Proposal (59) to South American Classification Committee


Split Celeus obrieni from C. spectabilis


Effect on South American CL: This proposal would elevate a taxon to species rank that we currently treat as a subspecies on our baseline list.


Background: Prior to 1973, Celeus spectabilis (Rufous-headed Woodpecker) was considered to be a single polytypic species consisting of two described subspecies:


C. s. spectabilis Sclater & Salvin, 1880 -- E Ecuador & NE Peru.

C. s. exsul Bond & Meyer de Schauensee, 1941 -- SE Peru, extreme W Brazil & N. Bolivia.


In 1973, Short described a new subspecies, C. s. obrieni, from a single unidentified specimen (pointed out to him by Charles O'Brien) in the AMNH: an adult female, collected 16 August 1926 by E. Kaempfer at Irućui, state of Piauí, Brazil, elevation 124 m, on the Rio Parnaiba. Short refers only generally to the habitat in which obrieni was collected, saying that it was from "dry forested country", and speculating that it was probably widespread in the Piauí-Maranhčo region.


Short's diagnosis of the specimen follows: "Differs from C. s. spectabilis and C. s. exsul in its smaller size (wings, tail, bill, tarsus), and in several color features, especially the greatly reduced barring dorsally, and reduced markings ventrally. Also whiter above and below; small outer rectrices mainly cinnamon (nearly all black in exsul and spectabilis); and secondaries paler, more buffy (less chestnut), especially on tertial feathers. The bill appears yellower, less white, compared with both older and more recently collected specimens of other races."


Measurements given by Short (1973) are as follows:


Wing Chord:
C. s. obrieni  136 mm
C. s. spectabilis (n = 2)  150 mm, 147 mm
C. s. exsul (n = 9) range of 138-153 mm



C. s. obrieni  95 mm
C. s. spectabilis (n = 2) 92 mm, 101 mm
C. s. exsul (n = 9) range 99-108


Exposed Culmen:

C. s. obrieni 24.3 mm
C. s. spectabilis (n = 2) 28.3
C. s. exsul (n = 9) range 28.7-31.2



C. s. obrieni 21.1 mm
C. s. spectabilis (n = 2) 23.4 mm, 23.1 mm
C. s. exsul (n = 9) range 22.1-24.2


Short concluded that the "new form clearly represents a race of C. spectabilis, rather than some other species of Celeus, by virtue of its fully rufous head, its black, shield-like breast patch, it mainly clear rufous secondaries, and its black, unbanded tail. Its bill, although small, matches spectabilis in the slight curvature of the culmen, and in the breadth across the nostrils. In its reduced markings obrieni bears the same relation to other races of C. spectabilis that C. torquatus torquatus does to other races of C. torquatus, and that C. flavescens ochraceus does to other races of C. flavescens."


Subsequent authors have followed Short's lead in treating obrieni as a subspecies of spectabilis: Sibley & Monroe (1990), Sick (1993), Winkler et al (1995), Parker et al (1996), Clements (2000) and Winkler & Christie (2002).


F. C. Novaes conducted surveys in Piauí in the region of Urućui-una close to the type locality of obrieni in 1980, but was unsuccessful in relocating the taxon (Novaes 1992). Novaes described the dominant habitat of the area as cerrado intermixed with caatinga, but with other kinds of vegetation, such as dry scrub forest, low riverine scrub forest, and swamp vegetation dominated by burití palms (Mauritia flexuosa).


The holotype (and lone specimen) of obrieni remained the sole basis for the inclusion of C. spectabilis on the Brazilian list, until 1995, when A. Whittaker, as part of a Goeldi Museum expedition, located and tape-recorded several individuals fitting the plumage characters of C. s. exsul at various localities in the upper Rio Juruá drainage in Acre (Whittaker & Oren 1999). All of the Acre birds were found either in stands of bamboo, or in humid second-growth (Cecropia dominated) bordering rivers; the typical habitats in which the species is found in Peru and Ecuador (e.g. Winkler et al 1995, Parker et al 1996, Ridgely & Greenfield 2001, Winkler & Christie 2002). Whittaker & Oren (1999), commenting on the distinctiveness of obrieni, combined with the huge range disjunction from other populations of C. spectabilis, and its very different habitat, suggested that obrieni should be accorded full species status. They also suggested the English name of "Caatinga Woodpecker" for obrieni, to highlight its fairly unique habitat (among Celeus). Winkler and Christie (2002), noted that "obrieni differs significantly in plumage, and data on habitat indicate major distinction from other races; possibly a separate species, but no further information available." These authors later go on to say "it has to be assumed either that the taxon is extinct or that it represents a highly aberrant form of another species."


Analysis: This is a difficult case, primarily because of lack of information. The taxon obrieni is described from a single specimen, there are no tape recordings of voice, and habitat preference is essentially inferred from the general habitat surrounding the type locality. In general, obrieni appears to be smaller than other Rufous-headed Woodpeckers, although I am largely unimpressed by the degree of difference (except possibly for culmen length) given a sample size of N = 1. On the other hand, plumage distinctions between obrieni and the other two subspecies of C. spectabilis are striking (as illustrated by dorsal and ventral photos in Short 1973). Short's comments that the differences, although major, are in order of magnitude similar to those found between various subspecies in the C. torquatus and C. flavescens complexes are well-taken. However, in both torquatus and flavescens, the different subspecies, although strongly divergent in plumage, have intergrading populations that are known to be vocally similar or identical to one another. With obrieni, we are talking about a highly distinctively plumaged bird that is separated from the nearest known population of spectabilis by ca. 3,150 km, and which occurs in a dry forest-caatinga-cerrado biome, whereas both spectabilis and exsul are birds of humid forest habitats. Whittaker and I searched for obrieni near the type locality in February of this year (without luck): we saw no habitat approaching the habitats in which we have seen C. spectabilis in Peru or w. Brazil. Aside from a few remnant patches of humid (semi-deciduous) terra firme forest (which were occupied by C. flavescens ochraceus), the habitats encountered were notably xeric. It seems safe to assume that regardless of the precise habitat in which obrieni was collected, it was not even remotely similar to that occupied by C. spectabilis in the remainder of its range. Given the extreme range disjunction, very different habitat types, and quantum plumage differences of the taxa involved, I am somewhat surprised that obrieni was not described as a separate species in the first place, but Short (1973) makes no mention that this was even under consideration.


Recommendation: In spite of my general reluctance to describe anything on the basis of a single specimen (especially lacking vocal data), I recommend splitting obrieni as a separate species from C. spectabilis on the basis of a hugely disjunct range, occupation of a different biome, and quantum differences in morphology. The caatinga region of northeast Brazil is a noted region of endemism, and recognition of obrieni as a separate species-level taxon would fit an established biogeographic pattern. Although I don't believe that conservation considerations should drive taxonomic decisions, elevation of this obscure form as a species-level taxon would have the added benefit of focusing more attention on locating an extant population, and on conservation initiatives for the caatinga region of Piauí in general.


Literature Cited:

CLEMENTS, J. F. 2000. Birds of the world: A checklist. Ibis Publishing Company, Vista, California.

NOVAES, F. C. 1992. Bird observations in the state of Piauí, Brazil, Goeldiana Zool. 17:1_5.

PARKER, T. A., III, D. F. STOTZ, AND J. W. FITZPATRICK. 1996. Ecological and distributional databases. Pp. 132_436 in: STOTZ, D. F., J. W. FITZPATRICK, T. A. PARKER III, AND D. K. MOSKOVITS. Neotropical birds: Ecology and conservation. University of Chicago Press, Chicago, Illinois.

RIDGELY, R.S., AND P. J. GREENFIELD. 2001. The birds of Ecuador. Vol. 1. Status, distribution, and taxonomy. Cornell University Press, Ithaca, New York.

SHORT, L. S. 1973. A new race of Celeus spectabilis from eastern Brazil. Wilson Bulletin 85:465_467.

SIBLEY, C. G., AND B. L. MONROE, JR. 1990. Distribution and taxonomy of birds of the World. Yale University Press, New Haven, Connecticut.

SICK, H. 1993. Birds in Brazil. Princeton University Press, Princeton, New Jersey.

WHITTAKER, A., AND D. C. OREN. 1999. Important ornithological records from the Rio Juruá, western Amazonia, including twelve additions to the Brazilian avifauna. Bull. B. O. C. 119:235_260.

WINKLER, H., D. A. CHRISTIE, AND D. NURNEY. 1995. Woodpeckers: An identification guide to woodpeckers of the world. Houghton-Mifflin, Boston, Massachusetts.

WINKLER, H., AND D. A. CHRISTIE. 2002. Family Picidae (Woodpeckers). Pp. 296_558 in: DEL HOYO, J., A. ELLIOTT AND J. SARGATAL eds. Handbook of birds of the world. Vol. 7. Jacamars to Woodpeckers.


Kevin J. Zimmer, August 2003


P.S.: If the proposal passes, then I'll follow up with another one on the English names.




Comments from Robbins: "Although it seems likely that obrieni is a valid species given the distinct plumage and the large distributional disjunction from Celeus spectabilis, I'm against elevating it to species status based solely on a single specimen. I'd even support this if Kevin and Andy had seen birds that matched the description of obrieni, but at this point we can't rule out the possibility, however unlikely, that the holotype is an aberrant or hybrid individual."


Comments from Jaramillo: "YES. The great difference in plumage, habitat, and massive range disjunction all suggest that this is a good species. It is unfortunate that we have only one specimen to work from, and no recent sightings of this taxon. However, the hybrid theory for this specimen is exceedingly unlikely in my opinion. The question would be hybrid of what? Particularly with respect to the place it is known from. I don't know that part of the world but is Celeus flavescens the only possibility there? Is there any other Celeus there which could be the potential other parent? I think that in this case the possibility that this single specimen is a hybrid is so remote that at least for me it becomes a non-issue."


Comments from Silva: "YES. obrieni remains as a big problem. I have a student (Marcos Pérsio Santos) living in Piauí. He spent lots of time looking for this taxon and did not find anything. Certainly, it is a distinctive taxon and I do not think that it represents a case of hybridization. The type-locality is a mosaic composed by cerrado, caatinga and dry forests. Unfortunately, most of the tall (25 m) dry forests that once was widespread in this region has been already completely modified.  If obrieni is associated with this kind of habitat, it is certainly under very strong threat.  I agree with Kevin that obrieni does not belong to Celeus spectabilis, a completely different species.  Although the weak evidence (one specimen and no recent record) I will vote yes in this proposal."


Comments from Stotz: "YES. Celeus obrieni is clearly related to spectabilis. Given that spectabilis doesn't occur anywhere close to the type locality of obrieni and the very distinctive plumage of obrieni, it seems essentially impossible to me that obrieni could be explained as a hybrid or aberrant spectabilis. Although the published justification is not very strong, I think it is sufficient to hang our hat on, and recognize obrieni."


Comments from Stiles: "[YES] Somewhat reluctantly, I will recommend species status. Reluctantly, because I dislike recognizing taxa from single specimens, especially when no reliable field observations are available. However, to my knowledge Kaempfer was a reliable collector and the data on the label are specific as to date and locality. The morphological differences in combination with the wide range disjunction and presumably very different habitat (here is where I am least content with the lack of field data, but if the locality is accurate the habitat must be different from spectabilis) all point to obrieni being something different."


Comments from Schulenberg: "My vote is NO. But that is a vote on principle, pure principle.


Kevin Zimmer does a good job of convincing me that obrieni deserves species rank. My only concern is that Kevin's well-reasoned arguments are not published. I don't see any substantive difference between Peters, or a field guide author, publishing a taxonomic decision with little or no available rationale; and our committee publishing or reaching a taxonomic decision on the basis of "inside" information, with little or no available (to anyone outside of SACC) rationale. In both cases, from the perspective of anyone on the outside, the decision is made inside of a "black box."

Admittedly, outsiders might stumble across our SACC proposal website, or we might even direct people to it. But what is the long term prospect of the website? To the greatest extent possible, I'd prefer for our decisions to be based on published information and published rationales. Kevin's proposal looks to me to be a great first draft of a short note for publication on this taxonomic change; if such a manuscript had been submitted already, then I would vote in favor of this proposal in a heartbeat.


I too am uncomfortable, of course, with species known from only a single specimen. Hybridization seems very unlikely to me in the instance, however: what would the parent combination be? And outside of some well known instances of secondary contact (e.g., Colaptes), so far as I know woodpeckers are not particularly prone to hybridization anyway.


It would be interesting to know what other specimens were collected the same day and place as the type of obrieni, to see whether this might give some clues as to what habitat the bird might be found in (or once occupied?)."


Comments from Whitney: "Celeus spectabilis obrieni is known only from the type at AMNH. That proposal in the literature to split it as a species is ludicrous (nothing more than a statement that it should be elevated to species rank because it's apparently unique). This question deserves careful analysis, which Kevin should do and publish before the SACC accepts it as a species.


I looked at the specimen a few years ago, and I wish I could lay may hands on the video and notes I made, but it's late and this may be a done deal now. Anyway, it's a very strange specimen, I think. I was not convinced, on looking at it, that it could not be a mutant individual showing a combination of characters that are all individually present across the genus, or perhaps a cross or backcross between C. flavus and another species (flavescens, even torquatus potentially involved). C. flavus has a lot of rufous in the wing up in that region. Finally, I went to Urućuí back in about March 1999. If the bird is at all like spectabilis, it is almost certainly going to be in a river-created habitat. I saw no sign anywhere near (above or below) Urućuí of patches of Gynerium, bamboo, other primitive grasses, other river-edge veg that looked like it might be OK for something similar to C. spectabilis. Who knows what the river looked like when Kaempfer was there -- but I can report that the entire region is basically cerrado, often right down to the bank of the river, and it's in amazingly good shape, I imagine very much like it was back then. All of this means next to nothing, of course -- but I do believe that obrieni is best left alone until someone finds it alive or gets a molecular analysis of it done. At least find out what its mother was. If she was not any of the species we have on the books, then that's enough to convince me that it can be elevated to species rank. (Not that this info is recoverable from this specimen, but it certainly could be tried).


What's the point of elevating it without clear reason? Feel free to forward this message to others if you think appropriate."


Comments from Remsen: "NO. In view of tom and Bret's comments above, I change my vote from YES to NO. When someone with Bret's experience and instincts has actually handled the type specimen and has his doubts, that's enough for me, and Tom's philosophical points are on target."


New comments from Zimmer: "Both Tom and Bret's points are well-taken, and indeed, I think I conveyed my own general reluctance to recognize anything based on only a single specimen. I also want to state that I respect Bret's informed views on this particular situation. However, I think a few of his points call for clarification. I don't think that proposals in the literature to elevate obrieni to species status are "ludicrous". Rather, they seem natural given the uniqueness of the type specimen and the huge range disjunction from any known population of Celeus spectabilis, the species with which obrieni was lumped by Short. Short also examined the type of obrieni, and compared it with all other Celeus at AMNH. He obviously thought the specimen was different enough (i.e. not clearly a mutant or hybrid) to warrant the creation of a new subspecies. Given the range disjunction and the biome from which the specimen originated, I'm surprised that he decided to describe obrieni as a subspecies rather than as a distinct species, but my point is the same: an experienced taxonomist looking at the specimen against a backdrop of all other Celeus still decided it represented a distinctive taxon rather than a hybrid or mutant of some type.


“I think that all too often we tend to invoke the hybrid theory when confronted with single specimens of distinctive taxa. For some groups, in which hybridization is a proven frequent phenomenon, such speculation may be justified. But for most other birds, I think it is less likely that someone collected the ultra-rare hybrid than that the specimen is of a valid taxon whose range and/or microhabitat we simply haven't managed to pin down. The validity of Pithys castanea was constantly questioned until it was rediscovered by LSU personnel. Ditto for Hemitriccus inornatus, Hemitriccus (Todirostrum) senex and Pipra vilasboasi. What about Conothraupis mesoleuca? It remains known from a single specimen (less distinctive than Celeus obrieni). Is it ludicrous that we continue to recognize it as a distinct species? Had Short described obrieni as a distinct species rather than as a distinct subspecies, it would be in the same boat as the Conothraupis -- an enigma that no one has been able to refind despite much searching in the vicinity of the type locality.


“We also have to ask what are the likely parents that would produce this hybrid. The only Celeus that we found in the region was Celeus flavescens ochraceus. The other Celeus species that Bret mentions (torquatus, flavus) are humid forest or gallery forest bird, and this habitat is lacking from the region. To accept the hybrid theory, we must not only accept that this rare occurrence (i.e. two Celeus species interbreeding and producing a spectabilis-looking hybrid) took place, but that the resulting hybrid was some sort of vagrant that made its way out into inhospitable habitat that was inconsistent with the habitat occupied by its parent forms.


While Bret found the type specimen of obrieni odd (i.e. possibly a mutant, hybrid, or backcross), his examination of it obviously wasn't enough to deter him from making the effort to search extensively for the bird near the type locality (which isn't exactly on your way to anywhere). Based on my own time in that region, I agree with Bret's assessment that there is no habitat typical of Celeus spectabilis. This proves nothing, because we don't have any real reason to think that obrieni would be ecologically similar to spectabilis anyway. Short placed obrieni with spectabilis on the basis of plumage characters alone. I don't agree with Bret regarding the lack of any real alteration to the native habitat since the time of Kaempfer. As Jose Maria has noted, much of the tall forest from Piauí has disappeared. A similar assumption that habitat loss could not explain the disappearance of Spix's Macaw predominated for some time, based on the abundance of caatinga extant within its former range. It is now known that the macaw was intimately tied to a critical abundance of caraiba trees; a microhabitat within a habitat. For a big woodpecker, for which potential nest trees could be a limiting factor, it seems entirely conceivable that destruction of tall, semi-deciduous forest or systematic removal of larger trees from remaining stands of such forest could explain a massive decline in a specialized species. I found caatinga and dry forest to be present over large areas of Piauí. At the same time, I found only remnant stands of taller forest, which supports Jose Maria's observations.


“None of this proves that obrieni is a distinct species, but I do think that the only reason we are having this conversation is because of the "accident" that Short described it as only a subspecies in the first place. If the hybrid/mutant argument prevails, then I see no more justification for recognizing obrieni as a subspecies of spectabilis than I do for recognizing it as a distinct species. To subscribe to the hybrid/mutant theory or the "can't recognize anything based on a single specimen argument" would invalidate recognition of obrieni at any level. If the bird is a legitimate taxon, I don't see how it can possibly belong with spectabilis. To maintain the status quo until the bird is discovered in life, is to essentially validate Short's description of the bird as a distinct taxon. Once you've accepted that part, the elevation of obrieni to species-level is not such a leap.”


New comments from Silva: "After reading carefully all new comments, I think I will keep my vote on obrieni."


More comments from Bret: "I guess I'll add a few more thoughts on the Celeus obrieni proposal, especially since Kevin called for some clarification of my comments.  First, I apologize for the word "ludicrous".  When I saw it written back at me, it did look harsh.


Hey, I don't know what this specimen represents.  Without a molecular analysis of it, I doubt we are going to "find out".  Now, if someone does manage to prove that it exists in the wild, that's great.  But that hasn't happened despite some good observers looking at the single known locality of occurrence (and Kaempfer's locality *can be trusted*; as I recall, his notes showed that he was at the end of an extended stay at Urućuí when he collected it, and it looked to me like he must have been close to town). For sure, Kevin, I thought it was worth going to the type locality to look for the bird. It is a distinctive looking bird. And Celeus woodpeckers, especially those in relatively open habitats, are almost certainly going to be heard or seen within three days.  (However, I wouldn't call that specimen, or most other Celeus, a "big woodpecker").


Where does this leave us?  Well, someone could write a paper actually evaluating the specimen and all associated data and reasoning.  The conclusion could be that it is a species-level taxon, or that it is a mutant/hybrid.  I think that's it for possibilities (unless the conclusion is that it's not possible to pin it down on present knowledge, surprise).  Short decided it was a lot like spectabilis and called it a subspecies; nobody seems very happy with that.  But Tom's got it quite right when he says that elevating it to species level without anything more than "it's disjunct and looks distinctive" is too arbitrary (not that we can't be a little arbitrary in some cases). So, I continue to think that if anyone wants to elevate obrieni to species level, then they should publish a well-balanced, objective argument for it.  That argument will have nothing to do with whether or not it was originally described as a species or subspecies.


Some points to be considered, among many others: Yes, if it is a simple F1 hybrid, it cannot be explained.  I don't think it could be this kind of hybrid, and I'm not arguing that it is a more complicated (and infinitely rarer) type of hybrid.  I will comment, however, that a wandering individual torquatus or flavus in this region is not too far-fetched.  There is a mosaic of habitats, and I suspect that populations of these species could be found not terribly far away.  Should such an individual end up mating with a "resident" flavescens, and if this were to happen again with another species... dream on.  At least this can be checked!  So, let's see *who it's mother was* -- if it is clearly a named species, then the case for obrieni as a valid taxon at any level is out... with one exception.  If the mother was especially close to spectabilis, there is a good argument for either splitting obrieni as a species [disjunct, distinctive sounds good now] or leaving it as a well-marked subspecies, depending on where one draws that line (and it looks like most/all of us, including me, would call it a species).  I am not arguing that it is a mutant displaying an odd combination of Celeus-like characters, though I think that is possible.  I *am* arguing that we cannot be sure what it is, *especially in the absence of a good, detailed analysis of the situation*.


It should be noted that there is no caatinga anywhere near the type locality.  There is no "transition to caatinga" anywhere near the type locality.  This entire region of Piauí and Maranhčo is a mosaic of cerrado, cerradčo, palms, and somewhat more mesic woodland.  This is reflected clearly on vegetation maps and in the avifauna present there -- right, Kevin (or am I missing something)?  In 1999 (I think that's when I was there), it was all in pretty darned good shape, with extensive cleared areas only around Urućuí itself (along with some nice habitat): basically, I was impressed.  The area was on the verge of massive alteration, however.  The tractors were already at work.  Farmers from Paraná, in particular, had bought up vast tracts of land and were planning to level it all for soybeans within the next few years; highly mechanized operations.  I talked with several of these guys, independently, always by accident, always without really wanting to. (Never stop for a beer where there are 0 women present. Or only 1 woman. And certainly don't do it more than once.)  Thus, aligning this locality with caatinga areas of endemism and suggesting that obrieni could be a species that fits some caatinga-related pattern, calling it "Caatinga Woodpecker" is inappropriate at best.


The idea that we are worried about this at all is due to it being described as a subspecies doesn't look at the other side of the coin: One might well have submitted a proposal to suggest that there are some reasonably good points in favor of C. obrieni not being a good species. Lumping?  Yes, that's always right there besides splitting.


Finally, somewhat sheepishly:  I saw a female Conothraupis mesoleuca at P. N. Noel Kempff Mercado back in about March 1993.  It sat in good view for more than a minute.  This was not far from the type locality, and I feel confident that it couldn't have been a speculigera though it sure looked a lot like it, mostly being grayer, less greenish as I recall.  Since the female is undescribed, and I've been meaning to get back for a more careful look at the area, I've just let it slide.  I am a strong proponent of getting documentation for stuff like this; I still don't have it.  No further excuse.


Thanks to all for considered opinions, but this is the last from me on this one."


Comments from Nores: "[YES] Si. Las razones para reconocer a un nuevo taxa son similares, ya sea para especies o subespecies. Si uno acepta que el único ejemplar conocido tiene características diferentes como para aceptarlo, y estas características son a nivel de especies, me parece que lo lógico es considerarlo como especie."