Recognize Trogon [v.] ramonianus with T. [v.] crissalis as constituting a single species distinct from Trogon violaceus



Proposal (430) to South American Classification Committee




Effect on SACC:  This would split an existing species, Trogon violaceus, into two species.


Background:  Our current SACC note is as follows:


8. The subspecies ramonianus and caligatus were formerly (e.g., Cory 1919, Pinto 1937) considered separate species from Trogon violaceus, but Peters (1945) considered them all conspecific. Ridgely & Greenfield (2001) considered caligatus of Middle America and northwestern South America to be a separate species from Trogon violaceus, and this was followed by Hilty (2003); SACC proposal to recognize this split did not pass because of insufficient published data. Genetic data (DaCosta & Klicka 2008) indicate that caligatus is basal to a group that includes Amazonian T. violaceus, T. curucui, and T. surrucura (and that Amazonian violaceus may be paraphyletic with respect to the latter two species). SACC proposal passed to recognize caligatus as a species.


New Information:  Remsen, in Proposal #378 to the SACC, summarized the genetic work of DaCosta & Klicka (2008) as follows:


“DaCosta & Klicka (2008) published a gene-based phylogeny of the genus that included samples of caligatus (N=9) from Mexico, Honduras, Costa Rica, and Panama, as well as, I think, W Ecuador (a sample from “eECU” is presumably a typo for “wECU”), nominate violaceus from the Guianan Shield (N=2), and Amazonian ramonianus (N=4) from Ecuador, Peru, Venezuela, and Bolivia.  They sampled 1 mitochondrial gene, ND2, and 1041 base pairs, of which 557 were phylogenetically informative.

“They found that their three groups fell into three clades: (1) caligatus was basal to a group of taxa that included not only the other violaceus samples but also T. curucui and T. surrucura, with strong support (100% maximum likelihood bootstrap, 100% Bayesian support); (2) nominate violaceus and T. curucui are sisters, also with strong support (100% maximum likelihood bootstrap, 100% Bayesian support); and (3 Amazonian ramonianus is the sister to group 2 (83% maximum likelihood bootstrap, 86% Bayesian support).”


In Proposal #378, Remsen offered the following Analysis & Recommendation (I’ve boldfaced parts of this for emphasis):

“With genetic support from only a single, mitochondrial gene as the basis for the relationship, one could argue that the tree is only a gene tree, not a species tree, or that incomplete lineage-sorting confounds the result.  However, with the qualitative vocal data, I think that published evidence is sufficient for a change in species limits, so I tentatively recommend a YES.  From the plumage and genetic data, one could also make a case that ramonianus should also be elevated to species rank, but I think this should await more detailed vocal analyses as well as sampling crissalis from E Brazil.”

The SACC subsequently voted unanimously to recognize T. caligatus as a species distinct from T. violaceus.  At that time, I advocated the additional move of treating the Amazonian subspecies of ramonianus and crissalis as together constituting a species distinct from Guianan violaceus on the combined basis of genetic evidence from the DaCosta & Klicka (2008) study, and on vocal differences, which, although lacking published analysis, have been described qualitatively in print, and numerous examples of which are available for examination on various public-accessible archives (e.g. Macaulay Library, Xenocanto).  Following is my assessment of the differences, taken from my comments on SACC Proposal #378:

Comments from Zimmer: “YES, on the basis of genetic and plumage data, combined with qualitative vocal data.  However, I would go further and strongly suggest that ramonianus, together with crissalis, constitutes a species distinct from both nominate violaceus and the caligatus group of Central America and trans-Andean western South America.  The DaCosta & Klicka paper presents genetic data backing such a treatment for ramonianus, which, in my experience, is the most vocally distinct taxon in the entire group.  There is no published vocal analysis to prove this, but there are published qualitative descriptions, as well as published sample recordings of nominate violaceus, the caligatus group, and ramonianus/crissalis.  Examples are also searchable online at the Macaulay Laboratory website (probably also at Xenocanto).  For example, go to the Macaulay Library site, and do a search for Trogon violaceus recordings.  Check out LNS recordings #38963 (Ted Parker recording from Pando, Bolivia) and #11364 (Curtis Marantz recording from Amazonas, Brazil), both of which are representative of ramonianus.  You will see that the notes of the song have a diphthongal or nearly bisyllabic quality.  This squares with the description of the song of “Amazonian Violaceous Trogon” in Ridgely & Greenfield’s Birds of Ecuador, which the authors describe as “a fast but relatively short series of clipped “cow” notes, the notes often becoming doubled (“cadow-cadow-cadow..”).”  This is in marked contrast to not only the songs of trans-Andean caligatus, but also to Guianan/n Amazonia east of the rio Negro nominate violaceus, both of which sound much more like Blue-crowned Trogon (T. curucui) in having a longer, faster series of higher-pitched notes which are more reminiscent of the song of Ferruginous Pygmy-Owl (Glaucidium brasilianum).  Again, compare the two LNS recordings of ramonianus noted above to any LNS recordings of nominate violaceus from the Guianan region, or to any recordings of the caligatus group from Central America or western South America.  In my experiences, the differences noted (bisyllabic or diphthongal notes, fewer notes per song, slower pace and lower pitch for ramonianus versus single-syllable notes, many more notes per song delivered at faster pace and higher pitch for nominate violaceus) are absolutely consistent throughout their respective ranges.  Songs of crissalis, although possibly not identical to those of ramonianus, are at least distinctly similar, and are noticeably different from those of nominate violaceus.  I would argue that the available genetic, morphological and vocal evidence for splitting ramonianus/crissalis from nominate violaceus is at least as solid as the evidence for splitting the caligatus group from nominate, and that the vocal differences are much greater between ramonianus/crissalis and nominate, than between nominate and the caligatus group.  (Caution:  Do not be misled by some of the purported violaceus LNS recordings from Mato Grosso, Brazil, which sound like north bank (nominate) violaceus.  I am certain that these represent misidentifications of the songs of Trogon curucui, an easy and natural error for observers familiar with the songs of violaceus from Central America or the Guianan region to make.  In each such recording that was accompanied by a voice announcement, the recordist reported the recorded bird as unseen, but thought to be violaceus.)” 

“As regards English names, I think Van’s suggestions of “Gartered Trogon” for the caligatus group and “Violaceous Trogon” for nominate, are excellent.  HBW lists “Amazonian Trogon” as a name available for ramonianus, and I think that would be appropriate for the combined ramonianus/crissalis.”

From the comments made by the various SACC committee members at the time, there appeared to substantial support for this position of elevating ramonianus/crissalis to separate species status, but virtually everyone felt that it needed to be presented as a separate proposal.


Analysis and Recommendation:  The genetic and plumage data for splitting ramonianus/crissalis from violaceus are as strong as the corresponding data were for splitting out trans-Andean/Central American caligatus from violaceus.  As outlined above, the vocal distinctions between ramonianus/crissalis and violaceus are greater and more obvious than those between caligatus and violaceus, although neither comparison has been quantitatively analyzed.  As was the case with the vocal differences between caligatus and violaceus, the voices of ramonianus/crissalis and violaceus have been described qualitatively in print (e.g. descriptions in Ridgely & Greenfield 2001a, 2001b, Hilty 2002, Schulenberg et al. 2007) and are available for comparison on various internet sound archives noted above.  As previously noted, both caligatus and ramonianus were historically treated as separate species distinct from violaceus  (e.g. Cory 1919, Pinto 1937), before being lumped without comment by Peters (1945).  Splitting out the ramonianus group based on the genetic analysis of DaCosta & Klicka (2008) and admittedly qualitative vocal differences would have the effect of restoring the historic plumage-based taxonomy that was changed without justification by Peters (1945), while keeping us consistent with our recent restoration of caligatus to separate-species status.  This split also makes sense on biogeographic grounds, with an essentially Guianan violaceus versus a widespread, southern and western Amazonian ramonianus/crissalis paralleling the respective distributions of recently recognized splits in barbets (Capito niger versus auratus group) and puffbirds (Notharchus macrorhynchos versus hyperrhynchus/paraensis).


My strong recommendation, therefore, is to split ramonianus (with crissalis) from violaceus, and to treat them collectively (ramonianus has priority) as a separate species Trogon ramonianus.  Given the widespread Amazonian distribution of this group, as opposed to the essentially Guianan distribution of violaceus, I think the English name of “Amazonian Trogon” makes the most sense for ramonianus.  One could also make an argument that the English name of violaceus should be changed for purposes of clarity.  I’m ambivalent about this – on the one hand, the name “Violaceous Trogon” is well established, and, it agrees with the Latin epithet.  On the other hand, I’m not at all sure that the name “Violaceous Trogon” is more strongly connected in either the birding or ornithological communities to the Guianan birds (versus the Central American/trans-Andean caligatus group or the Amazonian ramonianus group), and retention of it could cause confusion in the literature.  If a change in the English name of violaceus is deemed desirable, then “Guianan Trogon” would be my suggested choice.



Literature Cited:

DaCOSTA, J. M., AND J. KLICKA. 2008. The Great American Interchange in birds: a phylogenetic perspective with the genus Trogon. Molecular Ecology 17: 1328-1343.


HILTY, S. L. 2003. Birds of Venezuela. Princeton University Press, Princeton, New Jersey.


RIDGELY, R. S., AND P. J. GREENFIELD. 2001a. The birds of Ecuador. Vol. I. Status, distribution, and taxonomy. Cornell University Press, Ithaca, New York.

RIDGELY, R. S., AND P. J. GREENFIELD. 2001b. The birds of Ecuador. Vol. II. Field Guide. Cornell University Press, Ithaca, New York.

SCHULENBERG, T. S., D. F. STOTZ, D. F. LANE, J. P. O'NEILL, AND T. A. PARKER III. 2007. Birds of Peru. Princeton University Press, Princeton, New Jersey.


Kevin J. Zimmer, April 2010


Comments from Nores: “YES, aunque con dudas. Aunque los datos genéticos, morfológicos y de vocalizaciones muestran claramente que se trata de una especie diferente de T. violaceus, hay en el árbol algo no muy claro en el árbol. En el primer clade, ramonianus está separado pero relacionado con el clado compuesto por violaceus, curucui y surrucura, pero en la parte inferior del clade hay especímenens del eEcuador (=ramonianus) que no está relacionado con estas especies sino que es sister de otro de Costa Rica, Honduras, Mexico y Panama (=caligatus).”


Comments from Robbins: “YES.  All data sets support the recognition of ramonianus/crissalis as a species and  “Amazonian Trogon” is an appropriate English name.  I also agree with Kevin’s suggestion of using “Guianan Trogon” for nominate violaceus.”


Comments from Stotz: “YES.  I would be inclined to favor Guianan Trogon for the remnant T. violaceus if ramonianus is recognized.  Most observers think of ramonianus or even caligatus as the Violaceus Trogon.  I think that continuing to use Violaceous for this Guianan species would be confusing.  Because of the misfortune of violaceus referring to the Guianan taxon, the Notharchus solution in which the widespread Amazonian taxon retained the English name in broad use (White-necked) is not appropriate here.  So Amazonian Trogon (not my favorite, but probably better than Ramon’s Trogon) for ramonianus and Guianan Trogon for violaceus.”


Comments from Jaramillo: “YES.  This seems like the logical course of action given the data summarized by Kevin.  I think this is not a controversial move; more controversial may be the choices of English names.  In particular, I think retaining Violaceous is not a good move as it will cause confusion.  My thought is to split, use Amazonian for ramonianus, and use Guianan for violaceus.  To me Ramon’s Trogon sound like the name of a store somewhere in South Tucson that sells trogons, so I can see the lack of appeal for that name.”


Comments from Pacheco:  “YES.  Se os dados genéticos e de plumagem para desmembrar ramonianus /crissalis de violaceus são tão fortes quanto os dados correspondentes foram para dividir caligatus de violaceus, não resta melhor solução ao SACC.


Comments from Stiles:  “YES - evidence from (female) plumages and vocalizations looks solid.”


Comments from Pérez-Emán: “NO. I might be overcautious here, but even when I think the evidence for ramonianus being a different taxon is substantial, I am not quite sure about both crissalis and violaceus.  I did not find in the literature clear distributional limits between both crissalis and ramonianus.  Peters indicated crissalis is distributed only in Brasil (but calling attention to the need of revision of such distribution), something that Collar kept in HBW.  However, Zimmer and Friedman published at the same year (1948) different versions about the identity of Venezuelan southern populations (with Zimmer even including populations from what is currently known ramonianus as crissalis).  With this distributional confusion (which might indicate the need of doing a careful study of plumage variation in these populations), I heard vocalizations in Macaulay Library as Kevin indicated.  In the potential distributional range of ramonianus, songs are clearly different, and such songs are also found in southern Venezuela (some Schwartz recordings).  However, there are many songs within the same range, identified as violaceus that are clearly similar to Guianan violaceus.  There might be some ID mistakes (as Kevin pointed out), but there are many.  To get more confused, I heard the only Pará recording, supposed to be crissalis, and to my naive ear, it sounds more similar to violaceus than ramonianus.  So, I feel a bit uneasy, without underestimating Kevin’s knowledge, on taking final decisions here.  Additionally, there is not genetic information on the eastern distributional range of ramonianus/crissalis (which might correspond to crissalis populations).  My final comment is in relation to Guianan populations of violaceus.  Songs of this taxon are quite similar to curucui, and the fact that Kevin indicates that violaceus misidentifications are probably curucui songs, together with the close genetic relationships shown by the ND2 phylogeny of Da Costa & Klicka, and the clearly labile plumage color patterns (in relation to phylogeny), make me think that there is a need for a more thorough study of species limits in this group.”

Comments from Cadena: “NO, based on Jorge's comments. The group seems ripe for a formal analysis of species limits combining the existing molecular data with careful analyses of vocalizations, ideally linked to voucher specimens.”

Additional comments from Zimmer: “In addressing Jorge's concerns, I would just like to reiterate and clarify a few points.  First, I can only reiterate that I have found the described vocal differences between ramonianus/crissalis relative to Guianan violaceus to be absolutely consistent throughout the range of the former, and that includes numerous localities on the South Bank east of the rio Tapajós, which corresponds with what we know of the range of crissalis.  I have never recorded the ramonianus/crissalis voice within the known range of Guianan violaceus.  Nor have I ever encountered "Violaceous"-type trogons on the South Bank of the Solimões/Amazon singing Guianan-type songs.


“There are two likely explanations for the seemingly contradictory vocal samples that Jorge mentions.  One explanation, which I mentioned in the proposal, is that recordists familiar with the songs of "Violaceous" Trogon from the Guianas or from west of the Andes are quite likely to mistake the songs of Blue-crowned Trogon (T. curucui) for those of "Violaceous" Trogon. Trogons (particularly in Amazonia) are often a pain to see (and therefore get visual confirmation of identity), but are easy to hear and record.  I checked many archived recordings of fast, non-diphthongal trogon songs from south of the Amazon that were identified as violaceus, and found that those that had data indicated that the bird was not seen but was identified as violaceus on the basis of voice.  I would maintain that most, if not all of these represent misidentifications of curucui by recordists not familiar with the latter species, and whose template for what "Violaceous" Trogon should sound like is based on experience with Guianan violaceus or trans-Andean caligatus, both of which are much more similar vocally to curucui than either are to ramonianus/crissalis.


“The other likely explanation for some of the seeming contradictions that Jorge mentions, is our incomplete understanding of the distributional boundaries of the three taxa (ramonianus, crissalis, and violaceus), and the distinct possibility that violaceus may contact both ramonianus and crissalis.  Jorge notes that there are Schwartz recordings of ramonianus from southern Venezuela, but that "there are other songs from the same range, identified as violaceus that are clearly similar to Guianan violaceus."  He doesn't provide localities, but I assume these are from the upper Orinoco region.  It would not surprise me to find that ramonianus and violaceus have narrow zones of sympatry or parapatry in that region.  Curtis Marantz is documenting some pretty interesting patterns in that same region regarding the distributions of Xiphorhynchus guttatus guttatoides versus X. g. polystictus, and I know that Luciano Naka has been documenting all kinds of interesting taxon-replacements across narrow contact zones in Roraima and the upper rio Negro region of Brazil.  Isler et al (1999) postulated a possible contact zone between Myrmotherula surinamensis and M. multostriata in nearby eastern Colombia, and noted that "better knowledge of the geographic relationship between surinamensis and multostriata would contribute to an understanding of the pattern of interface between Guianan and Amazonian birds, which is among the least known aspects of Amazonian avian biogeography.


“Similarly, I would not be surprised to find that Guianan violaceus crosses the lower Amazon into south bank Pará, which could account for the single Pará recording that Jorge mentioned (assuming that this isn't another misidentified curucui recording) that sounded to his ears like Guianan violaceus rather than crissalis.  We know that at some south bank localities in Pará, Guianan taxa are represented rather than the expected south bank counterparts.  This is certainly the case at Caxiuanã for Hylopezus macularius, where the form present sounds like the Guianan form and not the very different south bank form found west of the Xingu at Alta Floresta.  Such is also the case with Celeus undatus, Cyanicterus, and other essentially Guianan birds that are known to extend south of the lower Amazon in Pará.


So, lack of precise knowledge of the distributional boundaries of the three taxa, and consequently, of potential contact zones, can cause confusion in sorting out the vocal differences, but that doesn't mean that the discussed vocal differences aren't real.  I would also agree with Jorge that there is need for a more thorough study of species limits in this group, especially given the vocal similarity of Guianan violaceus to curucui.  There is even some marked vocal variation within curucui: the subspecies peruvianus of w Amazonia (east to rio Tapajós) is the one that is vocally so similar to Guianan violaceus, whereas nominate curucui of northeast Brazil (south and east of the Amazon Basin) is quite different sounding, at least in pace.  However, just because there are some relationships in the species complex that are unclear, I don't think that is reason for ignoring evidence for others that are.  Vocally and genetically, the evidence for ramonianus/crissalis being distinct from Guianan violaceus, is, in my opinion, solid.


Comments from Remsen: “YES, but somewhat reluctantly.  The data are substandard, in my opinion, in terms of what has actually been published and analyzed.  However, they are probably no worse than those behind the other recent species splits that we have endorsed in Trogon.  Regardless of rigor, I enjoy having my previous misconceptions punctured concerning lack of diversity in Trogon – species limits in this genus are turning out to be just as complex and intriguing as in other Neotropical genera.”