Proposal (714) to South American Classification Committee

Recognize Cichlocolaptes mazarbarnetti as a valid species

Effect on South American CL: This proposal would add a recently described species to our main list.

Juan Mazar Barnett † and Dante Renato Corrêa Buzzetti (2014) recently described a new species of treehunter, Cichlocolaptes mazarbarnetti (Cryptic Treehunter) from Atlantic Forest of northeastern Brazil, more precisely from the ‘Pernambuco Center’ of endemism. The new species is known from only two sites: the type locality at Murici (09° 15’ S, 35° 50’ W) in the state of Alagoas, and Frei Caneca, Jaqueira (08º 43’ S, 35º 51’ W) in the state of Pernambuco.

This new treehunter differs from sympatric Philydor novaesi in its considerably heavier and longer body, various details in colorful plumage, and morphology of bill. It differs significantly from its only congener, Cichlocolaptes leucophrus, in having a uniform plumage that lacks buffy stripes on the ventral and dorsal regions of the body.

The differences in plumage between C. mazarbarnetti and C. leucophrus are quite considerable; however, there are other examples of sister species of foliage-gleaners in which one has a plain plumage and the other has a strongly streaked one: Simoxenops ucayalae and S. striatus, Syndactyla rufosuperciliata and S. dimidiata, and Automolus subulatus and A. cervicalis (Remsen 2003).

A comparison of songs and calls of Cichlocolaptes leucophrus holti, C. l. leucophrus, C. mazarbarnetti, Philydor novaesi, and P. atricapillus was made by authors. The song of Cichlocolaptes mazarbarnetti (sample: 27 songs, 129 calls) is markedly different from that of P. novaesi and it closely matches that of Cichlocolaptes leucophrus. However, the structure of the song of C. mazarbarnetti is similar to that of C. leucophrus, but the timbre and shape of the short notes are different.

Selected samples can be heard at:

http://www.xeno-canto.org/species/Cichlocolaptes-mazarbarnetti

The authors emphasize that the existence in the same area of a cryptic taxon resembling P. novaesi render past records of this species uncertain if not accompanied by a recording or detailed morphological or behavioral data. There are no recent observations of P. novaesi (see Lees et al. 2014).

Recommendation: I recommend a "YES" vote on accepting this treehunter as a new species to our list.

Literature Cited:

Lees, A.C., Albano, C., Kirwan, G. M., Pacheco, J. F. & Whittaker, A. 2014. The end of hope for Alagoas Foliage-gleaner Philydor novaesi? Neotropical Birding 14: 20-28.

Mazar-Barnett, J. & Buzzetti, D. R. C. 2014. A new species of Cichlocolaptes Reichenbach 1853 (Furnariidae), the ‘gritador-do-nordeste’, an undescribed trace of the fading bird life of northeastern Brazil. Revista Brasileira de Ornitologia 22: 75-94.

Remsen, J. V. 2003. Family Furnariidae (Ovenbirds). Pp. 162–357 in del Hoyo, J., Elliot, A. & Christie, D. A. (eds.) Handbook of the Birds of the World. Vol. 8. Broadbills to Tapaculos. Lynx Edicions, Barcelona.

José Fernando Pacheco, March 2016

_________________________________________________________________

Comments from Whitney: “The recent description of Cichlocolaptes mazarbarnetti from northeast Brazil requires particularly cautious analysis. The pertinent points of discussion should (in my opinion) start with, and center on:

1. There exists a single specimen, the holotype (#34530 in the Museu Nacional do Rio de Janeiro = MNRJ).

2. The specimen was not voice-recorded before collection.

3. The specimen has not been analyzed genetically, to place it in the extraordinarily taxon-complete phylogeny of the family Furnariidae.

4. The de facto syntopic, phenotypically similar Philydor novaesi also has not been analyzed genetically, and recordings of vocalizations attributed to that species now cannot be unequivocally ascribed to it because of the description of a syntopic, phenotypically similar furnariid.

5. Mazar Barnett and Buzzetti (2014) link some of their and others’ recordings and field observations to the single specimen entirely on assumptions, that soon become close to assertions (as they are convinced of what they’ve personally seen/heard/recorded), that cannot be independently verified by others (and, of course, not vouchered with the collected specimens).

6. The holotype is distinctively larger in several standard measurements (mainly bill size), and bears at least two distinctive plumage features, in relation to Philydor novaesi, as objectively verified by Claramunt (2014; not cited by Pacheco in this proposal).

7. Several independent observers are apparently convinced/satisfied that they have observed and recorded the new species in the field.

8. Despite numerous purported voice recordings of C. mazarbarnetti, there exists no image of it in life (at least, no one I have heard of is suggesting that any image of a furnariid from NE Brazil is or could be C. mazarbarnetti). There exist multiple images of presumed P. novaesi.

“I think there is a reasonable chance that Cichlocolaptes mazarbarnetti, after genetic analysis of the holotype, MNRJ 34531 (possibly also representing the new species, but possibly also “intermediate"), and the holotype and remaining specimens of Philydor novaesi, will prove to be a solid species. I’m not unduly troubled by the lack of any image of the bird in life, but it is strange that the bird was relatively easily tape-recorded on a number of occasions by people who carry cameras and would clearly understand the critical importance of obtaining photos/video of such a bird — especially in this case, where a specimen could not be legally collected, or even mist-netted. Vocalizations and DNA, or no, I am impressed by the photos of the specimen presented in the description and the morphological analysis of Claramunt (2014), which was quite a thorough analysis of the situation and presented several possible interpretations of the available evidence. And I guess that’s where we are, or must be, at the moment.

“In sum, until genetic analysis of these various MNRJ specimens can be undertaken, I recommend that the SACC and CBRO withhold judgment on the status of Cichlocolaptes mazarbarnetti. It’s worth pointing out that various homologous (in my opinion) vocalizations in the essentially identical repertoires of Philydor atricapillus (the type of the genus Philydor), presumed P. novaesi, Cichlocolaptes leucophrus + (very different) C. l. holti, and purported C. mazarbarnetti show remarkable similarity/overlap. The Derryberry et al. (2011, Evolution) phylogeny shows Cichlocolaptes leucophrus (I don’t know whether it was nominate or holti) basal to Heliobletus + P. atricapillus and its sister, P. pyrrhodes. It’s important to get the above-mentioned additional specimens included in the mix. I know it says in the description of C. mazarbarnetti that the authors couldn’t get permission to obtain DNA from the specimens at MNRJ, but this needs to be revisited before taking further taxonomic revision (in my opinion).  This is a prime example of one of the many critically important reasons that birds are sacrificed and stored in museums.”

Comments from Stiles: “NO. I agree with Bret that the problems regarding the identification and verification of mazarbarnetti are sufficient to put this proposal on hold for now. It would appear that these problems could be fairly easily resolved by a spot of judicious toepad clipping to sequence the problem taxa and a couple of diagnosable photos and published sonograms; hence, I prefer to await more evidence to fill in the holes. My vision of uses of specimens appears to conflict with that of the Río de Janeiro curator: if such use does not affect the integrity (identifiability) of the specimen (even if a type) and helps to resolve a difficult taxonomic decision, I would permit obtaining a genetic sample.”

Comments from Robbins: “NO, based on Bret’s comments it seems premature to recognize this taxon.”

Comments from Claramunt: “YES. This is a difficult case. I have personally examined the type series of P. novaesi and C. mazarbarnetti and compared the measurements I obtained from them with my database of Furnariidae morphometric data. For a full description and discussion regarding what I’ve found, see the article (available at: http://www4.museu-goeldi.br/revistabrornito/revista/index.php/BJO/article/viewFile/5703/pdf_2). My first impression when examining the type series was that the type of C. mazarbarnetti was a different bird that fits better the genus Cichlocolaptes. On the other hand, I failed to notice differences in plumage at that time, but the few differences pointed out by Mazar Barnett & Buzzetti seem correct. In a PCA, the type of C. mazarbarnetti appears within a cloud of specimens of Cichlocolaptes leucophrus and apart from the type series of P. novaesi, thus supporting the conclusion of Mazar Barnett & Buzzetti. Could the type of C. mazarbarnetti be an individual of P. novaesi with some sort of aberrant development? Maybe, but I haven’t seen anything like that in other species. Field observations of birds foraging on bromeliads and singing notes that sound like those of Cichlocolaptes, add to the hypothesis. At the end, I think that the existence of C. mazarbarnetti is the most plausible explanation for all this. An aspect that is not in the article, but that Juan mentioned to me long ago, is that the extreme similarity in plumage between P. novaesi and C. mazarbarnetti could be explained nicely by “social dominance mimicry”, a phenomenon that may explain this and many other cases of extreme plumage similarity among unrelated syntopic bird species (see Willis 1989 Rev. Bras. Biol. 49 and Prum 2015 Zool. J. Linn. Soc. 172).

“I agree that genotyping the specimens is feasible and could provide decisive information about this case. But in the meantime, we can show in the classification our informed assessment of the species diversity that is out there (or was, in this case). I recognize that the evidence is not unquestionable, but my judgment is that Cichlocolaptes mazarbarnetti is a valid species.”

Comments from Jaramillo: “YES – This is a tough one, and one with some uncertainties involved in its history. However, given the opinions and thoughts on this complex problem I am swayed more by Santiago’s analysis. As well, the taxonomy can be revised later if more information comes forward.”

Comments solicited from Curtis Marantz: “Although it is entirely possible that there are two very similar-looking species of furnariids in the mountains of Alagoas, I am not quite convinced that this can be proven. Having reviewed the manuscript by Mazar Barnett and Buzzetti multiple times prior to its publication, and having carefully studied in the field and recorded a variety of vocalizations by one of these species (search the Macaulay Library website for recordings of Philydor novaesi), I still cannot quite grasp the concept that two species are represented.

“Juan identified the bird in my recordings as the new Cichlocolaptes after I had independently noticed the close similarity of the sounds that I had recorded to those of C. leucophrys, and after Kevin Zimmer (and others?) had already suggested that the birds in Murici behaved more like Cichlocolaptes than Philydor, so I can certainly understand that there is indeed a Cichlocolaptes there, even if only one species is involved.

“I am likewise struck by the huge size of the one bird in the series relative to the others, and this is supported by a full suite of measurements; however, I am also troubled by the suggested juvenile specimen of the Cichlocolaptes, which effectively bridges the gap in measurements between the holotype and the series ascribed to the Philydor. I my work with woodcreepers, I found that a range of roughly 10% was not uncommon in measurements among a series of specimens, so the magnitude of the differences here are indeed striking, and especially so given that birds of the same sex were compared directly. I truly have no explanation for the size difference shown by these birds, other than that given such a small series of specimens it would be difficult to understand the full range of variation in the population.

“The eventual conclusion that I reached even before this manuscript was published was that there may indeed be two species involved, but that there was enough uncertainty for a variety of reasons that I felt it imperative that a genetic analysis was necessary to provide a definitive answer. Given the background, I was not overly troubled by the publication of this manuscript, but I would be most concerned about recognizing C. mazarbarnetti as a valid species without first having a definitive, genetic analysis.”

Comments from Remsen: “NO. I am persuaded by the comments of Bret and Curtis that the wisest course is to wait for genetic data – here’s a case in which a single specimen and its DNA can potentially resolve a problem. The specimen exists. There is sufficient doubt that I think waiting is best. I think Santiago will be shown to be correct.”

Comments from Areta: “YES. My last talk with Juan was about this bird while having a nocturnal tea in Salta, and I cannot move myself away from any subjective feeling on this issue. This is a very difficult case, and skepticism is necessary and healthy, in the face of the lack of definitely convincing evidence (i.e., genetic analyses of type specimens coupled to specimens/blood samples of sound-recorded individuals).

“First, that type specimens are not made available for genetic analyses violates the most basic principles justifying their very existence and seems untenable in my view. These extreme cases NEED sampling of the type specimens to solve the issue, lest we be swamped in another endless notorius/speluncae argument. I agree with Curtis' analysis, but I prefer the opposite decision: until these analyses are made, I will support Juan and Dante's view.

“Second, the size and plumage differences between novaesi and mazarbarnetti are striking as detailed in the description and corroborated by Santiago. I would like to offer a speculation here. I imagine that a single population of novaesi going extinct could exhibit abnormalities due to genetic problems, including gigantism. This gigantism could well result in other associated traits appearing (i.e., different plumage features). Given the shared genetic basis, resemblance to a related species might be expected. Alternatively, two species (novaesi and mazarbarnetti) may have hybridized during their demise, diluting any distinction or causing the appearance of strange phenotypes (both, morphologically and vocally). Mark Pearman told me that when Juan saw his illustrations of "novaesi", he thought that they exhibited intermediate features. Genetic analyses should take into account these possibilities.

“Third, vocalizations seem different and diagnostic most of the times, but the recording by Ciro Albano of a novaesi's song recalling that of mazarbarnetti casts enough doubt as to whether those differences are truly diagnostic, or they represent sexual differences or just variation within the same taxon.

“Finally, even with all the uncertainties mentioned by Bret and others, I support the recognition of mazarbarnetti until hard evidence (not just reasonable speculations) refuting its existence is provided. If Juan were here, he no doubt would illuminate us with other thoughts and arguments.”

Comments from Zimmer: “NO. The resolution of the classification of Philydor novaesi and Cichlocolaptes mazarbarnetti represents the most perplexing case of potential cryptic species that I have ever dealt with, and a case with which I have a lot of personal history, so I will apologize up front for the length of my comments on this proposal.

“My field experience with presumed Philydor novaesi at Murici (Alagoas) dates back to two encounters in January of 1999, which were the basis for Andy Whittaker and I questioning the taxonomic placement of ‘novaesi’ in Philydor, three+ years before 12 October 2002, which is when Juan Mazar Barnett (JMB) and Dante Buzzetti (DB) first encountered a bird at Murici that they initially identified as P. novaesi, but which they thought (based upon behavior, morphology and voice) should be reclassified as a Cichlocolaptes (Mazar Barnett & Buzzetti 2014). Based upon observations that Andy Whittaker and I made of foraging behavior, and, upon tape recordings of foliage-gleaners that we made from Murici between 1998 and 2004, our working hypothesis was that there was a single species of foliage-gleaner in question at Murici, described as Philydor novaesi, but which, in fact, should be more properly considered as a Cichlocolaptes, based upon size and structure, foraging behavior, and vocalizations. Sometime around 2001 or 2002, I communicated these thoughts to Curtis Marantz, who had recently returned (at that time) from having observed and tape-recorded presumed novaesi at Murici, and who had independently noticed the similarity of the vocalizations he had recorded to those of nominate Cichlocolaptes leucophrus. I went so far as to make spectrograms to compare recordings of presumed novaesi with C. leucophrus, but that was as far into it as I got before late 2003, when I first heard that Juan and Dante were talking about having discovered a cryptic ‘new’ Cichlocolaptes from Murici, following their discovery of what they subsequently referred to as “true P. novaesi” from Frei Caneca (Jaqueira, Pernambuco) in Feb 2003. Shortly thereafter, in January 2004, Andy and I ran into Bret Whitney (who, initially at least, was more than a little skeptical of both JMB’s two-species theory and our hypothesis of the birds we were seeing at Murici being a Cichlocolaptes rather than a Philydor) and Louis Bevier, first at Murici, and then, several days later, in Jequié, where we had the opportunity to sit down and compare notes and theories on what was going on, and to exchange tape recordings. Not long afterwards, Juan contacted several of us (Andy, me, Bret, and Curtis) in connection with the paper that he and Dante were preparing, to solicit any tape recordings or field observations that any of us had.

“At this point, I feel some detailed history of the genesis of the description of C. mazarbarnetti is called for, and rather than offer it up from memory of events and conversations of more than a decade ago, I think it is best here, to copy, verbatim, my original email response to Juan (cc’d to the others mentioned above), dated 4/6/2004:

“Hi Juan and everyone else! It’s time to throw my two cents in on this fascinating foliage-gleaner situation.

“Prior to 1998, we (Andy Whittaker and I) were unsuccessful in finding anything resembling Philydor novaesi at Murici. In January of 1998, Andy led a tour to Northeast Brazil that I was not a part of. It was on this tour that he encountered, for the first time, what he assumed was Philydor novaesi. He can speak better to the details of this encounter (since I wasn’t there), but as I remember it, it went down something like this: Andy spotted a rufous tail sticking out of a bromeliad and thought it was a Cichlocolaptes leucophrus (a bird that doesn’t occur at Murici). When the bird emerged from the bromeliad, he saw that it was similar to a larger version of Philydor atricapillus, and therefore *had* to be P. novaesi. Shortly thereafter, Andy heard an unfamiliar call, taped it, and in came the foliage-gleaner.

“This initial recording was of a vocalization that neither of us has heard since then, and it doesn’t sound typical of either a Philydor atricapillus or a Cichlocolaptes. In fact, it sounds stressed out, and since it obviously wasn’t given in response to playback, it could have been some mobbing call, possibly in response to the presence of a snake or some other predator. I believe we played this tape for Bret and Louis a couple of months ago when our paths crossed, and Bret may even have made a copy of it (I can’t remember).

“In any event, armed with Andy’s tape, we returned to Murici in January 1999. We trolled with the tape periodically, and after several tries, had a foliage-gleaner come roaring in overhead. The bird looked like a P. atricapillus, but with a less striking facial pattern, a brighter orange tail, and bigger proportions. I obtained about a minute of decent (but not great) video. It stayed high (subcanopy), hitching along open limbs, and looking at us, but did not vocalize. After a few minutes, it flew off. Almost immediately, we heard a loud, shrieking “REEP REEP REEP REEP!”, that was very reminiscent of the calls of Cichlocolaptes. We attempted to tape the vocalization, but the bird didn’t vocalize again. Later that same day, we hit on what was presumably another foliage-gleaner several hundred meters farther out on the trail. This one was with a mixed-species flock, and was foraging in the subcanopy (12-15 m abg). It was foraging like a Cichlocolaptes, concentrating its attention on rummaging around in bromeliads, with just its tail and hindparts sticking out. Tape playback of Andy’s “stressed-out” recording brought the bird roaring in overhead, and it was joined by a presumed mate. Once again, I got some decent but not great video, but the one time either bird vocalized (the shrieking ‘REEP REEP REEP…!”) was when neither of us was recording.

“During the course of the 1999, 2000 and 2002 tours, I was able to get a whopping 2 recordings of this bird. Each time it was several minutes after last playback. Each time it was the loud ‘REEP REEP REEP’ series (as an aside, this sounds the same to me as Juan’s recording #1. I think the recording that Bret refers to as sounding different and stressed out must be Andy’s original recording.). This bird, in our experience, has a habit of roaring in to tape playback, staying high overhead, remaining silent for several minutes, and then, lets rip with this startlingly loud call (similar to behavior we’ve noted for Cichlocolaptes leucophrus). Sometimes it waits 10 minutes or more before responding vocally, making it extremely hard to tape, especially when you’ve got an impatient tour group in tow, all of whom have ticked the bird and are ready to move on to other birds. Although I have only recorded the bird twice (Bret now has a copy of my best recording), and both times were delayed response to playback, I have heard the thing call perhaps 20 times, usually at some distance away, and never more than once in close succession. Each time we’ve heard it call spontaneously, Andy and I have whipped out our microphones and then stood there burning tape for 10 minutes or more without luck. Typically, the bird calls again about the time we’ve reholstered our microphones!

“Each time I’ve observed the bird, it has been at least 10 m above the ground, usually probing in bromeliads and other epiphytes. This, combined with the vocalizations we have heard (I have personally not heard anything other than the shrieking “REEP REEEP REEP” calls – the number of “REEPS” seeming to vary from one song to the next), led Andy and I to speculate that the bird seemed closer to Cichlocolaptes leucophrus than to Philydor atricapillus, despite the obvious plumage similarities. In my experience with Cichlocolaptes, the bird often adds a terminal rattle after the shriek notes, but usually as a response to tape playback (similar to Hyloctistes). Anyway, our assumption was that we were dealing with Philydor novaesi, but that perhaps (based on vocalizations and behavior) it was not a sister to atricapillus as assumed.

“Sometime around 2001-02, Curtis asked me if I had heard Philydor novaesi. When I said that I had, he asked what it reminded me of. I said “Cichlocolaptes”, and he said he had noted the same thing. He gave me a copy of one of his recordings, which as I recall, sounded like the things we had taped, but with some more elaborate rattling before or after the “REEP” notes. At this point, it never occurred to us that there might be more than one taxon involved. My thinking was simply that novaesi may have been misplaced in Philydor and was actually a Cichlocolaptes based on voice and foraging behavior. I got as far as making comparative spectrograms of the voices of P. atricapillus, presumed P. novaesi, and Cichlocolaptes leucophrus. But despite the obvious aural similarities of the Murici birds to Cichlocolaptes, the spectrograms of the two looked pretty different in terms of note shape.

“At about this time, Curtis informed me that Juan was working on this situation, and that he thought there were in fact two species involved. I have to admit skepticism at this point, because it seemed like a real stretch to think that there were two species of foliage-gleaners, both of which looked superficially like Philydor atricapillus, that were inhabiting the small remnant tract of forest at Murici.

“However, after running into Bret and Louis and comparing tapes and notes, the idea of two species seems more plausible. Bret’s recording from Murici was totally unlike anything that we have heard from foliage-gleaners at Murici. He indicated that he had never heard anything like our recordings (the “REEP REEP REEP” calls). Bret’s observations of foraging birds being in the understory and searching suspended dead leaves were also at odds with our observations.

“I have checked all of my tapes, and reviewed both cuts of videotape. I can’t find any single-note calls, nor any vocalizations at all besides the aforementioned “REEP” series. From Andy’s latest posting, I think he actually has Juan’s recordings reversed. Of the two cuts that Juan sent to all of us, the one that is identical to what we’ve recorded is the first cut (#1). The rattle in Cut #2 is not a vocalization that I have heard. As I said, I think Andy has simply reversed the two cuts in his latest e-mail.

“So that’s pretty much what I know about all of this. I don’t know the extent of Juan’s evidence for two species. I would caution that the most parsimonious explanation is that there is only a single species with a variable vocal repertoire. Although the vocalizations that I have taped remind me more of a Cichlocolaptes than of anything else, they are also somewhat suggestive of a loud series of shrieking notes given by Philydor atricapillus. My experience with atricapillus has been that you regularly hear the loudsong (which could be characterized as a descending trill or rattle), and occasionally hear another vocalization (alarm or aggression?) that begins with 2 loud, upslurred “WHEEEK WHEEEK” notes (purer in tone, less harsh than Cichlocolaptes) followed by a rattle. This vocalization is often given in response to tape playback, but is also given spontaneously. I have not compared spectrograms of this vocalization of atricapillus to those of the Murici birds. But my point is this: although the Murici birds sound more like Cichlocolaptes to me, it is not inconceivable that they are in fact related to atricapillus and have two similar vocalizations, a rattle or trill-like loudsong (the thing Bret has taped and gave me a copy of), and an aggression/alarm call (the shrieking “REEP REEP REEP”) that is analogous/homologous to the aforementioned “WHEEEK” calls of P. atricapillus. It could be that during January/February (the time of all of my visits to Murici) these foliage-gleaners are simply not that vocal (accounting for our never hearing the rattle-like vocalization), and are simply responding to our tape playback by giving only the aggression call. It is also possible that, in the absence of other foliage-gleaners, competitive release has allowed these Murici birds to expand their foraging niche to include both dead-leaf searching in the understory and bromeliad-searching in the midstory/subcanopy, accounting for the discrepancies of observed foraging behavior.

“I’m not saying that this is what I think is going on, but it does remain the most parsimonious explanation, and I think, Juan, that you must be prepared to address this possibility in your paper. As I said, I haven’t seen your evidence for two species, and perhaps, if I had, I would be as convinced as you are. But I do think you have to realize that it is a lot to accept that there are two cryptically similar and unrelated species of foliage-gleaners occurring in the same tiny patch of forest, and that both of them look amazingly similar to a third species of foliage-gleaner (atricapillus). Not only that, but one of these species has seemingly disappeared from the forest at the same time that people are just beginning to find the other species. I also agree with Bret that if you are to make the case for the Murici birds being a Cichlocolaptes, you also have to come to grips with the fact that Cichlocolaptes as currently known, has differentiated vocally into two populations (leucophrus and holti) that, although displaying some morphological differences, are still cut from the same cloth in terms of general plumage pattern. The Murici birds would represent a radical departure in terms of plumage.” (KJZ email dated 4/6/04).

“This was the beginning of a back-and-forth correspondence and exchange of tape recordings and information with Juan that lasted from April-August of 2004. I think that Van may have been privy to some of this exchange, because it was coming on the heels of his authoring of the chapter on Furnariidae for Handbook of Birds of the World. Anyway, during this correspondence, Juan initially asserted that “true” Philydor novaesi had disappeared from Murici, and that all recent (1999-2004) records of foliage-gleaners from Murici referred to the undescribed Cichlocolaptes. He had discovered a small population of “true” Philydor novaesi at Frei Caneca (Jaqueira, Pernambuco), where species novum was not then known to occur, in February 2003 (as cited in Barnett & Buzzetti 2014), and it was perceived differences between those birds and the birds at Murici that had led to the “two species hypothesis”. As we went back and forth on this, and shared our recordings with Juan, the story evolved, and became completely baffling to me. Juan identified all of our (mine and AW’s) recordings from Murici as being of “true” Philydor novaesi, in spite of his earlier assertion that novaesi had disappeared from Murici before 1998, and in spite of the similarity of the vocalizations we had recorded to those recorded by Curtis, which Juan identified as being of the Cichlocolaptes! In one email, Juan questioned whether Andy and I had ever actually seen the treehunter, this, in spite of the fact that we were postulating that the Murici foliage-gleaner was a Cichlocolaptes three years before he and Dante had come to the same conclusion. To accept Juan’s identification of all of our recordings as pertaining to “true” novaesi, is to make the convoluted argument that Andy and I had mistakenly concluded that the foliage-gleaners in Murici were more properly classified as being a Cichlocolaptes (based on structure, voice and behavior) than a Philydor, but all of our observations and recordings were, in fact of a Philydor that no one had recorded from the small remnant forest patch in years, but where there did just happen to be an undescribed species of Cichlocolaptes in that same forest patch that we managed never to see! I tried to point out the absurdity of this tortured reasoning to Juan, but his arguments became increasingly circular to my thinking, and without access to the series of P. novaesi specimens (including the holotype of C. mazarbarnetti), to see for myself the morphological distinctions, and, having seen only the birds at Murici (all of which looked, sounded, and behaved the same), I found myself at a dead-end, and eventually distanced myself from any further involvement in the still developing manuscript.

“Fast forward to February 2006, when Andy and I spent a week surveying Frei Caneca (Jaqueira). On 15 February, armed with our own tape recordings of the “REEP REEP REEP” calls from Murici, the “stressed out” yelping calls that Andy recorded at Murici in 1998, and a copy of Bret’s recording (recorded sometime between the mid-1990s and 2004) of a rattle song/call from Murici, we managed to lure in a pair of novaesi-type foliage-gleaners. We both made several audio recordings of this pair, and I made some video recordings of one of the birds as well. Unlike in all of our prior experiences with birds at Murici, these birds were extremely vocal in responding to our playback. They also repeatedly came in low (never more than 3-4 meters above ground), which, again, was very different from our experience with the birds at Murici, which always remained more than 10 m above ground, and typically waited to respond vocally until after they had flown off. Each time that the pair grew disinterested and flew off, we pulled them back in with our pre-recorded tapes, and they responded strongly to each of the 3 pre-recorded vocalizations. The birds gave multiple vocalizations, and none of them matched our (KJZ & AW) joint recordings from Murici, but one of their rattled vocalizations was a pretty close match for the cut we had copied from Bret. We did hear the shrieking “REEP REEP REEP” calls on a couple of occasions during our week at Frei Caneca, but each time, they were heard only very distantly (I accidentally got one very poor recording in the background when I was taping something else in front of me.), and never did we hear this vocalization from the pair of foliage-gleaners that we taped in on 15 Feb. At the time, my understanding from corresponding with Juan, was that the ‘treehunter’ was known only from Murici, and that only “true novaesi” was known from Jaqueira. The fact that the pair of birds that we encountered at Jaqueira responded to each of our three different prerecorded vocalizations (including the “REEP REEP REEP”), combined with the fact that we heard the “REEP REEP REEP” vocalization (albeit only distantly), further reinforced our stance that only one taxon was involved.

“I didn’t do anything further with this, preferring to wait until Juan and Dante published their paper, which, as we all know, ended up taking several more years to complete, in no small part due to Juan’s tragic illness and subsequent death. With publication of that paper and an accompanying morphometric analysis of the type series of P. novaesi (including the designated holotype of C. mazarbarnetti) by Claramunt (2014) in the same volume, the publication by Lees et al (2014) of a paper detailing the probable extinction of P. novaesi, and, finally, by SACC Proposal #714 and the comments it has already generated, I was stimulated to dig up and review all of my audio and video recordings of foliage-gleaners from Murici and Frei Caneca, and analyze them with ‘fresh eyes’, and, with the benefit of the published photographs of the type series of P. novaesi and the holotype of C. mazarbarnetti in front of me, as well as all of the descriptive material in Mazar Barnett & Buzzetti (2014) and Claramunt (2014). I also reviewed all of the audio recordings designated as either Philydor novaesi (13) or Cichlocolaptes mazarbarnetti (9) from the Xeno-Canto archives.

“My recordings of the earlier-referenced pair of birds from 15 February 2006 (at Frei Caneca) contain 4 different vocalizations: a single-note “squick”, also frequently doubled; an abbreviated, ‘rubbery’ rattle (‘brrrrup’); a series of 2-3 loud, upward-inflected “WEEEK” notes very reminiscent of the agonistic/alarm calls of Philydor atricapillus; and some disjunctly delivered minor-key yelping notes. All but the last of these vocalizations seemed analogous/homologous to different vocalizations that I have recorded of P. atricapillus. At no time did I record any vocalization that I took to be the primary song of the species. I will note however, that the most frequently recorded vocalizations designated as being P. novaesi in the Xeno-Canto archives (all of which are from the vicinity of Jaqueira, Pernambuco) are of a longer rattle that is very suggestive of the most common song-type of P. atricapillus, and which I feel pretty certain does represent the primary song of P. novaesi. This is contra the assertion in Lees et al (2014) that Andy Whittaker “made the first recording of the species’ loudsong” in 1998, because Andy’s original 1998 recording was of an excited string of yelping notes that sounded nothing like a rattle, nor like the song of P. atricapillus, but does match some excited calls purported to be of C. mazarbarnetti in the XC archives [And, it should also be noted, that Andy’s recording from 1998 is what we used in 1999 to bring in what appeared to be the Cichlocolaptes.). The “WEEEK” notes were quite different in quality from the shrieking “REEP REEP REEP” calls/songs that we recorded at Murici between 1999-2004. The latter were raspier and more frequency-modulated, and were delivered at a faster pace and a more constant frequency (not upward-inflected and more widely spaced like the calls of the Frei Caneca birds). At no time during my various encounters with foliage-gleaners at Murici did I ever hear or record any vocalizations other than the “REEP REEP REEP” series. As stated previously, this was very much at odds with our experience with the Frei Caneca pair, which gave a nearly continuous string of multiple types of calls in response to playback.

“My video recordings are even more suggestive. None of my videos were hi-definition, and all are handheld (not tripod mounted), so the quality is less than stellar. However, the footage of the Frei Caneca bird(s) is pretty decent and reveals several things of interest. For starters, the bird strikes me as looking a bit larger and heavier billed than P. atricapillus, but not markedly so. At the time, Andy and I both commented that the birds did not look as distinctly large as the birds we had seen previously at Murici, but it had been a couple of years since either of us had seen the Murici birds, so it was hard to be certain. [Keep in mind, that this was before we had seen photographs of the type-series of novaesi, including the distinctly larger holotype of C. mazarbarnetti. This was also before the publication of the description of C. mazarbarnetti or Claramunt (2014), so we had no knowledge of what the alleged plumage distinctions separating the two taxa were.] But my videotape from Frei Caneca does clearly show the distinct presence of a bold, buffy eyering, with a bold buffy supercilium above, and a shorter, parallel buffy stripe below, very similar to the highly distinctive facial pattern of P. atricapillus. The video also shows that the rump and uppertail-coverts are the same darker rufous color as the tail, which also matches the description of “true” P. novaesi as laid out by JMB & DB (2014). Furthermore, my video captures one of the Frei Caneca birds hitching through a low tangle and probing a curled dead-leaf, from which it extracted some unidentified prey item. In both the height above ground (3-4 m) and the dead-leaf-searching behavior, this bird appeared to be an ecological counterpart for P. atricapillus, and very different from any Cichlocolaptes (leucophrus/holti), as well as from the birds we had previously encountered at Murici between 1998-2004. Finally, I reviewed my two video sequences of foliage-gleaners from Murici, made on the same day in January 2000, and presumably of two different individuals. The quality of these videos was not as good, because the forest at Murici was taller and denser (equating to less light), and the birds remained much higher in the trees, so I was not as close. Nonetheless, the videos are instructive. Each of these two birds looks distinctly large, flat-crowned, and heavy-billed, much more so than any P. atricapillus, and more so than the Frei Caneca birds. At one point in one of the cuts, you can clearly hear one of our participants exclaim, “It’s huge!” That’s a couple of years before there was any talk of two species, and before Andy and I were talking to people about any taxonomic controversy, that this guy was struck by how big this foliage-gleaner looked to him. Furthermore, when I freeze the video, I am unable to see any sign of a buffy eyering, or of a buffy stripe below the eye, both of which were very apparent in my video of the Frei Caneca birds. My videos from Murici show birds with a narrow buffy supercilium, but no distinct patterning lower down on the face. Also, it appears that the rump and uppertail-coverts are contrastingly browner than the bright rufous tail. Each of these features match those described by JMB & DB in their description of C. mazarbarnetti, and are in contrast to the birds that I videotaped at Frei Caneca.

“So, in summation, my review of my own audio and video tapes, as well as of the archived XC recordings of others, supports (but does not confirm) the contention of Juan Mazar Barnett and Dante Buzzetti that there were, in fact, two cryptically similar taxa of foliage-gleaners occurring syntopically at Murici and Frei Caneca as recently as 1998–2006. Every bird that I ever encountered at Murici from 1999–2004 (my last visit to that site) gave the same “REEP REEP REEP …” vocalizations (and nothing more), behaved the same way (remaining 10+ m above the ground, burying its head in bromeliads when foraging, etc.), and looked the same way (large, with a very muted facial pattern compared to Philydor atricapillus, and with a tail that seemed distinctly brighter and more contrasting relative to the rump and lower back). Having never seen any of the pertinent specimens or the Frei Caneca birds (prior to 2006), Andy and I accepted these birds as representing novaesi, while thinking they would be a better fit for Cichlocolaptes. In looking through the Xeno-Canto archives, I note that all 13 archived recordings purported to be of Philydor novaesi were recorded from Jaqueira between 2003 and 2010. Of the 9 recordings purported to be of Cichlocolaptes mazarbarnetti, 7 were from Murici, and 2 were from Jaqueira, and all were recorded between 2002–07. This squares with my personal experience, as well as with JMB’s original assertion that novaesi had disappeared from Murici by 1999, and that all subsequent records from there pertained to the treehunter.

“Despite all of this, there are some nagging issues that cannot be ignored. One is that the pair of “true novaesi” that Andy Whittaker and I found at Frei Caneca in 2006 were strongly responsive to audio playback of treehunter-like vocalizations from Murici that are believed to represent Cichlocolaptes mazarbarnetti. Santiago and others have invoked interspecific social dominance mimicry (ISDM) as a possible explanation for the evolution of two cryptically similar foliage-gleaners in this region, but it makes no sense to me that the smaller of the two forms (the mimic) would respond aggressively to playback of the larger, dominant taxon (the model). It also seems anomalous that “true novaesi” more closely approximates the plumage of its smaller, presumed sister, the allopatrically distributed P. atricapillus, than it does the plumage of C. mazarbarnetti, if, in fact ISDM is to be invoked as the mechanism driving the similarity of the two forms.

“More importantly, there are the central issues raised by Bret in his comments (Points #2, 4 and 5, with which I agree wholeheartedly). The case for two species currently rests on a foundation of circumstantial evidence and underlying assumptions, the single most important being that the Cichlocolaptes-like vocalizations can be pinned to a cryptic taxon represented by the distinctive holotype (MNRJ #34530) of C. mazarbarnetti. But that specimen was not voice-recorded prior to collection, so, the link, although supported by field observations of several observers, cannot be confirmed through physical evidence. The possible existence of two syntopically occurring cryptic species also throws into question the identity of all birds previously identified in the field as P. novaesi, as well as any associated audio recordings or behavioral observations. All of which means that any attempts to distinguish between intraspecific vocal repertoires and interspecific vocal differences after the fact are going to rest on shaky ground. Bret was mistaken in stating: “Despite numerous purported voice recordings of C. mazarbarnetti, there exists no image of it in life (at least, no one I have heard of is suggesting that any image of a furnariid from NE Brazil is or could be C. mazarbarnetti).” My videotapes of 2 different birds from Murici in January 2000 (detailed above) are of purported C. mazarbarnetti, and they do support the plumage and structural distinctions noted by JMB & DB in their description, and appear to match well the photos of the MNRJ #34530. The existence of this video was made clear to Bret, Juan, and Curtis in my email of 4/6/2004 (copied above).

“When all is said and done, I think that JMB and DB were correct in their assessment of the situation, and that both Philydor novaesi and Cichlocolaptes mazarbarnetti are valid species (both of which, sadly, now appear to be extinct). However, I don’t think this interpretation has been adequately proven, and there are enough questions to at least give pause to any rush to judgment. The obvious remedy is there, begging to be applied. I would join all of the other voices in asking that the MNRJ revisit its position regarding allowing tissue from across the type series of P. novaesi (including MNRJ #34530) to be applied to a genetic analysis that would resolve not only the validity of C. mazarbarnetti, but also the taxonomic affinities of P. novaesi, both of which must currently be considered to remain in doubt. Until then, I think it is best to hit the “pause button” on any taxonomic revisions in this group.”

Comments from Bret Whitney: “In 2004, Juan contacted me and several other people including Andy Whittaker and Kevin, about the "undescribed Cichlocolaptes”. Juan and I exchanged a couple of e-mails, after which I suggested that a larger group of people should be involved in the exchange of messages, because I knew they had field experience with P. novaesi at Murici. That was when we all realized that it was going to be important to get everybody up-to-speed on what evidence existed for what. Communications started circulating more widely at that point, and Kevin sent the 4/6/2004 message he cited in the proposal here. This is the only message I can find on my computer that mentions his two January 1999 videos (or January 2000 — he cites his Murici videos both ways). As is clear from his message, in 2004 Kevin considered the videos to be of P. novaesi, and, having heard and recorded only the “REEP-REEP-REEP” vocalization from all foliage-gleaners he had personal experience with at Murici, he thought that P. novaesi might belong in Cichlocolaptes. Kevin now tells us, however, that his two videos of birds at Murici “are of purported C. mazarbarnetti, and they do support the plumage and structural distinctions noted by JMB & DB in their description, and appear to match well the photos of MNRJ #34530. The existence of this video mas made clear to Bret, Juan, and Curtis in my e-mail of 4/6/2004.”

“Well, this is obviously critically important — new — information. Suddenly, we have more to go on than a plea for toepad scrapes of the specimens I listed in my message at the top of this proposal. I urge Kevin to make available his three videos of these foliage-gleaners cited here — 2 from Murici of purported Cichlocolaptes mazarbarnetti from January 1999/2000, and one of purported Philydor novaesi from 15 February 2006, all with original audio tracks — for permanent archival at Macaulay Library and/or the Internet Bird Collection.

“I have a brief but reasonably good video of a single Philydor novaesi at Murici, from about 1997. Mario Cohn-Haft was my co-leader for a Field Guides tour of northeast Brazil. I found that bird by whistling an imitation of contact calls of Thamnomanes caesius in an attempt to locate an understory mixed-species flock. The Thamnomanes and also a Philydor novaesi started singing, which we tape-recorded.  In response to playback of the recording, the Philydor came in immediately, about 12 meters above ground. It moved around us for a moment or two, giving calls, which were also recorded (and which are audible on the video tape, as I recall). All of these vocalizations were strongly reminiscent of (supposedly homologous) vocalizations of P. atricapillus. Thus, I have absolutely complete confidence that those recordings and videotape represent Philydor novaesi, at its type locality. I haven’t seen that videotape in many years, but I believe it has been sent up to Macaulay Library. I will track it down and get it archived at ML and IBC, and send around an aviso when it is publicly available."

Comments from Stotz: “NO. Very interesting and confusing case. I think that the evidence that there are (were) two species of similar foliage-gleaner in these forest patches is not incontrovertible, so I think the best treatment currently is not to recognize C. mazarbarnetti.”