Proposal (#329) to South American Classification Committee

Recognize Scytalopus diamantinensis (Diamantina Tapaculo) as a valid species

 

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

Bornschein, Maurício, Belmonte-Lopes, Mata and Bonatto (2007) recently described a new species of tapaculo, Scytalopus diamantinensis (Diamantina Tapaculo) from Bahia, Brazil. The new species is known exclusively from the Chapada Diamantina geomorphological unit between 850 and1600m (8 localities with vouchers), mainly its eastern slopes. The new species is a member of the taxonomically complex S. speluncae group (see Mauricio 2006), within which it is most closely related to S. novacapitalis, S. pachecoi, and Scytalopus sp. nov. from the southern Espinhaço Range. (A nomenclatorial controversy exists here. This taxon was treated as the genuine S. speluncae by Raposo et al. 2006, but not by Bornschein et al. 2007). S. diamantinensis is diagnosable by vocal, plumage, and molecular characters from all other Brazilian taxa. Although its song differs little from those of the closely related species, its calls are notably distinct.

The Scytalopus populations occupying the Chapada Diamantina (Bahia) was regarded as representing the same taxon found in the southern part of Serra do Espinhaço (Minas Gerais) by Raposo et al. 2006, based on 4 specimens. However, recognition of the population of Diamantina as a distinct taxon by Bornschein and colleagues is supported by analysis of 5 specimens (5 males, 1 female, all with tissue samples examined) and 369 samples of voices of at least 18 individuals, as well as the examination of a much larger series of the related taxa.

The authors believe that S. diamantinensis, S. pachecoi, and Scytalopus sp. nov. would probably be lumped into a single polytypic biological species. Therefore, they recognized all four allopatric taxa with accelerating songs as distinct (at least) phylogenetic species. Furthermore, the mean differences in songs and consistent differences in calls in this clade are strongly correlated not only with clear genetic distinctions, but also with plumage differences.

Recommendation: I recommend a "YES" vote on accepting this tapaculo as a new (indeed biological) species to our list, based mainly on the genetic distinctions found among the related taxa. A member of this group (S. pachecoi) is already treated as valid by SACC (see Proposal #196).


Literature Cited:

BORNSCHEIN, M. R., G. N. MAURÍCIO, R. BELMONTE-LOPES, H. MATA AND S. L. BONATTO. 2007. Diamantina Tapaculo, a new Scytalopus endemic to the Chapada Diamantina, northeastern Brazil (Passeriformes: Rhinocryptidae). Rev. Bras. Orn. 15(2):151-174.

RAPOSO, M. A., R. STOPIGLIA, V. LOSKOT AND G. M. KIRWAN. 2006. The correct use of the name Scytalopus speluncae (Ménétriés [sic], 1835), and the description of a new species of Brazilian tapaculo (Aves: Passeriformes: Rhinocryptidae). Zootaxa 1271: 37-56.

José Fernando Pacheco, Jan. 2008


=========================================================================

Comments from Thomas Donegan: "The proposal to recognise Scytalopus diamantinensis (Diamantina Tapaculo) at first may appear a difficult one to accept.  Species rank was assigned to S. diamentinensis in the description with reference to diagnosable calls (scolds) and immature plumage under a PSC concept.  However, only mean differences in songs exist.  Songs are generally considered of primary importance in differentiating species in this genus.  Krabbe & Schulenberg (1997, 2003)'s taxonomy for Scytalopus (on which the SACC sequence is based) treated taxa with diagnosable calls, but not songs, as (BSC) subspecies.  I am not sure which species concept SACC employs.  However, a "No" vote would plausibly leave S. pachecoi as the only biological species level taxon of this three-taxon group.
 
"No published example (of which I am aware) exists of two Scytalopus species on the SACC list which vary diagnosably in their calls but not in their songs.  However, if reference can be made to soon-to-be-published work, a "Yes" vote is likely to prove to be the better long-term decision.  S. griseicollis and S. spillmanni replace one another sympatrically (actually in elevational parapatry and at distances of 10 m horizontally / 2 m vertically) at montane forest-paramo ecotones in the East Andes of Colombia.  Jorge Avendano and I have an MS in review which shows these species to differ diagnosably (per the Isler diagnosablility test for antbird songs, but for fewer variables) in their calls (scolds) and plumage but not in their songs - the same scenario presented by S. diamentinensis/pachecoi.  Although results of the East Andes study are not published yet - expected for Spring 2008 - they will bolster the case for recognition of S. diamentinensis, the undescribed Brazilian taxon and possibly some other Scytalopus populations as species under biological species concepts."

Comments from Stiles: "A tentative YES. Clearly this group of closely related forms is at least in the process of speciating and given the consistency of morphological, genetic and vocal differences among them is probably best united at the superspecies level rather than at the species level. Convergence in songs to reinforce interspecific territoriality among ecologically similar forms with parapatric distributions is not unknown and probably applies to the case mentioned by Thomas. The present case is rather one of limited divergence rather than convergence, but one could envision selection favoring conservatism in long-distance vocalizations like song in the absence of contact zones, while being relaxed regarding call notes which in any case tend to be more graded, close-range signals."

Comments from Robbins: "NO. Until the Donegan paper is published and available for review, I vote "no" for recognizing this taxon as a species."

Comments from Stotz: "NO. Given that Krabbe and Schulenberg is in my view the current state of the art on Scytalopus taxonomy, and the authors conclude that based on those criteria, this taxon would be considered part of a polytypic species and that the genetic details for this taxon, including the relationships among the various taxa of Brazilian tapaculos, are still to be published, I think it is premature to recognize this as a distinct species."

Comments from Nores: "NO. Teniendo en cuenta que no hay diferencias en coloración de los adultos, ni tampoco en el canto (sólo en los llamados), me parece que es más probable que sea solo una subespecie que difiere en tamaño."

Comment from Niels Krabbe: "I would vote no. One cannot use a phylogenetic species concept for some species and a biological species concept for others. No matter what the genetic distance, similar songs suggest they would interbreed if in contact.

"There are already at least two similar cases, where a population with a distinctive homologous call is not given biological species rank.

"These include S. spillmanni from the Western Andes of Colombia (see p. 44 in Krabbe et al. 2006, Orn Col. 4: 37-48), and Scytalopus canus opacus from southern Ecuador (see p. 72 in Krabbe and Schulenberg 1997, Orn. Monogr. 48).

"Genetically these two cases differ, the spillmanni from the West Andes of Colombia are almost identical to populations in the Central Andes of Colombia and Ecuador, whereas south Ecuador opacus differs from more northerly populations by about as much as diamantinensis does from its closest relatives.

"One may speculate what happens if two populations with different calls meet. One call might disappear, or both calls might remain as alternate calls (intermediate calls seem unlikely). I know of at least one case (a Scytalopus from ApurÌmac) of a form with two different calls that I suspect could have such an origin."

Comments from Jaramillo: "NO - I will await genetic data to see what is upheld and rejected then. Since our recognized taxa in Scytalopus are largely based on the work of Schulenberg and Krabbe, it seems logical to apply their criteria to this taxon. It appears that under their criteria they would not consider this a species."

Comments from Nacho Areta: ""I think that the main problem is that there is a new question regarding the functions of voices in the Scytalopus that needs to be answered. We know, largely thanks to the work by Krabbe et Schulenberg, that different songs indicate phylogenetic and biological species in Scytalopus. But we do not know  with the same degree of certainty what the phylogenetic/biological meaning of differences in calls could be. For example, are calls more important than song in maintaining pair-bonds? What if female choice is based on calls and not in songs? Another case of similar/identical songs but (apparently) differing calls and certainly differing phenotypes was presented by Chendo Fraga and I in the last NOC in Venezuela comparing the true Scytalopus magellanicus and a population living in Mendoza and Chile. This population is so different from nominate magellanicus that it was considered to be fuscus by some previous authors. I think that we should strive to understand the biological meaning of the calls to make a valid taxonomic judgment on this proposal. It is not only a matter of comparing how different the voices of are, but of what are the effects of these differences for real/potential interbreeding of the forms."

Comments from Cadena: "YES. This is a hard one for the reasons pointed out by several committee members, but I think that Nacho has made a good point: we simply don't know what is the function of calls in mate choice in Scytalopus. This doesn't mean that calls are not important and only songs are. Calls might even be crucial, as it seems to be the case with the two Hypocnemis ex. cantator that are sympatric in Amazonian Peru, whose songs are remarkably similar but they retain their integrity as separate species, possibly because their calls differ. Also, as noted by Bornschein et al., it is important to realize that variation in songs among species of Scytalopus that are already recognized by SACC and other authorities as distinct (e.g. S. iraiensis and S. speluncae) is minor, and the only substantial vocal differences betweeen such taxa are in their calls. Thus, I would argue that the status quo in this group is based on criteria that are not as stringent as the ones that are now being applied for this particular taxon. Because S. diamantinensis is diagnosable based on plumage and calls, and because it is genetically distinct (Alvaro, the paper describing it already has molecular data), I think that in the context of how Brazilian species in this group have been delimited and accepted by SACC in the recent past, the data are most consistent with the idea that S. diamantinensis should be considered a separate species."

Additional comment from T. Donegan: "The paper referred to above has now been published. http://www.ornitologiacolombiana.org/oc6/doneganyavendano.pdf ."

Comments from Schulenberg: "YES, largely for the reasons outlined by Daniel. I recognize that this may cause some inconsistencies, if Andean taxa are defined more on the basis of songs than of calls. But, inconsistencies happen. These may reflect differences in the biology of Brazilian vs Andean Scytalopus (i.e., differences in the "significance" of different vocalizations), or the inconsitencies may suggest that perhaps we should revisit the Andean taxa with calls as well as songs in mind. The much greater amount of genetic information on Scytalopus that is emerging also will be shining some light on the issue. As Daniel mentioned, the genetic data from Brazil already point to species status for these taxa."

Comments from Remsen: "NO. Diamantinensis is clearly a valid, diagnosable taxon, by three suites of characters. The question is at what rank? Plumage differences define other Scytalopus taxa recognized as subspecies, so this is of no help. No diagnosable differences in song are known, so that points towards subspecies rank to be consistent with Krabbe-Schulenberg scheme for Andean taxa. The differences in calls are of interest and merit further study. If differences in call notes are found to be associated with cessation in gene flow, then I will change my vote. To be consistent with my vote, I also consider other members of this Brazilian group that we currently treat as species to better fit the rank of subspecies. [Someone who knows more about this group should consider a proposal for change.]

"Concerning 'genetic differences' --- of course they differ genetically. If we assume that plumage and calls have a genetic basis, then we know they differ genetically even if we can't measure the genes responsible for the differences. Genetic differences per se do not count for anything in evaluating taxon rank of allopatric populations. Yes, genetic differences between parapatric and sympatric taxa are obviously critical to assessing gene flow, but when comparing allopatric populations, I think they are next to useless. Any two populations isolated for a certain number of generations are going to begin accumulating genetic differences. We can often detect this in the pitiful number of fast-evolving genes that we can sample (relative to the 25,000 or so genes out there in the bird genome), but using this as any sort of yardstick lacks conceptual or (so far) empirical underpinnings. North American wood-warblers, flycatchers, and others with clear plumage pattern or vocal differences and widespread syntopy are sometimes barely distinguishable genetically, but "our" ongoing work at LSU reveals tropical species with up to 10% sequence divergence (mtDNA) for which there are no detectable phenotypic differences. Thus, the data themselves on genetic divergence and speciation show nearly as large a range as we can measure between congeners -- they range from near 0% to more than 10%. Even restricting the comparisons to sedentary, tropical congeners (as in Scytalopus) will likely reveal a wide range of degrees of genetic divergence. Why? The latter depends not only on time-since-separation, but also on initial population size and structure, subsequent bottlenecks, effective population size, etc. Pending further empirical data, I think it is safe to say that the correlation between degree of divergence in, say, mtDNA sequences and "speciation" remains unknown at best, and that use of comparative % sequence divergence data provides little crucial information, despite its quantitative and 'modern' appeal. Part of that appeal is due to misinterpretation of the term "genetic distance", which implies to many that we've measure some genome-wide difference when in fact, the typical "genetic distance" involves differences in 1-2 mitochondrial genes, not the genome."

Comments from Zimmer: "YES. I have been delinquent in voting or commenting on this proposal until now, and although the proposal has already been rejected, I would like to weigh in. The debate surrounding this proposal has expanded to call into question the rank of the other taxa in the complex, and, as much as anything, I'd like to comment on this. The conflicting views expressed by others have made some excellent points. In particular, I find myself agreeing with much of what Van had to say about the use of genetic differences in determining species limits among allopatric populations, and about the general misinterpretation of the term "genetic distance". I also think that Daniel and Tom have made some valid points regarding our lack of understanding of the relative importance of calls versus songs as to how they function in mate choice.

"Let me just summarize my understanding of the Brazilian Scytalopus situation as it applies to this present debate. Historically, the Brazilian taxa have been considered to break down into two distinct groups: 1) indigoticus/psychopompus, with distinctly white ventral regions, rufous flanks, blue-gray upperparts and similar songs (a distinctly "froggy" trill) that are completely different from any vocalizations found among taxa in the other group (and which I won't further consider in the present discussion); and 2) the so-called speluncae group, comprised of several basically gray taxa whose songs consist largely of a lengthy, repetitious series of similar notes. More recent workers (e.g. Maurício 2005, Bornschein et al. 2007) have demonstrated that the speluncae group itself breaks down into two distinct subgroups: 1) consisting of three taxa (speluncae [which consists of a larger northern form with faster-paced songs, and a smaller southern form with slower-paced songs, the two of which, arguably, represent distinct taxa], iraiensis, and an-as-yet-unnamed taxon from the Serra da Ouricana, Serra das Lontras and surrounding highlands of SE-SC Bahia = species novum #1) in which the adult males are fairly uniformly dark gray and lack rufous-and-black barring on the flanks and vent, and, in which the songs consist of the aforementioned repetitious series of single notes without any differentiated accelerated trill at the end; and 2) a group consisting of four taxa (novacapitalis, pachecoi, diamantinensis, and an-as-yet-unnamed population from the southern part of the Espinhaço Range in Minas Gerais = species novum #2) that are generally paler gray than the speluncae/iraiensis subgroup (particularly on the median underparts) with rufous flanks/vent that are more or less barred, even in adult males, and, in which exists a distinctive song type with a differentiated ending that accelerates into a trill (this song type possibly given only or mostly by females of the various taxa, but which nonetheless is completely unknown from the speluncae/iraiensis subgroup).

"[TIME OUT: The foregoing discussion does not reflect an-as-yet-unresolved nomenclatural controversy within this group. Raposo et al. (2006) consider that the name speluncae has been historically misapplied to the dark gray birds of the Serra do Mar. Their examination of photographs of the holotype of speluncae (as well as of the type description and accompanying plate, in conjunction with the peculiar type locality (which is outside the current known range of the dark gray coastal-range birds) have led them to conclude that the holotype of speluncae is actually one of the paler-breasted, rufous-flanked birds of the southern Espinhaço range (treated by Maurício 2005, Bornschein et al. 2007 and other authors as "Scytalopus sp. novum"), and that therefore, the epithet speluncae is referable only to birds of the Espinhaço range, including birds from the northern part of the range in Bahia, subsequently recognized by Bornschein et al. 2007 as diamantinensis. Concomitantly, Raposo et al. (2006) concluded that the uniformly dark gray populations of the Serra do Mar (= speluncae by all other authors) lacked a name, and redescribed that population as Scytalopus notorius. These conclusions have been refuted by Bornschein et al. (2007), and the debate rages on. In the interest of not making this discussion any more complicated than it already is, I am not going to deal with this nomenclatural issue further, and will continue to use the nomenclature advocated by Bornschein et al. (2007), which coincides with current SACC taxonomy.]

"The immediate issue at hand is whether or not to treat the newly described (Bornschein et al. 2007) diamantinensis as a distinct species (as advocated by the authors), or as a subspecies of speluncae. The proposed new species (diamantinensis) is diagnosable from all other Brazilian taxa by some combination of morphological and/or vocal characters, as well as by genetic characters. Bornschein et al. (2007), in a phylogenetic analysis in which all known species of Brazilian Scytalopus were represented, consistently recovered (with high statistical support) a monophyletic group formed by diamantinensis, novacapitalis, pachecoi, and Scytalopus sp. nov. #2 (southern Espinhaço Range in Minas Gerais). Within this clade, diamantinensis was distinct from other taxa, but closest to novacapitalis (observed sequence divergence of 3.5%), followed by species novum (4.5%) and pachecoi (5.0%). Trouble arises when one considers the morphological and vocal distinctions. Although readily distinguishable morphologically from both novacapitalis and sp. novum #2 (both of which are distinctly paler on the median underparts), diamantinensis is indistinguishable from pachecoi in both its adult male and adult female plumages. There are apparently some diagnostic mensural differences (several characters for females; only bill depth and culmen length for males). Immature males of the two taxa are separable on the basis of differing barring patterns on the upper wing coverts. Vocal differences are also far from clear-cut. Diamantinensis has 2 distinct calls that are not found in any other taxa in the larger speluncae group, but the songs, while differing diagnosably (no overlap) in two characters from songs of novacapitalis, are apparently not diagnosably different from songs of either pachecoi or sp. novum #2 (Songs of diamantinensis do differ on average in several characters from songs of pachecoi and species novum #2, but there is overlap in all characters.)

"SACC debate to this point has generally centered on whether diamantinensis meets the criteria of differentiation established for biological species of Andean-distributed Scytalopus by the seminal Krabbe & Schulenberg (1997) paper, the authors of which recognized taxa as distinct at the species level only when their songs differed diagnosably. Advocates for using the Krabbe & Schulenberg criteria have voted against species-ranking for diamantinensis on the basis that its songs are not diagnosably different from those of pachecoi or species novum. Note that in the original proposal, Fernando states: The authors believe that S. diamantinensis, S. pachecoi, and Scytalopus sp. nov. would probably be lumped into a single polytypic biological species. Therefore, they recognized all four allopatric taxa with accelerating songs as distinct (at least) phylogenetic species.

"I think it is important to clarify that the authors did not state that they did not believe that the taxa involved did not represent biological species, but, rather, that they would not be accepted as biological species if held to the same criteria advocated by Krabbe & Schulenberg (1997).

"Based on the foregoing question, Van raised the logical next question, which goes beyond species status for diamantinensis, and questions whether the other Brazilian taxa under discussion merit species status or whether they should just be recognized at the subspecific level. Something that has not been made clear in the various discussions is that while many of these taxa are allopatrically distributed, some of the taxa occur syntopically at multiple sites. Maurício (2005), in his formal description of pachecoi, listed and mapped multiple localities in the Planalto Meridional of northeastern Rio Grande do Sul (primarily in the vicinity of Cambará do Sul) and in adjacent Santa Catarina in which both pachecoi and the southern form of speluncae occurred syntopically. He reported strong interspecific segregation in these areas, with speluncae being generally confined to steep slopes, surrounded by territories of pachecoi (in at least one case, pairs of the two forms were only 10-20 m apart). Subsequently, scattered territories of iraiensis have also been discovered in marshes within the highlands of the region, with the result that in the Cambará do Sul/Itaimbezinho & Aparados da Serra region, pachecoi, speluncae, and iraiensis can all be found within a few kms of one another (Indeed, we have seen and tape recorded all three taxa within a few kms of one another in this region.). Within these three sympatric taxa, adults (but not immatures or subadults) of pachecoi are distinctive in having paler gray underparts with consistently rufous & black-barred flanks/vent, and in having the accelerated song type with a terminal trill, as well as in having some diagnostic calls. The taxa speluncae and iraiensis differ more subtly from one another in morphological characters, and their songs are very similar, but they do differ diagnosably in their calls (iraiensis has a monosyllabic or bisyllabic call that Maurício termed "very distinct" from calls of all other members of the speluncae group). All three sympatrically occurring taxa (pachecoi, speluncae, iraiensis) exhibit ecological differences in terms of microhabitats occupied.

"Given that pachecoi, speluncae and iraiensis pass the test of sympatry, it would seem that their treatment as separate biological species is secure. The molecular analysis of Bornschein et al. (2007) establishes the monophyly of pachecoi, novacapitalis, diamantinensis, and species novum #2 as a clade distinct from speluncae, iraiensis and species novum # 1. So the issue then becomes whether pachecoi, diamantinensis, and sp. novum #2 should be considered separate species, or merely subspecies of novacapitalis, which has priority. As reiterated above, diamantinensis differs diagnosably (no overlap) in two characters of its song from novacapitalis (along with possessing 2 diagnostic call types and differing in plumage characters), and so, would seem to fit the Krabbe/Schulenberg criteria for recognition as separate species. Songs of pachecoi and novacapitalis also seem to differ diagnosably from one another in pace, frequency and note length (Table 2, Maurício 2005); pachecoi has a unique call not found in novacapitalis; and the two forms differ in plumage characters. So, again, these two taxa would seem to meet the Krabbe/Schulenberg criteria for recognition as separate species.

"The remaining pairwise comparisons to consider are diamantinensis versus pachecoi, and each of those forms relative to species novum #2. To reiterate, diamantinensis differs diagnosably in plumage characters from species novum #2, but is indistinguishable from pachecoi (except for immature males, which can be diagnosed on plumage characters). Since species novum #2 hasn't been formally described (except by Raposo et al. 2006), the SACC does not recognize it at any level, so it would seem that for the time being it is a moot point as to whether or not it is distinct from diamantinensis [The known ranges of the two forms are separated by more than 500 km; they occupy different habitats, with diamantinensis being strictly a forest bird and species novum #2 being found in both forest and non-forest habitats, but more often in campo rupestre; they differ diagnosably in plumage characters; they differ diagnosably in at least one call type; and the observed sequence divergence in mitochondrial DNA was 4.5%] or pachecoi. The diagnosability of diamantinensis relative to pachecoi rests on non-overlapping differences in wing covert pattern of immature males, diagnostic differences in calls, diagnostic differences in pace of the transitional segment of the accelerating song type, and in the 5.0% mitochondrial DNA sequence divergence (Bornschein et al. 2007).

"I would submit that the sum of the distinctions between diamantinensis, pachecoi and species novum #2, although perhaps not meeting the Krabbe/Schulenberg criteria for biological species, are, in fact, consistent with separate species status. They are as different from one another (differing ecologically, morphologically, and in calls, but with overlapping song characters) as are speluncae and iraiensis, two taxa in the same genus that despite subtle morphological differences and overlapping song parameters, exist sympatrically while maintaining their genetic integrity. The yardstick supplied by the speluncae/iraiensis example would seem to answer the question of whether diagnostic differences in calls can be as important as diagnostic differences in songs when it comes to mate choice and pre-mating barriers to reproduction. As pointed out by Daniel, differences in calls have been shown to be more important than songs in sorting out the species-limits in the Hypocnemis cantator group of antbirds. Isler et al. showed that the taxa peruviana and collinsi/subflava, although not differing diagnosably in song characters, did have distinctly different call types, and mated assortatively in areas of syntopy. Conversely, within the Schistocichla group of antbirds, calls did not vary appreciably between taxa, whereas songs did. I realize that we are talking about two different genera, but the point remains the same - different vocal characters may have different relative importance from one group of birds to the next.

"As Van points out, the sequence divergences reported by Bornschein et al. 2007 do not prove anything in and of themselves, but they are least consistent. Given the geographic distances involved (500+ km) between the nearest known inter-taxon populations in the diamantinensis/novacapitalis/pachecoi/species novum #2 clade, combined with the sedentary nature of Scytalopus in general, it would seem pretty clear that these taxa are on independent evolutionary trajectories. The only taxon in the greater speluncae/novacapitalis group that I don't have personal field experience with (including the two undescribed forms) is diamantinensis, so I can't speak for how different it is from "species novum #2", but the remaining members of the group all seem like pretty different critters to me, at least relatively speaking (given the general conservative nature of plumage and vocal differences within the genus as a whole). So, my vote would be YES on Proposal #329, and I would urge "no" voters to reconsider. I would also vote against any proposal to demote any of the Brazilian Scytalopus currently recognized in our list as distinct species to subspecies status."

Additional comments solicited from Niels Krabbe: "Kevin makes a fine summary of the Brazilian forms of Scytalopus (and points out that the authors of diamantinensis considered it a phylogenetic, but not a biological species).

"To sum up: The core issue is whether sp.nov.#2 (Espinhao range) and diamantinensis are the same biological species despite diamantinensis being closer related to novacapitalis than to sp.nov.#2. Following the song criterion, they are.

"Kevin brings forward two new arguments: habitat and geographical distance.

"As for habitat, I cannot help but think of Scytalopus latrans. Eastern and western forms in Ecuador and Colombia occupy distinct habitats (western birds semihumid to humid, including heavily disturbed, forest in a broad elevational range, eastern birds humid-wet, more pristine forest in a narrow elevational belt. The two have rather different songs; their calls vary a lot, but can sound fairly similar. They would qualify for being two species, were it not for the form subcinereus in southwestern Ecuador and northwestern Peru. Subcinereus occupies an exceptional wide range of habitats, from humid to dry forest, and appears to grade into both the eastern and western forms. Genetically, birds from the core area of subcinereus are closest to the eastern form.

"500 km may sound like a lot of geographical distance between diamantinensis and sp.nov."2, but both are in the same mountain range (Espinhao), so it might not take a big climatic change to bring them into contact again.

"So my vote remains NO."

Additional comments by Remsen: "YES, based mainly on Kevin's comments."

Additional comments by Robbins: "YES, based on Kevin's comments."