Southern Trogon” and “Amazonian Trogon” are misleading names from the perspective of the genus as a whole. Southern Black-throated Trogon, as noted by Dickens et al., resurrects a historical name, thus providing continuity with older literature.

Also, “Graceful” is somewhat unsatisfactory for tenellus (Latin for “delicate”, fide Jobling), with or without Black-throated, although as noted by Dickens et al., it resurrects the historical name used in Middle American bird literature from at least Ridgway (1911) on, until Eisenmann (1955) changed it to Black-throated after its lump (Peters 1945) into T. rufus. Some people don’t care about historical continuity, but for researchers, it is helpful, e.g. “Graceful Trogon” brings up 597 hits on Google, including Gould’s monograph on trogons, and 13 hits in Google Scholar.

“Kerr’s” is an eponym, which will anger the anti-eponymous zealots; however, Kerr was a female collector, which was highly unusual for the era (1912!), an American living in Colombia and collecting birds, and she collected the type specimen. In my opinion, not only deserves the honor but also calls attention to her generally overlooked contributions. See the attached provided through Gustavo Bravo, admirably dug out by Andrés Cuervo. (You may have lost your chance to buy Ingram’s Milkweed Cream, unfortunately).

At this point, for English names, let’s keep it simple: A YES means acceptance of the proposed names, at least for those we end up recognizing, and a NO means something else, either no compound names or modifications of the modifiers or both.

[Revision 15 June 22: To really keep things simple, let’s go Y/N on compound names vs. simple names for now, and worry about modifiers later]

References:

DICKENS, J. K., P.-P. BRITTON, G. A. BRAVO, AND L. F. SILVEIRA. 2021. Species limits, patterns of secondary contact and a new species in the Trogon rufus complex (Aves: Trogonidae). Zoological Journal Linnean Society: 1–42.

Van Remsen, September 2021

Comments from Areta:

“A. YES. Plumage and genetics support this split. See comments on C regarding vocalizations.

“B. YES. Mostly based on the apparently distinctive vocalizations (despite methodological shortcomings), marked change in tail hue over a short distance near the zone of geographic proximity between them and less so based on the phylogenetic information (the single sample of cupreicauda is quite away from the southern limit of tenellus).

“C. NO to recognizing muriciensis as a separate species. The data is very limited and unsatisfactory to make this move. The shallow genetic divergence, lack of diagnostic plumage features, and similar vocalizations to chrysochloros indicate to me, at most, subspecific status. Regarding the vocalizations, I have trouble in seeing the most basic parameters of the spectrograms in Figure 4 (e.g., time and frequency values), but these seem to have been built using different scales, thereby presumably distorting the similarities and differences among vocalizations. It also seems to me that there are important vocal differences among sexes in Trogons (fleetingly disregarded by the authors) that were not taken into account when making the comparisons; this can easily be heard and seen in spectrograms when couples of birds duet or respond to each other. I have trouble in matching the presumably "typical" songs depicted in the spectrograms to results of the discriminant analyses: songs which look very different overlap widely, suggesting that measurements were not able to capture key features of the sounds or that the spectrograms are not so typical. Finally, several recordings of chrysochloros sound exactly like muriciensis.

“D. YES to splitting chrysochloros from rufus. Plumage, vocal and genetic data agree in this split.”

Comments from Donsker: “E. My formal vote is NO. I think that we can come up with better English names than those proposed.”

Comments from Bonaccorso:

“A. YES, but it makes more sense to accept 921B.

“B. YES. T. tenellus and T. cupreicauda seem to have reached enough phenotypic differentiation (especially tail color; I love the tail-hue figure!) and vocal differences. However, I am not impressed by their genetic differences because the range of the species along Colombia was not covered in the phylogenetic analysis, thus, differences could result from isolation by distance. However, the possibility of parapatry and lack of intermediate individuals supports the potential lack of gene flow.

“C. “YES, but a bit hesitantly (I agree with Nacho that the data are sparse). Phenotypically, the Alagoas population is not so different from chrysochloros. However, vocally (and I am no expert), it seems to be very different. Still, the low sample size (N = 5) may not reveal the variation spectrum of vocalizations in the Alagoas population. Genetic differentiation is weak and based on one sample, whereas genetic sampling of chrysochloros does not cover localities closest to the Alagoas population. In short, I think the decision should be based on the acoustic data if some of you (experts on the subject) think the level of differentiation merits species status.

“I disagree with the position that no special considerations need to be made when the decision is about endangered species. I think there has to be asymmetry when considering species that may or may not be endangered. If, in the end, they are no good species, some money and effort may be lost, but at least the habitat of the species may get some level of protection for a while (which, needless to say, will benefit many other species). It is much worse to err in the other way. Not recognizing an endangered species and waiting for more data may have much worse consequences, especially in cases like this, where the range seems tiny. Also, from my experience with the Blue-throated Hillstar, the fact that we recognize new, endangered species does not guarantee immediate incorporation into IUCN lists (even after repeated requests to BirdLife International). Meanwhile, deforestation continues, and funds are not available for research, research-based conservation, or protection. So, our delay in recognizing these species may have fatal consequences under the current circumstances.”

“D. YES. In this case, vocal and genetic evidence seems to support species status. I don´t think phenotypic differentiation is so strong or useful in defining species status in this case. T. sulphureus seems very different from rufus and still, they hybridize.”

Comments from Pacheco: “A, B and D. YES. Plumage, genetics, and vocal repertoire provide a good endorsement for treating these taxa at the species rank.

“C. A vacillating YES, for the exact reasons explained by Elisa. Obtaining more data (vocalizations, genetics) and conservation measures for this population becomes dramatic races against the clock.”

Comments from Stiles: “YES to all except C.”

Comments from Josh Beck on Part E (voting for Claramunt on English names): “I have witnessed a lot of mild confusion among both friends and strangers in terms of which Trogon species carries which name, as so many of the names could readily apply to multiple species. I also at times have to stop and think for a second which of the white-tailed Trogons is called White-tailed, which of the green-backed is called Green-backed, and which of the Amazonian Trogons is called Amazonian. Across all this, despite vocal differences between the potential daughter species, Black-throated is a name that no one confuses as it corresponds to one of the more visually and vocally distinct Trogons (seas currently defined). For this reason, despite the proposed compound names being awkward, I am highly in favor of retaining Black-throated in the name. Coining 5 novel names in a group of birds that all look very similar might make for more elegant titles in field guides but will not serve birders or other English name users, and I really think that “will these names best serve common name users” needs to be a driving principle. As far as the modifiers, I feel that Graceful is not very helpful as it provides zero identification / location information. I would suggest Central American or Middle American is an obvious possibility that is far more useful. As for Kerr’s, as deserving as Kerr may or may not be, and independent of one’s views of patronyms, the name provides no information or help locating or identifying the species. Choco seems, at first glance, the obvious choice for a far more useful name. This situation is analogous, to me, to the names for the Daggerbills. Within the two, fortunately one of them bears a useful geographic modifier so it is easy to keep them straight, as Geoffroy’s doesn’t really tell the average birder or English speaker anything. If they hypothetically were named Graceful Daggerbill and Geoffroy’s Daggerbill there would be a lot of people, myself included, frequently pausing for a moment to remember which one is which.”

Comments from Jaramillo: “YES to A, B, C, D. I guess it would be good to give a set of opinions and viewpoints, but they have all been made by others so far. I do think I am giving benefit of the doubt to the researchers who put this together in painstaking detail, and have provided us a framework. In a sense, I do not feel qualified to counter their arguments on these opinions.

“E – NO. The compound names do not do much here that is beneficial in my mind. If we are to have a compound name, it would be better for it to be “yellow-bellied trogon” to separate from the red-bellied species, I am not serious, but in the end that would be more useful to English speakers. At this point, with such a massive split, the name “Black-throated Trogon” seems to be destined to be forgotten and in this case, likely that is a good thing given that none of these forms are best served by that name?”

Comments from Lane:

“A. YES.

“B. YES.

“C. No. The characters distinguishing its voice from T. chrysochloros seem weak, and weakened even more by the recording on X-c recorded in 2021 <https://xeno-canto.org/628115> that has a song with more notes. I just don't think this taxon will have significant defining characters once it is given more scrutiny.

“D. YES.

“E. NO. I am not all that in favor the "XXX Black-throated Trogon" because I am not usually very impressed by compounded names of this nature, particularly when there are so many syllables involved (kind of a "Black-backed Three-toed Woodpecker" sort of situation in my mind). "Black-throated" is not at all a defining character for any member of the complex, so it is a name I'm happy to lay aside in preference for something like "Graceful Trogon" (T. tenellus), "Kerr's Trogon" (T. cupreicauda), "Sulphur-bellied Trogon" (T. rufus and related taxa), and "Saffron-bellied Trogon" (T. chrysochloros incl. muriciensis)... or something along those lines.”

Comments from Nigel Collar (voting for Bonaccorso on English names):

“E. YES for compound names, although I’d go for:

Northern Black-throated Trogon T. tenellus

Western Black-throated Trogon T. cupreicauda

Central Black-throated Trogon T. rufus

Eastern Black-throated Trogon T. muriciensis

Southern Black-throated Trogon T. chrysochloros

“Reasons: Just keeping to the points of the compass (plus Central—admittedly the hardest to stomach) homogenizes everything and makes it seem much fairer (more equal) and simpler; they all get two extra syllables with the emphasis on the first—slips off the tongue. It doesn’t matter to me that the Northern is more western that Western. Everyone’ll get used to it quick enough.

“(Personally I don’t buy muriciensis as a species, and I’d treat it in Southern. (You could almost dare to call Central “Eastern” in that case, but let’s not go there…)”

Additional comments from Josh Beck (voting for Claramunt on English names):

E. [YES for compound names].”I've mulled this over quite a bit and asked around a wide range of Neotropical birders and guides for their opinions, so this is my vote but based heavily on the opinions of a range of American (US), South American, and European birders and guides. I do think that Dan and Alvaro make good points about the lack of uniqueness or elegance of the compound names. I also dislike them. Really, there are almost no good names in the Trogons. But new novel names coined will be equally useless and have no history, further disrupting stability and losing the affinity of these 4-5 species. The daughter species here are a phylogenetic group and "black-throated" is one of the better understood/better known names among a sea of not terribly informative nor memorable Trogon names. Removing the black-throated part would be acceptable if new, useful, elegant, uniquely identifying names were available, but barring divine inspiration, any and all names for these birds are going to fail to be uniquely identifying or inspiring or memorable. For that reason, I really think, and all but one of the people I asked agreed, that despite the ugliness and awkwardness, it's far more useful to retain the black-throated descriptor. For what it is worth, most people I asked also thought that the modifiers to "Black-throated Trogon" should be geographically based as anything else is, again, not really helpful in distinguishing the daughter species nor remembering the names.”

Comments from Remsen: YES to all except C, for which the evidence for species rank is weak, as noted in the proposal and several comments. As for E (compound names), I find Josh’s comments persuasive. If someone had come up with a compelling set of one-word “first names”, then that would be one thing, but in the absence of this, I think the compound names are more useful in that they help sort out a bunch of generally poor trogon names by branding one species complex with a name that provides historical continuity.”

Comments from Whitney (voting for Robbins/Schulenberg on E):

“E. I favor retaining “black-throated” as part of the English names. My name is Bret, and (mostly in Latin America) I am occasionally asked if it’s a nickname, or, what Bret means. Heck, I have no idea, so I tell them, “It’s meaningless, but my mom gave it to me, and it’s stuck.” I expect this is exactly what would happen with any of the names proposed by Dickens et al. I think the names they proposed are fine. “Graceful” may leave some wringing their hands, but heck again, it doesn't matter, because it will serve the intended purpose of labeling a geographically defined entity (an entity that has borne that name in the past). Now, if we remove “black-throated” from the names, we do, in fact, lose valuable information, as the "black-throated trogon” (T. rufus complex) everywhere is distinguished from all other (especially sympatric) trogons in being a lowland, terra firme understory (~3-9 m), yellow-bellied, barred-tailed, eyering-lacking, largely insectivorous Trogon (you notice I didn’t mention “black-throated” because that feature is truly useless — but again, that’s beside the point, it’s just a name, and a name that’s worked well forever). It is useful to ornithologists and birders to retain “black-throated” because it immediately categorizes these birds across their wide geographic radiation, and it provides a logical (and phylogenetically consistent) order among the multiple species of trogons sharing the range.

“I think it’s important to recognize that compound naming was not applied in the T. viridis split, which has led to confusion that would have been avoided by retaining “white-tailed”. I think it would make sense to revisit those names, and to apply compound names to future splits/descriptions of trogons (and, imho, most other splits, such as was done with the Hypocnemis cantator complex, which are now all “warbling-antbirds”… which, yes, don’t warble… but no one really cares [I don’t think…]).

“As to the status of T. muriciensis, I would vote NO from a purely scientific perspective, no emotion in the mix. I think there is a strong likelihood that having an appropriate geographic sample of tissues from the northern end of the range of chrysochloros would further close, but not erase, the relatively minor genetic gap with the single sample from Alagoas. At this point, there are multiple species restricted to the "Pernambuco Center of Endemism”, including a highly distinctive curassow that has just recently been reintroduced there (project in very early stages). These multiple species and subspecies, and especially the forests where they occur, are clearly “on the map” for Brazilian and international conservation priorities, and I believe the Alagoas/Pernambuco region was the focus for conservation funding raised at a recent Global Birdfair in the UK. Even if the region were not receiving much global attention, I would not recommend species status for the Alagoas bird — given its very close similarity to chrysochloros — without a more robust analysis.”

Comments from Robbins: “Although the authors have gone into extraordinary detail from a plumage, soft part, and vocal standpoint, the genetic data are weak and in the case of proposal B, potentially suffer from the lack of sampling. I would prefer to have more genetic data before making a definitive recommendation; however, that might not be forthcoming for some time. The reason I put emphasis on that data set is that there are not consistent differences in morphology and vocalizations across this complex. It should go without saying that differences in plumage among most of these taxa are relatively small, with male tenellus and rufus sulphureus standing out in dorsal tail coloration. Moreover, much of those dorsal tail differences are represented within the Amazonian populations, i.e., amazonicus, nominate, and sulphureus. Thus, if one treats Amazonian populations as a single species then dorsal tail pattern becomes questionable as a species-defining character.

“A) NO, because of the above concerns with the genetic data coupled with overlap between tenellus and the Amazonian group of rufus in vocalizations. Also, see comments under B.

“B) NO, we need data where tenellus, at least theoretically, comes into contact with cupreicauda at the Panamanian/Colombian border. Note that there is only a single genetic sample of cupreicauda, and it is from Esmeraldas, far from the potential contact zone. That interface might shed light on the importance of dorsal tail pattern (given the differences between these two) and the importance of vocalizations. Depending on those data, I could see where tenellus/cupreicauda are treated as a species and it is considered a separate species from cis-Andean populations.

“C) NO, data are too limited to assess taxonomic status.

“D) NO, although given current data, I lean more towards recognizing this as a species, primarily because of differences in vocalizations as outlined in Fig. 4 in Dickens et al. and the disjunct distribution. However, to be consistent with what I have outlined above I vote No. Note that plumage and perhaps soft part colors are not unique to chrysochloros, and the limited genetic data indicate that it is close to Amazonian populations (no surprise).”

Comments from Zimmer:

“A. YES. Despite lack of diagnostic vocal differences, I have a hard time accepting that there are two conspecific taxa whose ranges are widely disjunct, and separated not only by the Andes, but by the presence of another taxon in the same group (cupreicauda), that is demonstrably different from both the trans-Andean taxon and the cis-Andean taxon, and that cupreicauda is parapatric with tenellus without evidence of intergradation between the two.

“B. YES. Although it is not listed as a subset of this Proposal, it would also follow that I would also vote “YES” on treating cupreicauda as a separate species from rufus/amazonicus/sulphureus. As Van noted (in the Proposal), phenotypic/plumage characters are all over the map in “Black-throated Trogon” (sensu lato), and even within some subspecies groups (the Amazonian/Guiana rufus-group), and really don’t seem to work as isolating mechanisms. Given that, to quote Van “That leaves vocalizations as the best proxy for estimating gene flow or lack of it.” Accordingly, to my ears, songs of cupreicauda are distinct from those of tenellus to the north and west, and those of the various Guianan/Amazonian cis-Andea forms to the east and south. The absence of phenotypic evidence of gene flow between these two parapatric populations is further supporting evidence, which, admittedly, is weakened somewhat by the lack of sampling from closer to the potential contact zone.”

“C. NO, at least not for now. Sample sizes are just too small in my opinion, particularly given the absence of data from chrysochloros from Bahia to Rio de Janeiro, more proximate locales, which, when added to the mix, could well narrow any apparent gaps regarding potential differences in vocalizations, plumage, and genetics. I would note, however, that from the vocal samples I’ve listened to, muriciensis is closer (at least in song characters) to Amazonian populations than to fellow Atlantic Forest taxon chrysochloros. This would certainly fit an established pattern of several Atlantic Forest taxa from the Alagoas-Pernambuco center of endemism having their closest relatives distributed in SE Amazonian rather than the S Atlantic Forest. For example, think Automolus lammi (closer vocally and genetically to A. paraensis than to A. leucophthalmus), or Thamnophilus aethiops distans, or Hemitriccus griseipectus naumburgae. There’s also the broader pattern of widespread Amazonian taxa having highly isolated but closely related populations in the northern part of Brazil’s Atlantic Forest (south through Espírito Santo to N Rio de Janeiro for many of them) – think Cinereous Antshrike, Cinereous Mourner, Ringed Woodpecker, Bright-rumped Attila, Thrush-like Wren, White-winged Potoo, etc. Obviously, there are other species-pairs where the close affinities are between Pernambuco regional specialties and SE Atlantic Forest birds (Myrmotherula snowi/unicolor comes rapidly to mind), but the point is, that there are numerous examples of taxa from the forest fragments of Alagoas/Pernambuco being more closely related to Amazonian counterparts than to any taxa from farther south in the Atlantic Forest. So, it’s not implausible to me that muriciensis could be distinct from chrysochloros, but I do think we need more data points to be confident that is the case.”

“D. YES. This one is the most different vocally (perhaps, along with cupreicauda) from any of the others, and those vocal differences are congruent with phenotypic distinctions (even if it turns out these aren’t important as isolating mechanisms), ecological differences, genetic differences, and established biogeographic patterns.”