Proposal (814) to South American Classification Committee

Recognize Turdus murinus and T. arthuri as species distinct from T. ignobilis

Turdus murinus and T. arthuri are two dull, grayish-brown thrushes of the Guiana Shield, whose taxonomic status has long been conjectural. T. arthuri was described (as Planesticus arthuri) by Chubb in 1914 from the Abery River of British Guiana; it is a lowland species, widely distributed in white-sand areas of the Guianas, southern Venezuela, extreme northwestern Brazil and eastern Colombia. T. murinus was described by Salvin (1885) from Mt. Roraima and the Merumé Mountains of southern Venezuela and adjacent areas of Brazil and British Guiana; it is a highland species, restricted to the middle and upper elevations of the tepuis and other mountains of this area. Both taxa have usually been considered subspecies of the similarly dull-colored T. ignobilis since Chapman (1917) and Hellmayr (1934), although Chubb (1921) considered at least murinus to represent a distinct species. However, studies of the phylogeny of the Turdus thrushes by Cerqueira et al. (2016) showed that murinus was only very distantly related to ignobilis, such that its inclusion in ignobilis rendered this taxon polyphyletic. Nevertheless, they continued to include murinus as a basal subspecies of ignobilis, based on its more montane distribution, considered to be like that of nominate ignobilis and different from that of T. i. debilis which they separated as a distinct species from ignobilis because of its wide lowland distribution; they included arthuri here because it is also a lowland species (however, they did not have genetic material for nominate ignobilis). Their study was criticized by Avendaño et al. (2017) based upon extensive genetic data from Colombia. They found that the three subspecies of ignobilis (nominate ignobilis, debilis and goodfellowi) formed a compact clade (thus negating the separation of debilis as a distinct species) and reaffirmed the distinctness of murinus, recommending that it be accorded status as a distinct species; this eliminates the polyphyly of ignobilis. They also found that arthuri was part of the ignobilis group, but considerably more distantly related to ignobilis and suggested that it too merited species status, although they did not take this step because they had no evidence for sympatry between arthuri and T. i. debilis. However, a detailed examination of specimens from eastern Colombia by Stiles & Avendaño (2019) definitely established sympatry between arthuri and debilis at two localities, thus justifying the status of arthuri as a separate species. Hence, this proposal is to recognize both Turdus murinus and T. arthuri as species distinct from T. ignobilis. I divide the proposal into two parts, for both of which I recommend a YES:

A1. Recognize Turdus murinus as a distinct species, given its very distinct genetic relationship: it is not sister to the ignobilis group.

A2. An appropriate English name would be Tepui Thrush, given that it is effectively restricted to the middle and upper parts of the tepuis and adjacent mountains.

B1. Recognize Turdus arthuri as a distinct species, justified by sympatry without intergradation at two Colombian localities.

B2. I consider that the name suggested by Cerqueira et al. for this taxon, Campinas Thrush, is appropriate as it describes its habitat over the greater part of its distribution.

Pertinent literature:

Cerqueira, P. V., M. P. Dantas Santos & A. Aleixo. 2016. Phylogeography, inter-specific limits and diversification of Turdus ignobilis. Mol. Phylog. Evol. 97:177-186.

Avendaño, J. E., E. Arbeláez-Cortés & C. D. Cadena. 2017. On the importance of adequate geographic and taxonomic sampling in phylogeography: a reevaluation of diversification in a Neotropical thrush. Mol. Phylog. Evol. 111:87-97.

Stiles, F. G. & J. E. Avendaño. 2019. Distribution and status of Turdus thrushes in white-sand areas of eastern Colombia, with a new subspecies of T. leucomelas. Zootaxa 4567(1): 161-175.

F. Gary Stiles, March 2019

Comments from Areta: “A1. YES, although I would like to have biological data to help me understand what murinus does in life.

“A2. Sounds good. We are now having a ton of Tepui something. I note that Avendaño et al. (2017) proposed Pantepui Thrush.

“B1. YES, but I have some questions and elaborations. I see there is a growing body of evidence suggesting that arthuri is a valid species, but I would feel more comfortable with more data. Maybe I am exaggerating, but I would like to see biological data relevant to mating added to species splits. Also, this is not, I think, just another easy split. Here are my points:

“1) Do any of these thrushes migrate or experience seasonal movements? Do we know? This is relevant when evaluating the meaning of the sympatry of arthuri and debilis; also, when sympatric, do they occur in distinct habitats? At the two sympatric localities reported by Gary and Jorge, is there any evidence of breeding? Gonadal data from specimens? I presume arthuri would have been collected at white-sand areas, while debilis would not at their sympatry localities (I don´t know): if this habitat segregation is maintained in BREEDING sympatry, I would be convinced to recognize arthuri as a species. Also, if I understand correctly, didn´t Cerqueira et al 2016 show sympatry of arthuri (cururensis) with debilis? Or at least they seem to show that the distribution of arthuri (cururensis), in part, embedded within that of debilis? (see their Figure 1, showing pink spots of arthuri (cururuensis) amidst green areas of debilis). If so, this would add another argument in favor of recognizing arthuri as a valid species.

“2) How do all these recently diverged species sing? If, for example, we focus on Turdus albicollis, the species seems to have a remarkable consistency in song across its wide geographic range (at least to my ears, birds from Henri Pittier and PN Yacambú were easily identified as pertaining to the species albicollis, in comparison to subspecies contemptus and paraguayensis from Argentina). I haven´t found recordings of songs of arthuri to compare against ignobilis-debilis-goodfellowi and allies. Maybe those recordings are tossed among recordings attributed to ignobilis et al.? Has anyone knowingly recorded this thrush singing?

“3) Also, I am not sure that the genetic evidence unequivocally indicates that arthuri is a species. Note for example the haplotype network in Avendaño et al. (2017) which shows that maranonicus shares haplotypes with ignobilis and with debilis, whereas the distance between some haplotypes of goodfellowi to debilis-ignobilis-maranonicus is equivalent to the distance between the latter and arthuri. Similarly, the haplotype network in Stiles and Avendaño (2019) shows similar distances between nominate leucomelas to the rest of the recognized subspecies, resembling the distances between arthuri and other members of the ignobilis group. I am trying to think based on which data we have (I of course recognize that in the speciation continuum, some species will have deep divergences whereas” others will not, and I do not advocate using genetic distance as an infalible tool to delimit species). The split of arthuri would keep contributing to the paraphyly of ignobilis (which was already paraphyletic by the inclusion of maranonicus within it). This is quite interesting, as I suppose that nobody thinks that maranonicus is not a distinct species.

“4) Finally, besides the minor differences in plumage described by Stiles & Avendaño (2019) there is apparently a difference in tarsus length between arthuri and at least debilis (but we lack data for other taxa in the complex).”

Comments from Claramunt:

“A1. YES. Not even closely related to ignobilis.

“B1. YES. Even if not overwhelming, the evidence indicates the presence of two different species. Plumage differences match genetic differences, as the mtDNA trees show reciprocal monophyly (Cerqueira et al., or at least no evidence of non-monophyly Avendaño et al.) and coalescent analyses suggests separation of lineages despite inclusion nuclear genes and samples near areas of sympatry or parapatry. If arthuri were not intrinsically isolated from ignobilis, we would have found character clines and mismatches, given that there are no geographic barriers separating them.”

Comments from Jaramillo:

“A1. YES seems like a solid decision.

“A2. YES, although I think we are getting to the saturation point on using Tepui XXX as a name.

“B1. YES, sympatry counts for a lot!

“B2. YES, although note that HMW calls it Campina Thrush. What is the proper version, Campina or Campinas?”

Comments from Pacheco: “YES to A-B. The available data are robust to support both proposals.”

Comments from Stiles: “Here, a comment to hopefully ease Nacho’s doubts..

“I have numerous observations and specimens of T. i. debilis from much of Eastern Colombia, where it seems to be common year-round and sedentary. Data are fewer for T. arthuri, but are scattered through the year, also no evidence for migratory behavior. The breeding season of debilis is from late January or February through June or July; specimens of arthuri in breeding condition from April and June. Molting specimens of adult debilis are from about July through at least October, while a molting specimen is from October. Hence, the annual cycles of the two approximately coincide, but no suggestion of intermediate specimens, hence reproductive isolation seems likely. They segregate by habitat: arthuri is always in matorral or savanna over white-sand soils; debilis only occurs in or near such habitats where human disturbance is long-standing (at least 40-50 yr at Araracuara), and the Inírida área has both white-sand savanna (whence arthuri was taken and alluvial soils from the white-water Guaviare river (exact collecting site not known but probably near the town). Whether the two differ vocally remains to be determined; I leave this for someone with a better ear than mine).

“However, any song recorded from a white-sand locality (among the various recordings ascribed to debilis) is likely to be arthuri.”

Comments from Remsen: “A1. YES. Required by genetic data. A2. NO. I’m making a unilateral decision as Chair to stay with Pantepui as in Avendaño et al., and as mentioned by Nacho and Alvaro. Also, this is not a true parent-daughter split, so no need to worry about creating new name for ignobilis.

“B1. YES. Evidence of true sympatry is sufficient evidence for species rank; Gary’s comments also clarify the possibility of seasonal-only overlap, which I mentioned in. my original comments on the earlier version. B2. YES. This is a true parent-daughter split, but the ranges of ignobilis sensu lato and arthuri differ so vastly in areal extent that we can justify retaining Black-billed for the much more widespread and familiar ignobilis. However, as per Alvaro’s point above, I’m making another unilateral decision to go with Campina because (1) that name is evidently already in use, and (2) the singular is more appropriate, e.g. we don’t go with “Punas” Tinamou, “Selvas” Cacique, “Campos” Troupial, etc., much less “Brushlands” Tinamou etc. Proposals to revert to Tepui or Campinas are welcomed.”