Proposal (868) to South American Classification Committee

Treat Lepidocolaptes layardi as a subspecies of L. fuscicapillus

Effect on South American CL: If adopted, this proposal would lump two taxa of the genus Lepidocolaptes presently considered species and would thus recognize the polytypic species L. fuscicapillus (with nominate and layardi subspecies).

Background: Proposal 620 (2013) proposed the elevation to species rank of all taxa included in the species previously named Lineated Woodcreeper (Lepidocolaptes albolineatus). This proposal was adopted in 2014 and has subsequently been taken over in several worldwide taxonomic lists. (For the background until that date, see the proposal.)

New information:

1. Proposal 620 states in the background info: “taxa fuscicapillus and layardi being very distinct vocally, hence suggesting that the polytypic L. albolineatus may include more than a single species (Marantz et al. 2003).” However, what is described in this reference for fuscicapillus is in fact the voice of the (then undescribed) fatimalimae. On the contrary, they describe the voice of madeirae (=present fuscicapillus) and layardi as being the same.

2. Proposal 620 states in the new information: “Vocally, these five molecular clades/taxa have also proved to be very distinct, further supporting their treatment as independent species (Rodrigues et al. 2013).” However, the ‘proof’ given in Rodrigues et al. (2013) for vocal difference between fuscicapillus and layardi is nowhere to be found in the text (this publication is in fact about the description of the new taxon fatimalimae, and only the vocal difference of this taxon vs. all others is detailed. There is, however, a single picture of ‘representative sonograms’ that suggests a difference in voice between fuscicapillus and layardi (see also further) and a Table 1 in Supplementary Information. I am not sure if this supplementary information is still available on-line somewhere, for which I copy it here:

From this Table, it is clear that the measured sound parameters show no significant differences for the taxa fuscicapillus and layardi (at most a slight difference in song duration with considerable overlap).

3. Boesman (2016) made a brief analysis of the voice of the Lineated Woodcreeper complex. As no clear vocal differences could be found between fuscicapillus and layardi, the specific sonogram depicted in Rodrigues et al. (2013) was traced back to the original recording. When depicting with appropriate amplitude levels, the note shape appeared to be quite different and more in line with other sonograms of both fuscicapillus and layardi (with note shapes slightly dependent on excitement level of the bird).

4. Independently, Minns (2016) also catalogued all existing xeno-canto recordings. He states: “The song of layardi is similar to that of its neighbour fuscicapillus and I was not able to readily distinguish between them.”

From the above, I believe there is at present no indication at all that voice of fuscicapillus and layardi shows anything more than some minor differences.

Rodrigues et al. (2013) also supplied genetic information (albeit, similar to the vocal information, also in a very condensed way without providing much detail). From the figure 1 it appears that fuscicapillus and layardi are the two (sister) taxa which have the smallest genetic divergence.

Analysis and Recommendation: We have thus here two taxa which at most show some minor morphological and vocal differences, and which have a considerable mtDNA divergence, which, however, is the smallest compared to other members of the complex (the latter additionally DO show significant vocal differences among them as confirmed in Boesman (2016) and Minns (2016) !).

From the comments of the SACC members on proposal 620, I deduce that the vocal difference was the main argument to split Lepidocolaptes albolineatus into four species. With the new information presented here, this vocal difference is no longer present for the taxa fuscicapillus and layardi.

Whether the genetic divergence in itself is sufficient argument to maintain both taxa as full species obviously depends on which criteria are used. Accepting BSC species level purely on a moderate mtDNA divergence for two populations separated by a geographical barrier seems at least rather adventurous and may open the discussion what to do with many other similar cases.

It is clear that the Tapajos (and Tele Pires) river is an important geographical barrier, which has led to isolation and speciation in quite some cases. This in itself however can’t be an argument. After all, Duida Woodcreeper L. duidae also occurs on both sides of the Rio Negro, an equally important geographical barrier while at the other hand, it is only separated from L. albolineatus by the ‘small’ Rio Branco.

Therefore, I propose to reconsider the decision taken in 2013, would recommend lumping the two taxa and recognize the polytypic species L. fuscicapillus (with nominate and layardi subspecies).

As for the English name, del Hoyo and Collar (2016) lump both taxa as a single species and have given it the name Dusky-capped Woodcreeper (reflecting its scientific name).

Literature cited

Boesman, P. (2016). Notes on the vocalizations of Lineated Woodcreeper (Lepidocolaptes albolineatus). HBW Alive Ornithological Note 84. In: Handbook of the Birds of the World Alive. Lynx Edicions, Barcelona. https://static.birdsoftheworld.org/on84_lineated_woodcreeper.pdf DOI: 10.2173/bow-on.100084

del Hoyo, J., Collar, N.J. (2016). HBW and BirdLife International illustrated checklist of the birds of the world. Volume 2: Passerines. Lynx Edicions, Barcelona.

Rodrigues, E. B., Aleixo, A., Whittaker, A., and Naka, L. N. (2013). Molecular systematics and taxonomic revision of the Lineated Woodcreeper complex (Lepidocolaptes albolineatus: Dendrocolaptidae), with description of a new species from southwestern Amazonia. In J. del Hoyo, A. Elliott, J. Sargatal & D. Christie (Eds.), Handbook of the Birds of the World. Special Volume: New Species and Global Index, pp.248-252. Lynx Edicions, Barcelona.

Marantz, C., A. Aleixo, L. R. Bevier and M. A. Patten (2003). Family Dendrocolaptidae (Woodcreepers). Pp. 358-447 in: del Hoyo, J., A. Elliott, and D. A. Christie (eds.) (2003). Handbook of the Birds of the World. Volume 8: Broadbills to Tapaculos. Lynx Edicions, Barcelona.

Minns, J. (2016) Delineating the Lepidocolaptes albolineatus complex. https://www.xeno-canto.org/article/198

Peter Boesman, July 2020.

Comments from Remsen: “YES (depending on comments by others). The evidence above convinces me that the evidence for the split was over-stated and not independently analyzed by us as thoroughly as it should have been. With respect to genetic differences, when restricted to a few neutral loci from a few individuals within taxa of a monophyletic group, in my viewpoint they are irrelevant to taxon rank.”

Comments from Whitney: “I appreciate with this well-explained argument of Peter’s, and certainly concur that layardi is minimally differentiated (if at all) from fuscicapillus. This case is similar to those of other "HBW woodcreeper" cases (Campylorhamphus, Dendrocolaptes) in which sample sizes of data used in diagnostic analyses, phenotypic and genetic, were inadequate (regardless of the conclusions drawn from them).”

Comments from Claramunt: “YES. The previous splitting of this species was done without critical evidence.”

Comments from Robbins: “YES, given Peter Boesman's assessment.”

Comments from Stiles: “YES. Strong vocal similarity and weak genetic separation without evidence of reciprocal monophyly combine to make its species status unsustainable.”

Comments from Areta: “YES. The Lepidocolaptes seem to be quite geographically variable in plumage, but these differences are not paralleled by differences in vocalizations. This provides an interesting comparison to the case of Lepidocolaptes squamatus/falcinellus in relation to the level of isolation and the roles of rivers in promoting/maintaining differentiation.

Comments solicited from Jorge Pérez-Éman:

“This proposal aims to lump Lepidocolaptes fuscicapillus and L. layardi into a single polytypic species (L. fuscicapillus), partially reverting the decision to split L. albolineatus into five species, including recently described L. fatimalimae (Rodrigues et al. (2013), Proposal 620). In this proposal, Boesman argues that data to consider layardi different from fuscicapillus is unavailable, and makes the following points:

“1. Early statements about the large differences between fuscicapillus and layardi vocalizations were influenced by the known distribution of both taxa by that time (Marantz et al. 2003). The currently known distribution for each taxon in the entire L. albolineatus complex is now better understood by the work of Rodrigues et al. (2013) and, as pointed out by Boesman, such differences referred to fatimalimae (and not fuscicapillus) and layardi.

“2. Rodrigues et al. (2013) provided a molecular phylogeny and stated that the five monophyletic lineages obtained from this hypothesis proved to be very distinct vocally. However, Boesman rightly points out that Rodrigues et al. (2013) focused only on the differences between fatimalimae and the other taxa, and no detailed analyses were done on the other four taxa (albolineatus, duidae, fuscicapillus and layardi). He goes further and indicates that no clear vocal differences were found between fuscicapillus and layardi, and back this up with some general and qualitative comparative vocal analyses done within this complex (Boesman 2016, Minns 2016).

“I totally agree with Boesman in relation to these general comments. Rodrigues et al. (2013) focused on the description of fatimalimae and consequently, a proposal to split L. albolineatus into five taxa was not really well supported by their data and analyses. I think the authors did a great job to better understand the diversity and distribution of taxa within this species complex, but I don´t know if by omission or by constraints provided by HBW publishers, there is not enough information to evaluate the validity of such split. As early reviewers of Proposal 620 indicated, I was really frustrated trying to find details associated to detailed information on skin vouchers, recordings and tissues used in the study. First, the map included in the article is not clear and details are impossible to see and, second, I am still looking for the Supplementary Information (besides the Table S1 provided by Boesman). This information should be readily available to readers/reviewers and it is particularly relevant when taxonomic changes are proposed. In particular, I missed a more complete treatment of the morphological variation (although some of the Zimmer’s [1934] concerns on morphological variation were partly solved by description of fatimalimae), including patterns in contact zones among different taxa. Same goes for vocal and molecular variation. For example, available information for both morphological/molecular (Rodrigues et al. 2013) and vocal data (Minns 2016) suggest sympatry in Bolivia, but hybridization has not yet been assessed. A more thorough specimen evaluation in this region would have provided with a stronger support to some of their proposed splits by confirming sympatry without hybridization. Similarly, what are the patterns on both sides of the Madeira, Tapajos or Teles Pires rivers? These areas have been known to show complex patterns of phenotypic/molecular variation in other taxa and clear definition of species/taxon limits should take this into account.

“Boesman´s criticism on vocalizations analyses done by Rodrigues et al.(2013) are worth to expand here. The description of L. fatimalimae vocalizations and their differences to all other taxa within the L. albolineatus complex are very detailed and complete. However, there is no similar analysis done with other taxa vocalizations and the only results shown are included both in Table S1 and a figure with spectrograms of songs for each taxon. I am providing plots for each of the vocal characters included in Table S1 to better visualize patterns (Figure 1).

“Figure 1. Vocal characters used to compare taxa within the Lepidocolaptes albolineatus species complex (Rodrigues et al. 2013). Number of notes, song length (seconds) and pace (number of notes per second) data are extracted from Table 1 included in the Supplementary Information of the HBW Special Volume. Dots represent means and error bars one standard deviation. Taxa abbreviations: fat = fatimalimae, alb = albolineatus, dui = duidae, fusc = fuscicapillus, lay = layardi).

“As you can see, and clearly pointed out by Boesman, these data show no differences between fatimalimae and albolineatus, or between fuscicapillus and layardi vocalizations for any of the three variables. It is also interesting that the first three taxa show a larger variation in these characters than the last two and, given this variation, duidae slightly overlaps with both fuscicapillus and layardi. In fact, songs of these three taxa were found to be similar by Marantz et al. (2003) describing them as “slowing, descending series of clear whistles” differing mainly in number of notes and pace. Moreover, characters such as note shape, frequency, inter-note intervals and the acceleration/deceleration pattern, useful in the comparisons with fatimalimae, were not used by Rodrigues et al. (2013) in these other taxa but certainly should prove useful. Thus, the authors only “proof” for differences are the “representative” spectrograms included in their Figure 3. I use quotation marks in representative because we are just forced to believe or to doubt if those spectrograms are really typical or not (as no data on variation is included in the article). In fact, spectrograms provided for fuscicapillus and layardi are clearly different but, as discussed for Boesman (2016), such differences depend on how they are presented. As Boesman, I tracked both recordings in Macauley Library and they clearly are not as different as presented in Rodrigues et al. (2013) (Figure 2).

Figure 2. Spectrograms of songs of L. fuscicapillus (left) and L. layardi (right).

“Song by fuscicapillus shown in Figure 2 (ML51900) is the same depicted in Figure 3 of Rodrigues et al. (2013). It is not just different but it is also more representative of this taxon songs (which could be checked reviewing recordings available both in Macaulay Library and XenoCanto). On the other hand, I am showing a recording of layardi that is not the same as shown in Rodrigues et al. (2013) as the one shown by the authors is from a bird on breeding condition (courtship, display or copulation) based on information provided by the recordist (Curtis Marantz), and such condition might exaggerate some of the note shapes. When both song spectrograms are visually compared it is clear they are not as different as portrayed in Rodrigues et al. (2013).

“Consequently, based on morphological and vocal information, Rodrigues et al. (2013) do not provide enough information to make any evaluation of the status of the complete L. albolineatus complex and, more specifically for this proposal, of the separation of fuscicapillus and layardi. Boesman (2016) and Minns (2016) qualitative comments are not complete either. Besides statements/descriptions suggesting that vocalizations of fuscicapillus and layardi are not clearly distinguishable, no formal vocal analysis was included. Major comparisons are done focusing mostly on albolineatus and fatimalimae, but I am missing an analysis among all taxa, and quantitative and qualitative comparisons among duidae, fuscicapillus and layardi are needed. Such comparisons should have a clear criteria to define what we consider a species under the BSC used by this committee. Vocalization thresholds are not fixed and could vary depending on the species groups compared. Additionally, some taxa might not differ in their loudsongs but their calls might be important to keep reproductive isolation.

“The last point I would like to mention refer to the phylogenetic hypothesis and its interpretation. In this proposal, Boesman states “it appears that fuscicapillus and layardi are the two (sister) taxa which have the smallest genetic divergence”. Similarly, Rodrigues et al. (2013) indicated in their Figure 1 “numbers refer to posterior probabilities values and genetic distances between sister groups associated with the labeled nodes”. It is important to understand that the phylogenetic hypothesis included in Rodrigues et al. (2013) do not show any sister relationships among fatimalimae, duidae, fuscicapillus and layardi. It only shows albolineatus as potential sister taxon to a unresolved group (polytomy) formed by these four taxa. Additionally, regardless of the relevance (or lack of) of genetic divergence for taxonomic decisions, fuscicapillus and layardi do not show the smallest genetic divergence, as such divergence is the same among all taxa included in the polytomy, as clearly indicated in Figure 1 of Rodrigues et al. (2013). The lack of evidence for sister taxon relationship between fuscicapillus and layardi is a clear limitation to lump these taxa into a single species. What if fuscicapillus, for example, turned out to be closer related to duidae? Would we be asking the same questions?

“As molecular variation within lineages was not shown by Rodrigues et al. (2013), I downloaded from GenBank all sequences used in this study plus the ones used by Arbelaez-Cortes et al. (2012, Zoologica Scripta) for their study on Lepidocolaptes molecular systematics, to run a Maximum Likelihood phylogenetic analysis (Figure 3). The phylogenetic hypothesis for the genus Lepidocolaptes, using same gene and sequences (plus the ones for the rest of species), clearly suggests that the story is more complex. L. albolineatus might be related to other taxa within the genus though there is a total lack of resolution at the base of the tree with few robust relationships among taxa. Each lineage, however, is strongly supported, and layardi shows a well-supported structure with two groups diverging in 2% for that gene (ND2). Unfortunately, I don´t have information about localities to explore if there is any geographical signal in such pattern. Thus, not only the evolutionary history of this group is more complex than we thought but also the lack of strong support for any phylogenetic relationships suggests caution before any taxonomic rearrangement is proposed, both for the L. albolineatus species complex and, particularly, regarding the taxonomic status of both fuscicapillus and layardi.

Figure 3. Phylogenetic hypotheses of the genus Lepidocolaptes based on the ND2 gene (left) and Ultraconserved Elements (UCEs) (Harvey et al. in review; right). The ND2 hypothesis shows in color each of the five currently recognized taxa for the L. albolineatus species complex. Both hypotheses show bootstrap support values based on Maximum Likelihood analyses.

“Additional but still unpublished evidence is provided by upcoming suboscine phylogeny by Harvey et al. (in review) (kindly provided by Mike Harvey). A Maximum Likelihood analysis based on genomic sequences (UCEs) resulted in a well-resolved phylogeny showing a polyphyletic L. albolineatus species complex and suggesting that fuscicapillus and layardi are sister taxa (Figure 3). This phylogenetic hypothesis is, by the way, congruent with the previous two-species taxonomy of this group (albolineatus/fuscicapillus; Cory & Hellmayr 1925, Catalogue of Birds of the Americas). Though genomic data are not immune to analytical problems, and this phylogeny only includes one individual/taxon, at least we have some evidence that fuscicapillus and layardi might be sister taxa. I say “might” because until duidae is not included we are missing an essential part of the picture.

“In summary, splitting or lumping fuscicapillus and layardi into two or just a single species requires evidence that has not been available to review. A thorough morphological and vocal analyses should be made available considering the geographical distribution of each taxa, contact zones (if any), and the potential geographical variation suggested by the ND2 gene hypothesis (for layardi). Evidence to split the complete species complex was not thorough or available and, definitely, neither morphological nor vocal evidence were available to consider these two taxa as different species. Lumping them back into a single species also requires evidence. Even when more thorough analyses might eventually show they do not differ either morphologically or vocally, there is no available information indicating fuscicapillus and layardi are sister taxa (until duidae is included in phylogenetic analysis). Additionally, geographical distribution of layardi is congruent with many other Amazonian taxa and a large genetic divergence (even for just one mitochondrial gene) suggest some sort of isolation, a pattern that needs to be further explored (looking for potential causes of the lack of congruence between phenotypic/molecular characters). Unfortunately, there is no basis to keep the current treatment (i.e., different species) and lumping them appears to be the available practical solution for the time being (meaning that a change in taxonomic status is not based on demonstration they belong to one species or that each lineage does not represent a single species).”

Comments solicited from Mario Cohn-Haft: “I don’t really have much of substance to add on this one based on my personal experience. Basically, I believe that the songs are virtually indistinguishable between layardi and fuscicapillus. But as I reread the Rodrigues et al. paper and other evidence called to bear, I am a bit mystified by: 

“1) the rather different looking sonograms between the two in the Rodrigues et al. HBW paper— I’m curious (but not enough to look them up) to hear those particular cuts that are graphed. The fuscicapillus is pictured very faintly even in the strongest parts of the song, so it’s actually conceivable that the higher peak in layardi notes are actually made in that cut too, but were filtered out of the figure! And even if not, not sure how different that would sound. In any case, most cuts of both spp. sound pretty much identical to me, and in my days of greater activity in se Amazonia, I never made a distinction between e and w of the Tapajós birds, but was acutely aware that the 4 quadrants of the Amazon (formed by Negro, Solimões/Amazon, Madeira) each had different sounding albolineatus types— i.e., the 4 taxa that everyone’s comfortable with.

“2) I don’t understand the phrase "the fuscicapillus group is vocally heterogeneous, with taxa fuscicapillus and layardi being very distinct vocally, hence suggesting that the polytypic L. albolineatus may include more than a single species” in the intro of Rodrigues et al. Is that supposed to mean that the two differ from one another, or that they (as a unit) differ from the rest? It seems like an odd statement to put in the intro as is, and yet seems to have been picked up in the SACC deliberations repeatedly as if it were a conclusion.

“It’s not all that unusual for allospecies to have similar songs, but to differ in calls or in overall repertoire or in the use of the same repertoire in different behavioral contexts, so having the same song doesn’t put me off species distinctiveness automatically. But it definitely puts up a red flag.

“The genetic difference, however, seems convincing to me. Yes they’re the most recent split. but the data, as presented, indicate two nice reciprocally monophyletic groups with a substantial percent divergence. That suggests they really are not mixing and that to me is pretty strong evidence of species status. But similar cases have proved with subsequent genomic analyses to have lots of gene flow, and reinterpretation has been necessary. So, I’d feel a lot more comfortable seeing genomic analyses.

“Sampling is obviously an important issue. In the pdf of the article, I can’t make out from the map which localities were sequenced. I’d sure like to see more samples and from more strategic localities. Even not being able to distinguish the different kinds of sample points in the figure, I can see that the potential contact zone was not well sampled, which seems like an important lacuna for concluding anything.”

“Finally, even if the species hold up as distinct, it’s not clear whether the names are right. The type locality of fuscicapillus, if I’m reading the map right, is from near to where a contact zone should be expected. Interestingly, it's also not far from localities with both (!) fatimalimae and fuscicapillus. That sympatry in itself is interesting (as mentioned by one of the reviewers of the proposal) and could point either to sympatry strengthening the species status of the former, or to a zone where potentially all 3 taxa come into contact and could mix, and where modern analyses should be focused.

“In all, rejecting species status for layardi is basically a conservative move and seems reasonable to me. Not because there’s strong evidence that it’s not a species, but because there is still a lack of strong evidence that it is.”

Comments solicited from Jason Weir: "I have not analyzed the contact zone between Lepidocolaptes [fuscicapillus] layardi and L. [f.] fuscicapillus, so I can only offer an opinion based on other contact zones we have analyzed for woodcreepers in this region. Subspecies of Dendrocincla fuliginosa, Glyphorynchus, as well as Xiphorhynchus elegans/spixii that come into contact in this headwater region (Rio Teles Pires) all have very narrow hybrid zones. I consider them all to be excellent biological species on the basis of intrinsic postzygotic isolation rather than premating isolation which, based on vocal divergence is presumably lacking (Dendrocincla fuliginosa) or weak to moderate (Xiphorhynchus elegans/spixii), and does not prevent extensive interbreeding at the contact zone. Each of these species pairs is > 5% GTR-gamma distance in cytochrome b. In contrast, genomic data I generated for a younger pair (<4% GTR-gamma distance in cytochrome B) of woodcreeper subspecies from the Xiphorhynchus guttatoides group demonstrates a very broad hybrid zone here and while they would be considered a phylogenetic species and have been considered species in the past, it certainly is not reproductively isolated. Contact zones are unlikely to occur for all Amazonian taxa and studying those that do exists takes considerable effort, thus I think it is prudent to extrapolate evidence we have for the timeline to postzygotic reproductive isolation observed in currently published hybrid zones to taxa like L. [fuscicapillus] layardi. Given my hybrid zone data, I suspect that even in the absence of vocal differentiation, I would recognize woodcreeper taxa in this region that are > 5% diverged as likely possessing strong intrinsic postzygotic isolation and I would have no hesitation defining them as likely representing biological species on this basis. Less than 5% divergence would require more careful consideration and in the absence of strong vocal differentiation I would favor taxonomic lumping."

Comments from Pacheco: “YES. I agree that the split of these two taxa was a misinterpretation and that the available vocal data points to a very close relationship.”

Comments from Bonaccorso: “YES. The evidence supports lumping these two taxa into one.”

Comments from Zimmer: “YES, but with reservations about how to interpret the genetic data and phylogenetic hypothesis, given that, as correctly pointed out by Jorge, the paper by Rodrigues et al. (2013) does not resolve the issue of sister relationships among fatimalimae, duidae, fuscicapillus and layardi. The evidence presented does not unambiguously make the case for splitting fuscicapillus and layardi, but neither does it prove that they should be lumped. We are left in the position of not knowing what we don’t know. I think the interpretation that Rodrigues et al. (2013) were focused primarily on documenting the distinctiveness of fatimalimae is correct. Remember that there was confusion in the literature over the subspecific identity of populations from west of the Madeira – these were attributed by Peters to fuscicapillus, despite the fact that fuscicapillus was described in 1868 from Mato Grosso, east of the Madeira. The population from the Madeira-Tapajós interfluve, was described in 1919 from Porto Velho, on the east bank of the Madeira in Rondônia, as subspecies madeirae. Andy Whittaker hit on the fact that birds from the west bank of the Madeira in Brazil sounded the same as birds on the rio Javari (Brazil-Peru border) [all purported to be fuscicapillus according to Peters] but distinctly different from birds from the east bank of the Madeira in Mato Grosso and Rondônia, the respective type localities for fuscicapillus and “madeirae”. Songs of birds from those type localities sounded identical, and it didn’t make sense from the standpoint of biogeography that there were two identical sounding subspecies occurring on the east bank of the Madeira in Mato Grosso and Rondônia. Marantz, et al. (2003 in HBW Volume 8), had apparently reached that conclusion, and restricted the range of fuscicapillus to west of the Madeira, seemingly overlooking that the type specimen was from the east bank. All of this led Whittaker and his eventual co-authors to conclude that “madeirae” was a junior synonym of fuscicapillus, that it occurred only east of the Madeira, and that the form with a different voice that was widespread west of the Madeira, lacked a formal name, and that is the taxon novum which Rodrigues et al. (2013) described as fatimalimae. I would agree with Mario Cohn-Haft that the Lepidocolaptes occurring in the four quadrants of Amazonia: 1) albolineatus in the NE; 2) duidae in the NW; 3) fatimalimae in the SW; and 4) fuscicapillus + layardi in the SE all sound distinctly different from one another (although there was no adequate analysis in Rodrigues et al. [2013] to demonstrate this), but that there is no real distinction between the voices of birds on either side of the Tapajós (fuscicapillus to the west and layardi to the east). Given that, I’m fine with taking a conservative approach and lumping layardi into fuscicapillus (which has priority), despite my aforementioned reservations regarding the genetic data and associated phylogenetic hypotheses.”

Comments by Lane: “YES. I think Peter has laid out a strong case for lumping L. layardi into L. fuscicapillus.”