Proposal (1016) to South American Classification Committee

Treat Cnemoscopus chrysogaster as a separate species from C. rubrirostris

Note: This is a high-priority issue for WGAC.

Background: Our SACC note on this is as follows:

12c. The southern subspecies chrysogaster was formerly (e.g., Berlepsch 1912) considered a separate species from Cnemoscopus rubrirostris. Hellmayr (1936) treated them as conspecific, and this treatment has been followed in all subsequent classifications. Del Hoyo & Collar (2014) treated chrysogaster as a separate species (Golden-bellied Tanager).

The two taxa show a classic biogeographic pattern for Andean cloud-forest birds, with the break between them at the North Peruvian Low/Marañon:

C. r. rubrirostris: w. Venezuela and Colombia south to n. Peru (Cajamarca)

C. r. chrysogaster: n. Peru (Amazonas) south to s. Peru (Cuzco)

The two taxa are little greenish-yellow-grayish tanagers with few phenotypic characters to work with. Here is the plate by Hilary Burn in HBW (Hilty 2011):

Here are photos of LSUMNS specimens (with an epic face-off of Donna Schmitt vs. Tristan Davis skins):

Here are a couple of photos from Macaulay to give you a feel for the birds (rubrirostris by Sandy & Bob Sipe, from Napo, and chrysogaster by Juan Martinez, from Cuzco)

Hellmayr (1936) seems to have been the first one to treat them as conspecific, and he wrote: “Similar to the nominate race, but larger; bill and legs dark horn color instead of reddish; breast and abdomen brighter (clearer) yellow, etc.” and “Some of the Ecuadorian specimens [rubrirostris], by slightly larger size and darker bills, form the passage to the Peruvian C. r. chrysogaster.” Thus, Hellmayr saw specimens of rubrirostris from Ecuador that in his opinion had some of the characters of chrysogaster.

Zimmer (1947) did not discuss the situation much in part possibly because at that time rubrirostris had not yet been recorded in Peru; he only mentioned that Hellmayr’s measurements might not have properly reflected his assessment of size. Zimmer wrote: “There is thus little difference in the dimensions of the two forms.”

Isler & Isler (1987) pointed out a difference in song duration based on Ted Parker recordings at LNS (Macaulay): 3-4 sec for rubrirostris and less than 2 sec for chrysogaster.

Schulenberg et al. (2007) implied that call notes are the same described the songs as follows:

a. rubrirostris: “Squeaky series of lisping notes, with repetitive phrases”

b. chrysogaster: “similar but may lack the repetitive quality”

New information:

Del Hoyo & Collar (2014) treated chrysogaster as a separate species from Cnemoscopus rubrirostris based on the Tobias et al. point scheme as follows (provided by Pam Rasmussen):

There is no Boesman analysis of this pair.

Discussion and Recommendation: I recommend a NO on this one for now. A published analysis of vocal differences is what holds me back. I would say that chances are high that these songs will differ distinctly and that those differences will be sufficient for us to state that they have diverged to the point associated with barriers to gene flow in tanagers. This is the classic Allotaxa Dilemma. I have a hunch that those differences in bill color and size, iris color, and lack of overlap tail length indicate something meaningful in terms of mate selection and foraging ecology.

But my NO is based on what we know now, not what we might predict for the future. As scientists, I object to making decisions based on hunches about what future data might or might not show. Our job, as I see it, is to assess the data available AT PRESENT. In this case, we only have hints on voice. As for chronic pleas for genetic data, the only outcomes that might be useful are (1) that the degree of divergence is so great that it is beyond that know for sister species of tanagers or (2) so weak that the wise thing to do is continue to classify then as subspecies until we have better data on song.

English names: BirdLife International used “Gray-hooded Tanager” for rubrirostris and “Golden-bellied Tanager” for chrysogaster. If the proposal passes, we’ll need a separate proposal on this, The current name used by us and everyone else is Gray-hooded Bush Tanager”. By our SACC guidelines on English names, this is a case in which the daughters, with their roughly equal ranges, should both get new names, leaving the parental name associated with the former taxonomy. The “Bush Tanager” part comes from the former idea that Cnemoscopus was related to Chlorospingus, formerly all called “Something Bush Tanager”. As pointed out by Isler & Isler (1987), “Bush” is inappropriate. Hellmayr (1936) used Red-billed Cnemoscopus and Golden-bellied Cnemoscopus, so the Golden-bellied provides a link to the name in use from 1936 until Meyer de Schauensee (1966) changed Cnemoscopus to Bush-Tanager, and Golden-bellied links the English name to chrysogaster. The problem with these names is that they could be swapped between the two taxa and be equally apt, i.e. both have “gold” bellies (a slight exaggeration) and gray hoods. Given the limited number of phenotypic differences, I’m not sure we can do much better, but certainly “Red-billed” ought to be considered for rubrirostris given that this is perhaps the biggest difference between the two; “red” might not be the perfect description but …. . Is “Cnemoscopus” intractable as an English name? It sure would set these birds apart. Derivation (fide Jobling): “cnemo” for mountain slope, and “scopus” for searcher.

References: (see SACC Bibliography for standard references)

Van Remsen, July 2024

Comments from Robbins: “NO. I concur with Van, I would like to see an analysis, even if not published (i.e., if simply presented as an addendum to this proposal), of vocalizations before splitting these two.”

Comments from Areta: “NO. Not an easy call for me, as the bill and leg colours impart distinctive looks to these beasts. I´ve found photographs of birds from Tabaconas (see grey eye, pale bill and greyish legs) and Tapichalaca (reddish eye, pale bill and greyish legs) that look intermediate between rubrirostris and chrysogaster. However, birds like this also appear far from the potential contact zone in Colombia and Venezuela, suggesting that these could be young rubrirostris as pointed out by Tom Schulenberg. Regarding the songs, I don´t think the statements in Schulenberg et al. (2007) are strong enough to clarify vocal differences (indeed, the text seems to have been written with great care not to imply obvious distinctions that may not exist). I vote NO to the split for now, given the possible transitional phenotypes exactly at places in which one would expect them to appear and the lack of proper analyses of their vocalizations. Note also that it seems easy to turn a rubrirostris into something pretty much like chrysogaster simply by arresting color change through development.”

Comments from Andrew Spencer (voting for Remsen): “YES. I think these are a good split. To add to the background information Van gave in the proposal, these birds seem different in life, at least to me. Northern rubrirostris is a notably easier to find bird, hanging out in the open more often, moving more slowly, just a general more friendly bloke. Southern chrysogaster on the other hand is seemingly quite a bit less common, shyer, moves through the habitat faster, and seems to want nothing to do with you. Or at least with me, in my experience. When I was living in Ecuador and making the occasional trip to Peru, I was struck every time in how my experiences of the birds differed.

“Diving into the vocals a bit, I do feel I have additional information to add. Overall both of these birds seem to sing very rarely, which the paucity of recording bears out. I personally have never heard chrysogaster sing, and have only heard rubrirostris do so a couple of times. From perusal of available material, both appear to have two song types: a prolonged, simple dawn song, and a fast and more variably delivered day song. I briefly touch on the differences between populations in the Birds of the World account, for which I wrote the vocal section a couple of years ago during an update. As a quick summary, rubrirostris appears to have much simpler dawn songs, alternating two notes, whereas chrysogaster incorporates more variety. For day songs, it appears that rubrirostris may deliver them with longer, more jumbled, faster phrases than in chrysogaster. HOWEVER, all of the above on songs is from an extremely small sample size and nearly no personal experience. I am including two screenshots, one of daytime songs, and one of dawn songs. In each, the top examples are of rubrirostris, and the bottom of chrysogaster.

“Day songs, rubrirostris ML21791 above, chrysogaster ML17336 below:

“Dawn songs, rubrirostris ML245129 and ML246632 above, chrysogaster ML40057 and ML240571 below:

“Where I believe the vocal smoking gun is located is in their much, much more commonly heard contact calls. I first noticed this many years ago when in Peru, being somewhat surprised that what I was hearing from chrysogaster (example ML77202801) was different from the dry spit note I was used to from rubrirostris (example ML69350901) . The call of chrysogaster is a higher-pitched, noticeably differently shaped, polyphonic tink note. Although this can sound a bit subtle if you're not listening closely, it's really obvious once learned, and a night and day difference on the spectrogram. I am including a screenshot of a composite of every single recording of contact calls from Xeno-Canto, plus a couple of additional chrysogaster from other sources. These cover nearly the entire range of both forms (chrysogaster on the left, rubrirostris on the right). There are way fewer available from chrysogaster (further evidence of their behavioral differences I believe), but many from rubrirostris. Note that there is no intermediacy between the calls, and very little variation within each taxon.”

“Contact calls, six including all available examples from Xeno-canto and ML of chrysogaster, left, and nineteen including all available examples from Xeno-canto of rubrirostris, right. Note that there may be a bit of pseudo replication in the rubrirostris examples since I didn't carefully check to make sure some recordings weren't of the same individual, but this is minor.

“The same calls as in the previous example, using the Adobe Audition color palette for the spectrogram. Some details on some of the calls may be easier to see here, while others are probably easier in the grayscale example.

“A final note on the plumage/bare parts color of these birds. To me the images Nacho links are more likely to be a bit of variation due to age, and honestly do not look at all off for rubrirostris to me. They definitely don't give me any chrysogaster vibes.

“if this proposal did pass, my vote for English names would be Red-billed Cnemoscopus and Black-billed Cnemoscopus. I like the symmetry of using Cnemoscopus with using Chlorospingus. Helps that both are relatively easy to say by scientific name standards, and also avoid having to even think about the rules on hyphens!”

Comments from Stiles: “YES, again tentative; morphometric data with especially a larger N for chrysogaster would be good to have. However, the data on vocalizations look strong to me, even though sample sizes are small, and Andrew’s brief descriptions of differences in behavior are interesting; that of rubrirostris pretty well describes my experience with it (but I have no personal experience with chrysogaster).”

Additional comment from Remsen: “After seeing the vocal differences documented by Andrew, I change my recommendation on this one from NO to YES. Those call note differences are the clinchers for me, and their documentation herein constitutes sufficient evidence, in my opinion, for a split.”

Comments from Lane: “YES. Andrew had drawn to my attention the differences in voice between the two taxa, and frankly, these are distinctive enough in plumage and soft part colors that I’m rather surprised that they were lumped, allotaxa or no! Unfortunately, I have limited experience with them both as they are rarely around long enough to really study long when one encounters a mixed flock… but I think the differences outweigh the superficial similarities.”

Additional comments from Areta: “YES. OK, it is great that Andrew dove into the calls of these birds and came up with such a neat result. I would still like to see a more comprehensive study, but the vocalizations (a previously missing piece) add more evidence favouring the split. And assuming that birds with different leg and bill colours are juveniles, I can accept the split.”

Additional comments from Robbins: “YES. Although I would have liked to see a more in-depth analysis of song, what Andrew presents on call notes appears to be consistently different between these two. That coupled with the obvious bill color (apparently there may not be size differences [?]) leads me to now treat chrysogaster as a species. I wonder if there are differences in irides color. In the HBW plate, the iris in rubrirostris is depicted as reddish-brown vs. grayish-brown in chrysogaster.”

Comments from Claramunt: “YES. Bill and iris color are not superfluous differences, and vocal traits matching the split. Now, how is this case different from the Rauenia darwinii case?”