Proposal (617) to South American Classification Committee


Recognize newly described Nystalus obamai and split Nystalus striolatus into two species



Effect on South American CL: If adopted, this proposal would add a newly described species of Nystalus to the list and elevate to full species status one subspecies of N. striolatus (torridus).


Background: The Striolated Puffbird, Nystalus striolatus is an Amazonian endemic species with two currently recognized disjunct subspecies, one from Mato Grosso, Brazil westward to the foothills of the Andes (nominate striolatus), and another in eastern Pará, Brazil (torridus). According to Bond and Schauensee (1940), N. s. torridus is distinguished from the nominate form by its overall darker coloration, larger size, color of the throat, more heavily streaked breast and underparts, and spotted interscapular region.


New information: As a result of a multi-character taxonomic revision of the Nystalus striolatus complex, Whitney et al. (2013) uncovered a new phenotypically and genetically distinct population in south-central Amazonia (to which the name striolatus actually applies), prompting the naming of the westernmost population as N. obamai. This study also supported a separate species status for the phenotypically and genetically distinct easternmost population (torridus).


Analysis/Recommendation: The fact that the Bayesian phylogeny recovered N. obamai as monophyletic with high statistical support, and that it is separated from both N. s. striolatus and N. s. torridus by similar uncorrected genetic p-distances (respectively, 3 and 3.5%), in conjunction with morphological and vocal diagnoses (Whitney et al. 2013), supports its recognition as valid species level taxon.  Similarly, the same can be inferred for N. s. striolatus and N. s. torridus, which are separated by 3.2% of uncorrected sequence divergence in addition to the aforementioned vocal and morphological diagnoses. Hence, we recommend the recognition of three species in the N. striolatus complex: Western Striolated-Puffbird (N. obamai; distributed west of the Madeira river in Brazil and both banks of the Ucayali / Marañon in Peru and Ecuador, as well as along the foothills of Andes in Bolivia and Peru; Fig. 1); Natterer’s Striolated-Puffbird (N. striolatus; distributed in the Madeira - Tapajós interfluve in lowland Amazonian Brazil and northern Depto. Santa Cruz, Bolivia); and Eastern Striolated-Puffbird (N. torridus; occurring east of the Tapajós and south of the Amazon rivers to western Maranhão and northern Tocantins in Amazonian Brazil).


Literature Cited


Bond, J. and R. M. Schauensee (1940). Description of a new puff-bird from the lower Amazon. Notulae Naturae 50:1-2.


Whitney, B.M., Piacentini, V.Q. Schunck, F., Aleixo, A., Souza, B.R.S., Silveira, L.F., Rego, M.A. 2013. A name for Striolated Puffbird west of the Rio Madeira with revision of the Nystalus striolatus (Aves: Bucconidae) complex. Pp. 240–244 in: del Hoyo, J., A. Elliott, J. Sargatal, and D.A. Christie (eds.) (2013). Handbook of the Birds of the World. Special Volume: New Species and Global Index. Lynx Edicions, Barcelona.


Carlos Eduardo B. Portes and Alexandre Aleixo, December 2013





Comments from Remsen: “NO.  The vocal difference between obamai and the other two taxa is clear and impressive, and I would vote for a revised proposal that focuses solely on that taxon.  But ranking torridus as a species is a stretch in my opinion.  The sonograms of obamai and the recordings I listened to on Xeno-canto show clearly the stuttered introduction noted in the text, which then states that torridus is most like obamai, despite their geographic separation by nominate striolatus:


Of the other two taxa, the song of N. s. torridus is most similar in that it often features a slight break in the first part almost producing a bisyllabic effect.”


“However, one of the two sonograms (F) presented for torridus does not show the stutter, and notice that the text states “often”.  So, without larger N and documentation of individual vs. taxon variation, I’m not convinced.  Figure 1 seems to indicate that only 6 recordings of torridus and only 3 of nominate striolatus were analyzed, so that concerns me when the differences are, to me, subtle.  Although I trust Bret and his co-authors on these sorts of judgments, I think that more thorough sampling and quantitative analyses are required before treating torridus as a species.

“The plumage differences are small and basically irrelevant to species limits as far as we know (tangentially, a fundamental flaw in my opinion with the “tally up the differences” approach of the Tobias et al. scoring system for determining species rank).  That leaves the genetic data, the significance of which is vastly overstated in the paper as well as in the proposal.  As Gary notes above, we have a grand total of 4 individuals from a single mitochondrial gene!  There is really no point in discussing monophyly when you only have 4 individuals and 3 taxa.  The small (by tropical forest bird standards) genetic distances among the individuals would not be unexpected within a single taxon (by anyone’s standards) that occurs over such a large range, simply by virtue of isolation by distance.

            “I recommend that this proposal be overhauled, with a part A (obamai) and part B (torridus).”


Comments from Cadena: “NO. I believe obamai may well be a good species, but I find that the information presented in the paper describing it is insufficient to demonstrate it clearly. Vocalizations do sound different, but there are no quantitative analyses and the authors mention the existence of vocal variation within obamai, so this makes me wary about accepting the conclusions on vocal differentiation (it would be easier for me if at least some indication of the number of recordings examined and their geographic distribution were provided). I realize that most of the vocal differences are qualitative and not quantitative (i.e. the introductory stutter in obamai), but I also wonder about their mentioning that obamai responds (albeit less strongly) to songs of torridus. Could it be that the stutter is not relevant for species recognition? The genetic data say little to nothing about species status for allopatric populations, and evidently too much is made in the description about reciprocal monophyly with such an extremely limited data set (4 individuals, 2 of which are obamai, possibly from the same locality? I could not find locality data for these in the paper). Plumage variation is slight (arguably similar to plumage variation existing among subspecies) and, although there is no formal statistical analysis, obamai does not appear morphometrically distinct relatively to the other two taxa.”


Comments from Stiles: “NO, at least for now. The vocal differences are not great, but might be diagnosable with statistically meaningful samples from all three populations over their entire distributions. A photo or two showing representative specimens of all three forms should also be included, as well as a thorough analysis of morphometrics.  The genetic differences are appreciable, but with such small samples “reciprocal monophyly” is not demonstrated, just assumed: all we can really say is that the specimens differ by x %.”


Comments from Nores: “NO, for the reasons given by Van, Gary and Daniel.”


Comments from Zimmer: “NO.  I also think that this proposal needs to be split, so that we can deal separately with obamai versus splitting striolatus.  I also would like to see some bigger sample sizes, with more geographic breadth, and a true vocal analysis.”


Comments from Bret Whitney: Before the voting reaches quorum on this proposal (and others from the Special Volume of HBW), I suggest that you all take into consideration the information on specimens examined, sample sizes, and spread of localities available in Supporting Information on the IBC website (click on the emblem in the middle of their home page).  This information is appropriately cited in each of the descriptions, yet it is obvious that no one who has voted thus far has bothered to consult it.  For example, Nystalus obamai was given a fairly thorough vocal vetting, with n = 4 for striolatus, 14 for torridus, and 30 for obamai.  These n should have been included in the description, for sure; I neglected to put them in there.  However, the map (Fig. 1) indicates all localities examined for each taxon, whether vocal or skin.  There, sample size of vocalizations examined is possible to infer:  4 for striolatus, 7 for torridus, and 20 for obamai.  Some of these localities were represented by more than one individual, which becomes clear in the SI.  The description should have made it clear that there was no overlap in these three vocal types as mapped in Fig 1 (all it says is that vocalizations were “diagnostic”, which is not a well-supported statement without context of the SI, and I should have provided a stronger statement).  We did mentioned that there is some variation in vocalizations of obamai from across its range, and included a couple of examples exemplifying it, but I guess I should have added a phrase to make it clearer that the vocal type is still 100% diagnostic of obamai everywhere it is mapped in Fig. 1 (Van, you’ve understandably distorted the bit about similarity to torridus — no, the two are not at all similar, it’s just that torridus is slightly more similar to obamai than is the geographically most proximate taxon, striolatus).  Morphological differences are minor at best (does that tip them to subspecies?), and the genetic analysis is clearly nothing close to being definitive, but, quite honestly, it is “better than nothing”, which is why we included it — basically just to show that we probably are dealing with three discreet populations that are quite clearly indicated by the geographic spread of diagnostic vocalizations.  The key word in that sentence as concerns the genetic data is, of course, “probably”.  Personally, I do believe it is “probable” that we have three monophyletic groups, which is the best we can do for many/most of these taxonomic decisions, mainly because of poor to very poor (like this case) sampling of both n and geographic spreads.  I’m optimistic, however, that we will be able to gather adequate, perhaps even robust, samples of voice-recorded specimens from lots of critical regions (in Brazil, at least) over the coming years.  Then, and really only then, will we be able to make proposals for “new species” that actually have meaningful data behind them — especially true when we eventually determine which genes influence and control characteristics relevant to mate choice, and how those genes are shaped by abiotic forces.  For the time being, however, taxonomists are bumping heads in a dimly lit swirl of real and artificial trees."


Comments from Robbins: “NO. Given all that has been stated, including Brett underscoring actual samples that were examined, it seems that at least obamai deserves species status. So, I agree with remarks by other committee members that a separate proposal should be made recognizing obamai. Yes, torridus may indeed merit species status, but until there is additional sampling (genetic, vocal) I think the prudent course of action is to recognize only obamai as a species”